Exploiting Our Contemporary Understanding of the Molecular Pharmacology of the Estrogen Receptor to Develop Novel Therapeutics By

Home , Tamoxifen

Exploiting Our Contemporary Understanding of the Molecular Pharmacology of the Estrogen Receptor to Develop Novel Therapeutics

Kaitlyn Jo Andreano

Department of Pharmacology and Cancer Biology Duke University

Date:______Approved:

______Donald McDonnell, Supervisor

______Kris Wood, Chair

______Cynthia Kuhn

______James Alvarez

______Jeffrey Marks

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pharmacology and Cancer Biology in the Graduate School of Duke University

2020

ABSTRACT

Exploiting Our Contemporary Understanding of the Molecular Pharmacology of the Estrogen Receptor to Develop Novel Therapeutics by

Kaitlyn Jo Andreano

Department of Pharmacology and Cancer Biology Duke University

Date:______Approved:

______Donald McDonnell, Supervisor

______Kris Wood, Chair

______Cynthia Kuhn

______James Alvarez

______Jeffrey Marks

An abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pharmacology and Cancer Biology in the Graduate School of Duke University

2020

Abstract

The estrogen receptor (ER/ESR1) is expressed in the majority of breast and gynecological cancers. As such, drugs that inhibit ER signaling are the cornerstone of pharmacotherapy for these malignancies. Treatment strategies include the Selective

Estrogen Receptor Modulator (SERM) tamoxifen, which acts as a competitive antagonist, and aromatase inhibitors (AIs) drugs that inhibit the enzyme responsible for the production of 17-β estradiol (E2), the most biologically important estrogen. However, the clinical utility of these treatment strategies are limited by the development of de novo and acquired resistance. The mechanisms underlying resistance to these endocrine therapies are varied and complex include activating genomic alterations in ER (amplification, translocations, and mutations), cell cycle dysregulation and activation of alternative growth factor signaling pathways. Interestingly, it has been observed that ER signaling remains engaged and targetable in the majority of these tumors at all stages of disease.

As such, the selective estrogen receptor downregulator (SERD) fulvestrant, which is both a competitive antagonist and downregulator of ER, is often used to treat tumors progressing on AIs or tamoxifen. However, the unfavorable pharmacokinetic properties of this drug have largely limited its use as a monotherapy creating a need for additional

ER-modulators.

The field has put much effort into developing orally bioavailable, next-generation

SERDs to replace fulvestrant in advanced breast cancer. However, many early efforts to optimize compounds for their degradation activity has not yielded clinically useful drugs.

Notwithstanding issues related to drug exposure which may have impacted efficacy there is significant data to suggest that “antagonist activity” is the primary driver of SERD efficacy. To address the need to replace or optimize fulvestrant therapy for advanced breast cancer we undertook both unbiased and biased approaches to define new therapeutic strategies that target ER.

In the first set of studies, we investigated the impact of mutations in ESR1, which occur in metastatic lesions, may have on receptor pharmacology. Specifically, activating point mutations within the ligand binding domain (LBD) of ESR1 have presented as a mechanism of acquired resistance to AIs in metastatic breast cancer; as well as in both de novo and acquired resistance in primary gynecological cancers. Interestingly, these mutations are also resistant/partially resistant to many clinically relevant SERMs and

SERDs, including tamoxifen and fulvestrant. Therefore, we undertook a study to elucidate the molecular mechanism(s) underlying ESR1 mutant pharmacology in relevant models of breast cancer. These studies revealed, unexpectedly, that the response of ESR1 mutations to various ligands was dictated primarily by the relative coexpression of ERWT in cells. Specifically, altered pharmacology was only evident in cells in which the mutants were overexpressed relative to ligand-activated ERWT. Importantly, while undertaking an

unbiased approach to evaluate all clinically relevant antagonists for activity on the ESR1 mutants, we made the serendipitous discovery that the antagonist activity of the SERM lasofoxifene was not impacted by mutant status. This finding has led to its clinical evaluation as a treatment for patients with advanced ER-positive breast cancer whose tumors harbor ESR1 mutations, with additional studies in patients with gynecological cancer patients likely to be undertaken in the near future.

In addition to the unbiased approach outlined above we also approached the problem of resistance taking a candidate approach to evaluate structurally distinct SERDs, as monotherapy and in combination with CDK 4/6 inhibition, in relevant models of advanced breast cancer. G1T48 is a novel orally bioavailable, non-steroidal small molecule antagonist that we demonstrated both in vitro and in vivo has the potential to be an efficacious oral antineoplastic agent in ER positive breast cancer. While G1T48 can effectively suppress ER activity in multiple models of endocrine therapy resistance, this compound still displayed partial resistance to the ERmuts.

Together, our data supports the hypothesis that novel compounds targeting ER should be optimized based on antagonist potential and not on degradative activity per se.

As such, the results of these studies will inform the development of next-generation therapeutics for endocrine therapy resistant cancers, especially those harboring ESR1 mutations.

Dedication

This thesis is dedicated in honor of my late grandmother, Brenda Lee Dunn, whose battle with cancer inspired me to become a cancer researcher.

vii

Contents

Abstract ...... iv

List of Tables ...... xiii

List of Figures ...... xiv

Abbreviations ...... xvi

Acknowledgements ...... xxi

1. Introduction ...... 1

1.1 Thesis Research ...... 1

1.2 The Estrogen Receptor ...... 2

1.3 The role of ER in malignancies of the female reproductive system ...... 6

1.3.1 The Role of ER in Luminal Breast Cancer ...... 6

1.3.2 The Role of ER in Gynecologic Malignancies ...... 7

1.4 ER as a therapeutic target for malignancies of the female reproductive system .... 8

1.4.1 Standard of Care therapeutic strategies to target the ER signaling axis ...... 8

1.4.2 Optimization of treatment strategies to combat endocrine resistance ...... 10

1.4.3 ESR1 mutations ...... 13

1.4.4 Next-generation SERDs for the treatment of endocrine resistant cancers ...... 17

1.4.5 Factors that govern SERD efficacy: implications for future drug selection ...... 21

1.4.6 High affinity SERMs as a viable therapeutic option ...... 24

1.4.7 Problems that will be addressed ...... 26

2. Allelism dictates ESR1 mutant pharmacology in breast cancer ...... 27

viii

2.1 Introduction ...... 27

2.2 Results ...... 30

2.2.1 The expression of clinically relevant ERmuts does not alter the pharmacology of ER ligands in cells expressing ERWT...... 30

2.2.2 The antagonist potency of SERDs and SERMs is reduced in cells expressing ERmuts alone...... 39

2.2.3 ER ligands exhibit subtle differences in their ability to facilitate the interaction of ERmuts with coregulators...... 42

2.2.4 The altered pharmacology of ERmuts is only evident when their expression in cells exceeds that of the WT receptor...... 46

2.3 Discussion ...... 53

3. Discovery and treatment of ESR1 mutations in gynecological cancers ...... 60

3.1 Introduction ...... 60

3.2 Results ...... 62

3.2.1 ESR1 genomic profiles in gynecological malignancies ...... 62

3.2.2 Clinical relevance of ERmuts in gynecological malignancies and response to treatment ...... 67

3.2.3 ERmuts confer partial resistance to endocrine therapy in ovarian cancer cells .. 70

3.3 Discussion ...... 74

4. Characterization of a novel SERD for the treatment of endocrine progressing breast cancer ...... 80

4.1 Introduction ...... 80

4.2 Results ...... 82

4.2.1 G1T48 is similar to fulvestrant in its ability to downregulates the estrogen receptor and inhibit estrogen signaling in breast cancer cells ...... 82

4.2.2 G1T48 inhibits the growth of ER positive breast cancer cells ...... 88

4.2.3 G1T48 inhibits estrogen signaling in endocrine-resistant breast cancer models ...... 90

4.2.4 Evaluation of the in vivo therapeutic efficacy of the SERD G1T48 and the CDK4/6 inhibitor lerociclib using breast cancer xenograft models of estrogen- dependent MCF7 and tamoxifen-resistant (TamR) ...... 93

4.2.5 Evaluation of the combined efficacy of lerociclib and G1T48 in a xenograft tumor model of resistance to estrogen deprivation in vivo ...... 97

4.2.6 Evaluation of the combined efficacy of lerociclib and G1T48 in a Patient Derived Xenograft Model harboring the ERY537S Mutation ...... 98

4.3 Discussion ...... 101

5. Conclusions: Future Directions and Implications ...... 106

5.1 Remaining mechanistic questions that will have implications for preclinical drug development ...... 106

5.1.1 Investigating the role of active ERWT to “normalize” the activity of the ERmuts ...... 106

5.1.2 Mechanistic explanation for the lack of an impact of ESR1 mutants on lasofoxifene potency...... 108

5.1.3 ERmut pharmacology as a mediator of tumor cell/immune cell crosstalk and the promotion of metastasis ...... 110

5.1.4 The role of growth factors in future SERD (or SERM) development ...... 112

5.2 Clinical Implications ...... 113

5.2.1 Investigating the single cell allelic frequency of ERmuts in patient tumor samples ...... 113

5.2.2 Dissecting the differences between diagnosis of breast or gynecological cancers harboring the ERmuts ...... 114

5.2.3 Impact of this thesis work on current Clinical Trials ...... 116

Appendix: Materials and Methods ...... 117

A.1 Chemicals and Ligands ...... 117

A.2 Generation of ERmut expression constructs ...... 117

A.3 Cell Culture ...... 118

A.4 Luciferase Reporter Assays ...... 119

A.5 siRNA Transfection Assay ...... 120

A.6 Cofactor Profiling ...... 120

A.6.1 Transfection experiment ...... 120

A.6.2 Cofactor Profiling Peptide Acquisition ...... 122

A.7 Proliferation ...... 124

A.8 In-Cell Westerns ...... 125

A.9 Immunoblots ...... 126

A.10 Identification of ERmuts in Gynecological Cancers ...... 126

A.10.1 Comprehensive Genomic Profiling: ...... 126

A.10.2 Identification of ESR1 mutations in public databases ...... 128

A.10.3 Clinical Evaluation of Gynecologic Malignancies with ERmuts ...... 129

A.11 qPCR and RNA profiling ...... 129

A.12 Radioactive Binding Assay ...... 130

A.13 Chromatin Immunoprecipitation (ChIP): ...... 131

A.14 Animal Studies ...... 133

A.14.1 MCF7 Naïve Tumor Studies ...... 133

A.14.2 TamR Tumor Studies ...... 133

A.14.3 LTED Tumor Studies ...... 135

A.14.4 PDX Tumor Study ...... 136

A.15 Statistics ...... 137

A.15.1 Dose response curve statistics ...... 137

A.15.2 Animal Statistics ...... 137

References ...... 138

Biography ...... 157

xii

List of Tables

Table 1: Mechanisms Associated with Resistance to Frontline Endocrine therapies ...... 11

Table 2: Transcriptional IC50 values (M) of antagonists in MCF7B Cells ...... 34

Table 3: GI50 values (M) of antiestrogens in MCF7B Cells ...... 35

Table 4: GI50 values (M) of antiestrogens in T47D cells ...... 38

Table 5: Transcriptional IC50 values (M) of antiestrogens in SKBR3 cells ...... 42

Table 6: Transcriptional IC50 values (M) of antiestrogens in MCF7I cells ...... 49

Table 7: Types and frequency (%) of ESR1 alterations identified in gynecologic malignancies by primary site ...... 63

Table 8: ERmuts identified in gynecologic malignancies by histological subtype ...... 64

Table 9: Clinical Characteristics of patients identified with ERmuts in gynecological malignancies ...... 70

Table 10: IC50 and IC90 values of antagonists (pM) ...... 74

Table 11: ER degradation IC50 values in MCF7 cells ...... 85

Table 12: Radioactive Binding Assay IC50 values ...... 88

Table 13: GI50 Values of antiestrogens in Breast Cancer Cell Lines ...... 89

Table 14: ERmut and ERWT transcriptional IC50 (M) Values ...... 92

Table 15: ER Targeted-Compounds Currently in Clinical Trials for Endocrine Therapy Resistant Breast Cancer ...... 116

Table 16: Primer sequences for generation of ERmut expression constructs ...... 118

Table 17: Experimental Conditions for Luciferase Reporter Assays ...... 119

Table 18: siRNA sequences ...... 120

Table 19: Peptide Sources and Sequences ...... 122

xiii

List of Figures

Figure 1: Schematic Illustration of ER modular structure...... 3

Figure 2: Mechanisms of ER activation...... 5

Figure 3: Classes of antagonists that target ER signaling...... 9

Figure 4: Single cell receptor allelism likely impacts response to therapy...... 16

Figure 5: Structural determinants of SERD classifications...... 18

Figure 6: Structural comparison of Raloxifene and Lasofoxifene to other clinically relevant SERM and SERDs...... 26

Figure 7: Cells expressing both the ERWT and ERmuts have similar pharmacological responses to antiestrogens when compared to cells only expressing ERWT...... 33

Figure 8: Proliferative responses to antiestrogens is indistinguishable in MCF7B cells. .. 35

Figure 9: Albeit differences in clonal variability, there is indistinguishable differences in T47D proliferation in response to antiestrogens...... 37

Figure 10: Validation of ER mutation status in engineered cell lines...... 39

Figure 11: ERmuts confer antiestrogen resistance when expressed alone...... 41

Figure 12: Differential cofactor recruitment reveals modest changes in overall receptor conformation between the WT and mutant receptors...... 45

Figure 13: The altered pharmacology of ERmuts can be manipulated by their expression level...... 48

Figure 14: The altered pharmacology of ERmuts is only evident when expressed at a level higher than the WT receptor...... 52

Figure 15: Schematic overview of ERmuts identified in gynecologic malignancies...... 66

Figure 16: Clinical relevance of ERmuts in gynecologic malignancy...... 68

Figure 17: ER LBD mutations confer constitutive transcriptional activity and alter receptor sensitivity to SERMs/SERDs...... 73

xiv

Figure 18: G1T48 is a potent SERD...... 84

Figure 19: G1T48 is a complete estrogen receptor antagonist...... 87

Figure 20: G1T48 inhibits ER- positive breast cancer cell growth...... 89

Figure 21: G1T48 inhibits ER signaling in models of endocrine therapy resistance in vitro...... 92

Figure 22: Combination strategy of G1T48 and the CDK4/6 inhibitor lerociclib inhibit in vivo breast cancer xenograft models of estrogen-dependent MCF7 and tamoxifen- resistant (TamR)...... 95

Figure 23: Analysis of intratumoral ESR1 protein levels in harvested tumor tissue (TamR)...... 96

Figure 24: Combination strategy of G1T48 and the CDK4/6 inhibitor lerociclib in vivo in an estrogen deprived xenograft model...... 98

Figure 25: Evaluation of the combined efficacy of lerociclib and G1T48 in a Patient Derived Xenograft Model harboring the ESR1 Y537S Mutation...... 100

Figure 26: Receptor dimerization is one possible mechanism by which activated ERWT normalizes cellular response to antiestrogens in ERmuts expressing cells...... 107

Figure 27: Dimerization Hypothesis Experimental Design...... 108

Abbreviations

3’UTR Three Prime Untranslated Region

4-OHT 4-hydroxytamoxifen

AACR American Association of Cancer Research

AF Activation Function

AI Aromatase Inhibitor

β-gal Beta-Galactosidase

BSA Bovine Serum Albumin

CBD Clinical Benefit Duration

CBP CREB-binding protein

CA125 Cancer Antigen 125

CBX Cell Based Xenograft cDNA Complementary Deoxyribonucleic Acid

CFS Charcoal- Stripped Bovine Serum

CGP Comprehensive Next-Generation Genomic Profiling

ChIP Chromatin Immunoprecipitation

Cmax Maximum achievable serum concentration

CRGA Clinically Relevant Genomic Alterations

Ct Threshold Cycle

CMC Carboxymethyl Cellulose

xvi

CoA Coactivator

CONFIRM Comparison of Fulvestrant 250mg and 500mg in Postmenopausal Women with Estrogen Receptor Positive Advanced Breast Cancer Progressing or Relapsing After Previous Endocrine Therapy

COSMIC Catalogue of Somatic Mutations in Cancer

CoR Corepressor

CORNR Corepressor/Nuclear Receptor Interaction Motif

CYP19A1 Cytochrome P450 Family Member 19 Subfamily A Member 1 (Aromatase)

DBD DNA Binding Domain

DNA Deoxyribonucleic Acid

E2 17 β-estradiol

EFECT The Evaluation of the Efficacy and Tolerability of Fulvestrant and Exemestane in Hormone Receptor Positive Postmenopausal Women with Advanced Breast Cancer

EIP Estrogen Receptor Interacting Peptide

ELAINE Evaluation of Lasofoxifene versus Fulvestrant in Advanced or Metastatic ER+/HER2- Breast Cancer with an ESR1 Mutation

ER Estrogen Receptor

ERα/ESR1 Estrogen Receptor Alpha (protein and gene abbreviation)

ERβ/ESR2 Estrogen Receptor Beta (protein and gene abbreviation)

ERE Estrogen Responsive Element

ERmut Ligand Binding Domain Mutants of Estrogen Receptor Alpha

xvii

ERWT Wild-type Estrogen Receptor Alpha

FBS Fetal Bovine Serum

FERGI Study of GDC-0941 or GDC- 0980 (pan-PI3K inhibitors) with Fulvestrant versus Fulvestrant in Advanced or Metastatic Breast Cancer Participants Resistant to Aromatase Inhibitor Therapy

FFPE Formalin Fixed Paraffin Embedded Tissue blocks

GENIE Genomics Evidence Neoplasia Information Exchange

GI50 Drug concentration giving a 50% reduction in growth or proliferation

GRE Glucocorticoid Response Element

GSK-3 Glycogen Synthase Kinase-3

H12 Helix 12 of the Estrogen Receptor Alpha Ligand Binding Domain

IC50 Drug concentration giving a 50% reduction in transcriptional activity

IC90 Drug concentration giving a 90% reduction in transcriptional activity

IUCAC Institutional Animal Care and Use Committee

LBD Ligand Binding Domain

LTED Long-Term Estrogen Deprivation MCF7 model

MAF Mutant Allele Frequency

MAPK Mitogen-Activated Protein Kinase

MCF7B MCF7 subclones derived in Myles Brown’s lab to express wild- type and mutant Estrogen Receptor

xviii

MCF7I MCF7 cell lines derived to induce wild-type or mutant Estrogen Receptor expression in response to Doxycycline

MONALESSA-3 Study of Efficacy and Safety of LEE0011 (Ribociclib) in Men and Postmenopausal Women with Advanced Breast Cancer

MONARCH-2 A Study of Abemaciclib (LY2835219) Combined with Fulvestrant in Women with Hormone Receptor Positive HER2 Negative Breast Cancer mTOR Mammalian Target of Rapamycin

NCOR Nuclear Receptor Corepressor

PALOMA-3 Palbociclib (PD-0332991) Combined with Fulvestrant in Hormone Receptor- Positive HER2- Negative Metastatic Breast Cancer After Endocrine Therapy Failure

PBST Phosphate Buffered Saline with Tween 20

PCR Polymerase Chain Reaction

PDX Patient Derived Xenograft

PEARL Postmenopausal Evaluation and Risk-reduction with Lasofoxifene

PEG Polyethylene Glycol

PI3K Phosphoinositide 3-Kinase

PI3KCA PI3K Catalytic Subunit Alpha

PNK T4 Polynucleotide Kinase

PVP Polyvinylpyrolidone

SERCA Selective Estrogen Receptor Covalent Antagonist

SERD Selective Estrogen Receptor Degrader

xix

SERM Selective Estrogen Receptor Modulator siRNA Small Interfering Ribonucleic Acid

SMRT Silencing Mediator of Retinoic Acid and Thyroid Hormone Receptor

SRC Steroid Receptor Coactivator

STAR Study of Tamoxifen and Raloxifene for the Prevention of Breast Cancer in Postmenopausal Women

START South Texas Accelerated Research Therapeutics

TamR Tamoxifen Resistant MCF7 Model

TCGA The Cancer Genome Atlas

TF Transcription Factor

TFF1 Trefoil Factor 1

Acknowledgements

I am incredibly grateful for the support and encouragement I received both inside and outside the lab throughout my Ph.D. training. Donald McDonnell has been a great mentor. I am extremely thankful for all the opportunities gain experience within drug development that have been provided to me as result of working in Donald’s lab. I want to thank my committee members, Kris Wood, James Alvarez, Cindy Kuhn and Jeff Marks for their helpful suggestions and insightful discussions on my many projects throughout my training. I thank all the members of the lab for their insightful discussions throughout my graduate school career. I especially thank the most senior members of the lab, John

Norris, Suzanne Wardell, Ching-yi Chang and Rachid Safi, for their guidance throughout this journey. I want to thank my collaborators on the projects presented in this thesis, especially John Norris, Suzanne Wardell and Stephanie Gaillard, for their contributions to this body of work.

Thank you to my family and friends for their undying support and encouragement. I want to especially thank my “lab bestie” Taylor Krebs for being such an amazing friend and support system throughout this journey, I mean it when I say I would have never made it without you! A special shout out to my amazing parents, Jeff and Kelly

Andreano, for their endless support, encouragement, and sacrifices throughout my life as

I pursued my many dreams, I owe all my success to you!

xxi

1. Introduction

1.1 Thesis Research

The overarching goal of my thesis research was to leverage our most contemporary understanding of the mechanisms that determine estrogen receptor (ER) pharmacology to define new therapeutic strategies for the treatment of cancers that impact the female reproductive system. A central theme was to understand how ER ligand binding domain mutations, found in patients with advanced metastatic disease who have progressed on frontline endocrine therapy, impact ligand pharmacology and therapeutic response. Specifically, the molecular mechanisms underlying the unique pharmacology of these mutants was dissected with a view to informing the identification of the next generation of ER modulators for use in the treatment of tumors that harbor one or more of these mutations. In addition, following our discovery that these ER mutations are also present in gynecological cancers, we expanded our studies and exploited our knowledge of mutant activity in breast cancer to develop therapeutics approaches for an expanded array of estrogen regulated cancers. A secondary objective of this work was to define the key mechanistic features that contribute to antagonist efficacy on both wild-type and mutant ER. This work culminated in the identification of a new utility for a previously discovered drug and its repurposing as a breast cancer therapeutic.

1.2 The Estrogen Receptor

The physiologically relevant estrogens include 17 β-estradiol (E2), estrone and estriol.

These hormones regulate processes of biological importance including the growth and development of the human reproductive, neuroendocrine, skeletal, adipose and cardiovascular systems. E2, the most potent of these estrogens, is synthesized primarily in the ovary [1]. Testosterone is the immediate precursor of E2, the conversion of which is catalyzed by the CYP19A1 (aromatase) enzyme [2]. In the liver and other peripheral tissues, E2 can be converted to estrone and estriol and is frequently conjugated by esterification to sulfates for elimination. After the ovaries cease to produce E2 following menopause, estrone, produced through the conversion of androstenedione in fat tissues, becomes the predominant estrogen [1, 3].

The Estrogen Receptors (ER) are members of the nuclear hormone receptor superfamily of ligand activated transcription factors [4]. There are two genetically distinct isoforms of ER; ERα (ESR1) and ERβ (ESR2). ERα and ERβ are structurally similar and contain an N-terminal A/B domain, a central zinc finger DNA binding domain (DBD), a hinge region (D Domain), and the C-terminal E domain, which contains the ligand binding domain (LBD) (Figure 1) [5, 6]. The LBD domain consists of a helical sandwich, formed by 12 α- helices linked by loop regions to form a ligand binding pocket. The highly

conserved DBD and LBD enable ERα and β to bind the same ligands and bind to identical

DNA response elements within target genes[5, 6].

AF-1 DBD Hinge LBD/ AF-2 ERα A/B C D E F

ERβ 18% 97% 24% 58% 12%

Figure 1: Schematic Illustration of ER modular structure.

Percentages indicate homology between ERα and ERβ

The transcriptional activity of ER is facilitated by specific “activation function” domains (AF). ERα contains two AF domains (AF-1 located in the A/B region, and AF-2 located in the LBD) while ERβ only contains the AF-2 region located in the LBD [7].

Beyond these differences, the actions of ERα and ERβ are distinguishable by their tissue distributions; ERα is widely expressed while ERβ expression is more restricted and is found primarily in the uterus, blood, monocytes, and colonic and pulmonary epithelial cells [8, 9]. The roles of ERβ in biological and pathological contexts remain controversial; therefore, the focus of this work has been ERα (which will henceforth be referred to simply as ER).

ER can be activated by both ligand-dependent and -independent mechanisms [10, 11].

Ligand-dependent mechanisms involve conformational changes within the LBD. In the absence of ligands (apo-receptor), the LBD remains in the cytoplasm in an inactive conformation. However, in response to estrogenic ligands, ER undergoes a conformational change that results in its homodimerization [11]. The structure of the LBD is thus stabilized and the terminal helix (H12) folds over the ligand binding pocket, allowing the receptor to bind DNA directly through estrogen response elements (ERE) or indirectly through interaction with other DNA bound transcription factors [11]. Ligand- independent activation can occur through phosphorylation of the receptor by several cellular pathways, including MAPK and PI3K [10]. Figure 2 summarizes the ligand- dependent and independent mechanisms of ER activation.

A) CoA CoA ER ER

ER ER ER ER ER Estrogens B) TF

C) Phosphorylation CoA CoA ER ER K ER

Figure 2: Mechanisms of ER activation.

(A) Classically, ER is activated by the action of estradiol (E2), bound intracellularly, leading to the dimerization and nuclear transport of ER. (A, B) Once nuclear, ER can interact with coactivators (CoA) on directly on DNA, or it can associate indirectly with DNA via binding to other transcription factors (TF). (C) ER action can also be influenced by the activation of intracellular kinases (K), which can phosphorylate ER or its interacting cofactors.

Regardless, of the mechanism of activation, a key defining feature of ER action is its ability to interact with transcriptional coregulators and nucleate the assembly of regulatory complexes that positively or negatively regulate transcription. The primary transcriptional coactivators of ER are the p160 family of proteins (SRC-1, SRC-2 and SRC-

3), p300 and CREP binding protein (CBP) [12-14]. ER can repress transcription by recruiting corepressors such as silencing mediator of retinoic acid and thyroid hormone

receptors (SMRT) and nuclear co-repressor (N-CoR). The orientation of H12, which is influenced by the nature of the bound ligand and its structure, is a key determinant of receptor coregulator preferences and that which determines receptor pharmacology [13,

15, 16]. Coregulators can bridge interactions with the general transcriptional machinery or modulate the chromatin landscape to influence downstream ER activity. It is now appreciated that cofactors recruitment process is regulated by cell-context and thus the same ligand can yield different responses in different cells [17].

1.3 The role of ER in malignancies of the female reproductive system

1.3.1 The Role of ER in Luminal Breast Cancer

According to the American Cancer Society, breast cancer is the most common cancer diagnosis among women with one in eight developing breast cancer in her lifetime

[18]. Unfortunately, despite advances in early detection strategies and treatment options, breast cancer remains the second highest contributor of cancer related deaths in women behind lung cancer. Breast cancer is a genetically and phenotypically diverse set of diseases and which are characterized by their receptor (ER, PR, Her-2) status. These subtypes include luminal A, luminal B, Her-2 enriched, basal-like and normal-like [19, 20].

The luminal cancers are marked by the expression of ER and its target gene the progesterone receptor (PR). These are the most common breast cancers, accounting for

roughly 70% of all cases [21]. Canonical ER target genes have been identified to regulate cancer cell proliferation and survival, highlighting how dysregulation of the ER pathway may contribute to cancer growth. As such, drugs that inhibit the ER signaling axis remains the cornerstone for breast cancer pharmacotherapy.

1.3.2 The Role of ER in Gynecologic Malignancies

Gynecological malignancies include cervical, ovarian, uterine, vaginal and vulvar cancers. Uterine and Ovarian Cancers are the two biggest contributors to the overall population of ER-positive gynecological malignancies. Specifically, according to the

American Cancer Society (2018), uterine cancer (also called endometrial cancer) is the fourth most commonly diagnosed and the sixth highest contributor of cancer related deaths in women [18]. In endometrial cancer, the progression of Type I tumors (90% of tumors) is associated with unimpeded estrogen signaling. Normally, the progesterone receptor (PR) will negatively regulate this signaling axis and this level of regulation is often lost in these cancers.

Although not as common as endometrial cancer, ovarian cancer is the fifth largest contributor of cancer related deaths in women [18]. High Grade Serous Carcinoma is the most commonly diagnosed Ovarian Cancer and 80% of these tumors express ER, making it a viable therapeutic target for many ovarian cancer patients as well [22].

1.4 ER as a therapeutic target for malignancies of the female reproductive system

1.4.1 Standard of Care therapeutic strategies to target the ER signaling axis

Among the interventions most commonly used to target the ER signaling axis are aromatase inhibitors (AIs), Selective Estrogen Receptor Degraders (SERDs), and Selective

Estrogen Receptor Modulators (SERMs). AIs (letrozole, anastrozole, or exemestane) are competitive inhibitors of aromatase (CYP19A1), the enzyme that converts androgens into estrogens [23]. SERDs (like fulvestrant) are drugs that function primarily as competitive inhibitors of ER, but also induce a conformational change that targets the receptor for proteasomal degradation [24, 25]. SERMs (like tamoxifen) are drugs which function as ER antagonists in breast cancer cells but can function agonists in other tissues (e.g. bone) [26-

28] (Figure 3).

A) B) C)

Cytoplasm Nucleus

Aromatase X CoA CoA ER ER ER ER ER Androgens Estrogens

Marks receptor AIs For degradation SERDs SERMs

Figure 3: Classes of antagonists that target ER signaling.

There are currently three methods to target the ER signaling axis. (A) First, AIs work by blocking the conversion of androgens to estrogens thus depleting the receptor of its agonists. (B) Secondly, SERDs bind to the receptor and mark it for degradation. (C) The third are SERMs that act as antagonists in the breast and agonists in other tissues such as the bone. AIs and the SERM tamoxifen are frontline therapies, while the SERD fulvestrant is second line.

In luminal breast cancer, it is now standard practice to use AIs as frontline endocrine therapy in postmenopausal patients or in high-risk premenopausal patients combined with ovarian suppression [23]. Although previously a standard of care treatment for breast cancer, tamoxifen is now primarily used for the adjuvant treatment of premenopausal breast cancer patients at low-risk for recurrence with or without ovarian suppression [26, 27]. These treatment options are still widely applicable to gynecological cancers, however endocrine therapies are normally used as a second line therapy after chemotherapy and surgery to remove the tumor and organ. Endocrine therapies for gynecologic cancers include tamoxifen, AIs and progestins [29, 30].

The SERD fulvestrant is used in patients who progress on frontline endocrine

therapies and is given as monotherapy or in combination with other targeted therapies

[25]. Currently, fulvestrant is the only clinically approved SERD for cancer therapy. This high-affinity ligand is a very effective inhibitor and downregulator of ER expression in preclinical models. However, its clinical utility is limited by its poor pharmacokinetic and pharmacodynamics properties [25, 31, 32]. The optimization or replacement of fulvestrant as the standard of care for endocrine progressing cancers will be further discussed throughout the rest of this chapter.

1.4.2 Optimization of treatment strategies to combat endocrine resistance

While AIs and tamoxifen have had a very significant impact on disease-free and overall survival in patients with ER-positive breast cancer, de novo and acquired resistance to either type of drug remains a noteworthy clinical issue. Specifically, resistance to these therapies occurs in 20% and 33% of cases (for tamoxifen and AIs, respectively)[21, 33-35].

Table 1 summarizes the known mechanisms of resistance to tamoxifen and AIs [21, 33-

36].

Table 1: Mechanisms Associated with Resistance to Frontline Endocrine therapies

Pathways Examples Tamoxifen or AI resistance ER signaling ER loss Tamoxifen and AI ER amplification AI ER mutation or translocation AI ER truncation Tamoxifen ER phosphorylation Tamoxifen ER methylation Tamoxifen Expression of other NRs Tamoxifen ER associated AP1 overexpression Tamoxifen cofactors and Novel interaction with GRHL2 Tamoxifen transcription NF-κB activation Tamoxifen factors Aberrant expression or mutation in ER Tamoxifen and AI coregulators (Examples include: SRC-3, CBP, p300) Growth Factor EGFR overexpression or mutation Tamoxifen and AI Signaling ERBB2 amplifcation, de-repression or mutation Tamoxifen and AI IGFR1 overexpression or mutation Tamoxifen and AI FGFR overexpression Tamoxifen MAPK Signaling Mek and Erk activation Tamoxifen and AI PI3K Signaling PI3KCA mutation AI PTEN loss or mutation Tamoxifen and AI Akt (AKT1) activation, mutation or Tamoxifen and AI overexpression SRC Signaling SRC Activation Tamoxifen Cell Cycle RB, p16 and p18 loss AI CCND1 amplification AI TP53 mutation AI MDM2 amplification AI EMT and CSC Notch, Hedgehog, WNT and TWIST1 AI Snail and Slug Tamoxifen Tumor dormancy AI Apoptosis and BCL-2 and survivin activation AI Senescence Telomerase activation AI Tumor ECM: fibronectin and collagen Tamoxifen and AI Microenvironment Interactions with immune cell populations Tamoxifen and AI

Observations that ER remains engaged in the regulation of processes of importance in cancers that have escaped frontline endocrine interventions has led to further exploitation of this receptor as a therapeutic target, specifically through the optimization or replacement of fulvestrant therapy. The EFECT trial (NCT00065325) investigated a low- dose 250 mg fulvestant therapy compared to the AI exemestane in patient tumors resistant to other AIs [37, 38]. This study found that giving fulvestrant was not any better or worse than treating these tumors with exemestane, a disappointing result given that the patients had already progressed during AI treatment. However, the CONFIRM trial

(NCT00099437) demonstrated that simply by increasing the dose of fulvestrant from 250 mg to 500 mg improved survival rates in breast cancer patients whose tumors have progressed on other endocrine therapies [39, 40].

Another method of improving fulvestrant efficacy is its inclusion in combination therapies. Clinical trials using a combination of AI or fulvestrant with pan-PI3K or mTOR inhibitors have been promising but inconclusive, and toxicity is a hurdle to dose escalation[40-43]. Therefore, the combination strategy of fulvestrant and CDK 4/6 inhibitors are often used to combat endocrine progressing cancers. The PALOMA-3

(NCT01942135), MONALESSA-3 (NCT02422615), and MONARCH-2 (NCT02107703) clinical trials were all designed to test the utility of adding CDK 4/6 inhibitors to fulvestrant for the treatment of endocrine progressing cancers [44-47]. The first of these,

the PALOMA-3 trial evaluated the efficacy of combination of CDK 4/6 inhibitor palbociclib with fulvestrant compared to fulvestrant alone, and the overwhelming benefit observed led to fast approval by the FDA. The results of this study demonstrated both an overall survival benefit and a significant progression free survival rate.

1.4.3 ESR1 mutations

Whereas the mechanisms underlying resistance are diverse, it is now clear that gain of function point mutations within the ligand binding domain (LBD) of ESR1 likely contribute to the development of resistance to AIs and signify an important clinical problem [48-53]. Although rare in primary breast tumors, mutations in the ESR1 LBD

(ERmut) occur in up to 40% of metastatic lesions, a finding that is consistent with their selection in conditions of estrogen deprivation [49-51, 53-60]. The relevance of these mutations in gynecological cancers will be discussed in Chapter 4. Interestingly, progression of endocrine therapy resistance tends to result in mutations in ER as opposed to other commonly mutated resistance drivers including PIK3CA mutations [51].

Hot spots for these mutations are amino acids 536, 537, and 538 within the ligand binding domain. The Y537S and D538G mutations (ERY537S and ERD538G) are the most prevalent and account for 70% of all cases [49-51, 53-60]. Preclinically, it has been demonstrated that these mutations drive ER-dependent transcription, proliferation, and

tumor cell migration in the absence of hormone [48-53]. Additionally, it has been demonstrated that these mutations regulate a neomorphic gene set, not associated with the wild-type ER (ERWT), that drives a metastatic phenotype [48, 61]. Importantly, several reports have demonstrated that these mutant ERs may exhibit partial resistance

(decreased potency) to standard hormonal therapies, including tamoxifen and fulvestrant

[48-52]. This resistance was first uncovered in 1997 when Benita Katzenellenbogen’s lab characterized mutations in the Y537 amino acid to understand the structure activity relationship of ER and ER ligands. At this time, it was suggested that these mutations result in constitutive activity and decreased the potency of the tamoxifen metabolite 4- hydroxytamoxifen, (4-OHT)[62]. However, these mutations were not further investigated or appreciated until 2013, when two independent reports emerged indicating that receptor mutations were present in metastatic lesions of patients who progressed on AIs

[50, 51]. Since then, the field has been focused on developing appropriate pharmaceutical approaches to target these mutations.

Early reports that informed our current understanding of the pharmacology of ERmuts in breast cancer cells were performed in model systems in which the mutants were expressed in the absence of ERWT [49-52, 63]. Specifically, in ER-negative cell lines, ERWT or

ERmut cDNA plasmids were overexpressed with a reporter gene under the control of an

ER response element to evaluate transcriptional repression in response to relevant

ligands. Recognizing that this approach does not take into account the heterogeneity of

ERWT/ ERmut expression in advanced ER-positive breast tumors cells that results from the selective pressure of endocrine therapy, later reports utilized ER-positive cells that had been genetically engineered to concurrently express both ERWT and ERmut [48, 64-68]. As these mutations are mainly prevalent in patients that have undergone AI therapy, these later reports studied this biology in the absence of E2 activation of ERWT. Therefore, these studies left untested the possibility that the in the presence of E2, the relative expression levels of the ERWT and ERmut may dictate response to therapeutics.

The clinical data that has investigated the role of these mutations in resistance to fulvestrant therapy suggests the importance of receptor allelism. First, the findings from the FERGI trial (NCT01437566) demonstrated that ERmut status did not impact progression-free survival in response to fulvestrant therapy when the median ESR1 mutant allele frequency was low (0.45%) [57]. Conversely, in the PALOMA- 3 study

(NCT01942135), there were observed differences in fulvestrant progression-free survival in response to mutation status (allele frequency of 10%) [59, 60]. This study also suggested that ERmut containing clones were a small fraction of the whole tumor and as such the low allele frequency estimate was not representative of each individual cell (Figure 4). Finally, the likely importance of ERmut allelism was suggested in a recent study that revealed a propensity for a loss of heterozygosity (LOH) of ERWT when an ERmut is also present in the

tumors of patients on endocrine therapies[69]. Specifically, in breast cancer patients that harbored ESR1 mutants, LOH of the WT allele drove 78% of ESR1 mutant specific allele balance, while background loss of allele for non-mutant containing tumors also on endocrine therapy was only 30%. These data suggest that the ERWT is important in determining ERmut response to therapy and that tumors having a lower expression ERWT have a survival advantage.

WT WT WT WT

WT Mut WT WT WT Mut WT Mut WT Mut Mut Mut

Mut Mut Mut Mut

Figure 4: Single cell receptor allelism likely impacts response to therapy.

It is not clear how receptor allelism is represented on the cellular level. As such, drug discovery efforts must be geared to targeting all potential expression patterns.

Taken together, these studies highlight the need to better understand the role that functional ERWT plays in the resistance of ERmuts to endocrine therapy. Furthermore, given the pharmacokinetic limitations of fulvestrant therapy, any decreased potency for ER will

likely render fulvestrant irrelevant in the targeting these mutations. As such, these findings highlight the need to further explore therapeutic options, both existing and novel antiestrogens, to better target these receptors regardless of ER status. Studies geared to address these important points will be discussed in subsequent chapters.

1.4.4 Next-generation SERDs for the treatment of endocrine resistant cancers

Fulvestrant is currently the only approved SERD for the treatment of breast cancers that have progressed on frontline endocrine therapies [25]. However, the poor pharmacokinetic and pharmacodynamic properties of fulvestrant limit its clinical utility, especially in the context of its decreased potency against ERmuts [31, 32]. Thus, it has been of great interest in the field to develop orally bioavailable antagonists with SERD activity.

SERDs are generally split into classes that share common chemical features, including (a) a steroidal backbone (fulvestrant), (b) an acrylic acid side chain (GW5638, GW7604, GDC-

0810, AZD9496) or (c)s basic side chains (e.g. bazedoxifene, RAD1901) (Figure 5)[24, 25,

63, 70-74]. Interestingly, both acidic SERDs and basic SERDs are orally bioavailable provide alternatives to fulvestrant, which is administered by intra-muscular depot injection.

Fulvestrant AZD9496 GDC-0810

HO HO O O

H H F OH H N F F H F F N O F F H H Cl S F HO F

N N H

RAD1901 Bazedoxifene GW5638 GW7604

O OH O HO O H N HO HO HO N N O

OH HN

Figure 5: Structural determinants of SERD classifications.

Steroidal SERDs (black) include fulvestrant and contain a steroidal ring structure. Acidic SERDs (red) are classified based on their acrylic side-chain. Basic SERDs (blue) are characterized by having amine-containing side chain.

The first acidic SERDs to be characterized were GW5638 and its higher affinity 4- hydroxylated metabolite GW7604 [72, 74-76]. These compounds are nonsteroidal derivatives of tamoxifen. In vitro GW5638 can inhibit both the agonist activity of E2 and the inverse agonist activity of fulvestrant. Interestingly, the McDonnell laboratory previously demonstrated that the mechanism of ER degradation of GW5638 is functionally distinct from fulvestrant [72]. Unfortunately, these compounds were not developed further as a result of a corporate merger and subsequent portfolio decisions.

Next-generation acidic SERDs include AZD9496 and GDC-0810 are both structurally related to GW5638 [65, 70, 77, 78]. They have been characterized to be potent inhibitors and degraders of ER in breast cancer models. They also demonstrated good pharmacokinetic profiles in murine models. However, while they have been shown to have efficacy in late stage breast cancer (AZD9496 (NCT02248090, NCT03236874) and

GDC-0810 (NCT01823835)), the clinical development of these compounds was halted, due to unanticipated toxicities in the clinic [79, 80]. Currently no data on the effectiveness of these compounds in gynecological cancers is available.

Nonsteroidal ER antagonists that possess basic side chains (e.g. bazedoxifene and

RAD1901) may prove useful in breast cancer. Bazedoxifene is an orally bioavailable, high- affinity SERM that in breast cells displays SERD and antagonist activity, while in other tissues like the bone acts as an ER agonist [63, 81]. It is currently approved for the treatment of osteoporosis or, in combination with conjugated-equine estrogens, for the prevention of menopausal symptoms [27]. Importantly for breast cancer, it displays antagonist activity comparable to fulvestrant in treatment naïve, and endocrine resistant breast models (Tamoxifen Resistant and Long Term Estrogen Deprived/LTED which mimics AI treatment) [63, 81]. Notably, bazedoxifene decreases the risk of endometrial cancer [82]. Bazedoxifene was previously in clinical trials as a treatment option for endocrine progressing breast cancers with palbociclib (NCT02448771). Early studies

suggest that this treatment regimen is well-tolerated and demonstrated significant clinical activity, although further progress on this work remains to be determined and may be deterred by limited patent life.

Preclinically, RAD1901 can inhibit naïve and resistant models of breast cancer [71, 83].

Importantly, RAD1901 demonstrates a very complex and unique pharmacology. This compound demonstrates a “U-shaped” pharmacology such that at low doses it has an agonist profile, while at higher doses it displays both SERD and antagonist activity [71].

Interestingly, RAD1901 is unique among synthetic ER antagonists in its ability to cross the blood-brain-barrier, and as such, is a potential therapeutic option for vasomotor instability (hot flushes) and breast cancer brain metastasis [71]. RAD1901 (NCT02338349) is currently in Phase III clinical development, making it the oral SERD currently closest to

FDA approval.

Unfortunately, recent reports have investigated the clinical utility of many of these compounds (bazedoxifene, RAD1901, AZD9496 and GDC-0810) against ERmuts and have found that they still demonstrate partial resistance (decreased potency) to these antagonists [52, 65, 83, 84]. However, it is important to note that there were limitations of these models, as was previously discussed. Therefore, these data further reinforce the argument that additional work is needed to determine viable therapeutic options for patients with ER mutations. Interestingly, while the field has focused much effort into

developing novel SERDs, it has been previously reported in several contexts that the degradative function of SERDs is actually uncoupled from their antagonist activity.

Additionally, the mechanisms by which different classes of SERDs degrade the receptor are unique, leaving the question open if a novel SERD has any advantage compared to a high affinity antagonist that lacks SERD activity.

1.4.5 Factors that govern SERD efficacy: implications for future drug selection

Most drug discovery efforts that have been undertaken have exploited the use of

ER degradation as the primary mechanism of efficacy. However, one of the most hotly debated issues in the field is whether or not SERDs have any advantage over high affinity competitive ligands without agonist or SERD activity. Thus, understanding the molecular mechanisms that dictate these differences is of great clinical significance.

There is currently mechanistic data in the literature that supports the notion that

ER degradation is not the main driver of SERD mediated antagonism. Previous work in the McDonnell laboratory demonstrates that fulvestrant mediated degradation is a saturable process that is uncoupled from antagonism. Specifically, in overexpression models of breast cancer, Wardell et. al 2011 demonstrates that increased concentrations of

ER lead to a saturation of degradation capacity but not of ER target gene inhibition [85].

Additionally, data from multiple breast cancer cell lines that elucidated the saturable

nature of this process suggested that expression of components of degradation machinery affects receptor turnover potential, which suggests a cell to cell and patient to patient variability in this process. It is important to note that overexpression of ER in these models did not affect E2 (the most effective ER degrader) mediated turnover or activation of the receptor or fulvestrant’s ability to act as a competitor against E2 for ER binding.

Additionally, ER turnover was equally dispensable for fulvestrant inhibition of growth factor signaling. The data from this manuscript supported the hypothesis that high affinity for ER and effective competitive inhibition, albeit ER turnover, are the primary mechanistic drivers of fulvestrant inhibition of ER [85]. This notion was later confirmed with other antagonists with known SERD activity such as bazedoxifene and GW7604 [72,

81].

Recent work from our laboratory using in vivo mouse models have further validated this hypothesis [86]. In this study, we treated animals with a clinically relevant dose of fulvestrant (25 mg/kg), along with the generally used dose (200 mg/kg) and intermediate doses. Interestingly, the low clinically relevant dose of fulvestrant exhibited comparable anti-tumor efficacy, but not robust ER turnover, compared to the much higher widespread used dose. In this low clinically relevant dose, ER turnover actually varied widely between samples despite anti-tumor response, which is something that is observed when comparing pre- and post-treatment biopsies in the clinic [87]. Additionally, when

next-generation SERDs such as AZD9496 and GDC-0810 were compared head-to-head for their anti-tumor efficacy in these models, there was a significant difference in ER turnover

[86]. However, no differences in anti-tumor efficacy, reinforcing that these two mechanisms are uncoupled.

Finally, recent work from Genentech provides a potential mechanistic explanation for these phenomena [88]. In a recent study, Metcalf et al compared their first clinically relevant acidic SERD (GDC-0810), a newer basic SERD (GDC-0927) and a non-degrader that shares structural similarities to GDC-0927 (GNE-274) with other clinically relevant ligands including fulvestrant, 4-OHT and AZD9496. These mechanistic studies highlighted several points relevant to the argument that SERD activity is not the key determinant in antagonist efficacy. First, degradation does not guarantee full antagonist profile, as GDC-0810 displayed partial agonist activity in cellular models of breast cancer.

This compound has also been demonstrated to have its degradation uncoupled from its transcriptional profile. Second, structurally similar compounds that had differential effects on ER turnover such as GDC-0927 and GNE-274 have similar antagonist efficacy on breast cancer cell proliferation. Interestingly, the SERDs fulvestrant and GDC-0927 had a unique impact on ER mobility (specifically through the attenuation of intranuclear diffusion of ER) relative to other compounds including GDC-0810 and AZD9496.

Additionally, partial agonists such as GNE-274 and 4-OHT increase chromatin

accessibility, while full antagonists such as GDC-0927 and fulvestrant do not. Given the rapid nature of this phenomena, ER turnover is unlikely to account for these differences

[88]. These observations are in agreement with previous studies that suggest that fulvestrant mediated cellular recompartmentalization of ER precedes its degradation [72].

Taken together these data suggest that ER immobilization dictates both antagonism and

ER turnover. Therefore, optimization of compounds based solely on their ability to degrade ER is not sufficient to identify compounds that will reliably antagonize the receptor. Taken together this body of work from our lab and others has provided the impetus to develop high affinity ligands that stabilize ER in an antagonist conformation as opposed to optimization of SERD activity.

1.4.6 High affinity SERMs as a viable therapeutic option

As much work in the field has resolutely searched for compounds that exhibit both antagonist and degradative functions, high affinity antagonists have often been overlooked. Originally developed as a treatment for climacteric symptoms associated with menopause, raloxifene and lasofoxifene have also been shown preclinically to have inhibitory effects on breast tumors in relevant animal models [27, 73, 81, 89]. Structures of these compounds compared to previously described SERMs and SERDs are in displayed in Figure 6. The STAR trial (NCT00003906) compared the preventative effects of

tamoxifen versus raloxifene in postmenopausal women who were at an increased risk of breast cancer [90]. This trial demonstrated that raloxifene is slightly less effective than tamoxifen at reducing the risk of invasive breast cancer and more effective than tamoxifen at preventing endometrial cancer. Importantly, this trial also demonstrated an overall decreased adverse side-effect profile which led to its approval for the prevention of breast cancer.

Interestingly, the PEARL trial (NCT00141323) demonstrated that in women with osteoporosis, lasofoxifene significantly reduced the risk of breast cancer (in women with average risk) but did not have an effect on the occurrence of endometrial cancer [91]. Like other clinically relevant ligands previously discussed, it has been shown that raloxifene has a reduced potency against the mutant receptors in breast, and the low bioavailability of this compound would complicate its use in this setting; however, the effect of lasofoxifene on these mutations has not been explored [48].

Fulvestrant Bazedoxifene

N H N F F O O F H H S F HO F OH

Tamoxifen Raloxifene Lasofoxifene

N N O o

O N O OH

HO S HO

Figure 6: Structural comparison of Raloxifene and Lasofoxifene to other clinically relevant SERM and SERDs.

Raloxifene and lasofoxifene have an amine side chain like basic SERDs.

1.4.7 Problems that will be addressed

The studies presented in subsequent chapters investigate the importance of relative ERWT and ERmut expression and activity in determining antagonist pharmacology.

Particular emphasis will be directed towards studying novel and existing high affinity antagonists, regardless of their ability to function as a SERD. These studies are intended to inform the discovery of next-generation ER antagonists for use as cancer therapeutics.

2. Allelism dictates ESR1 mutant pharmacology in breast cancer

This chapter represents the work that will be published in the journal Molecular Cancer Therapeutics in the cited manuscript [92].

Andreano, K.J. et al. The dysregulated pharmacology of clinically relevant ESR1 mutants is normalized by ligand- activated WT receptor. Molecular Cancer Therapeutics. Accepted.

2.1 Introduction

ER (ESR1) is a member of the nuclear hormone receptor superfamily of ligand- activated transcription factors and is expressed in the majority of luminal breast cancers

[4, 21]. Upon binding an estrogenic ligand, this transcription factor regulates the expression of genes required for cancer cell proliferation and survival. Not surprisingly, drugs that inhibit estrogen actions are the cornerstone of pharmacotherapy of breast cancers that express ER [4]. Among the interventions most commonly used are the SERM tamoxifen, a drug which functions as an ER antagonist in breast cancer cells, and AIs

(letrozole, anastrozole, or exemestane), competitive inhibitors of CYP19A1 (aromatase), the enzyme that converts androgens into estrogens [23, 27]. Whereas both classes of drug effectively inhibit ER signaling in breast cancer, it is now standard practice to use AIs in the adjuvant setting as frontline endocrine therapy in postmenopausal patients or in high- risk premenopausal patients when combined with ovarian suppression [23]. Tamoxifen is primarily reserved for the adjuvant treatment of premenopausal breast cancer patients at

low-risk for recurrence with or without interventions to achieve ovarian suppression [26,

27]. These endocrine therapies have had a very significant impact on disease-free and overall survival in patients with breast cancer, although de novo and acquired resistance to either type of drug remains a significant clinical issue [93-96]. However, the observation that ER remains engaged in the regulation of processes of importance in cancers that have escaped frontline endocrine interventions have led to the continued exploitation of this receptor as a therapeutic target [39].

Fulvestrant, a SERD, is used in patients who progress on frontline endocrine therapies and is given as monotherapy or in combination with targeted therapies [25].

Drugs of this class function primarily as competitive inhibitors of agonist binding to ER, but their inhibitory activity is reinforced by a drug-induced conformational change that targets the receptor for proteasomal degradation[24, 25]. Currently, fulvestrant is the only clinically approved SERD. Whereas this drug is a very effective inhibitor and downregulator of ER expression in cellular and animal models of breast cancer, its clinical utility is limited by its poor pharmaceutical properties and by the need to administer it as a large bolus intramuscularly [31, 32]. Further, it is not clear to what extent ER within tumors is occupied by fulvestrant at the maximum doses that can be delivered to patients

[97]. This has driven the search for oral SERDs (or SERMs) that are as effective as fulvestrant in inhibiting ER activity but which have tissue exposure levels sufficient to

saturate the receptor. From these efforts emerged the first generation oral SERDs GW5638,

AZD9496 (NCT02248090, NCT03236874) and GDC-0810 (NCT01823835), all of which demonstrated efficacy in late state disease but whose development has been discontinued

[65, 70, 75, 77-79]. Other oral SERDs, like RAD1901 (NCT02338349), are currently in clinical development [71, 83].

Whereas the mechanisms underlying resistance to endocrine therapies are varied and complex, it is now clear that gain of function point mutations within the LBD of ESR1 that permit it to exhibit constitutive transcriptional activity can confer resistance to AIs

[49-51]. Although rare in primary breast tumors, ERmuts occur in up to 40% of metastatic lesions, a finding that is consistent with their selection by conditions of extreme estrogen deprivation [49-51, 54-57, 98]. Two of the most common mutations, ERY537S, and ERD538G, account for roughly 70% of all ESR1 mutations identified in patients with metastatic breast cancer [49-51, 54-57, 98]. In addition to constitutively activating transcription, these mutations also exhibit distinct neomorphic activities that likely contribute to disease progression [48, 61]. Notwithstanding these important differences, most attention has been focused on how these disease-associated mutations reduce the ER binding affinity of some clinically important antagonists, an activity that may limit their therapeutic utility

[49-51, 54-57, 98]. The development of most SERDs was initiated before the prevalence of

ERmuts was fully appreciated, and it is now apparent that, as with fulvestrant, the affinity

of ERmuts for even the most contemporary SERDs is substantially reduced (~one order of magnitude) [52, 65, 83]. Thus, in addition to addressing whether inhibition of ER with these drugs is a viable approach to inhibit ER-positive, endocrine therapy-refractive disease, there remains an open question as to their efficacy in cancers expressing the ERmuts

[49-51]. Thus, the primary goal of this study was to define the impact of ERmuts on the pharmacology of ER ligands with a view to prioritizing existing drugs for clinical evaluation in patients. Additionally, elucidation of the molecular mechanisms underlying the dysregulated pharmacology of ERmuts was also undertaken with the goal of informing the identification of the next generation of ER modulators for use in the treatment of advanced breast cancer.

2.2 Results

2.2.1 The expression of clinically relevant ERmuts does not alter the pharmacology of ER ligands in cells expressing ERWT.

Prior studies that informed our current understanding of the pharmacology of

ERmuts in breast cancer cells were performed in model systems in which the mutants were expressed absent ERWT [49-52, 63]. Whereas this may be an appropriate way to model the pharmacology of compounds in cells homozygous for the mutants, this approach does not take into account the heterogeneity of ERWT/ ERmut expression in advanced ER-positive breast tumors cell that results from the selective pressure of endocrine therapy [49]. To

address this issue, we performed a comprehensive analysis of ER ligand pharmacology in cellular models in which ERWT is expressed alone or in combination with ERmut, the latter a scenario that is likely to represent what occurs within the majority of tumor cells in patients with metastatic disease.

To enable the evaluation of ERmut pharmacology, we created MCF7 cell derivatives that express ERWT alone (MCF7B-WT) or both ERWT and individual ER mutants (MCF7B-Y537S and MCF7B-D538G) [61]. The structures of the antagonists evaluated in this study include the most of the clinically relevant SERMs and SERDs that are available. As expected, basal

(ligand-independent) ER transcriptional activity, assessed using an ERE-luciferase reporter, was higher in both MCF7B-Y537S and MCF7B-D538G cells when compared to the isogenic MCF7B-WT cells. Further, as observed in MCF7B-WT cells, treatment with 17β- estradiol (E2) increased ER-dependent transcriptional activity in both MCF7B-Y537S and

MCF7B-D538G cell models (Figure 7A). Notably, however, no significant shift in potency or efficacy was observed for any of the ER ligands tested in this assay when comparing either

MCF7B-Y537S or MCF7B-D538G with MCF7B-WT (Figure 7B-I, Table 2). Importantly, a similar result was observed when cell proliferation, as opposed to transcription, was used to monitor ER activity (Figure 8,Table 3). Previous studies which demonstrated shifts in ligand potency in similar models were performed in hormone stripped media where the activity of ERWT is blunted [61, 68].

A) B) C)

stradiol (E2) Fulvestrant 4-hydroxytamoxifen 5000000 1.5 WT 2.0 ER ER WT Y537S Y537S U 4000000 U ER ER L L D538G 1.5 D538G R R

1.0 ER ER

d 3000000 d e U e z z

L 1.0 li li

R WT a 2000000 a ER 0.5 rm Y537S rm 0.5 o 1000000 ER o N ERD538G N 0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 -14 -12 -10 -8 -6 Log [M] Log [M] Log [M] D) E) AZD9496 GDC-0810 F) RAD1901 1.5 WT 2.0 WT 2.0 WT ER ER ER Y537S Y537S Y537S U U ER ER ER L L D538G 1.5 D538G 1.5 D538G R R

1.0 ER ER ER U d d L e e z z R 1.0 1.0

li li d a a e

0.5 z rm rm li 0.5 0.5 o o a N N rm

0.0 o 0.0 0.0 0 -12 -10 -8 -6 N 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M] G) I) Raloxifene H) Bazedoxifene Lasofoxifene 2.0 WT 1.5 WT 2.0 ER ER ER WT Y537S Y537S Y537S U U ER ER U ER L L 1.5 D538G D538G L 1.5 D538G R R R

ER 1.0 ER ER d d d e e e z z 1.0 z 1.0 li li li a a a 0.5 rm rm 0.5 rm 0.5 o o o N N N

0.0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M]

Figure 7: Cells expressing both the ERWT and ERmuts have similar pharmacological responses to antiestrogens when compared to cells only expressing ERWT.

(A-I) MCF7B cells were plated in phenol red free media supplemented with charcoal stripped media for 48 hours and then transfected with an estrogen responsive reporter gene (7X- ERE-TATA-LUC). After 5 hours, cells were treated with E2 (0.1 nM) alone or in the presence of ER antagonists (10-12M to 10-6M). Firefly luciferase activity was assessed and normalized to to β- galactosidase transfection control (Y-Axis). Data points are the mean of three technical replicates, and error bars are the standard deviation of these replicates. Data presented is a representative of three independent experiments. Two-way ANOVA was utilized, comparing the logIC50 of all three independent experiments, to determine if there were significant differences between the WT and mutant receptors. No significant differences (p-value < 0.05) were determined.

Table 2: Transcriptional IC50 values (M) of antagonists in MCF7B Cells

Compound ERWT ERY537S ERD538G Fulvestrant 2.05E-10 1.75E-10 3.52E-10 4-hydroxytamoxifen 7.54E-10 7.04E-10 1.28E-10 AZD9496 4.69E-10 3.97E-10 7.51E-10 GDC-0810 9.56E-09 1.69E-08 4.87E-09 RAD1901 3.65E-09 6.02E-09 7.18E-09 Raloxifene 2.54E-10 2.22E-10 4.67E-10 Bazedoxifene 2.88E-10 5.39E-10 8.42E-10 Lasofoxifene 5.61E-10 4.67E-10 5.00E-10

Fulvestrant 4-hydroxytamoxifen AZD9496 t WT t

n WT 1.5 n 1.5 WT 1.5 e e

ER ER t ER Y537S t Y537S Y537S

ER ER on ER D538G on D538G D538G C C

1.0 ER 1.0 ER 1.0 ER DNA DNA

d d e 0.5 e 0.5 0.5 z z li li a a rm rm Normalized DNA Content o 0.0 o 0.0 0.0 N 0 -12 -10 -8 -6 N 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M]

RAD1901 Raloxifene Bazedoxifene t t

t n 1.5 WT 1.5 WT n 2.0 WT n e e t

ER t ER e ER t Y537S Y537S Y537S on ER ER on ER

on D538G 1.5 D538G

D538G C C

C 1.0 1.0 ER ER ER DNA

DNA 1.0

DNA

d d d e 0.5 0.5 e z e z z li li 0.5 li a a a rm rm rm o 0.0 0.0 o 0.0 o N N

N 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M]

Lasofoxifene t WT

n 1.5

e ER t Y537S ER on D538G C 1.0 ER DNA

e 0.5 z li a rm

o 0.0 N 0 -12 -10 -8 -6 Log [M]

Figure 8: Proliferative responses to antiestrogens is indistinguishable in MCF7B cells.

MCF7B cells were grown in DMEM-F12 media containing 2% FBS for 7 days while being treated with ER antagonists (10-12 – 10-6 M). Cellular proliferation was assessed by measuring DNA content (Hoechst stain) and DNA content is normalized to vehicle control. Data presented is a representative of three independent experiments.

Table 3: GI50 values (M) of antiestrogens in MCF7B Cells

Compound ERWT ERY537S ERD538G Fulvestrant 5.22E-10 1.01E-09 1.26E-09 4-hydroxytamoxifen 8.37E-10 1.27E-09 2.15E-09 AZD9496 1.22E-10 2.75E-10 2.76E-10 RAD1901 2.11E-10 3.40E-09 5.69E-09 Raloxifene 3.40E-10 4.40E-10 1.02E-09 Bazedoxifene 4.73E-10 1.08E-09 1.43E-09 Lasofoxifene 4.25E-10 5.93E-10 9.14E-10

We, and others, have reported extensively on the role of cell context in regulating

ER pharmacology, a likely consequence of differences in coregulator expression. Thus, we extended our studies to evaluate ER pharmacology in a second model, in which ERWT expressing T47D cells were engineered to express ERY537S or ERD538G in addition to endogenous ERWT (Figure 9 A and B, Table 4)[48]. Interestingly, of the five mutant clones tested, only ERY537SA displayed resistance to any of the compounds analyzed (Figure 9A).

We also confirmed that the expression of ERWT and ERmut did not change over time and were maintained under the conditions of our in vitro assays (Figure 10)[99]. Interestingly, the ERY537SA clone that displays partial resistance has a higher allelic frequency of ERY537S compared to the ERY537SB clone that does not show resistance (Figure 9 and 10). However, our results appear to conflict with in vitro studies from others in which it was determined that ERY537S and ERD538G display an altered response to clinically relevant SERMs and

SERDs [48-52, 63].

Fulvestrant 4-hydroxytamoxifen t t

n 1.5 n 1.5

e WT e WT

t ER A t ER A WT WT

on ER B on ER B C C 1.0 ERWT C 1.0 ERWT C ERY537S A ERY537S A DNA DNA Y537S Y537S

d ER B d ER B

e 0.5 e 0.5 z ERD538G A z ERD538G A li li a ERD538G B a ERD538G B

rm D538G rm D538G o 0.0 ER C o 0.0 ER C N 0 -12 -10 -8 -6 N -14 -12 -10 -8 -6 Log [M] Log [M]

AZD9496 RAD1901 GDC-0810 t t t n n 1.5 1.5 n 2.0 e e WT WT e WT t t ER A ER A t ER A WT WT WT on on ER B ER B on ER B 1.5 C C C

1.0 ERWT C 1.0 ERWT C ERWT C ERY537S A ERY537S A ERY537S A DNA DNA DNA 1.0

Y537S Y537S Y537S d d ER B ER B d ER B e e 0.5 0.5 e z z ERD538G A ERD538G A z ERD538G A li li li 0.5 a a ERD538G B ERD538G B a ERD538G B rm rm D538G rm D538G D538G o o 0.0 ER C o 0.0 ER C 0.0 ER C N N 0 -12 -10 -8 -6 N -14 -12 -10 -8 -6 -14 -12 -10 -8 -6 Log [M] Log [M] Log [M]

Raloxifene Bazedoxifene Lasofoxifene t t t n n 1.5 2.0 n 1.5 e e WT WT e WT

t ER A t ER A t ER A WT WT WT on on

ER B ER B on 1.5 ER B C C C

1.0 ERWT C ERWT C 1.0 ERWT C ERY537S A ERY537S A ERY537S A DNA DNA

1.0 DNA

Y537S Y537S Y537S d d ER B ER B d ER B e e 0.5 e 0.5 z D538G z D538G ER A ER A z ERD538G A li li

0.5 li a a ERD538G B ERD538G B a ERD538G B rm rm D538G D538G rm D538G

o ER C o ER C 0.0 0.0 o 0.0 ER C N N 0 -12 -10 -8 -6 -14 -12 -10 -8 -6 N -14 -12 -10 -8 -6 Log [M] Log [M] Log [M]

Figure 9: Albeit differences in clonal variability, there is indistinguishable differences in T47D proliferation in response to antiestrogens.

T47D cells were grown in RPMI media containing 2% FBS for 7 days while being treated with ER antagonists (10-12 – 10-6 M). Cellular proliferation was assessed by measuring DNA content (Hoechst stain) and DNA content is normalized to vehicle control. Data presented is a representative of three independent experiments.

Table 4: GI50 values (M) of antiestrogens in T47D cells

Compound ERWT ERWT ERWT ERY537S ERY537S ERD538G ERD538G ERD538G A B C A B A B C Fulvestrant 5.21 2.07 5.70 7.66 3.41 1.44 4.06 9.75 E-10 E-10 E-10 E-10 E-10 E-09 E-09 E-10 4-OHT 7.53 3.98 7.19 1.22 8.06 9.83 6.74 4.70 E-10 E-10 E-10 E-08 E-10 E-09 E-09 E-09

AZD9496 2.24E 9.84E 2.13E 8.55E 2.70E 1.09E 1.23E 4.48E -10 -11 -10 -09 -10 -09 -09 -10 RAD1901 1.25E 1.00E 2.23E 2.07E 1.05E - 6.34E 9.25E 4.00E -08 -08 -08 -07 07 -08 -08 -08 GDC-0810 8.94 1.20 1.62 5.95 9.42 5.800 8.016 6.53 E-10 E-09 E-09 E-08 E-10 E-09 E-09 E-09 Raloxifene 2.42 2.70 3.09 6.48 1.96 2.79 2.88 1.32 E-10 E-10 E-10 E-09 E-10 E-09 E-09 E-09 Bazedoxifene 5.69 3.30 5.00 1.60 6.46 1.46 8.86 2.21 E-10 E-09 E-09 E-08 E-10 E-08 E-09 E-09 Lasofoxifene 1.26 1.08 1.45 1.30 1.08 9.60 9.79 4.56 E-09 E-09 E-09 E-08 E-09 E-09 E-09 E-09

A) B)

MCF7 B T47D 15 ERWT 25 ERWT ERY537N 20 ERY537N 10 ERY537S T ERY537S T C D538G 15 ER C

a D538G

t ER a l t l e e D 10

5 D

0 0 T S G W A R 537 D538 T B T C S A B A B C E Y WT S G G G R W W R E R R R 537 537 E E E E Y Y D538 D538 D538 R R R R R E E E E E

Figure 10: Validation of ER mutation status in engineered cell lines.

The previously published, genetically engineered MCF7B and T47D cell lines were confirmed to have both the WT and corresponding mutant receptor using a previously published detection method. A primer towards the Y537N mutation was used as a negative control as this mutation should not be in any of the lines. The delta Ct value (compared to a control probe targeting another region of the ER LBD) below 5 suggests that the particular allele is present in the sample.

2.2.2 The antagonist potency of SERDs and SERMs is reduced in cells expressing ERmuts alone.

To reconcile the discrepancies between our results presented here and those reported by others, we employed an overexpression model comparable to those that had been used previously to evaluate ERmut pharmacology[48-52, 63]. Vectors expressing ERWT,

ERY537S or ERD538G, together with an ERE-luciferase reporter, were co-transfected into ER- negative SKBR3 breast cancer cells. Using this model system, we demonstrated, as was observed in MCF7B cells, that both ERY537S and ERD538G exhibited constitutive transcriptional activity (Figure 11A). However, while the efficacy of fulvestrant and 4-

hydroxytamoxifen were comparable for all three receptors, the antagonist potency of these two clinically important compounds in cells expressing ERY537S or ERD538G was reduced by approximately one order of magnitude when compared to ERWT (Figure 11 B and C, and Table 5). The acidic SERDs, AZD9496 and GDC-0810, were found to be inactive as antagonists on ERY537S, and indeed the latter compound functioned as a partial agonist in this assay (Figure 11 D and E, and Table 5). This is similar to our previous finding demonstrating that GW7604, a structurally distinct acidic SERD that is the 4-hydroxylated analog of GW5638, had reduced efficacy when assayed in ERY537S expressing ovarian cancer cells [100]. Additionally, the potency of RAD1901, raloxifene, and bazedoxifene were reduced with subtle differences in the pharmacology noted when assayed on either

ERY537S or ERD538G (Figure 11 F-H, and Table 5). One of the most interesting findings in this study was that lasofoxifene, a SERM originally developed for the treatment/prevention of osteoporosis, was the only compound found to be as effective and potent an antagonist when evaluated in cells expressing ERY537S or ERD538G when compared to ERWT (Figure 11I and Table 5). This latter observation is in agreement with the findings of a recent study from our group showing that lasofoxifene was as effective an inhibitor of ERmuts as ERWT in cellular models of gynecological cancers (to be discussed in Chapter 3)[100]. These findings have important clinical implications that could inform the optimal selection of

ER antagonists for the treatment of patients with ERmuts in advanced disease.

A) B) C)

17 - Estradiol (E2) Fulvestrant 4-hydroxytamoxifen 300000 1.5 1.5 ERWT ERWT ERWT U Y537S U ERY537S ERY537S L ER L R R D538G D538G

D538G 200000 ER 1.0 ER 1.0 ER d d

e e U z z L li li a a R * 100000 0.5 0.5 * rm rm o o * N N

0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 D) Log [M] E) Log [M] F) Log [M] RAD1901 AZD9496 GDC-0810 ERWT 2.0 3.5 2.0 ERWT ERY537S ERWT 3.0 U U ERY537S U ERD538G ERY537S L L 1.5 L 1.5

2.5 R R D538G R D538G

ER ER d d d e e e 2.0 * z z 1.0 z 1.0 li li li 1.5 a a * a * rm rm 0.5 rm 1.0 0.5 o o o N N N 0.5 0.0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M] G) Raloxifene H) Bazedoxifene I) Lasofoxifene 2.0 2.0 1.5 ERWT ERWT ERWT U U U Y537S Y537S Y537S ER ER L ER L L 1.5 1.5 R R

R D538G D538G D538G

ER ER 1.0 ER d d d e e e z z z 1.0 1.0 li li li a a a 0.5 rm rm 0.5 0.5 rm o o *

* o N N * * N 0.0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Log [M] Log [M] Log [M]

Figure 11: ERmuts confer antiestrogen resistance when expressed alone.

SKBR3 cells were plated in phenol red free media and transfected with an estrogen responsive reporter gene (3X-ERE-tata-Luc) in the presence of ERWT, ERY537S or ERD538G. After 5 hours, cells were treated with E2 (1 nM) (A) and ER antagonists (10-12 M to 10-6 M) (B-I). Firefly luciferase activity was assessed and normalized to β-galactosidase transfection control (Y-Axis). Data points are the mean of three technical replicates, and error bars are the standard deviation of these replicates. Data presented is a representative of three independent experiments. Two-way ANOVA was utilized, comparing the logIC50s of all three independent experiments, to determine if there were significant differences between the WT and mutant receptors. Significant differences (p-value < 0.05) of the mutant IC50s when compared to that of the WT that were determined by this analysis are represented with a star. For GDC-0810 and AZD9496 on ERY537S the highest dose tested (10-6 M) was used as a surrogate, as the IC50 is greater than this value. The only compound that did not reach a significant difference for either mutant isoform was lasofoxifene.

Table 5: Transcriptional IC50 values (M) of antiestrogens in SKBR3 cells

Compound ERWT ERY537S ERD538G Fulvestrant 1.22E-09 4.04E-08 2.97E-08 4-hydroxytamoxifen 1.14E-09 3.27E-08 1.62E-09 AZD9496 4.17E-09 >1E-06 5.76E-09 GDC-0810 6.23E-10 4.15E-08 5.20E-09 RAD1901 2.07E-08 6.27E-07 9.31E-08 Raloxifene 6.23E-10 4.15E-08 5.20E-09 Bazedoxifene 1.47E-09 3.50E-08 1.17E-08 Lasofoxifene 6.45E-10 1.58E-09 1.98E-09

2.2.3 ER ligands exhibit subtle differences in their ability to facilitate the interaction of ERmuts with coregulators.

We next embarked on studies to define the molecular basis of the differences in the pharmacology of ERWT, ERY537S or ERD538G. Resolution of this issue, we anticipated, would allow for the optimal use of existing endocrine therapies, and inform the development of the next generation of ER modulators for breast cancer. Receptor conformation has emerged as the primary mechanism by which information flows from a ligand through the receptor to the transcriptional machinery [73, 76, 101]. Similarly conformed ER ligand complexes can exhibit diverse activities in different cell contexts as a consequence of the cell-selective expression and differential recruitment of functionally distinct coregulators. Further, subtle changes in ER structure, induced by structurally similar ligands, can result in different transcriptional outputs on individual target genes

[102-104]. Thus, it is possible that differences in the pharmacology of the ERmuts noted in

cellular models of ER-positive (MCF7B and T47D cells) or ER-negative (SKBR3 cells) breast cancer could result from differences in cofactor expression and their differential recruitment by ERWT, ERY537S or ERD538G upon ligand activation.

Given the primacy of receptor structure in determining pharmacological output on ER, we evaluated the impact of SERMs and SERDs on the conformation of different receptor-ligand complexes using a cofactor peptide binding assay, the utility of which we have described previously [75, 102, 105, 106]. In this assay, short peptides identified using combinatorial peptide phage display, and peptides derived from the receptor interaction domains of validated coactivators (CoA) and corepressors (CoR), are expressed as GAL4-

DBD peptide fusions (Figure 12A). Additionally, a control peptide that interacts with ER in the presence of any ligand (αII) was also utilized. ERWT, ERY537S or ERD538G were modified to contain a VP-16 acidic activation domain at their amino termini (Figure 12A). The interaction of the VP-16-ER proteins with the GAL4-peptide fusions in the presence of each ligand was assessed by measuring transcriptional activity on a GAL4-responsive luciferase reporter.

As expected given their constitutive activity, both mutant receptors interact in a ligand-independent manner, albeit to different degrees, with CoA-like peptides

(designated with red brackets), and these interactions are further elevated upon the addition of E2 (Figure 12B). The CoA interaction profiles of E2-activated ERWT, ERY537S or

ERD538G are surprisingly indistinguishable. Importantly, the constitutive interaction of the mutants with CoA peptides is only partially attenuated upon the addition of

SERMs/SERDs (Figure 12 C and D). Notably, raloxifene and lasofoxifene appear to be the two most effective inhibitors of CoA-peptide binding to ERY537S or ERD538G. Subtle quantitative differences in the binding of CoR-peptides (blue brackets) to the receptors in the presence of different ligands were also noted but there were no obvious differences in peptide binding preferences. One exception is the robust interaction of the

RAD1901/ERY537S complex with a subset of the CoR peptides. We infer this to mean that this particular ligand-receptor complex may have an increased ability to recruit corepressors to ERmut. However, when taken together, it appears that the ERWT can adopt different conformational states upon binding different ligands and these interactions are substantially similar in each of the mutant receptors. Taking into account the limitations of this study (i.e. surfaces on ER not probed with our current technology), we concluded that it is unlikely that the differences in the pharmacology of the ERmuts observed in different cells can be attributed to differential coregulator binding alone.

A) ER Peptide

VP-16

GAL4 Luciferase

5X GAL TATA

WT Y537S B)ER C)ER D)ER D538G II I II I II I V V V 13 16 13 16 13 16 / bI2 / bI2 / bI2 1 1 1 bT1 NR bT1 NR / bN2 / bT1 NR bN2 / bN2 7 7 LX23 7 LX23 P LX23 P P EIP EIP EIP397 EIP420 EIP484 EIP104 EIP786 EIP793 EIP EIP397 EIP420 EIP484 EIP104 EIP786 EIP793 EIP397 EIP420 EIP484 EIP104 EIP786 EIP793 T7-ASC2 T7-ASC2 T7-ASC2 GRI GRI GRI T7- PGC1a T7- PGC1a T7- PGC1a ACTR 621-821 ACTR 621-821 ACTR 621-821 SRC1 NR Box 1 ACTR 400-1000 SRC1 NR Box 1 ACTR 400-1000 ACTR 400-1000 SRC1 NR Box 1 4 Vehicle 5 RAD1901

Fulvestrant Raloxifene Bazedoxifene Lasofoxifene AZD9496 GDC-0810 4-hydroxytamoxifen 17 -Estradiol

Figure 12: Differential cofactor recruitment reveals modest changes in overall receptor conformation between the WT and mutant receptors.

(A) A mammalian two-hybrid assay was used to evaluate ligand-dependent recruitment of peptides that mimic ER coregulators. (B) Hep- G2 cells were co-transfected with VP-16 tagged WT or mutant ER, Gal4DBD tagged peptides and a Gal4-responsive reporter gene and pCMV β- gal. 24 hours later cells were treated with saturating concentrations of ligands (10 µM) and incubated for 48 hours. Normalized response, which was obtained by normalizing luciferase activity to β-gal activity, was used as input for Ward hierarchical clustering. Heat maps of mutant ERs are re-ordered to match the WT receptor. Results demonstrated a change in receptor conformation in response to ER activating mutations. At the top of the graph, there are three classes of peptides: ligand indiscriminate (black), peptides associated with receptor inhibition (blue) and receptor activation (red). Data presented is a representative of two independent experiments.

2.2.4 The altered pharmacology of ERmuts is only evident when their expression in cells exceeds that of the WT receptor.

One of the key differences between MCF7B (and T47D cells) and SKBR3 models is that in the latter cell line ERY537S or ERD538G are expressed in the absence of ERWT. We considered it possible that in the MCF7B (and T47D) cell background, ERWT pharmacology has primacy and normalizes the transcriptional activity of the mutants. We considered it likely, therefore, that by overexpressing the mutants relative to ERWT in MCF7 cells, that the altered mutant pharmacology apparent in SKBR3 cells would emerge. To test this hypothesis, we generated MCF7 cells in which ERWT, ERY537S or ERD538G expression was regulated in a doxycycline-inducible manner, allowing titratable expression of these proteins (MCF7I) over endogenous ERWT. Considering the pharmacology noted in SKBR3 cells, we selected fulvestrant (potency shift observed with both mutants), AZD9496 (loss of efficacy as an inhibitor of ERY537S) and lasofoxifene (potency and efficacy unaffected by mutation status) for analysis in these model systems. The transcriptional activity and pharmacology of receptor combinations were assessed using a transfected ERE-luciferase reporter gene (Figure 13 and Table 6). Relative to its activity on ERWT, it was noted that increased expression of ERD538G resulted in a reduction in fulvestrant potency, and a trend towards reduced potency was also noted with ERY537S (Figure 13A and Table 6). Likewise, the potency of AZD9496 on ERY537S was considerably reduced upon its overexpression in

MCF7I cells (Figure 13B and Table 6). The pharmacology of lasofoxifene was unaffected

by receptor expression levels (Figure 13C and Table 6). No changes were noted in the activity of any ligand in cells overexpressing ERWT alone. These data suggest that the altered pharmacology of ERmuts may only be manifest when they are expressed at a higher level than ERWT.

A) Fulvestrant B) AZD9496 C) Lasofoxifene

5 ng/mL Doxycyline 5 ng/mL Doxycyline 5 ng/mL Doxycyline 2.5 WT 2.0 WT 1.5 WT ER ER ER

Y537S U Y537S U U 2.0 ER ER Y537S ER L L D538G L 1.5 D538G R R R D538G

ER ER 1.0 ER d d 1.5 d e e e z z z 1.0 li li li a a 1.0 a 0.5 rm rm rm

0.5 o o 0.5 o N N N

0.0 0.0 0.0 -14 -12 -10 -8 -6 -14 -12 -10 -8 -6 0 -12 -10 -8 -6 Fulvestrant (Log [M]) AZD9496 (Log [M]) Lasofoxifene (Log[M])

20 ng/mL Doxycyline 20 ng/mL Doxycyline 20 ng/mL Doxycyline 2.5 WT 2.0 WT 1.5 WT ER ER ER Y537S Y537S Y537S U U U 2.0 ER ER ER L L L D538G 1.5 D538G D538G R R R

ER ER 1.0 ER d d d 1.5 e e e z z z 1.0 li li li a a a 1.0 * 0.5 rm rm rm 0.5 o o o 0.5 N N N

0.0 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 0 -12 -10 -8 -6 Fulvestrant (Log [M]) AZD9496 (Log [M]) Lasofoxifene (Log[M])

Figure 13: The altered pharmacology of ERmuts can be manipulated by their expression level.

(A-C) MCF7 cells were engineered to express the ERWT or ERmuts in a dose-dependent manner in response to doxycycline treatment over the endogenous ERWT. Cells were plated in phenol red-free media for 48 hours with doxycycline 5 or 20 ng/ml as indicated and then transfected with an estrogen responsive reporter gene (7X-ERE-tata-Luc). After 5 hours, cells were treated with 17β-estradiol (0.1 nM) and ER antagonists (10-12 M to 10-6 M). Firefly luciferase activity was assessed and normalized to β-gal (Y-Axis). Data points are the mean of three technical replicates, and error bars are the standard deviation of these replicates. Data presented is a representative of three independent experiments. Two-way ANOVA was utilized, comparing the logIC50 of all three independent experiments, to determine if there were significant differences between the WT and mutant receptors. Significant differences (p-value < 0.05) of the mutant IC50s when compared to that of the WT that were determined by this analysis are represented with a star.

Table 6: Transcriptional IC50 values (M) of antiestrogens in MCF7I cells

Compound ERWT ERY537S ERD538G Fulvestrant 2.77E-09 3.42E-09 3.32E-09 (0ng/mL doxycycline) Fulvestrant 2.71E-09 1.82E-09 2.44E-09 (5ng/mL doxycycline) Fulvestrant 2.32E-09 7.48E-09 2.60E-08 (20ng/mL doxycycline) AZD9496 1.80E-09 2.06E-09 2.26E-09 (0ng/mL doxycycline) AZD9496 1.58E-09 4.00E-09 1.911E-09 (5ng/mL doxycycline) AZD9496 9.97E-10 3.70E-08 1.89E-09 (20ng/mL doxycycline) Lasofoxifene 3.93E-09 2.10E-09 1.62E-09 (0ng/mL doxycycline) Lasofoxifene 2.04E-09 4.98E-09 3.63E-09 (5ng/mL doxycycline) Lasofoxifene 1.62E-09 2.72E-09 1.64E-09 (20ng/mL doxycycline)

A series of experiments were designed to examine whether the altered pharmacology of select ERmuts was solely an artifact of their overexpression or if overexpression was required to outcompete a normalizing effect of ERWT. To this end, the impact of altering the expression of ERWT relative to ERmuts was evaluated initially in

SKBR3 cells. Consistent with results presented in Figure 11, the potency of fulvestrant and

AZD9496 was reduced in SKBR3 cells expressing ERY537S or ERD538G alone (Figure 14A and

B). However, as the expression of the ERWT was increased to comparable levels with ERmut, the pharmacology of fulvestrant and AZD9496 was normalized to their activity on ERWT

(Figure 14A-B). Lasofoxifene antagonist efficacy remained unchanged as the expression levels of ERWT and ERmut were altered (Figure 14C). It is important to note that in the absence of ligand, the constitutive activity of the mutant receptors is observed even when

ERWT is present. Thus, in the absence of hormone, the mutant is functionally in excess indicating that ERWT activity, and not its expression alone, is required to achieve the normalization of ER pharmacology noted. This finding supports previous data in the literature that demonstrate that under conditions of extreme hormone deprivation, the resistance of ERmuts to ER ligands is not affected by the presence of ERWT [61, 68]. To support these findings, we performed an analogous experiment in MCF7I cells. In this context, we expressed ERmuts in cells expressing endogenous ERWT and consistent with our prior observations, the pharmacology of cells expressing ERWT or ERmuts were found to be indistinguishable. However, when the expression of the endogenous receptor was reduced using an siRNA directed against the 3’ UTR of the ER mRNA, the mutant pharmacology emerged (Figure 14 D-H). Doxycycline-induced expression of ERWT and

ERmuts and the effectiveness of the siRNA mediated knockdown of endogenous ER protein levels were confirmed. As observed in SKBR3 cells, the constitutive activity of the mutant receptors was not diminished by coexpression of ERWT. Together, these results indicate that activated ERWT can normalize the pharmacology of ERY537S and ERD538G, and that response to ER ligands following aromatase inhibitor therapy will depend on the relative

co-expression of ERmuts and ERWT in breast cancer cells. For reasons yet to be determined the pharmacology of the SERM lasofoxifene is not affected by mutant status.

A) B) C) Fulvestrant AZD9496 * Lasofoxifene ) ) 40 40 ) 40 M M M

9 9 9 - - - 0 0 * 0 x1 x1 x1 ( ( (

20 20 20 50 50 50 C C C I I I

0 0 0 WT Y537S WT Y537S WT Y537S

AZD9496 Lasofoxifene Fulvestrant 25 25 * 25

) 20 )

) 20 20 M

M 9 M

9 15 0 - - 15 15 0 0 x1 ( x1

( 10 (

10 10 50 50 C 50 I

C 5 C I I 5 5 0 0 0 WT D538G WT D538G WT D538G D) E) F) Lasofoxifene Fulvestrant AZD9496 * 10 10 10 ) ) 8 ) 8 8 M M M

9 9 9 - - - 6 6 0

6 0 0 x1 x1 x1 ( ( (

4 4 4 50 50 50 C C C I I 2 I 2 2

0 WT 0 0 ER + + - - Y537S + + - - + + - - ER - - + + - - + + - - + + Control siRNA + - + - + - + - + - + - 3’ UTR siRNA - + - + - + - + - + - + Fulvestrant AZD9496 Lasofoxifene 10 10 10 )

8 ) 8 ) 8 M M M

9 9 9 - - 6 6 - 0 6 0 0 x1 x1 x1 ( ( (

4 4 4 50 50 50 C C C I I 2 2 I 2

0 0 0 WT ER + + - - D538G + + - - + + - - ER - - + + - - + + - - + + Control siRNA + - + - + - + - + - + - 3’ UTR siRNA - + - + - + - + - + - +

Figure 5 Figure 14: The altered pharmacology of ERmuts is only evident when expressed at a level higher than the WT receptor.

(A-C) SKBR3 cells were transfected with an estrogen responsive reporter gene (3X-ERE- tata-Luc) in the presence of different ERWT to ERmut construct ratios. (D-F) MCF7I cells were plated with doxycycline and siRNA and then transfected with 7X-ERE-tata-Luc. The IC50s of each dose response curve are plotted. Two-way ANOVA was utilized, comparing the logIC50s of all three independent experiments, to determine if there were significant differences between the WT and mutant receptors. Significant differences (p-value < 0.05) of the mutant IC50s when compared to that of the WT that were determined by this analysis are represented with a star. Data presented is a representative of three independent experiments.

2.3 Discussion

The goal of this study was to define the molecular basis for the altered pharmacology exhibited by the most clinically relevant ERmuts, information we anticipate could inform the selection of existing drugs for use in patients with advanced ER-positive breast cancer whose tumors harbor these mutations. In cell-based models of breast cancer, we made the important observation that when compared to ERWT, the pharmacology, most notably antagonist potency, of these mutants was significantly impacted by the relative co-expression of ERWT and ERmut. Previously, we and others have observed that the potency of existing ER antagonists was reduced in cells expressing either of the two most frequently occurring ERmuts (ERD538G and ERY537S)[50-52, 63, 100]. In this study, we have demonstrated that such differences are dependent on the relative expression level of both the ERWT and ERmuts and are only apparent under conditions where ERmuts are substantially overexpressed relative to ERWT. Given the previously reported neomorphic activities of the ERmuts, it is possible that the differences in response to ER ligands may only be manifest on select endogenous target genes [48, 61]. However, in our system, the activity of the mutants in the transcriptional reporter assays mirror their activities, in the presence of various ligands, when cell proliferation is used as the readout. These findings are significant as prior studies that have informed our current understanding of the

importance of ERmuts in the pharmacotherapy of breast cancer were performed in cells expressing only ERmuts in the absence of the ERWT [49-52, 63].

Whereas we have been able to confirm using several experimental models that

ERWT normalizes the activity of coexpressed ERmuts, the mechanism(s) by which this activity occurs is elusive. One possibility is that ERWT preferentially dimerizes with ERmuts and simply outcompetes ERmut homodimers. However, such a simple mechanism would require that the ERmuts would exhibit reduced homodimerization/ heterodimerization activity. It is more likely that in cells where ERWT and ERmut are present, and assuming no differences in dimerization ability, that the majority of the receptor (75%) would exist in an ERWT/ERWT or ERWT/ERmut complex and that the presence of the WT receptor normalizes the response (potency) to ligands. The recent cryo-EM structure of the ER coregulator complex is informative as to how ERWT may normalize the activity of the mutant [107].

Specifically, it was observed that the establishment of a productive transcription complex requires each monomer in an ER dimer to engage a p160 coregulator (i.e. SRC-3) to establish a platform upon which p300 can be recruited. Thus, in an ERWT/ERWT or

ERWT/ERmut complex, the conformational change(s) induced in ERWT by antagonists would result in the expulsion of one or two SRC proteins from the complex and a productive transcriptional complex could not form. Using peptide binding/cofactor binding studies, we have demonstrated that the interaction of ERmuts with coactivators is substantially

inhibited upon the addition of saturating concentrations of most antagonists, explaining why the efficacy of existing inhibitors is not affected by the most commonly occurring mutations.

There are several immediate clinical implications of this work. It is clear that selection for ERmuts by estrogen deprivation (aromatase inhibitor) manifests as resistance.

However, given that most mutants would be expected to be co-expressed in breast cancer cells with ERWT, it was unclear how they would impact the pharmacology of fulvestrant and other clinically important SERDs and SERMs. Our findings suggest that in ERWT expressing cells the presence of a mutant receptor is unlikely to have any significant impact on response to existing antagonists unless its expression vastly exceeds that of

ERWT (or in cells in which it is solely expressed). Some clinical data supports that assertion

[57]. Specifically, baseline and on treatment evaluation of ERmuts in circulating tumor DNA was evaluated and correlated to fulvestrant response (alone or in combination with PI3K inhibition) as a part of the Phase II FERGI study (NCT01437566). The findings of this study demonstrated that median ESR1 allele frequency was low at 0.45% and as such progression-free survival was not different in patients with ERmuts compared with ERWT patients. Conversely, in the PALOMA- 3 study (NCT01942135), there were observed differences in fulvestrant progression-free survival in response to mutation status [59, 60].

Interestingly, these studies report a higher whole tumor allele frequency of ERmuts, with a

reported expression fraction of 0.10 (or 10%). This study also suggested that ERmut containing clones were a small fraction of the whole tumor and as such the low allele frequency estimate was not representative of each individual cell. It is clear that cells expressing ERY537S emerged in the fulvestrant only arm of the PALOMA-3 arm and this has been taken as definitive evidence that this mutation reduces the potency of fulvestrant. We propose the alternative hypothesis that fulvestrant exposure is not sufficient to efficiently occupy ERWT, or the ERmuts, and that cells expressing the constitutively active mutants have a fitness advantage.

In our study, dysregulated ERmut pharmacology is only manifest when the expression of the mutant receptor(s) exceeds that of its ERWT counterpart. It is not clear how often this occurs in individual tumor cells and further research is needed to adequately assess allelism at the cellular level. Mutations in ESR1 can be detected in clinical tumor samples and circulating tumor DNA using next-generation sequencing and ddPCR [50, 51, 54-57, 59, 60, 98]. However, these assays are not designed to establish allelic frequency (homozygous versus heterozygous ESR1 alleles) on a single cell basis. The likely importance of ERmut allelism was suggested in a recent study that revealed a propensity for a loss of heterozygosity of ERWT when an ERmut is also present in the tumors of patients on endocrine therapies [69]. The inability to assess ERWT/ERmut allelism in a facile manner reinforces the need to understand the relationship between ER expression

level and ligand potency/efficacy as a means to select/develop pharmaceuticals for use in the treatment of patients whose mutants harbor ERmuts.

It is likely that even in situations where the ERmuts is expressed at a higher level compared to ERWT, it is only of significance when potency is a limiting property of a drug

(i.e. fulvestrant). However, our work suggests that most of the liabilities of the mutants can be mitigated by increasing the dose (assuming dose-proportional exposure and tolerable side-effect profile) of individual drugs as antagonist efficacy is not compromised by the expression of the most commonly occurring ERmuts. This highlights the importance of drug exposure when considering new/existing drugs for use in the treatments of patients with mutant receptors. One approach that has been developed to address the reduced affinity of the mutants is to develop Selective Estrogen Receptor Covalent

Antagonists (SERCAs)[108, 109]. The first of this new class of drugs, currently in clinical development, essentially converts tamoxifen into a covalent ER binder, thus mitigating the impact of the mutation on binding affinity. One SERCA is currently in clinical trials for metastatic breast cancer patients progressing on endocrine therapy

(NCT03250676)[108]. However, it is likely that locking the receptor in a “tamoxifen- induced conformation” is going to result in the selection of cancer cells which support the partial agonist activity of tamoxifen, an activity that is associated with acquired resistance

[36, 75, 110].

The SERM lasofoxifene appears to have attributes that would make it particularly useful in patients where there is concern as to the contribution of ERmuts to drug response.

In this study, and an earlier study in gynecological cancers, we demonstrated that this drug is an efficient antagonist whose actions are not influenced by mutant status [100].

Lasofoxifene was initially developed for the treatment of climacteric symptoms and osteoporosis associated with menopause. It is currently under evaluation in the ELAINE trial (NCT03781063) to assess its efficacy compared to fulvestrant, post aromatase, and

CDK4/6 inhibitor therapy, as a treatment for patients whose tumors harbor ERmuts. We also noted that the acidic SERDs (represented by GDC-0810, and AZD9496) are ineffective inhibitors of ERY537S. Thus, it is likely that the efficacy of this class of drugs will be diminished as the allelic frequency of ERY537S increases in patients. Currently, there are numerous new ER modulators, including LSZ102 (NCT202734615), AZD9833

(NCT03616586), GDC-9545 (NCT03332797), SAR429859 (NCT03284957), G1T48

(NCT03455270) and Zn-C5 (NCT03560531) under evaluation in the clinic [111].

Notwithstanding the potential impact of ERY537S on the response to acidic SERDs, it appears as if the mutant status of tumors may not be a significant issue for drugs that achieve significant exposure to offset the decreased potency noted (for all but lasofoxifene). We believe the studies presented herein should emphasize approaches to

achieve maximal drug exposure in tumors as opposed to developing new molecules that demonstrate increased affinity for the mutant receptors.

3. Discovery and treatment of ESR1 mutations in gynecological cancers

This chapter represents a collaboration effort with Dr. Stephanie Gaillard, a gynecological oncologist at Johns Hopkins University (formerly Duke) as well as a group of other very talented clinicians and scientists. This work has been published as cited below[100]. The right to reproduce this article in a thesis or dissertation of an author of this publication is maintained in the copyright agreement.

Gaillard, S.L., Andreano, K.J. et al., Constitutively active ESR1 mutations in gynecologic malignancies and clinical response to estrogen-receptor directed therapies. Gynecol Oncol, 2019. 154(1): p. 199-206.

3.1 Introduction

Gynecologic malignancies commonly express ER and/or PR and endocrine therapy is often considered as treatment for advanced, potentially hormone sensitive gynecologic cancers, especially low-grade endometrial and ovarian tumors. Endocrine therapy blocks ER signaling through a variety of strategies, most commonly estrogen- deprivation, as with AIs, or direct-antagonism of ER through SERMs or SERDs.

Failure of endocrine therapy occurs through intrinsic or acquired resistance mechanisms. Mutations in ESR1 are a mechanism of resistance to endocrine therapy commonly observed in metastatic breast cancer. These mutations occur predominantly in the LBD of the receptor and result in constitutive activation of ER in the absence of

estrogen. Activating ERmuts most frequently occur in a hotspot region encompassing amino acids 536–538 within the ER LBD, with a smaller number occurring at other LBD sites, namely E380, V422, S463, and L469 [51, 52, 54, 57, 112]. In breast cancer ERmuts arise in 25–

50% of patients who receive endocrine therapy, especially AIs, but are relatively rare (3%) in primary tumors [50, 51, 53, 57, 113]. Thus, ERmuts are primarily an acquired resistance pathway to endocrine therapy that may also account for rare cases of intrinsic resistance.

ESR1 amplification has also been suggested as a mechanism of resistance to endocrine therapy resulting in worse outcomes [114, 115]. ESR1 amplification is reported in early pre-cursor lesions of endometrial cancer [116-118]. However, the clinical impact of ESR1 amplifications is controversial and detection methods may result in overcalling of this genomic alteration [119].

A study evaluating ER as a predictive biomarker in endometrioid endometrial cancer identified 19 cases of ERmuts in 1034 samples (1.8%)[120]. This study focused solely on mutations arising in codons 536–538, was limited to endometrioid endometrial cancers, and did not provide clinical information regarding prior endocrine therapy or response to therapy.

The purpose of the current study was to quantify the frequency of ESR1 genomic alterations, including ERmuts, identified by comprehensive genomic profiling in gynecologic malignancies. We present our clinical experience treating tumors with both

de novo and acquired ERmuts. The effect of selected SERMs and SERDs on the transcriptional activity of individual ERmuts was examined.

3.2 Results

3.2.1 ESR1 genomic profiles in gynecological malignancies

Of the 9645 clinical samples from gynecologic malignancies evaluated with CGP in this study, 285 (3%) samples contained a total of 295 ESR1 alterations, including substitutions, amplification, and rearrangements. Ten cases exhibited two separate alterations each. The types and frequency of alterations by site of disease origin are listed in Table 2. Substitutions were the most common ESR1 alteration identified (194/295, 66%), with 44% (86/194) of those occurring in ESR1 codons expected to result in ER constitutive activity. Of the activating variants, alterations in codons 536–538 were most common accounting for 25% (75/295) of all ESR1 alterations and occurring in 0.8% (75/9645) of cases. Another 12 cases (0.1%) contained ERV422del, ERS463P, or ERL469 mutations. ESR1 amplifications were the next most common genomic alteration identified (80/295, 27%) and were present in 0.8% of cases. Median ESR1 copy number in amplified cases was 8

(range 6–38). ESR1 rearrangements were present in 0.2% of cases (Table 7).

Table 7: Types and frequency (%) of ESR1 alterations identified in gynecologic malignancies by primary site

Type of Frequency Ovary/FT Uterus Cervix Vulva/Vagina alteration N=9645 N=5594 N=3101 N=720 N=216 Total 295 (3.1) 120 (2.1) 160 (5.2) 9 (1.2) 6 (2.8) Amplification 80 (0.8) 45 (0.8) 34 (1.1) 1 (0.1) - Deletion 1 (<0.1) - 1 (<0.1) - - Fusion 2 (<0.1) 1 (<0.1) - - 1 (0.5) Rearrangements 18 (0.2) 9 (0.2) 9 (0.3) - - Total 194 (2.0) 65 (1.2) 116 (3.7) 8 (1.1) 5 (2.3) Substitution Variants Codon 536-538 75 (0.8) 18 (0.3) 56 (1.8) 1 (0.1) -

Other mutants 11 (0.1) 3 (0.0) 6 (0.2) - 2 (0.9) “-“: none present

Types and frequency of ERmuts by histological subtype are listed in Table 8. ERmuts were more common in uterine cancers (63/3101, 2%) compared to other primary sites

(24/6530, <1%, p<0.0001). ERmuts were enriched in carcinomas with endometrioid histology: 4.4% (24/548) in uterine endometrioid vs 0.2% (1/446) in uterine serous carcinomas (p<0.0001) and 3.5% (5/144) in ovarian endometrioid compared to 0.3%

(12/3502) in ovarian serous carcinomas (p = 0.0004). Two uterine endometrioid carcinomas exhibited the following co-occurring ESR1 mutations: ERY537N with ERY537S, and ERL536H with ERY537C, respectively. An ovarian serous carcinoma exhibited both ESR1 ERY537S and

ERD538G. Grade was not available for any of the uterine cases and absent for 89% of the ovarian cases, thus an assessment based on grade could not be performed. Uterine

endometrial stromal sarcomas (ESS) had a proportionally higher frequency of ERmuts than uterine leiomyosarcomas (LMS) [3/103 (3%) vs. 3/421 (0.7%), respectively], though this did not reach statistical significance (p=0.09).

Table 8: ERmuts identified in gynecologic malignancies by histological subtype

Primary site and N ERmuts N (%) ERY537S or Other activating histology ERD538G variants Cervix 720 1 (0.1) Clear Cell 15 1 (6.7) 1 Ovary/Fallopian Tube 5594 21 (0.4) Carcinoma NOS 1079 4 (0.4) 2 2 Endometrioid 144 5 (3.5) 2 3 Serous 3502 12 (0.3) 12 Uterus 3101 63 (2.0) Carcinoma NOS 1063 27 (2.5) 10 17 Endometrioid 548 24 (4.4) 10 14 Clear Cell 78 1 (1.3) 1 Papillary serous 446 1 (0.2) 1 Carcinosarcoma 303 4 (1.3) 4 Leiomyosarcoma 421 3 (0.7) 2 1 ESS 103 3 (3.0) 3 Vulva/Vagina 216 2 (0.9) SCC 134 1 (0.7) 1 Adenocarcinoma 32 1 (3.1) 1 Total 9645 86 (0.9) 45 (52.3) 41 (47.7)

To further examine the prevalence of ERmuts in human ovarian and endometrial cancers, publically available databases were explored using the cBioPortal and COSMIC

[121-129]. In total 41 gynecologic malignancies with ERmut were identified. The majority

(37/41, 90%) occurred in uterine tumors including 2 cases of endometrial stromal sarcoma.

No cases were identified in the cervical cancer TCGA database or the ovarian cancer

TCGA database, which comprises exclusively high-grade serous cancer cases. One endometrial case had 2 ERmut alleles (ERV422del and ERY537H). Interestingly, the majority of the ERmuts occurred in endometrioid tumors (34/41, 83%). An additional uterine endometrioid case was identified with an ERP535H mutation. This mutation has not been functionally characterized and it was not included in the analyses.

The overall frequency of cases with ERmuts in the publically available databases was similar to that seen in the CGP dataset. In the AACR Project GENIE databases, ERmuts were present in 2.3% (32/1363) of uterine endometrial cancers, 0.1% (2/1733) of ovarian cancers,

0.4% (1/279) of cervical cancers, and 0.9% (2/234) of uterine sarcomas. Similarly, ERmuts were enriched in endometrioid endometrial cancers [30/654 (4.6%) endometrioid vs 0/244 serous, p = 0.0001] and endometrioid ovarian cancers [2/70 (2.9%) endometrioid vs. 0/838 high-grade serous, p=0.006]. ERmut was enriched in ESS compared to LMS (2/16 vs 0/129, respectively, p = 0.012) in the AACR Project GENIE uterine sarcoma dataset. Of the 248 cases with mutation data in the TCGA uterine corpus dataset, 5 (2.0%) contained ERmuts, all with endometrioid histology; however, the comparison between endometrioid and serous histology did not show a statistically significant difference [5/200 (2.5%) endometrioid vs 0/44 serous, p=0.59].

In sum, combining our CGP dataset and the publically available datasets, 125 gynecologic malignancies with 129 individual ERmuts were identified. Variants most commonly occurred in the known hotspot region, with 29/129 (22.5%) in codon 536, 59/129

(45.7%) in codon 537, and 24/129 (18.6%) in codon 538. Figure 15 presents a schematic overview and frequency distribution of the ERmuts identified. ERY537S (42/129, 32.6%) and

ERD538G (24/129, 18.6%)were the most common individual ERmuts identified.

A) 59 Y537S/N/C

L536H/P/R/V 27 23 D538G S463P12 V422del3

NH2 NTD DBD hinge LBD COOH 18 26 30 1 552 595 0 3 2

Figure 15: Schematic overview of ERmuts identified in gynecologic malignancies.

(A). Distribution of mutations identified (B). Frequency of individual variants identified. N=123, DBD: DNA Binding Domain, LBD: Ligand Binding Domain

3.2.2 Clinical relevance of ERmuts in gynecological malignancies and response to treatment

Detailed clinical information was available for 8 patients with ERmuts. Figure 16 and

Table 9 illustrates the clinical course for patients identified with ERmuts tumors. In 5 of 8 cases, ERmut was identified after AI exposure. Serial sampling performed in 3 showed the emergence of ERmut (Patients A and G) or an increase in allelic frequency (Patient D).

Patient D had two exposures to AI: one of short duration (1 month) and a subsequent exposure lasting 7 months. Because no pre-AI sample is available, when the ERmuts first evolved cannot be determined. However, the allelic frequency of ERY537S was increased after the second exposure (37% post-7-months vs 4% post-1-month, 9.25-fold increase) compared to KRAS (40% post-7 months vs 23% post-1 month, 1.74-fold increase).

Figure 16: Clinical relevance of ERmuts in gynecologic malignancy.

Clinical course of patients identified with gynecologic malignancies harboring ERmuts (designated here as mutESR1). Eight individual patients (A-H) with ERmuts were identified. Each box or wide arrow delineates a treatment received. White is surgery, dark pink is chemotherapy, light blue is fulvestrant, peach is AI, teal is tamoxifen, gray is immunotherapy and dark blue is other targeted therapy. The width reflects duration of therapy. A wide arrow represents ongoing therapy. Hashed boxes/arrows represent combined therapy. The triangle reflects when the sample evaluated by CGP was procured.

Of the 8 cases, 6 experienced a greater duration of benefit with ER directed therapy than chemotherapy. In some cases, ER-directed therapy led to extended benefit. In particular, a 58-year-old woman (Patient C) with Stage IIIC primary peritoneal low-grade serous carcinoma had no tumor response to neoadjuvant chemotherapy. She underwent cytoreductive surgery and CGP revealed an ERY537N mutation. The patient had no prior history of endocrine therapy. Given the concern for intrinsic resistance to AI therapy conferred by the mutation, treatment with fulvestrant was started and within 4 months she experienced a major CA125 biochemical response. She has had prolonged clinical benefit of N4 years with minimal residual disease based on radiologic imaging. In other cases, switching therapy to an alternate SERM or SERD after progression on the initial ER- targeted therapy also provided clinical benefit (Patients A and B). Patient A experienced a combined total of 20 months CBD with tamoxifen followed by fulvestrant compared to

2.5 months CBD with the antecedent chemotherapy. Therefore, despite the presence of an activating ESR1 mutation, therapy with ER-targeting agents can be beneficial for some patients with ERmuts associated gynecologic malignancies.

Table 9: Clinical Characteristics of patients identified with ERmuts in gynecological malignancies

Patient Age Disease site Prior AI Benefit of SERM/SERD A 24 Ovary No Yes B 35 Synchronous No Yes endometrial and ovary C 58 Primary No Yes Peritoneal D 43 Ovary Yes No E 59 Ovary Yes Yes F 40 Ovary Yes No G 59 Ovary Yes Yes H 59 Endometrial Yes Yes

3.2.3 ERmuts confer partial resistance to endocrine therapy in ovarian cancer cells

In breast cancer cell lines, ERmuts exhibit decreased sensitivity to tamoxifen and fulvestrant. We evaluated the transcriptional activity in response to SERMs or SERDs of the most common ERmuts identified in our series using a reconstituted ERE-luciferase reporter assay in CAOV2 ovarian carcinoma cells. ER constructs were generated containing 3 different amino acid substitutions at position 537 (ERY537C, ERY537N, ERY537S) and 1 at position 538 (ERD538G). ERWT was only activated in the presence of E2. Each of the mutants exhibited substantial constitutive activity in the absence of E2 when compared to the activity of ERWT (Figure 17). The constitutive activity of each of the receptors was similar to the maximally E2-stimulated activity of ERWT, with the exception that ERY537S

showed statistically significant increased activity (134% compared to WT receptor, p =

0.0137).

The ability of clinically available SERMs and SERDs (4-OHT; raloxifene; bazedoxifene; lasofoxifene; fulvestrant and GW7604) to inhibit ERmut transcriptional activity was evaluated in ovarian cancer cells. In the CAOV2 ovarian cancer cell model, each of the drugs was able to effectively inhibit ERmut transcriptional activity, albeit with reduced potency when compared to ERWT. Differences in potency between ERWT and

ERmuts varied considerably by drug. For example, the 4-OHT IC50 required for ERY537C is 9X higher than for ERWT, while the fulvestrant IC50 required for ERY537C was >1600X higher that required for ERWT.

Interestingly, the mutant receptors themselves exhibited differential responses to the drugs (Figure 17). The IC50 for each of the drugs on the ERY537N and ERY537C mutants were typically similar and within ~2 fold of each other. In contrast, the IC50 for the ERY537S and ERD538G mutants were more likely to be similar and frequently higher than the IC50 for either of the other two mutants.

The IC50 and IC90 of each receptor/ antagonist pair was compared in our assays

(Table 10) to the antagonists reported maximum achievable blood concentration in humans (Cmax) [130-133]. Because the Cmax of the tamoxifen metabolite, 4-OHT, was not available, the median serum concentration of 4-OHT measured in patients receiving

tamoxifen 20 mg daily was used; the typical dose of tamoxifen used for the treatment of gynecologic malignancies is 20 mg twice a day [134]. Figure 17C is a visual representation of the Cmax/IC90 and shows that the concentration to reach the IC90 is achievable with each of the antagonists for the ERY537N and ERY537C mutations, while only lasofoxifene would be expected to achieve the IC90 for the ERD538G mutation and none reach the IC90 for ERY537S.

A) B) Constitutive Activity of Mutant Receptors Fulvestrant 4-hydroxytamoxifen 5 1.5 1.5 ERWT ERWT

U 4 Y537S Y537S L

U ER U ER L L R

Y537N Y537N R 1.0 ER R 1.0 ER d 3 d d e

e Y537C e Y537C z ER ER z z li li li

a 2 a ERD538G a ERD538G rm

rm 0.5 rm 0.5 o 1 o o N N N

0 0.0 0.0 T S C G W E2 0 -12 -10 -8 -6 0 -12 -10 -8 -6 + Y537 Y537N Y537 D538 WT Log [M] Log [M]

Raloxifene Bazedoxifene 1.5 2.0 ERWT ERWT Y537S Y537S

U ER U ER 1.5 C) L Y537N L Y537N R 1.0 ER R ER d d

e ERY537C e ERY537C z z 1.0 li li 100 a ERD538G a ERD538G

50 rm 0.5 rm o o 0.5 10 N N 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6

0 8 9 Log [M] Log [M] C /I 6 ax

m Lasofoxifene GW7604 C 4 1.5 2.0 ERWT ERWT Y537S Y537S

2 U ER U ER 1.5 L Y537N L Y537N R 1.0 ER R ER d d

0 e ERY537C e ERY537C z z 1.0 T N C S G li li

a D538G a D538G W ER ER

R Y537 Y537 Y537 D538 rm 0.5 rm E R R R R o o 0.5 E E E E N N

Fulvestrant Raloxifene 0.0 0.0 0 -12 -10 -8 -6 0 -12 -10 -8 -6 4-hydroxytamoxifen Bazedoxifene Log [M] Log [M] Lasofoxifene

Figure 17: ER LBD mutations confer constitutive transcriptional activity and alter receptor sensitivity to SERMs/SERDs.

(A) Luciferase assay measuring transcriptional activity of ERWT and ERmuts in the presence of E2. (B) Inhibition dose response curves for each anti-estrogen. All inhibition curves were done in the presence of 10–9 (1 nM) E2. Data normalization is performed with respect to the -14 data point of each treatment group for each individual receptor. These plots include data from five independent experiments and each value is an average of triplicates from each experiment. (C) Cmax for each ER antagonist were identified in the literature and divided by each receptor’s IC90 as determined by the dose-response curves in the luciferase assays. The values are reported on a logarithmic scale.

Table 10: IC50 and IC90 values of antagonists (pM)

Fulvestrant 4-OHT Raloxifene Bazedoxifene Lasofoxifene GW7604 IC50 (IC90) IC50 (IC90) IC50 (IC90) IC50 (IC90) IC50 (IC90) IC50 (IC90)

ERWT 0.53 (4.76) 102.7 (924) 13.27 (119) 1.867 (16.8) 0.7425 (6.68) 8576 (77200) ERY537N 830.9 1050 (9450) 46.98 (423) 268.1 (2410) 82.75 (745) 38290 (7480) (345000) ERY537C 887.8 923.1 (8310) 121.5 (1090) 166.1 (1490) 22.69 (204) 22270 (7990) (200000) ERY537S 9606 8877 (79900) 696.8 (6270) 4548 (40900) 1351 (12200) 218900 (86500) (1970000) ERD538G 8102 4393 (39500) 487.7 (4390) 2301 (20700) 208.5 (1880) 86870 (72900) (782000)

3.3 Discussion

This study demonstrated that activating mutations within the ESR1 LBD occur in gynecologic malignancies. This finding has important treatment implications. ER- dependent malignancies harboring these mutations are unlikely to respond to estrogen deprivation therapies, such as AIs, because these alterations confer constitutive transcriptional activity to ER. The frequency of ERmut in the current study is low, only 0.9% in unselected cases. However, these mutations are enriched in endocrine-responsive subtypes of gynecologic malignancies: specifically, the endometrioid histologic subtype of endometrial and ovarian cancers. Identification of ERmut in up to 5% of endometrioid endometrial cancers is higher than previously reported (1.8%)[120]. The difference may be due to sampling of more recurrent cases or post treatment cases in the current dataset.

Another endocrine-sensitive histologic subtype, ESS, also exhibited a higher frequency of

ERmut than its counterpart, LMS, however this was statistically significant in only one dataset. Four cases exhibited two ERmut consistent with reports of heterogeneity and polyclonality of ERmut in breast cancer [135]. The frequency of ESR1 amplifications in the current dataset, specifically in the endometrial cancer cohort, is much lower than previously reported [116-118]. The clinical relevance of ESR1 amplifications is still unknown.

Endocrine therapy is preferentially used in low-grade gynecologic malignancies.

Whether ERmut are enriched in low-grade gynecologic malignancies could not be determined because of the limited information regarding tumor grade in the current study. However, ERmut appear to be enriched in histologic subtypes likely to have a higher proportion of low-grade cases (i.e. endometrioid vs serous histology), supporting the hypothesis. Further studies comparing the prevalence of ERmut in high-grade and low- grade subtypes are needed. Nevertheless, these variants are more likely to be clinically relevant in the subset of tumors for which endocrine therapy is commonly utilized.

In breast cancer, ERmuts have been demonstrated to arise primarily as a resistance mechanism to estrogen deprivation therapy resulting in ligand-independent ER signaling.

The current study is limited by the lack of information regarding prior treatment history from the majority of patients whose tumors underwent CGP. Thus it is impossible to

determine whether ERmuts emerge in gynecologic malignancies as a result of exposure to endocrine therapy. However, at least three of the clinical cases support the hypothesis that these mutations are selected for or emerge as resistance mechanisms to AI therapy. This is consistent with a recent case report of a patient with low-grade serous cancer who was found to have a ERY537S mutation in a metastatic lesion that progressed after treatment with AI [136]. Interestingly, ERmuts in breast cancer appear more likely to develop after AI therapy in the metastatic versus adjuvant treatment setting [56, 137]. In a recent study of adjuvant AI for low-grade serous ovarian cancer, no ERmuts were identified in the small proportion of patients who developed recurrent disease[138]. Thus, ERmuts may be more likely to develop in patients treated with AIs for metastatic disease.

Conversely, three cases (Patients B, C and F) had no known history of endocrine therapy prior to identification of ERmuts. For Patient B, the mutation was present in tumor at diagnosis, suggesting that the mutation arose de novo and possibly played a role in the pathogenesis of the tumor. Similarly, the ERmuts identified in the TCGA cases were from samples collected at the time of diagnosis. However, whether patients may have had prior exposure to anti-estrogens for treatment of a separate malignancy, such as breast cancer is not known. Nonetheless, these mutations may develop independent of AI exposure and result in intrinsic resistance to estrogen-deprivation therapies.

Because of the constitutive activity conferred by the mutation, tumors with ERmuts

are unlikely to respond to AI therapy. The in vitro studies support prior reports that higher doses of anti-estrogens are required to inhibit mutant ER[51, 81]. However, ERmuts breast cancers can respond to tamoxifen and fulvestrant supporting that efficacious doses can be achieved clinically [54, 57]. For Patient C, early identification of the mutation has led to long-term clinical benefit with fulvestrant, extending past 52 months at the time of submission of this manuscript. Others also received clinical benefit, though this was not universal. Overall, SERM/SERD therapy has the potential to provide benefit despite the presence of a ERmut.

What accounts for the variable differences in benefit among the clinical cases of

ERmuts is unknown. Differences may be due to 1) the use of hormone therapy in a later phase of the disease course after development of multiple adaptive/resistance mechanisms, 2) the influence of co-occurring mutations; or 3. the specific ERmuts present within each tumor (e.g. ERY537N vs ERY537S). Supporting the third hypothesis, the transcriptional data demonstrated ER transcriptional activity with various SERMs/SERDS differs by individual mutation. This is similar to the findings of other groups who showed differences in response to SERMs/SERDs between the ERY537S and ERD538G mutations [48,

52]. The current data shows that different amino acid substitutions at the same site (ERY537C vs ERY537N vs ERY537S) also exhibit different IC50 to individual drugs. Furthermore, exposure (Cmax) of some agents exceeds the IC90 of some, but not all, mutated receptors.

It is important to recognize the limitations of this data: nuances in transcriptional response in vitro may not translate to clinical efficacy given the complexities of tumor proliferative signals and individual patient pharmacokinetic and pharmacodynamic considerations.

The finding of ERmuts in gynecologic malignancies has several important clinical implications. First, ERmuts may arise de novo in gynecologic malignancies in the absence of prior exposure to endocrine therapy. Endocrine therapy, especially AIs, may increase the prevalence of these mutations. Thus, traditionally endocrine-responsive gynecologic malignancies should be assessed for these mutations. This may be especially important for tumors that develop resistance or are refractory to endocrine therapy. Second, the functional and clinical data supports the use of alternative endocrine therapy, namely

SERMs or SERDs, in some patients with ERmuts gynecologic malignancies. Thus, the identification of a ERmuts does not preclude further treatment with selected endocrine therapy, though relative response may be affected by the individual mutation present.

Given the low frequency of these mutations in gynecologic malignancies, a large scale effort will be required to delineate the prevalence of ERmuts across gynecologic malignancies and conditions under which they arise, with an emphasis on malignancies considered for treatment with endocrine therapies. The recent development of technology that can be used to evaluate mutation status in plasma circulating tumor DNA may be useful to non-invasively monitor tumor mutation status over time and in response to

treatment. This approach has identified greater heterogeneity and polyclonality in the development of ERmuts than had been appreciated with tissue evaluation from individual mutation sites [54, 135, 139]. Finally, prospective studies using ESR1 mutation status to direct endocrine therapy should be undertaken to understand how these mutations may be used to guide therapeutic decision making and personalize care for patients with gynecologic malignancies.

4. Characterization of a novel SERD for the treatment of endocrine progressing breast cancer

This chapter represents a collaborative effort with Dr. John Norris, Dr. Suzanne

Wardell, Jennifer Baker, Taylor (Krebs) Desautels and other colleagues in the McDonnell

Laboratory as well as collaborators at G1 Therapeutics, Inc. and the University of Illinois at Chicago [140]. This article is Open Access and the Creative Commons License can be found at http://creativecommons.org/licenses/by/4.0/. This article remains largely unchanged other than formatting to fit with the other chapters in this thesis and the removal of some supplementary files.

Andreano, K.J.*, Wardell S.E.*, et al., G1T48, an oral selective estrogen receptor degrader, and the CDK4/6 inhibitor lerociclib inhibit tumor growth in animal models of endocrine- resistant breast cancer. Breast Cancer Res Treat, 2020. 180(3): p. 635-646.

4.1 Introduction

ER (ESR1) is expressed in a majority of breast cancers, and drugs that inhibit ER signaling are the cornerstone of breast cancer pharmacotherapy for ER-positive/HER2- negative disease [21]. These targeted approaches include the SERM tamoxifen that acts as a competitive ER antagonist in the breast, and AIs that inhibit aromatase, the enzyme responsible for estrogen production [23, 27]. However, the development of resistance limits the duration of meaningful therapeutic responses. Mechanisms of resistance to these endocrine therapies include cell cycle dysregulation and activation of alternative growth factor signaling pathways [21]. For example, activation of MAPK, PI3K, and GSK-

3 can result in increased phosphorylation of ER or its attendant coregulatory proteins leading to ligand-independent ER activity and resistance [141-144]. Recently, genomic alterations in the ER gene itself, including amplification, translocation, and ligand binding domain mutations (most frequently ERY537S and ERD538G) have emerged with AI therapy

[21, 34, 35, 53].

After progression during tamoxifen and AI therapy, other endocrine treatments including the steroidal SERD fulvestrant are generally used [25]. SERDs are a class of ER antagonists that in addition to competitively displacing estrogens, also trigger ER downregulation [24]. Although initially successful, the onset of resistance limits durable responses when used as a monotherapy. Therefore, in an effort to improve the therapeutic lifespan of endocrine treatments for metastatic breast cancer, combination regimens have been extensively studied. Clinical trials using a combination of AI or fulvestrant with pan-

PI3K or mTOR inhibitors have been promising but inconclusive, and toxicity often remains an impediment to dose escalation [40-43]. Therefore, CDK4/6 inhibitors have emerged as a favored option when considering combination endocrine therapies [44-47].

However, the poor bioavailability of fulvestrant, coupled with its intramuscular route of administration and the long time to steady state blood levels, compromises its clinical use

[31, 32]. Indeed, even at the higher clinical dose (500mg) of fulvestrant, pharmacodynamic imaging suggests incomplete receptor saturation [145].

Collectively, these data highlight an unmet need for a safe, orally bioavailable

SERD with appropriate pharmaceutical properties. Herein we describe the preclinical development of G1T48, an orally bioavailable, potent, and selective nonsteroidal ER antagonist and downregulator [146]. G1T48 was found to robustly inhibit ER activity in multiple in vitro models of endocrine therapy resistance, including those harboring ER mutations or growth factor activation. Importantly, G1T48 demonstrated robust antitumor activity in an animal model of early stage estrogen-dependent breast cancer and suppressed the growth of tamoxifen and estrogen deprivation resistant xenograft tumors with increased efficacy observed for the combination of G1T48 and lerociclib, a newly developed CDK4/6 inhibitor [147, 148].

4.2 Results

4.2.1 G1T48 is similar to fulvestrant in its ability to downregulates the estrogen receptor and inhibit estrogen signaling in breast cancer cells

Novel ER antagonists with SERD activity have recently been described, but clinical development of these compounds has thus far been limited due to unanticipated side effects or for undisclosed reasons [65, 70, 71, 78, 83, 88, 111, 149]. We sought to identify an orally bioavailable SERD using the chemical backbone of raloxifene as a starting point, since this SERM has demonstrated a favorable safety profile in the clinical setting of breast

cancer prevention and osteoporosis treatment [90, 150]. G1T48 incorporates an acrylic acid side chain (Figure 18A) and was the product of structure-guided investigations, driven by activity in breast cancer cell lines [65, 70, 73, 77, 78, 88, 111, 146]. G1T48 was first assessed for its ability to downregulate ER when compared to several benchmark SERMs and SERDs including fulvestrant [24, 89]. Using In-Cell Western assays, G1T48 was found to downregulate ER with an efficacy modestly more potent than steroidal and other

SERDs (fulvestrant, AZD9496; approximately 10% ER remaining after treatment) (Figure

18B and Table 11). These data demonstrate that in vitro G1T48 is a potent and efficacious

SERD.

A) O F O F OH S OH CF3 O

H F F OH O F H H S F H N N HO F O O H H HO Fulvestrant RU58,668 Tamoxifen 4-hydroxytamoxifen

O HO O OH

HO N HO O

N O N O

HO S OH GW5638 GW7604 Raloxifene Bazedoxifene HOOC HO O H O N o N H O NH H F F H N F O N F Cl F HO HO S O GDC-0810 AZD9496 G1T48 Lasofoxifene

B) Fulvestrant RU58,668 Tamoxifen* 4-OH GW5638* GW7604 Raloxifene Bazedoxifene GDC-0810 AZD9496 G1T48 Lasofoxifene

-12 -11 -10 -09 -08

-07 Log (M) Ligand -06

Figure 18: G1T48 is a potent SERD.

(A) Chemical structures of G1T48 and benchmark SERMs and SERDs. (B) MCF7 cells were treated with ER ligands (10-12 – 10-6 M) for 18 hours prior to fixation and detection of ER levels by In-Cell Western. *For tamoxifen and GW5638, dose response was 10-11 – 10-5 M.

Table 11: ER degradation IC50 values in MCF7 cells

Compound IC50 Fulvestrant 2.74E-10 RU 58668 2.02E-11 Tamoxifen 4.46E-08 4-hydroxytamoxifen 1.00E-10 GW5638 3.11E-07 GW7604 4.20E-10 Raloxifene 5.94E-11 Bazedoxifene 5.69E-10 GDC-0810 9.49E-11 AZD9496 1.08E-11 G1T48 3.30E-11 Lasofoxifene 1.89E-11

We next evaluated the ability of G1T48 to inhibit endogenous ER target gene transcription in MCF7 cells. As shown in Figure 19A, G1T48 suppressed estrogen- mediated activation of the Trefoil Factor-1 (TFF1) mRNA similarly to fulvestrant and additional antiestrogens (4-OHT, GDC-0810, AZD9496, raloxifene). The biochemical basis of G1T48-mediated ER antagonism was further evaluated using 3H-estradiol whole-cell competition assay. Results showed that G1T48 displaced radiolabeled agonist binding with potency greater than fulvestrant and similar to bazedoxifene (Figure 19B and Table

12).

The inability of some ER antagonists, notably SERMs, to completely oppose the actions of estradiol is seen as a liability when being considered for the treatment of advanced therapy-resistant breast cancer. While data from Figure 19B confirm G1T48 is

a SERD, the potential remains for SERM activity, as G1T48 was developed based on compounds that exhibit both SERM and SERD activity. Therefore, we next considered the impact of G1T48 treatment on ER target genes that are differentially regulated by SERMs and SERDs [151]. As shown in Figure 19C, compounds with SERM activity regulate these genes in a manner similar to the agonist estradiol, a reflection of their intrinsic agonist potential (red, green, and blue clusters). In contrast, G1T48 regulates these genes in a pattern that is consistent with compounds previously shown to downregulate ER (e.g.

GW7604, fulvestrant, GW5638, RU 565899, GDC-0810, and AZD9496; orange, teal and purple clusters).

When bound by estrogen, ER is recruited to target gene promoters to activate or repress target gene transcription through recruitment of coregulator (coactivator or corepressor) proteins that modify chromatin structure [152]. We assessed the ability of

G1T48 and benchmark SERMs or SERDs to inhibit estrogen-mediated recruitment of ER to the TFF1 promoter using chromatin immunoprecipitation (ChIP) assays. While estrogen and 4-OHT treatment significantly increased ER binding to the TFF1 promoter

(Figure 19D), G1T48 inhibited the binding of ER to this promoter, with or without estrogen, with efficacy similar to fulvestrant, supporting the idea that G1T48 is an efficient

ER antagonist in vitro.

A) 6 TFF1 5

Fold Induction 2

0 Veh -9 -10 -9 -8 -7 -10 -9 -8 -7 -10 -9 -8 -7 -11 -10 -9 -8 -10 -9 -8 -7 -10 -9 -8 -7 Log (M) E2 G1T48 +E2 Fulvestrant +E2 GDC-0810 +E2 AZD9496 +E2 Raloxifene+E2 4-OHT+E2

ER-RBA B) 1500 C) E2 Tamoxifen n i E2 e t 4-OHT

o 1000

r Fulvestrant p Lasofoxifene g RAL u /

M BZA Vehicle 500 P

C G1T48 Raloxifene Bazedoxifene 0 -12 -10 -8 -6 -4 GW7604 Log (M) ligand GW5638 G1T48 TFF1 Fulvestrant D) 2.0 Vehicle RU58,668 1.5 E2 GDC-0810 t u AZD9496 np I

1.0 %

0.5 T13 R SDK2 K SNX24 AGR2 RGNEF RAPGEL1 SLC6A14 STC2 SLC2A1 FHL1 YWHAZ SLFN5 PMP22 RFTN1 LRP2 0.0 t 6 High Low h 8 T 9 e 4 H V T tran 1 O D94 G 4- Z ulves A F

Figure 19: G1T48 is a complete estrogen receptor antagonist.

(A) MCF7 cells were treated with ER antagonists (10-10 – 10-7 M) plus estradiol (E2; 10-9 M) for 18 hours. TFF1 mRNA expression was analyzed by real-time PCR. GAPDH was used to normalize data. (B) MCF7 cells were treated with 10-10 M 3H-17β-E2 and competitor ligand (10-12 – 10-6 M) for 2 hours. Cells were collected and radioactive counts were monitored on a Beckman LS 6000SC Scintillation counter. Error bars indicate the SD of duplicate samples. (C) MCF7 breast cancer cells were treated with ER ligands (E2, fulvestrant, RU58,668, Raloxifene @ 100 nM; G1T48, 810, 9496, Lasofoxifene, 4-OHT, 7604, Bazedoxifene @ 1.0 µM; 5638, Tamoxifen @ 10 µM) for 24 hours. mRNA expression was analyzed by real-time PCR. GAPDH was used to normalize real- time PCR data. Heatmaps were generated from real-time PCR data after analysis with JMP pro software (SAS) using the Ward hierarchical clustering algorithm. (D) MCF7 cells were treated with ligand (E2: 5X10-10 M; ER antagonists: 10-6 M) as indicated for 45 mins. Cells were fixed with formaldehyde and chromatin was immunoprecipitated with anti-ER antibody. Real-time PCR was used to assess the relative amount of ER bound to the TFF1 gene promoter. Error bars indicate the standard deviation of triplicate samples.

Table 12: Radioactive Binding Assay IC50 values

Compound IC50 Estradiol (E2) 6.62E-11 Fulvestrant 2.99E-09 Raloxifene 2.26E-10 Bazedoxifene 7.27E-10 G1T48 8.54E-10

4.2.2 G1T48 inhibits the growth of ER positive breast cancer cells

To examine the therapeutic potential of G1T48, we performed cell proliferation assays using multiple ER positive breast cancer cell lines (Figure 20 and Table 13). G1T48 significantly inhibited estrogen-mediated growth of MCF7 cells demonstrating approximately 3-fold higher potency when compared to fulvestrant (Figure 20A).

Additionally, G1T48 and benchmark antiestrogens also inhibited the estrogen-mediated growth of ER positive BT474 and ZR-75-1 breast cancer cells while no growth inhibition was observed in ER negative MDA-MB-436 breast cancer cells (Figure 20B-D). Thus,

G1T48 selectively inhibits the growth of ER-positive, but not ER-negative, breast cancer cells.

A) MCF7 B) BT474 10 5

8 4 Fulvestrant Fulvestrant n n o o i i t t 4-OHT 4-OHT c c 6 3 u u GDC-0810 GDC-0810 nd nd I I

d d 4 AZD9496 2 AZD9496 l l o o F F G1T48 G1T48 2 1

0 0 -12 -10 -8 -6 -4 -12 -10 -8 -6 -4 Log (M) ligand Log (M) ligand

C) ZR-75-1 D) MDA-MB-436 5 6

4 Fulvestrant Fulvestrant n n o o i i t 4-OHT t 4 4-OHT c 3 c u GDC-0810 u GDC-0810 nd nd I I

d 2 AZD9496 d AZD9496 l l

o o 2

F G1T48 F G1T48 1

0 0 -12 -10 -8 -6 -4 -12 -10 -8 -6 -4 Log (M) ligand Log (M) ligand

Figure 20: G1T48 inhibits ER- positive breast cancer cell growth.

(A) ER positive MCF7. (B) ER-positive BT474. (C) ER- positive ZR-75-1 were treated for 7 days with 10-10 M E2 in addition to ER antagonists (10-11 – 10-5 M). D) ER- negative MDA-MB-231 cells were treated for 7 days with ER antagonists (10-11 – 10-5 M). Cellular proliferation was assessed by measuring DNA content (Hoechst stain) and is presented as fold induction over vehicle control. Error bars indicate the standard deviation of triplicate samples.

Table 13: GI50 Values of antiestrogens in Breast Cancer Cell Lines

Compound MCF7 BT474 ZR-75-1 MDA-MB-436 Fulvestrant 8.48E-10 1.15E-08 7.21E-08 >1.0E-05 4-hydroxytamoxifen 3.58E-09 9.99E-07 1.02E-08 >1.0E-05 GDC-0810 1.82E-09 1.20E-07 1.17E-07 >1.0E-05 AZD9496 4.58E-11 2.68E-09 2.28E-09 >1.0E-05 G1T48 2.57E-10 2.21E-08 9.62E-08 >1.0E-05

4.2.3 G1T48 inhibits estrogen signaling in endocrine-resistant breast cancer models

A key mechanism of resistance to aromatase inhibition are ERmuts that result in reduced potency for 4-OHT and fulvestrant as compared to ERWT [35, 48-53, 64, 66, 113].

To assess the activity of G1T48 on endocrine-refractory ERmuts, we utilized a reporter gene assay in ER negative SKBR3 breast cancer cells transfected with ER expression vectors

(ERWT, ERY537S or ERD538G) (Figure 21A-C and Table 14). G1T48 was found to be a potent and effective inhibitor of both ERWT and ERD538G transcription. As has been previously reported [50, 51], all antiestrogens tested, including G1T48, demonstrated reduced potency against ERmut activity when compared to ERWT (Table 14).

Dysregulated growth factor signaling has emerged as another primary mechanism of resistance to tamoxifen and AI therapy [34]. Activation of these signaling pathways can alter the pharmacology of compounds like tamoxifen, converting them from antagonists to agonists through phosphorylation of ER or its attendant coregulator proteins [141-144].

G1T48 and comparator SERMs and SERDs were evaluated for their ability to inhibit insulin-mediated MCF7 cell proliferation, a model for endocrine therapy resistance.

Compounds with SERD activity, including G1T48, effectively blocked growth factor- mediated cell growth, while compounds with SERM activity (4-OHT) were less effective

(Figure 21D). These data together support the potential for G1T48 to have efficacy in the

treatment of AI or tamoxifen resistant breast cancers having growth factor pathway activation.

A) B) ERWT ERY537S 300000 500000 Fulvestrant Fulvestrant 4-hydroxytamoxifen 400000 4-hydroxytamoxifen 200000 Raloxifene Raloxifene Bazedoxifene 300000 Bazedoxifene RLU RLU G1T48 G1T48 200000 100000 100000

0 0 -12 -10 -8 -6 -4 -12 -10 -8 -6 -4 Log (M) ligand Log (M) ligand

D538G C) ER D) MCF7 (20mM Insulin) 600000 80000 Fulvestrant Fulvestrant 4-hydroxytamoxifen 4-hydroxytamoxifen 60000 400000 Raloxifene Raloxifene Bazedoxifene Bazedoxifene 40000 RLU G1T48 RFU G1T48 200000 20000

0 0 -12 -10 -8 -6 -4 -14 -12 -10 -8 -6 -4 Log (M) ligand Log (M) ligand

Figure 21: G1T48 inhibits ER signaling in models of endocrine therapy resistance in vitro.

(A-C) SKBR3 breast cancer cells were transfected with an estrogen-responsive reporter gene together with ER (ERWT, ERY537S, or ERD538G) expression plasmids prior to 18 hours of treatment with E2 (1.0 nM) and ER antagonists (10- 11 – 10-5 M). Luciferase activity was assessed. Error bars indicate the standard deviation of triplicate samples. (D) MCF7 cells were treated for 7 days with insulin (20 nM) plus increasing dose of antiestrogens (10-12 – 10-7 M). Cellular proliferation was assessed by measuring DNA content (Hoechst stain) and is presented as relative fluorescence units. Error bars indicate the standard deviation of triplicate samples.

Table 14: ERmut and ERWT transcriptional IC50 (M) Values

Compound ERWT ERY537S ERD538G Fulvestrant 4.49E-09 2.02E-08 1.05E-08 4-hydroxytamoxifen 3.25E-09 1.10E-08 3.11E-09 Raloxifene 2.73E-10 4.79E-09 5.76E-09 Bazedoxifene 9.79E-10 2.90E-08 7.97E-09 G1T48 2.14E-09 3.45E-08 1.47E-08

We next evaluated the therapeutic potential of G1T48 in ER-positive primary and endocrine refractory breast cancer models in vivo. G1T48 was first assessed, as a monotherapy or in combination with the CDK4/6 inhibitor lerociclib, for its impact on the growth of naïve MCF7 xenograft tumors (Figure 22A). Ovariectomized estrogen-treated female nu/nu mice bearing MCF7 xenograft tumors were randomized to treatment with vehicle, lerociclib (50 mg/kg), and/or G1T48 (30 or 100 mg/kg). G1T48 treatment demonstrated dose-dependent repression of tumor growth. Combination of lerociclib and

G1T48 was more effective than either monotherapy, demonstrating an added benefit to using these agents together.

The TamR xenograft model exhibits tamoxifen-stimulated growth that can be inhibited by compounds with SERD activity with added benefit observed upon combination with CDK4/6 inhibitors [63]. Therefore, ovariectomized tamoxifen-treated mice bearing TamR xenografts were randomized to treatment with lerociclib (50 mg/kg or 100 mg/kg), G1T48 (30 mg/kg or 100 mg/kg), fulvestrant (200 mg/kg), or CDK4/6 inhibitor palbociclib (100 mg/kg) as monotherapies or a combination of lerociclib (50 mg/kg) and G1T48 (30 or 100 mg/kg). In this model system, lerociclib demonstrated efficacy equivalent to that of the mechanistic clinical comparator palbociclib (Figure 22B).

G1T48 was found to demonstrate dose-dependent inhibition of TamR tumor growth

(Figure 22C) albeit with less efficacy than fulvestrant. Interestingly, G1T48 treatment resulted in greater downregulation of intratumoral ER levels than fulvestrant despite less efficient inhibition of tumor growth (Figure 23). Finally, combination of G1T48 with lerociclib, using suboptimal doses of each inhibitor, resulted in tumor growth inhibition significantly greater than that observed for either compound as monotherapy (Figure

22D).

A) B) MCF7 in vivo tumor growth TamR in vivo tumor growth ) 3 ) 3 m 1000 Vehicle m 1000 (m Lerociclib(50 mg/kg) Vehicle (m e G1T48 (30 mg/kg) Palbociclib (100 mg/kg qd) e m 750

u Lerociclib(50 mg/kg qd) l m o G1T48 (100 mg/kg) u Lerociclib(100 mg/kg qd) l

V **** o **** r Lerociclib + G1T48 (30) v 500 o **** r 500 o m Lerociclib + G1T48 (100) u

m **** T u

250 t ** age age r r

e 0 e

v 0

0 10 20 30 40 v A 0 10 20 30 40 50 60 70 80 A Days of treatment Days of treatment C) D)

TamR in vivo tumor growth TamR in vivo tumor growth ) ) 3 3 m 1000 Vehicle m 1000 Vehicle Fulvestrant (200 mg/kg qw)

(m **** **** (m G1T48 (30 mg/kg qd)

G1T48 (30 mg/kg qd) e e Lerociclib (50 mg/kg qd) **** G1T48 (100 mg/kg qd) m m u u G1T48 (30 mg/kg) l l o o + Lerocicliib (50 mg/kg) v v

r **** 500 r 500 o o **** m n.s m u u t t age age r r e 0 e 0 v v

A 0 10 20 30 40 50 60 70 80

A 0 10 20 30 40 50 60 70 80 Days of treatment Days of treatment

Figure 22: Combination strategy of G1T48 and the CDK4/6 inhibitor lerociclib inhibit in vivo breast cancer xenograft models of estrogen-dependent MCF7 and tamoxifen-resistant (TamR).

(A) Ovariectomized estrogen-treated female nu/nu mice bearing MCF7 xenograft tumors were randomized to treatment with vehicle, lerociclib (50 mg/kg) or G1T48 (30 or 100 mg/kg), alone or together, daily for 28 days by oral gavage. 2-way ANOVA comparison of average tumor volumes throughout treatment, followed by Bonferroni multiple comparison test, indicated significant tumor growth inhibition by all treatments, as well as increased response to the combination of G1T48 (30 mg/kg) and lerociclib. Error bars represent standard error of measurement. (B, C, D) Ovariectomized tamoxifen-treated female nu/nu mice bearing TamR xenograft were randomized to treatment with vehicle, palbociclib (100 mg/kg), lerociclib (50 or 100 mg/kg) B), fulvestrant (200 mg/kg) or G1T48 (30 or 100 mg/kg) C), with lerociclib and G1T48 being tested alone and in combination D) Daily for 28 days. 2-way ANOVA comparison of average tumor volumes throughout treatment, followed by Bonferroni multiple comparison test, indicated significant tumor growth inhibition by all treatments, as well as increased response to the combination of G1T48 (30 mg/kg) and lerociclib. Error bars represent standard error of measurement.

Intratumoral ESR1

2.0 ) e l

c i Immunoblot analysis of intratumoral levels of estrogen receptor, pRb, and actin Immunoblotion 1.5 analysisImmunoblot of intratumoral analysislevels of intratumoral of estrogen receptor,levels of estrogenpRb, and receptor,actin pRb, and actin eh ss V

e Vehicle Vehicle G1T38, 50 mg/kgG1T38, 50 mg/kg

o Vehicle G1T38, 50 mg/kg r t p

Immunoblotx 1. analysis0 of intratumoral levels of estrogen receptor, pRb, and actin ed e ImmunoblotImmunoblot analysis of analysisintratumoral of intratumorallevels of estrogenlevels of receptor, estrogen pRb receptor,, and actin pRb, and actin

z li

Vehiclea G1T38, 50 mg/kg R1 Vehicle Vehicle G1T38, 50 mg/kgG1T38, 50 mg/kg Human ER Human ER Human ER Human ER Human ER Human ER

rm 0.5 ES no ( 0.0 Human ERpRb pRb Human ERpRb pRb pRb Humane ER ) HumanpRb) ER ) d Human) ER Human ER cl qd ehi g V /k Immunoblot analysis of intratumoral levels of estrogen receptor,mg pRb, and actin Immunoblot analysis of intratumoralβ-actin levelsβ-actin of estrogenβ-actin receptor,β-actin pRb, andβ-actin actin β-actin (30 pRb pRb pRb pRb pRb pRb Vehicle G1T38, 50Fulvestrant mg/kg Fulvestrant48 G1T48, 30mg/kgG1T48, 30mg/kg Vehicle G1T38,T 50 mg/kgFulvestrant G1T48, 30mg/kg G1T48, 100 mg/kgG1T48, 100 G1T48,mg/kg 100 mg/kg 1 G G1T48 (100 mg/kg qd Lerociclib (50 mg/kg qd β-actin Immunoblot Immunoblotanalysis of intratumoral analysis of intratumorallevels of estrogenβlevels-actin receptor, of estrogenβ-actin pRbFulvestrant receptor,β, -andactin (200 actin mg/kg pRbβ -qwactin, and actinβ-actin Human ER Human ER Human ER Human ER HumanHuman ER Lerociclib ER +G1T48 (30Human mg/kg ERq Human ER Human ER Human ER Fulvestrant Human ERG1T48, 30mg/kg Human ER Human ER Vehicle Vehicle G1T38, 50 mg/kgG1T38,Fulvestrant 50 mg/kgFulvestrant G1T48, 30mg/kgG1T48, 30mg/kgG1T48, 100 mg/kgG1T48, 100 mg/kgG1T48, 100 mg/kg Vehicle Lerociclib, 50 mg/kg Fulvestrant G1T48, 30 mg/kg G1T48, 100 mg/kg

pRb ERpRb pRb pRb pRb pRb pRb pRb pRb pRb pRb pRb pRb

Humanβ-actin ER Human ER Human ER HumanHuman ER ER Human ER Human ER Human ER Human ER Human ER Human ER Human ER

β-actin β-actin β-actin β-actinβ-actin β-actin β-actin β-actin β-actin β-actin β-actin β-actin β-actin Fulvestrant G1T48, 30mg/kg Fulvestrant G1T48, 30mg/kg G1T48, 100 mg/kgG1T48, 100 mg/kg pRb pRb pRbpRb pRbpRb pRb pRb pRb pRb pRb pRb pRb

β-actinHuman ER Human ER Human ER β-actinFigureHuman ER 23: Analysis ofβHuman β-intratumoral-actinactin ER β-actinβ-actin ESR1β-actin proteinβHuman-actin ERlevelsβ-actin in harvestedβ-actin tumorβ-actin tissueβ-actin β-actin Fulvestrant G1T48,(TamR). 30mg/kg Fulvestrant G1T48, 30mg/kg G1T48, 100 mg/kgG1T48, 100 mg/kg pRb pRb pRb pRb pRb pRb Preserved tumor tissues from fulvestrant treatment groups featured in Figure 6 were

processedHuman ER asHuman describedβ -ERactin (methods)Human ER andHumanβ- actinanalyzed ER by immunoblot.Human ER β-actinHuman Bands ER detected were β-actin β-actin β-actin quantitated (ImageJ) and normalized within each blot to the average expression detected for

triplicate vehiclepRb samples (the same 3 vehicle samples were included on each separate blot, pRb pRb pRb pRb pRb indicated by the red boxes).

β-actin β-actin β-actin β-actin β-actin β-actin

4.2.5 Evaluation of the combined efficacy of lerociclib and G1T48 in a xenograft tumor model of resistance to estrogen deprivation in vivo

Although AIs have seen rapid adoption in the adjuvant setting, de novo and acquired resistance remains a persistent impediment to sustained clinical responses. We have developed an ER positive model of aromatase resistance, termed long term estrogen deprived (LTED), to model this clinical situation [86]. In order to evaluate the combined efficacy of lerociclib and G1T48 in this model system, LTED xenograft tumors were orthotopically established in ovariectomized female nu/nu mice. When tumors measured

0.1-0.15 cm3 volume, G1T48 (5 mg/kg or 100 mg/kg) and lerociclib (50 mg/kg or 100 mg/kg) were administered alone and in combination, with fulvestrant (25 mg/kg) and palbociclib

(100 mg/kg) included for comparison. As previously observed with the MCF7 parental and TamR models, G1T48 demonstrated dose-dependent inhibition of tumor growth

(Figure 24A). Additionally, the tumor growth inhibition after treatment with G1T48 and lerociclib was comparable to their clinical comparators (fulvestrant and palbociclib, respectively) (Figure 24B). Combination of G1T48 with lerociclib suppressed tumor growth significantly compared to monotherapy (Figure 24C) and resulted in tumor regression for a majority of tumors receiving the combined therapy regimen.

A) B) C) LTED in vivo tumor growth LTED in vivo tumor growth LTED in vivo tumor growth ) ) ) 3 1.25 3 1.25 3 1.25 m m m c (c (c (

1.00 1.00 1.00 e e e m m m u u u ol 0.75 ol 0.75 ol 0.75 v v v

r r r o o o m 0.50 m 0.50 m 0.50 u u u t t t

e e e 0.25 0.25 0.25 ag ag ag r r r e e e v v v A 0.00 A 0.00 A 0.00 0 20 40 60 0 20 40 60 0 20 40 60 Days of treatment Days of treatment Days of treatment Vehicle Vehicle Vehicle Palbociclib (100 mg/kg qd) Fulvestrant (25 mg/kg q2w) G1T48 (5 mg/kg qd) Lerociclib (50 mg/kg qd) G1T48 (5 mg/kg qd) Lerociclib (50 mg/kg qd) Lerociclib (100 mg/kg qd) G1T48 (100 mg/kg qd) G1T48 (5 mg/kg) + Lerociclib (50 mg/kg)

Figure 24: Combination strategy of G1T48 and the CDK4/6 inhibitor lerociclib in vivo in an estrogen deprived xenograft model.

Ovariectomized vehicle-treated female nu/nu mice bearing LTED xenograft were randomized to treatment with vehicle, lerociclib (50 or 100 mg/kg) or palbociclib (100 mg/kg) (A) or G1T48 (5 or 100 mg/kg) or fulvestrant (25 mg/kg) (B), alone or together (C), Daily for 28 days. 2-way ANOVA comparison of average tumor volumes throughout treatment, followed by Bonferroni multiple comparison test, indicated significant tumor growth inhibition by all treatments, as well as increased response to the combination of G1T48 (30 mg/kg) and lerociclib. Error bars represent standard error of measurement.

4.2.6 Evaluation of the combined efficacy of lerociclib and G1T48 in a Patient Derived Xenograft Model harboring the ERY537S Mutation

As described above, mutations in the LBD of ESR1 are an emerging mechanism of resistance to AIs. The efficacy of G1T48, as a mono- and combination therapy with lerociclib, was evaluated using a Patient Derived Xenograft (PDX model) harboring the

ERY537S mutation (Figure 25A). Female athymic nu/nu mice were implanted with the

ST2177 LUMB PDX tumor and following treatment with G1T48, a dose-dependent decrease in tumor growth was observed [83]. Intriguingly, G1T48 alone (30 and 100 mg/kg) or in

combination with lerociclib was more efficacious than fulvestrant (Figure 25A). Survival curve analysis demonstrated that the combination of 30 mg/kg G1T48 with lerociclib was more effective than monotherapy using either drug alone (Figure 25B and C). Taken together these data highlight that G1T48 is either comparable or superior to fulvestrant in several models of endocrine therapy resistance, demonstrating its potential as a therapeutic agent.

ERMut LUMB breast tumor (ST2177) )

A) 3 1500 Vehicle Fulvestrant (5 mg/animal) 1000 G1T48 (30 mg/kg) G1T48 (100 mg/kg) Lerociclib(50 mg/kg) 500 Lerociclib + G1T48 (30) Lerociclib + G1T48 (100)

Average tumor volume (mm volume tumor Average 0 10 20 30 40 50 60 70 Days of treatment

B) ERMut LUMB breast tumor (ST2177) 110 100 Vehicle 90 G1T48 (30 mg/kg) 80 Lerociclib (50 mg/kg) 70 G1T48 (30 mg/kg) + Lerociclib 60 50 40 * * * 30 Percent survival Percent 20 10 0 0 10 20 30 40 50 60 70 Time

C) ERMut LUMB breast tumor (ST2177) 110 100 90 Vehicle 80 G1T48 (100 mg/kg) 70 Lerociclib (50 mg/kg) 60 * * G1T48 (100 mg/kg) + Lerociclib 50 * 40

Percent survival Percent 30 20 10 0 0 10 20 30 40 50 60 70 Time

Figure 25: Evaluation of the combined efficacy of lerociclib and G1T48 in a Patient Derived Xenograft Model harboring the ESR1 Y537S Mutation.

(A) Female nu/nu mice were engrafted with a START- PDX model, designated ST2177, harboring an ESR1 Y537S mutation. Mice were randomized to vehicle, fulvestrant (5 mg/animal), G1T48 (30 or 100 mg/kg), lerociclib (50 mg/kg), or the combination of G1T48 and lerociclib and treated for 60 days. (B, C) Kaplan-Meier analysis is presented as time for tumors to reach endpoint (2.5 times original tumor volume). * Kaplan Meier analysis followed by a Mantel-Cox test for significance demonstrated significantly greater tumor growth delay for these comparisons using an adjusted Bonferroni threshold of p < 0.012. Error bars represent SEM.

100

4.3 Discussion

Targeting ER activity using therapies that directly oppose the mitogenic action of estrogen or that block estrogen synthesis is a proven strategy for the treatment and prevention of breast cancer. In locally advanced or metastatic disease, resistance to these therapies frequently emerges within two years, at which time treatment options are severely limited. Fulvestrant, a potent ER antagonist and downregulator, was initially approved for the treatment of endocrine therapy-resistant disease and more recently as first-line therapy for advanced ER-positive, HER2-negative breast cancer not previously been treated with endocrine therapy. However, despite promising preclinical activity, the poorly controlled pharmacokinetics of fulvestrant remains a significant barrier to prolonged clinical efficacy. Clinical trials comparing high-dose (500 mg) to low-dose

(250mg) fulvestrant demonstrated superiority for the 500 mg dose in both first- and second-line settings, suggesting that increased target engagement can improve the outcome of ER degradation therapy [39, 40]. However, given its intramuscular route of administration, continued improvements in the clinical response to fulvestrant by further dose escalation appear unlikely. Therefore, development campaigns in this area have focused on the identification of orally bioavailable SERDs. The most active SERDs share common chemical features: either (a) a steroidal backbone (e.g. fulvestrant, RU58,668) or

(b) an acrylic acid side chain (GW7604, GDC-0810, AZD9496) [65, 70, 74, 77-79, 111].

101

Additional ER antagonists with novel chemical structures have also been reported to exhibit SERD properties, but none has yet gained FDA approval and some have been discontinued due to adverse effects or for undisclosed reasons [65, 70, 71, 77-79, 81, 83, 88,

111, 149]. We have identified a novel, orally bioavailable SERD, G1T48, that contains both a steroidal backbone and an acrylate side chain. G1T48 binds ER with low nanomolar affinity, inhibits estrogen-mediated target gene expression and breast cancer cell growth, and importantly blocks the tumor promoting effects of ER in both naïve and endocrine therapy-resistant animal models of breast cancer.

A hallmark feature of fulvestrant differentiating it from compounds like tamoxifen is a that fulvestrant is a true antagonist with no agonist activity yet recorded regardless of tissue context. By contrast, tamoxifen is a SERM, demonstrating robust antagonist activity in the breast, but mimicking the agonist effect of estrogen in bone, the endometrium, and serum lipid profiles [153, 154]. This mechanistic difference between tamoxifen and fulvestrant can also be observed in breast cancer cells, where transcriptional profiling studies revealed that tamoxifen can regulate a subset of genes in a similar manner to estradiol. Our ER target gene regulation studies confirm the agonist activity of tamoxifen, with stimulation of SDK2, AGR2, and RAPGEL1 expression similar to the effect of estrogen treatment [151]. Compounds with SERD activity such as fulvestrant and AZD9496 did not increase these transcripts, consistent with a lack of agonist potential. The transcriptional

102

profile in breast cancer cells of G1T48 is most similar to fulvestrant and other SERDs.

Interestingly, our studies revealed that there were modest differences in the transcriptional profiles even among the pure antagonist class of compounds, suggesting that they might engender different receptor conformations. The impact of these differences in ER target gene activation remain to be explored but could suggest that cross-resistance between different classes of SERDs can be avoided. Recent studies have indicated that in addition to receptor degradation, ER mobility is differentially impacted by sub-classes of SERMS and SERDs, and that compounds impeding mobility are more efficacious antagonists [88]. The impact of G1T48 on ER mobility is not currently known; however, our studies establish that G1T48 has very low intrinsic ER agonist activity.

Acquired resistance to endocrine therapy is complex and multifactorial; however, mutations in the ESR1 gene that result in ligand-independent receptor activity have emerged as a potential mechanism to account for approximately 30-40% of resistant disease following AI treatment [48-53, 64, 66]. It is significant therefore that G1T48 was found to suppress transcriptional activity (efficacy but not potency) attributed to the two most prevalent endocrine-refractory ER mutants, ERY537S and ERD538G. Long-term estrogen deprivation leading to a state of estrogen hypersensitivity is another means to model aromatase inhibitor therapy resistance. We have developed a new model of resistance to estrogen deprivation without ERmut [86]. Using this model system, treatment with low-

103

dose G1T48 (5 mg) resulted in incomplete tumor growth inhibition, while high-dose

G1T48 (100 mg) as monotherapy resulted in tumor regression in the majority of animals, similar to fulvestrant, demonstrating the effectiveness of SERD therapy in this setting of resistant disease.

The combination of SERDs with CDK4/6 inhibitors has been evaluated clinically, most recently in the PALOMA-3 trial comparing the co-administration of the CDK4/6 inhibitor palbociclib (Ibrance®) with fulvestrant to fulvestrant alone. The results of this study demonstrated an overall survival benefit (median survival 34.9 months compared to 28 months) and a significant progression free survival rate (9.5 months vs 4.6 months) for the combination arm [44, 45]. These noteworthy improvements led to the 2016 FDA approval of palbociclib and fulvestrant combination therapy for ER-positive, HER-2- negative breast cancers progressing on other endocrine therapies. Further trials

(MONALESSA-3 (NCT02422615) and MONARCH-2 (NCT02107703) have also demonstrated the utility of administering other CDK4/6 inhibitors with fulvestrant to improve patient outcomes [44-47]. The increased efficacy observed for the combination of

G1T48 and lerociclib, as compared to monotherapy administration, in multiple in vivo breast cancer models sensitive or refractory to endocrine therapy treatment supports the potential utility of this regimen as an intervention in multiple stages of breast cancer

104

treatment. Collectively, these data indicate that G1T48 has the potential to be an efficacious oral antineoplastic agent in ER+ breast cancer.

105

5. Conclusions: Future Directions and Implications

5.1 Remaining mechanistic questions that will have implications for preclinical drug development

5.1.1 Investigating the role of active ERWT to “normalize” the activity of the ERmuts

In Chapter 2, we described the results of studies which demonstrated in several experimental models that ERWT normalizes the activity of coexpressed ERmuts. However, the mechanism by which this activity occurs was not determined. Our favored hypothesis is that when ERWT and ERmut are present, that the majority of the receptor (75%) exists in an ERWT/ERWT or ERWT/ERmut complex. When an antagonist interacts with an ERWT monomer it disengages the coactivator binding surface and since an active ER-dimer complex requires both monomers to present a coactivator binding surface the requirements for activation are not met. Therefore, the presence of antagonist occupied

ERWT normalizes the cellular response (potency) to ligands through dimerization (Figure

26).

106

CoA CoA WT WT Antiestrogen Response

CoA CoA WT Mut ???

CoA CoA Mut Mut Antiestrogen Resistance

Figure 26: Receptor dimerization is one possible mechanism by which activated ERWT normalizes cellular response to antiestrogens in ERmuts expressing cells.

To test this hypothesis, we have developed new reporter gene constructs and DNA binding mutants, in order to look at how the dimerization may impact receptor interactions. First, it has been previously described that there are only three amino acids within the ER DBD that are required to change its ERE specificity to recognize a GRE

(glucocorticoid response elements)[155]. Therefore, DBD mutants with these amino acids

(203, 204, and 207) were developed. Next, three additional reporter gene constructs were developed, 1) a full site GRE reporter, 2) a half site ERE followed by a half site GRE and

3) the half site GRE followed by a half site ERE. Figure 27 depicts the hypothesis of how

107

these new DBD mutants should heterodimerize with another ER on one of the half site reporters. Unfortunately, however, the receptor constructs intended to have altered DNA binding maintained activity on the reporter genes; this is might be attributable to cryptic sites present to the native structure of the reporter. Therefore, a future direction could involve taking a global approach in which these constructs are stably expressed in MCF7s and compared for their effect on global changes on gene expression in response to ER ligands.

CoA CoA ER Mut

hGRE hERE

Figure 27: Dimerization Hypothesis Experimental Design.

In this experimental model, the ERWT has been mutated at three sites within the DBD domain (pink) and should in theory now recognize a GRE. If this DBD mutant (pink) was transfected into cells with a half-site reporter and a LBD mutant (orange), the half-site reporter should read on the activity of the receptor dimer pair. The labels on these receptors correspond to their LBD mutant status. The reverse of having the DBD mutants introduced with the LBD mutants was also attempted.

5.1.2 Mechanistic explanation for the lack of an impact of ESR1 mutants on lasofoxifene potency.

I hypothesize that the SERM lasofoxifene maintains its potency against ERmuts in all contexts tested as the mutants do not compromise ligand binding affinity. It has been

108

previously reported that some compounds (fulvestrant, 4-hydroxytamoxifen, bazedoxifene) have decreased affinity for the exogenously expressed LBD [84, 156].

However, no studies have been published with lasofoxifene. Additionally, I wanted to be able to test the affinities of all compounds in a whole cell system that we could manipulate the relative coexpression levels of the different receptors. Therefore, a radioligand binding assay was employed in the doxycycline-inducible MCF7 cell lines. In this experiment, cells were treated with a fixed concentration of titrated E2 and treated with increasing concentrations of the test ligands. It was expected that compounds that have a shifted potency will also have a shifted affinity in the overexpressed scenario while lasofoxifene will maintain its affinity. Therefore, if the ERmut and ERWT all have similar affinities to lasofoxifene, while the other compounds have a decreased affinity for the mutant receptors, the affinity could explain why only some compounds lose potency for the exogenous mutant receptors.

Because radiolabeled versions of all SERMs and SERDs of interest are not available, this assay was set up to measure competition with E2 as a surrogate for the affinity with which these ligands bind to the receptor. Unfortunately, because of the reduced affinity of the ESR1 mutants for E2 as compared to the ERWT, this experiment was not possible under those conditions. The affinity of E2 for the mutant receptors was too different from the WT; therefore, it was impossible to saturate the receptors a requirement for a

109

quantitative competitor ligand binding assay. One possible way to approach this experiment would be to develop the radiolabeled SERMs and SERDs. Another approach would be to revert to the LBD receptor experiments that have been previously published and test lasofoxifene in these conditions.

5.1.3 ERmut pharmacology as a mediator of tumor cell/immune cell crosstalk and the promotion of metastasis

Three concepts have emerged which together provide a general mechanistic explanation for ER pharmacology. (1) It has been shown that the overall conformation of

ER is determined by the nature of the ligand with which it interacts, (2) Studies have demonstrated that receptor conformation influences the ability of ER to interact with its attendant cofactors, and (3) It is now understood that the biological activities of different ligand/receptor/cofactor interactions are not equivalent [17]. Therefore, it was interesting that the modest changes that were described in the cofactor recruitment profiling experiments did not track with changes in any of the biological contexts tested

(transcription or proliferation). To that end, these ERmuts have been described to uniquely regulate immune signaling pathways in metastatic breast cancer [61]. I hypothesize that the conformational differences in response to antagonists that were observed between the

ERWT and ERmuts have impacts on tumor cell intrinsic and extrinsic (paracrine effects of the tumor cell on the immune cell) promotion of metastasis.

110

Therefore, a future direction of this study would be to understand how these observations related to the neomorphic cofactor interactions of ERmuts may apply to the cancer cell/immune cell crosstalk and the promotion of metastasis. Specifically, if resources were not a factor, the most elegant experiment would be to develop and validate syngeneic tumor models harboring different ratios of the murine equivalents of ERWT and

ERmuts. Then these cells could be injected into immune competent animals, and then treated with different ER antagonists. At the end of the study, primary tumors (and any metastatic tumors) could be prepared for flow analysis to determine what immune cell populations are impacted by the differences in alleles and antagonist treatment. A much simpler experiment would be to take spent media from those same lines that had been treated with antagonists in vitro, and add this media to immune cell populations in vitro for functional assays. Finally, to simply look at the impact of ER antagonists on tumor cell intrinsic promotion of metastasis, the already available human breast cancer cell lines could be treated with antagonists over the course of an in vitro metastasis assay. It is expected that the results of these studies will reflect the changes seen in the cofactor profiling experiment in Chapter 2.

111

5.1.4 The role of growth factors in future SERD (or SERM) development

In the introduction of this thesis, I introduced the question of whether the development of SERD was actually necessary or if a high affinity antagonist that induces a particular “antagonist” conformation was sufficient. Within the introduction, compelling evidence was provided that at the level of ERWT itself, SERD activity is not necessary. However, these studies have not been undertaken in the context of the ERmuts.

In Chapter 2, using unbiased approaches, it was determined that the SERM lasofoxifene was the only existing ligand that was investigated that did not display partial resistance for the ERmuts. In Chapter 4, biased approaches were undertaken to characterize a novel

SERD that had properties of both basic and acidic SERDs; this SERD did display situational reduced potency against the ERmuts like many other compounds previously tested. Additionally, findings in the TamR tumor study were in agreement with previous literature that SERD and antagonist activity did not track with each other. In the space of

ERmut pharmacology, work is still needed to determine the “ideal” ligand to target these receptors. One remaining concern, would be the impact that SERMs have on growth- factor signaling. Specifically, activation of growth-factor signaling pathways is another mechanism of endocrine therapy resistance [34]. Compounds with SERD activity, including G1T48, effectively block growth factor-mediated proliferation, while compounds with SERM activity are generally less effective. Of importance, it has been

112

demonstrated in the literature that ERmuts affect the anti-proliferative response and resistance to SERMs (specifically tamoxifen) through increased cross-talk with growth- factors[157, 158]. Therefore, growth-factor mediated proliferation would have to be studied in cells harboring different ratios of ERWT and ERmuts and the impact of lasofoxifene and other compounds studied. This may provide crucial information in determining the mechanistic determinants that need to be considered when developing new antagonists to target ERmuts.

5.2 Clinical Implications

5.2.1 Investigating the single cell allelic frequency of ERmuts in patient tumor samples

It was suggested in Chapter 2 that the inability to assess ERWT/ERmut allelism at the single cell level was an obstacle for future therapeutic development to treat patients whose tumors harbor ERmuts. However, MissionBio has recently developed a precision genomics platform called Tapestri that can achieve clonal resolution at the single-cell level to detect mutant alleles in cancer. This technology utilizes PCR- based sorting of barcoded cell populations followed by RNA-seq to determine the oncogene mutation status in tumors

[159-162]. Due to the novelty and prematurity of this technology, it has currently only been published using small sample sizes (1-12 samples) [163-166]. However, as this

113

technology matures and develops, it would be a very useful asset to utilize to further our knowledge on cellular allelism of ERmuts.

A future study could be to analyze the single cell allelism and correlate that to disease response to fulvestrant therapy in patients harboring ERmuts. To this end, Dr. Sarah

Sammons, who has been a collaborator on the work in Chapter 2, has access to over 120 patient tumor samples that we could potentially utilize if we were able to gain IRB approval. This study could be followed up by looking at sequential biopsies and determining if the populations that have a higher allele frequency of ERmuts were enriched as fulvestrant therapy continues. If resources were available, it would also be interesting to look at fulvestrant versus lasfoxifene treated cohorts from the ELAINE trial. I would hypothesize that the patients that responded well to fulvestrant would have lower single- cell allele frequencies of ERmuts and that through the treatment duration populations that harbored higher allelic frequencies would emerge. This phenotype would not be present in lasofoxifene treated groups, as there would be a universal response to therapy.

5.2.2 Dissecting the differences between diagnosis of breast or gynecological cancers harboring the ERmuts

One of the most interesting findings in Chapter 3, as compared to Chapter 2, is the fact ERmuts appear as a mechanism of both de novo and acquired resistance in primary gynecological cancers. It would be interesting therefore to look into whether these tumors

114

may eventually metastasize, as it has been shown that these mutations are associated with a worse prognosis. This has already been suggested preclinically. After the work in

Chapter 3 was published, it was demonstrated that the ERD538G allele in endometrial cancer cells drives a distinct transcriptional program compared to the ERWT and promotes tumor cell migration but not proliferation, which supports the hypothesis that the ERmuts would behave similarly in gynecological cancers compared to breast cancer cells [167]. Therefore, it would be interesting to investigate whether or not, in either animal models or as retrospective study in patients, cells harboring these ERmuts are more likely to metastasize and if we can utilize the wealth of information available from breast cancer patients to better tackle gynecological cancers. However, it is also possible that the metastatic phenotype has not emerged simply because of early diagnosis of patients with these ERmuts without prior endocrine therapy.

To that end, two of eight patients that were highlighted in Chapter 3 actually presented with these ERmuts prior to receiving endocrine therapy (de novo resistance) to treat their cancer and were at relatively young ages (35 and 58) where menopause would likely not be a factor. Unfortunately, much retrospective analysis of patients past history was unavailable. However, it would be interesting in the future to look into what are the factors that govern the emergence of primary gynecological cancers harboring the ERmuts at diagnosis. Specifically, if certain types of hormonal birth control (for example

115

combination or progestin-only) or receiving endocrine therapy as a preventative measure against breast cancer development has an impact on the emergence of these mutations.

These studies would be informative for our understanding of these ERmuts and the treatment and prevention of cancer harboring these mutations.

5.2.3 Impact of this thesis work on current Clinical Trials

There are currently several compounds in clinical development to target ER in endocrine resistant breast cancer, two of which are the direct result of the studies presented in this thesis. A summary of these trials is presented in Table 15.

Table 15: ER Targeted-Compounds Currently in Clinical Trials for Endocrine Therapy Resistant Breast Cancer

Compound Classification ClincialTrials.gov Identifier Lasofoxifene SERM NCT03781063 G1T48 SERD NCT03455270 RAD1901 SERD NCT02338349 LSZ102 SERD NCT202734615 AZD9833 SERD NCT03616586 GDC-9545 SERD NCT03332797 SAR429859 SERD NCT03284957 Zn-C5 SERD NCT03560531 ARV-471 ER- Targeted PROTAC (Protein Degrader) NCT04072952 H3B-6545 SERCA (Selective ER covalent antagonist) NCT04288089

116

Appendix: Materials and Methods

A.1 Chemicals and Ligands

Fulvestrant (1047) and Raloxifene (2280), were purchased from Tocris. Estradiol

(E8875) and 4-hydroxytamoxifen (H7904) were purchased from Sigma. Bazedoxifene

(S2128) was purchased from Selleckchem. Lasofoxifene (HYA0038K), RAD1901

(HY19822A), GDC-0810 (HY12864) and AZD9496 (HY12870) were purchased from

MedChem Express. GW5638, GW7604, and RU 58,668 were provided by Donald

McDonnell (Duke University). G1T48 was provided by G1 Therapeutics, Inc., as analytical grade compound.

A.2 Generation of ERmut expression constructs

Exsite mutagenesis was performed using the corresponding primers presented below (Table 7) on a pENTR2B ER WT construct using Pfu ultra taq polymerase and primers were PNK phosporylated. Following PCR amplification, products were digested with DpnI at 37°C for 1hr, followed by overnight ligation at 16°C. Ligated products were transformed into DH5α bacterial cells and grown on kanamycin resistant plates. The pENTR clones were verified by sequencing and then swapped into destination vectors

(pcDNA-DEST, VP16 or pLenti CMV TRE3G puro) using the Gateway system

(Invitrogen) for expression analysis.

117

Table 16: Primer sequences for generation of ERmut expression constructs

Primer Sequence ER Y537N For AATGACCTGCTGCTGGAGATG ER Y537N Rev GAGGGGCACCACGTTCTTGCA ER Y537S For GACCTGCTGCTGGAGATGCTG ER Y537S Rev GCTGAGGGGCACCACGTTCTT ER Y537C For TGTGACCTGCTGCTGGAGATG ER Y537C Rev GCTGAGGGGCACCACGTTCTT ER D538G For GGTCTGCTGCTGGAGATGCTG ER D538G Rev ATAGAGGGGCACCACGTTCTT

A.3 Cell Culture

SKBR3, HepG2 and parental MCF7 cells were purchased from ATCC. T47D cells were a gift from Dr. Steffi Oesterreich, University of Pittsburgh [48]. MCF7B cells were a gift from Dr. Myles Brown, Dana- Farber Cancer Institute[61]. Caov2 were a gift from Dr.

Susan Murphy, Duke University. MCF7 parental cells were maintained in DMEM/F12,

MCF7B cells were maintained in DMEM, and HepG2 were maintained in Basal Eagles

Media. All other cells were maintained in RPMI.

MCF7I cell lines were generated from parental MCF7 cells as follows. MCF7 cells were first infected with pLenti CMV rtTA3G blast (Addgene) followed by ER-expressing constructs in the backbone of pLenti CMV TRE3G puro (Addgene) using the following protocol. 293Tcells were transfected with pVSVG and psPAX2 (all gifts from Kris Wood,

Duke University School of Medicine) and the viral construct of interest using Fugene 6

118

(Promega E2691) per the manufacturer’s protocol. Media was removed and replaced with

DMEM containing 30% FBS 18 hours after transfection. Cells were allowed to produce virus for 2 days and then media was collected. Viral media was filtered and 4 µg/mL of polybrene was added. Viral containing media was added to MCF7 cells that had been split

1:3 24 hours prior. After 48 hours, antibiotic selection was added and cells were selected.

The cells used in these experiments are maintained in DMEM/F12 media supplemented with blastocidin (10µg/mL) and puromycin (1µg/mL) to maintain the constructs.

A.4 Luciferase Reporter Assays

Cells were co-transfected with an ERE-TATA luciferase reporter gene (see Table 8 for details) and expression constructs for either wild-type or mutant receptors using

Fugene transfection reagent (Promega). pCMV-β-gal was used as a control for transfection efficiency. Ligands (dose titration of antagonists in the presence of E2, see table for details) were added five hours post transfection. Cells were lysed 24 hours later and the luciferase and β-gal assays were performed as described previously[168, 169].

Table 17: Experimental Conditions for Luciferase Reporter Assays

Cell line Reporter E2 competitor dose SKBR3 3X-ERE-TATA-LUC 1 nM MCF7 7X-ERE-TATA-LUC 0.1 nM Caov2 7X-ERE-TATA-LUC 10 nM

119

A.5 siRNA Transfection Assay

For siRNA transfection experiments, cells were plated over aliquoted siRNAs targeting the 3’ UTR of ER (to knockdown endogenous ER) using Lipofectamine

RNAiMax (Thermo-Fisher Scientific) per the manufacturer protocol. The siRNAs used in this study were custom Stealth siRNA from ThermoFisher. Sequences are provided below

(Table 18).

Table 18: siRNA sequences

siRNA Sequence 3’ UTR Forward GUA GCC ACA ACA AUC CUG CAC AAG U 3’ UTR Reverse ACU UGU GCA GGA UUG UUG UGG CUA C Control Forward ACU ACG UUA GGU UGU GGU GCG UUA C Control Revers3:e GUA ACG CAC CAC AAC CUA ACG UAG U

A.6 Cofactor Profiling

A.6.1 Transfection experiment

HepG2 cells were seeded in 96-well plates and transfected with VP16-ER,

5XGal4Luc3, Gal4DBD-peptide fusion constructs (pM-peptides), and pCMV β-gal using

Lipofectin as previously described [75, 102, 105, 106]. Detailed description of peptide sequences and generation can be found in Section 6.6.2. Cells were then treated with saturating concentrations of ligands (10 µM for ER antagonists) for 48 hours. Luciferase assays were performed as described above. The data were standardized to avoid bias due to signal strength and clustered with the Ward hierarchical clustering method using JMP

120

Pro 13 (SAS). The hierarchical cluster dendrogram was ordered by the first principal component.

121

A.6.2 Cofactor Profiling Peptide Acquisition

Table 19: Peptide Sources and Sequences

Probe Cofactor/ Sequence Citation peptide αII Peptide SSLTSRDFGSWYASR [102] α/β V Peptide SSPGSREWFKDMLSR [102] α/β I Peptide SSNHQSSRLIELLSR [102] 7β16 Peptide HFLINQHLYKLLQDTDIVV [75] bN2 Peptide EYHEKRWLEGHIHHRIKSLLENS [105] bI2 Peptide EMEWMKALRQHISGELRRNYTEE [105] bT1 Peptide ELFDAFQLRQLILRGLQDDIPYH [105] LX23 Peptide RIHGYSPMLRALLLEEEAPK [106] EIP104 Peptide SSKSEFDSQLRHIIWTQLTDTPFEFSR This study EIP113 Peptide SSGPPWTFQLRNIIYQGLTHEAQPFSR This study EIP397 Peptide SRHFLINQHLYKLLQDTDIVVSR This study EIP420 Peptide SRHWLMEGHLEQLLHDQPLTLSR This study EIP484 Peptide SSTTPHSKELFAIIGENIVLRPHYVSR This study EIP786 Peptide SSGPPWTFQLRNIIYQGLTHEAQPFSR This study EIP793 Peptide SSKSEFDSQLRHIIWTQLTDTPFEFSR This study ACTR 400-1000 Cofactor amino acids 400-1000 NM_18165 9.2 ACTR 621-821 Cofactor amino acids 621-821 NM_18165 9.2 T7-ASC2 Cofactor amino acids 799-927 BAA11498. 2 T7- PGC1α Cofactor amino acids 19-193 NM_00133 0751.1 GRIP NR Cofactor amino acids 322-1121 NM_02115 0.3 SRC1 NR Box 1 Cofactor amino acids 621-765 NM_00374 3.4

122

Human ER was cloned into a baculovirus shuttle vector pDW464 to make an in- frame fusion of the ER with the biotin acceptor peptide (Science Reagents, El Cajon, CA).

Recombinant ER-baculovirus was generated using Bac-To-Bac Baculovirus Expression

System (Life Technologies Bethesda Research Laboratories) following the protocol provided by the manufacturer. ER-recombinant protein was produced in Sf9 insect cells following infection with recombinant baculovirus particles. A soluble extract of infected

Sf9 cells was used to affinity purify biotinylated ER-fusion protein with monomeric avidin resin (Promega Corp., Madison, WI). To select for ER interacting peptides (EIP), a modified protocol from that previously described was used [170]. Briefly, 2 pmol of baculovirus-expressed ER protein were diluted in 100 µL of PBST (137 mmol/L NaCl, 2.7 mmol/L KCl, 4.3 mmol/L Na2HPO4 (pH 7.3), 0.1% Tween 20) and applied to a single well in a DNA-coated cell culture plate. To prepare the DNA-coated plate, a 96-well Costar plate was first coated with 20 µg of Neutravidin Biotin-Binding Protein (Pierce

Biotechnology, Rockford, IL) overnight at 4°C, and 2 pmol of double-stranded 5′- biotinylated oligonucleotides (diluted in PBST) containing an ERE were added for 1 hour at room temperature. The protein-coated plate was incubated overnight at 4°C. The wells were then blocked with 150 µL of 2% milk in NaHCO3(pH 8.5) for an additional 1 hour at room temperature and washed five times with PBST to remove unbound proteins. Then,

25 µL of the phage peptide library (with 1010 phage particles) diluted in 100 µL of MPBS

123

(2% nonfat dry milk in PBS) was added to the wells, and the plate was sealed and incubated for 3 hours at room temperature. Construction of the CoRNR M13-phage library was described previously [170]. Nonbinding phage were removed by washing the wells five times with 300 µL of PBST. The bound phage were eluted with 100 µL of 0.1 mol/L HCl for 10 minutes. The eluent was neutralized with 50 µL of 1 mol/L Tris-HCl

(pH. 7.4). Phage eluted from the target were amplified in Escherichia coli DH5α F′ cells for

5 hours at 37°C, and the supernatant containing amplified phage was collected for use in subsequent rounds of panning; a total of five rounds of panning were done. A PCR reaction was then done using 1 µL of the unamplified eluent phage from both rounds 3 and 4, with mBax reverse and forward primers to amplify peptide inserts. PCR products were purified and digested with XbaI and XhoI before their cloning in a modified pM

(Gal4-DBD) vector. The peptide sequences were deduced by DNA sequencing.

A.7 Proliferation

MCF7 cells were plated in DMEM/F12 supplemented with 8% charcoal dextran treated FBS in 96-well plates (5K cells/well) for 48 hours. Cells were treated in DMEM/F12 with 2% FBS plus test compound (dose response; 10-12 to 10-6) for 6 days. Media was aspirated and cells were frozen at -80°C overnight. 100 µL H20 was added to wells and plates were incubated at 37°C for 1 hour and then frozen at -80°C overnight. 100 µL of

124

aqueous Hoechst 33258 (DNA dye; stock concentration 1mg/mL) was diluted 1:1000 in

TNE buffer (10mM Tris, 2 mM NaCl, and 1 mM EDTA) and fluorescence was measured

(360 nm excitation/460 nm emission).

A.8 In-Cell Westerns

MCF7 cells were plated in DMEM/F12 supplemented with 8% charcoal dextran treated FBS in 96-well clear bottom black plates (25K cells/well). After 48 hour incubation, cells were treated with hormone (dose response; 10-12 to 10-6 M) for 24 hours. Cells were fixed with formaldehyde (3.7%), permeabilized using PBS 0.1% TRITON X-100, and blocked in In-Cell Western blocking buffer (3% goat serum, 1% BSA, 0.1% cold fish skin gelatin, and 0.1% Triton-X 100 in PBS). Cells were then incubated with anti-ER antibody

(SP1, Fisher Scientific 1:500) in diluted In-Cell Western blocking buffer (1:3 in PBS) overnight. Cells were washed with PBS 0.1% Tween, and stained with secondary antibody

(Biotium CF770 goat anti-rabbit, 1:2000) in diluted In-Cell Western blocking buffer (1:3 in

PBS). ER protein expression was assessed using the LI-COR Odyssey imaging system.

DRAQ5 (DNA stain, 1:10,000, Thermo Scientific) was used to normalize ER protein expression. Data is reported as percent ER remaining after drug treatment by comparing to DMSO control.

125

A.9 Immunoblots

Cell pellets were resuspended in RIPA buffer (50 mM Tris, pH 8, 150 mM NaCl,

1% NP-40, 0.5% deoxycholate, 0.02% SDS, 1mM EDTA and protease inhibitors). 30 µg of protein was resolved by SDS-PAGE prior to transfer to nitrocellulose membrane and probed with anti-ER (D-12, Santa Cruz Biotechnology 1:1000) and goat anti-mouse HRP conjugated antibody (1:5000). Blots were visualized using ECL and film.

A.10 Identification of ERmuts in Gynecological Cancers

A.10.1 Comprehensive Genomic Profiling:

Comprehensive next generation sequencing-based genomic profiling (CGP) was performed for 9645 cases of gynecologic malignancies involving the ovary, fallopian tube, uterus, cervix, placenta, vulva, or vagina during the course of routine clinical care. The pathologic diagnosis of each case was confirmed on routine hematoxylin and eosin (H&E) stained slides and all samples forwarded for DNA extraction contained a minimum of

20% tumor nuclear area.

The sequencing methods used for CGP have been described in detail elsewhere

[171, 172]. Sample processing and sequencing analysis was performed in a Clinical

Laboratory Improvement Amendments (CLIA)- and College of American Pathologists

(CAP)-accredited laboratory (Foundation Medicine). In brief, samples undergo

126

pathologist review to ensure sufficient tumor material (minimum 20% tumor nuclei) and resolve any conflicts with provided histological description. From each sample, ≥50ng

DNA was extracted from 40 microns of tumor samples provided as formalin-fixed, paraffin-embedded (FFPE) tissue blocks. The samples were assayed using adaptor- ligation and hybrid capture (Agilent SureSelect custom kit) next-generation sequencing

(FoundationOne®) for all coding exons from 287 (version 1) or 315 (version 2) cancer related genes, plus select introns from 19 (version 1) or 28 (version 2) genes frequently rearranged in cancer. Sequencing of captured libraries was performed using Illumina

HiSeq technology to a mean exon coverage depth of >500X, and resultant sequences were analyzed using both an algorithmic pipeline and manual curation for base substitutions, small insertions or deletions, copy number alterations (focal amplifications and homozygous deletions), and select gene fusions, as previously described [171, 172].

Clinically relevant genomic alterations (CRGA) were defined as alterations that are targetable by anti-cancer drugs currently available on the market or in registered clinical trials. Germline variants documented in the dbSNP database (dbSNP142; http://www.ncbi.nlm.nih.gov/SNP/), with two or more counts in the ExAC database

(http://exac.broadinstitute.org/), or recurrent variants of unknown significance that were predicted by an internally developed algorithm to be germline were removed, with the exception of known driver germline events. Known confirmed somatic alterations

127

deposited in the Catalog of Somatic Mutations in Cancer (COSMIC v62) were highlighted as biologically significant. All inactivating events (i.e. truncating mutations and deletions) in known tumor suppressor genes were also called as significant. The bioinformatics processes used in this study included Bayesian algorithms to detect base substitutions, local assembly algorithms to detect short insertions and deletions, a comparison with process-matched normal control samples to detect gene copy number alterations and an analysis of chimeric read pairs to identify gene fusions. To maximize mutation-detection accuracy (sensitivity and specificity) in impure clinical specimens, the test was previously optimized and validated to detect base substitutions at a ≥5% mutant allele frequency

(MAF), indels with a ≥10% MAF with ≥99% accuracy, and fusions occurring within baited introns/exons with >99% sensitivity.

A.10.2 Identification of ESR1 mutations in public databases

Public datasets of endometrial and ovarian cancers were queried for ESR1 mutations. The cBioPortal was used to access genomic data from the cervical TCGA, endometrial TCGA, ovarian TCGA, uterine carcinosarcoma, and AACR Project GENIE datasets [123-128]. COSMIC (cancer.sanger.ac.uk) was used to identify an additional sample with an ESR1 mutation from an ovarian cohort [129].

128

A.10.3 Clinical Evaluation of Gynecologic Malignancies with ERmuts

Three cases were identified during review of medical records under an

Institutional Review Board protocol (with waiver of informed consent and a HIPAA waver of authorization) at the Duke University Medical Center aimed at evaluating the utility of evaluation of ER in the course of routine clinical care of patients with low-grade ovarian or endometrial cancers treated at Duke University between January 1, 2000 and

June 30, 2016. The other three cases were identified through the Clearity Foundation under an approved protocol including a waiver of informed consent and a HIPAA waiver of authorization, obtained from the Western Institutional Review Board. Demographic, tumor characteristics, treatment and response data were collected.

A.11 qPCR and RNA profiling

MCF7 cells were plated in DMEM/F12 supplemented with 8% charcoal dextran treated FBS for 48 hours. Cells were then treated for 24 hours with ligand and RNA was isolated using the AurumTM total RNA isolation kit (Bio-Rad, Hercules, CA). After cDNA synthesis (iScript kit, Bio-Rad) real-time PCR was performed using the Bio-Rad CFX384 real-time system. GAPDH mRNA expression was used to normalize all real-time data using the 2-ΔΔCT method [173]. For ER gene signature, the data were first normalized to the vehicle control within each gene. To avoid signal strength bias, the data were

129

standardized using the following equation; Ζ=Χ-µ/σ where Χ is the normalized signal, µ is the average signal for all conditions within a gene and σ is the standard deviation. The data are then clustered with the Ward hierarchical clustering method using JMP (SAS).

The hierarchical clustering dendogram is ordered by the first principle component.

A.12 Radioactive Binding Assay

MCF7 cells were plated in DMEM/F12 supplemented with 8% charcoal dextran treated FBS in 24 well plates (500K cells/well). After 48 hour incubation, cells were treated with 0.1 nM 3H-17β-E2 (PerkinElmer Catalog number NET317, lot number 2526124) and competitor ligand (dose response; 10-11 to 10-6 M) for 2 hours. Cells were washed 2X with

DMEM/F12 supplemented with 8% charcoal dextran treated FBS and 1X with 1X

Phosphate Buffered Saline (PBS). Cells were lysed with 200 µL of Lysis buffer (2%S SDS,

10% glycerol, 10mM Tris-Cl pH 6.8). Cell lysates were diluted with 300 µL of 10mM of

Tris- HCl pH 8.0. Scintillation vials were prepared with 3 mL of CytoScint. 300 µL of lysate was added to respective scintialltion vials. Samples were read on a Beckman LS 6000SC

Scintillation counter for 1 minute per sample.

130

A.13 Chromatin Immunoprecipitation (ChIP):

MCF7 cells were grown to 90% confluence in DMEM/F12 supplemented with 8% charcoal dextran treated FBS for 3 days, at which time cells were treated with ligand for

90 minutes and subjected to ChIP analysis. Each plate of cells was cross-linked with 1% formaldehyde PBS solution for a maximum 10 minutes and quenched with 125mM glycine solution containing 5mg/ml bovine serum albumin (BSA) for 5 minutes. Cells were then rinsed once and harvested with ice cold PBS, pelleted at 8000 rpm for 30 seconds at room temperature and snap frozen for storage at -80ºC. All solutions were supplemented with 10mM sodium butyrate and protease inhibitors. Cell pellets were thawed on ice, and then resuspended in 10mL Lysis Buffer 1 (50mM HEPES pH 7.5, 140 nM NaCl, 1 mM

EDTA 10% Glycerol, 5% NP-40, 2.5% Triton X-100, protease inhibitor cocktail). These solutions were incubated for 10 minutes rocking at 4°C. Samples were then spun down at

2000 rpm for 4 minutes, supernatant removed, resuspended in 10mL Lysis Buffer 2 (10mM

Tris pH 8.0, 200 nM NaCl, 1 mM EDTA pH 8.0, 0.5 mM EGTA pH 8.0, protease inhibitor cocktail), and then incubated for 5 minutes while rocking at 4°C. Samples were then spun again at 2000 rpm for 4 minutes, supernatant removed, and resuspended in 1mL Lysis

Buffer 3 (0.1% NaDeoxycholate, 0.5% N-laurylsarcosine, 1 mM EDTA, 0.5 mM EGTA, 10 mM Tris pH 8.0, 100 mM NaCl, protease inhibitor cocktail). Cell lysates then were sonicated using the Covaris S220 instrument according to the manufacturers’ instructions.

131

Triton X-100 was then added to the sheared chromatin samples to make a final concentration of 1%, and samples then centrifuged at 13000 rpm for 10 minutes at 4°C.

Supernatant was then transferred to a new tube and sheared chromatin was diluted using

Dilution Buffer (20 mM Tris [pH 8.0], 150 mM NaCl, 2 mM EDTA, 1% Triton X-100). ER

ChIP was performed by incubating sheared, diluted chromatin with 5 µg of D12 anti-ER antibody. Antibodies were allowed to bind overnight at 4°C while rotating and then captured on protein A/G magnetic beads (Pierce, Cat#: 88802) which had been previously washed three times with 5 mg/mL BSA. After 45 min of incubation with the beads, the immunoprecipitates were washed twice with Wash Buffer A (50 mM HEPES pH 7.8, 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% Deoxycholate, 0.1% SDS, protease inhibitor cocktail), twice with Buffer B (20 mM Tris pH 8.0, 1 mM EDTA, 0.5% NP40, 0.5% Na

Deoxycholate, 0.25 M LiCl), and twice with TE Buffer (20 mM Tris pH 8.0, 2 mM EDTA).

Following washes, precipitates were eluted in Elution Buffer (50 mM Tris pH 8.0, 1 mM

EDTA, 1% SDS). Crosslink reversal was performed by addition of 21 µL 5 M NaCl to each sample and incubated at 65°C over-night. 4 µL 0.5 M EDTA and 1 µL 20 mg/mL proteinase K was then added to each sample and incubated at 42° C for 1 hour. DNA was then purified using the Qiagen PCR purification kit (Cat#: 28104) according to manufacturer’s instructions. ER recruitment was then assessed via qPCR.

132

A.14 Animal Studies

All procedures were approved by the Institutional Animal Care and Use

Committee (IACUC) of Duke University or South Texas Accelerated Research

Therapeutics (START, San Antonio, Texas) prior to initiating the experiment.

A.14.1 MCF7 Naïve Tumor Studies

South Texas Accelerated Research Therapeutics (START, San Antonio, Texas) evaluated antitumor activity of G1 Therapeutics test (G1T48 and lerociclib) agents in a Cell-Based

Xenograft (CBX) model, MCF7, representing human ER positive breast cancer, in immune deficient mice. Female athymic nude mice (Crl:NU(NCr)-Foxn1nu) at 6-12 weeks of age were implanted subcutaneously with cultured MCF7 cells. Estrogen was supplemented via the drinking water to the animals. The study was initiated at a mean tumor volume of approximately 150-250 mm3. G1T48 and lerociclib were formulated in 50 mM citrate buffer, pH 4.3 and dosed at 10 mL/kg, PO/ qd x 28. Data collection endpoint for this study was completed on day 62

A.14.2 TamR Tumor Studies

All procedures were approved by the Duke University Institutional Animal Care and Use

Committee (IACUC) prior to initiating the experiment. 120 female nu/nu mice (~6 weeks

133

of age) were ovariectomized under anesthesia (isoflurane) and in the same procedure implanted sc (scapular region) with tamoxifen (Tam) treatment pellets (5 mg/60 days, ~3.3 mg/kg/d continuous release, Innovative Research of America) 24-48 hours prior to having an ~8mm3 section of tamoxifen-resistant (TamR) tumor tissue engrafted orthotopically

(right axial mammary fat pad) under anesthesia. Tumors were measured 3X weekly, concurrent with weight and behavior monitoring, until tumors reached ~0.1-0.15 cm3 volume (l x w2 x 0.5). Mice were then randomized (n = 8-10) to treatment with: Vehicle, fulvestrant (200 mg/kg), (palbociclib - 100 mg/kg), G1T48 (30 or 100 mg/kg), or the combination of G1T48 + lerocicilib. Vehicle used in this study was 9% PEG 400/0.5%

PVP/0.5% Tween 80/ 0.05% CMC for orally gavaged compounds. Fulvestrant was formulated in 5% DMSO/95% NF grade corn oil. Treatments were administered weekly

(fulvestrant) or daily as indicated for 10 weeks, and tumor measurement and weight monitoring continued as above throughout that time. 2-3 hours after the final treatment, animals were euthanized by CO2 exposure, followed immediately by cardiac puncture for blood collection and secondary method decapitation. Plasma and tumor tissues were cryopreserved for future analysis. Frozen tissues were pulverized prior to protein extraction in RIPA buffer (50 mM Tris, pH 8, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate,

0.02% SDS, 1 mM EDTA). 25 µg of cleared extracts were resolved by SDS-PAGE prior to transfer to PVDF membrane and immunoblot analysis by standard methods. Bands

134

detected were quantitated using ImageJ per standard methods

(http://lukemiller.org/index.php/2010/11/analyzing-gels-and-western-blots-with-image- j/).

A.14.3 LTED Tumor Studies

All procedures were approved by the Duke University Institutional Animal Care and Use

Committee (IACUC) prior to initiating the experiment. 120 female nu/nu mice (~6 weeks of age) were ovariectomized under anesthesia (isoflurane). 24-48 hours later, a ~8mm3 section of long-term estrogen deprived (LTED) tumor tissue engrafted orthotopically

(right axial mammary fat pad) under anesthesia. Tumors were measured 3X weekly, concurrent with weight and behavior monitoring, until tumors reached ~0.1-0.15 cm3 volume (l x w2 x 0.5). Mice were then randomized (n = 8-10) to treatment with: Vehicle (10 mM citrate buffer + oral vehicle 10:5:85 (below)), fulvestrant (25 mg/kg fulvestrant), palbociclib (100 mg/kg) G1T48 (5 or 100 mg/kg), lerociclib (50 or 100 mg/kg), or combination treatments. Treatments were administered bi-weekly (fulvestrant) or daily as indicated for up to 8 weeks, and tumor measurement and weight monitoring continued as above throughout that time. 2-3 hours after the final treatment, animals were euthanized by CO2 exposure, followed immediately by cardiac puncture for blood collection and secondary method decapitation. Plasma and tumor tissues were

135

cryopreserved for future analysis.

A.14.4 PDX Tumor Study

South Texas Accelerated Research Therapeutics (START) evaluated antitumor activity of

G1 Therapeutics test agents, G1T38 and G1T48, in a START Patient-Derived Xenograft

(START-PDX) model, designated ST2177, representing human ESR1 mutant, Y537S, ER+ breast cancer[83]. Female athymic nude mice (Crl:NU(NCr)-Foxn1nu) at 6-12 weeks of age were implanted subcutaneously with cultured MCF7 cells. The study was initiated at a mean tumor volume of approximately 150-250 mm3. G1T48 and lerociclib were formulated in 50 mM citrate buffer, pH 4.3 and dosed at 10 mL/kg, PO/ qd x 28.

Fulvestrant was administered through subcutaneous injection. Data collection endpoint for this study was completed on day 33. In order to account for tumor growth outliers, the mouse having the largest final tumor growth measurement in each group was excluded from the tumor growth plots. All animals were included in the survival curve analysis.

136

A.15 Statistics

A.15.1 Dose response curve statistics

Two-way ANOVA was utilized, comparing the logIC50 of all three independent experiments, to determine if there were significant differences between the WT and mutant receptors. Significant differences (p-value < 0.05) were appropriately noted.

A.15.2 Animal Statistics

Tumor growth data were subjected to exponential growth curve analysis constrained to share an initial value, and to two-way ANOVA analysis followed by

Bonferroni multiple comparison test. Significant difference as compared to the vehicle treated control (p<0.05) was detected for multiple groups at several time points (indicated on graphs). Groups showed equivalent variance (10-15% with normal distribution) throughout all time points, justifying the statistical analyses that were selected. % change in tumor volume was calculated per the following equation: % change = (final tumor volume/initial tumor volume) – 1. Comparison of % change in tumor volume between groups was done using a one-way ANOVA followed by Holm-Sidak multiple comparison test.

137

References

1. Shoham, Z. and M. Schachter, Estrogen biosynthesis--regulation, action, remote effects, and value of monitoring in ovarian stimulation cycles. Fertil Steril, 1996. 65(4): p. 687-701.

2. Subramanian, A., M. Salhab, and K. Mokbel, Oestrogen producing enzymes and mammary carcinogenesis: a review. Breast Cancer Res Treat, 2008. 111(2): p. 191-202.

3. Judd, H.L., W.E. Lucas, and S.S. Yen, Serum 17 beta-estradiol and estrone levels in postmenopausal women with and without endometrial cancer. J Clin Endocrinol Metab, 1976. 43(2): p. 272-8.

4. Mangelsdorf, D.J., et al., The nuclear receptor superfamily: the second decade. Cell, 1995. 83(6): p. 835-9.

5. Hall, J.M., J.F. Couse, and K.S. Korach, The multifaceted mechanisms of estradiol and estrogen receptor signaling. J Biol Chem, 2001. 276(40): p. 36869-72.

6. Kuiper, G.G., et al., Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology, 1997. 138(3): p. 863- 70.

7. Tzukerman, M.T., et al., Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol, 1994. 8(1): p. 21-30.

8. Palmieri, C., et al., Estrogen receptor beta in breast cancer. Endocr Relat Cancer, 2002. 9(1): p. 1-13.

9. Couse, J.F., et al., Tissue distribution and quantitative analysis of estrogen receptor- alpha (ERalpha) and estrogen receptor-beta (ERbeta) messenger ribonucleic acid in the wild-type and ERalpha-knockout mouse. Endocrinology, 1997. 138(11): p. 4613-21.

138

10. Smith, C.L., O.M. Conneely, and B.W. O'Malley, Modulation of the ligand- independent activation of the human estrogen receptor by hormone and antihormone. Proc Natl Acad Sci U S A, 1993. 90(13): p. 6120-4.

11. Beekman, J.M., et al., Transcriptional activation by the estrogen receptor requires a conformational change in the ligand binding domain. Mol Endocrinol, 1993. 7(10): p. 1266-74.

12. Hall, J.M. and D.P. McDonnell, Coregulators in nuclear estrogen receptor action: from concept to therapeutic targeting. Mol Interv, 2005. 5(6): p. 343-57.

13. Varlakhanova, N., et al., Estrogen receptors recruit SMRT and N-CoR corepressors through newly recognized contacts between the corepressor N terminus and the receptor DNA binding domain. Mol Cell Biol, 2010. 30(6): p. 1434-45.

14. Smith, C.L., Z. Nawaz, and B.W. O'Malley, Coactivator and corepressor regulation of the agonist/antagonist activity of the mixed antiestrogen, 4-hydroxytamoxifen. Mol Endocrinol, 1997. 11(6): p. 657-66.

15. Shiau, A.K., et al., The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell, 1998. 95(7): p. 927-37.

16. Hu, X. and M.A. Lazar, The CoRNR motif controls the recruitment of corepressors by nuclear hormone receptors. Nature, 1999. 402(6757): p. 93-6.

17. McDonnell, D.P., The molecular pharmacology of estrogen receptor modulators: implications for the treatment of breast cancer. Clin Cancer Res, 2005. 11(2 Pt 2): p. 871s-7s.

18. Cancer Facts and Figures 2018, A.C. Society, Editor. 2019

19. Perou, C.M., et al., Molecular portraits of human breast tumours. Nature, 2000. 406(6797): p. 747-52.

139

20. Sørlie, T., et al., Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A, 2001. 98(19): p. 10869-74.

21. Ignatiadis, M. and C. Sotiriou, Luminal breast cancer: from biology to treatment. Nat Rev Clin Oncol, 2013. 10(9): p. 494-506.

22. Lisio, M.A., et al., High-Grade Serous Ovarian Cancer: Basic Sciences, Clinical and Therapeutic Standpoints. Int J Mol Sci, 2019. 20(4).

23. Rydén, L., et al., Aromatase inhibitors alone or sequentially combined with tamoxifen in postmenopausal early breast cancer compared with tamoxifen or placebo - Meta-analyses on efficacy and adverse events based on randomized clinical trials. Breast, 2016. 26: p. 106-14.

24. McDonnell, D.P., S.E. Wardell, and J.D. Norris, Oral Selective Estrogen Receptor Downregulators (SERDs), a Breakthrough Endocrine Therapy for Breast Cancer. J Med Chem, 2015. 58(12): p. 4883-7.

25. Nathan, M.R. and P. Schmid, A Review of Fulvestrant in Breast Cancer. Oncol Ther, 2017. 5(1): p. 17-29.

26. Kim, H.A., et al., Adding Ovarian Suppression to Tamoxifen for Premenopausal Breast Cancer: A Randomized Phase III Trial. J Clin Oncol, 2019: p. JCO1900126.

27. Maximov, P.Y., T.M. Lee, and V.C. Jordan, The discovery and development of selective estrogen receptor modulators (SERMs) for clinical practice. Curr Clin Pharmacol, 2013. 8(2): p. 135-55.

28. McDonnell, D.P., Definition of the molecular mechanism of action of tissue-selective oestrogen-receptor modulators. Biochem Soc Trans, 1998. 26(1): p. 54-60.

140

29. Carlson, M.J., K.W. Thiel, and K.K. Leslie, Past, present, and future of hormonal therapy in recurrent endometrial cancer. Int J Womens Health, 2014. 6: p. 429-35.

30. Simpkins, F., A. Garcia-Soto, and J. Slingerland, New insights on the role of hormonal therapy in ovarian cancer. Steroids, 2013. 78(6): p. 530-7.

31. Robertson, J.F., et al., Pharmacokinetic profile of intramuscular fulvestrant in advanced breast cancer. Clin Pharmacokinet, 2004. 43(8): p. 529-38.

32. Howell, A. and F. Sapunar, Fulvestrant revisited: efficacy and safety of the 500-mg dose. Clin Breast Cancer, 2011. 11(4): p. 204-10.

33. Musgrove, E.A. and R.L. Sutherland, Biological determinants of endocrine resistance in breast cancer. Nat Rev Cancer, 2009. 9(9): p. 631-43.

34. Tryfonidis, K., et al., Endocrine treatment in breast cancer: Cure, resistance and beyond. Cancer Treat Rev, 2016. 50: p. 68-81.

35. Ma, C.X., et al., Mechanisms of aromatase inhibitor resistance. Nat Rev Cancer, 2015. 15(5): p. 261-75.

36. Cocce, K.J., et al., The Lineage Determining Factor GRHL2 Collaborates with FOXA1 to Establish a Targetable Pathway in Endocrine Therapy-Resistant Breast Cancer. Cell Rep, 2019. 29(4): p. 889-903.e10.

37. Chia, S., et al., Double-blind, randomized placebo controlled trial of fulvestrant compared with exemestane after prior nonsteroidal aromatase inhibitor therapy in postmenopausal women with hormone receptor-positive, advanced breast cancer: results from EFECT. J Clin Oncol, 2008. 26(10): p. 1664-70.

38. Mauriac, L., G. Romieu, and J. Bines, Activity of fulvestrant versus exemestane in advanced breast cancer patients with or without visceral metastases: data from the EFECT trial. Breast Cancer Res Treat, 2009. 117(1): p. 69-75.

141

39. Di Leo, A., et al., Results of the CONFIRM phase III trial comparing fulvestrant 250 mg with fulvestrant 500 mg in postmenopausal women with estrogen receptor-positive advanced breast cancer. J Clin Oncol, 2010. 28(30): p. 4594-600.

40. Di Leo, A., et al., Final overall survival: fulvestrant 500 mg vs 250 mg in the randomized CONFIRM trial. J Natl Cancer Inst, 2014. 106(1): p. djt337.

41. Krop, I.E., et al., Pictilisib for oestrogen receptor-positive, aromatase inhibitor-resistant, advanced or metastatic breast cancer (FERGI): a randomised, double-blind, placebo- controlled, phase 2 trial. Lancet Oncol, 2016. 17(6): p. 811-821.

42. Baselga, J., et al., Buparlisib plus fulvestrant versus placebo plus fulvestrant in postmenopausal, hormone receptor-positive, HER2-negative, advanced breast cancer (BELLE-2): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol, 2017. 18(7): p. 904-916.

43. Campone, M., et al., Buparlisib plus fulvestrant versus placebo plus fulvestrant for postmenopausal, hormone receptor-positive, human epidermal growth factor receptor 2- negative, advanced breast cancer: Overall survival results from BELLE-2. Eur J Cancer, 2018. 103: p. 147-154.

44. Cristofanilli, M., et al., Fulvestrant plus palbociclib versus fulvestrant plus placebo for treatment of hormone-receptor-positive, HER2-negative metastatic breast cancer that progressed on previous endocrine therapy (PALOMA-3): final analysis of the multicentre, double-blind, phase 3 randomised controlled trial. Lancet Oncol, 2016. 17(4): p. 425-439.

45. Turner, N.C., et al., Overall Survival with Palbociclib and Fulvestrant in Advanced Breast Cancer. N Engl J Med, 2018. 379(20): p. 1926-1936.

46. Slamon, D.J., et al., Phase III Randomized Study of Ribociclib and Fulvestrant in Hormone Receptor-Positive, Human Epidermal Growth Factor Receptor 2-Negative Advanced Breast Cancer: MONALEESA-3. J Clin Oncol, 2018. 36(24): p. 2465-2472.

142

47. Sledge, G.W., et al., MONARCH 2: Abemaciclib in Combination With Fulvestrant in Women With HR+/HER2- Advanced Breast Cancer Who Had Progressed While Receiving Endocrine Therapy. J Clin Oncol, 2017. 35(25): p. 2875-2884.

48. Bahreini, A., et al., Mutation site and context dependent effects of ESR1 mutation in genome-edited breast cancer cell models. Breast Cancer Res, 2017. 19(1): p. 60.

49. Jeselsohn, R., et al., Emergence of constitutively active estrogen receptor-α mutations in pretreated advanced estrogen receptor-positive breast cancer. Clin Cancer Res, 2014. 20(7): p. 1757-67.

50. Robinson, D.R., et al., Activating ESR1 mutations in hormone-resistant metastatic breast cancer. Nat Genet, 2013. 45(12): p. 1446-51.

51. Toy, W., et al., ESR1 ligand-binding domain mutations in hormone-resistant breast cancer. Nat Genet, 2013. 45(12): p. 1439-45.

52. Toy, W., et al., Activating ESR1 Mutations Differentially Impact the Efficacy of ER Antagonists. Cancer Discov, 2017. 7(3): p. 277-287.

53. Jeselsohn, R., et al., ESR1 mutations—a mechanism for acquired endocrine resistance in breast cancer. Nat Rev Clin Oncol, 2015. 12(10): p. 573-83.

54. Fribbens, C., et al., Plasma ESR1 Mutations and the Treatment of Estrogen Receptor- Positive Advanced Breast Cancer. J Clin Oncol, 2016. 34(25): p. 2961-8.

55. Leal, M.F., et al., Early enrichment of ESR1 mutations and the impact on gene expression in pre-surgical primary breast cancer treated with aromatase inhibitors. Clin Cancer Res, 2019.

56. Schiavon, G., et al., Analysis of ESR1 mutation in circulating tumor DNA demonstrates evolution during therapy for metastatic breast cancer. Sci Transl Med, 2015. 7(313): p. 313ra182.

143

57. Spoerke, J.M., et al., Heterogeneity and clinical significance of ESR1 mutations in ER- positive metastatic breast cancer patients receiving fulvestrant. Nat Commun, 2016. 7: p. 11579.

58. Wang, P., et al., Sensitive Detection of Mono- and Polyclonal ESR1 Mutations in Primary Tumors, Metastatic Lesions, and Cell-Free DNA of Breast Cancer Patients. Clin Cancer Res, 2016. 22(5): p. 1130-7.

59. O'Leary, B., et al., The Genetic Landscape and Clonal Evolution of Breast Cancer Resistance to Palbociclib plus Fulvestrant in the PALOMA-3 Trial. Cancer Discov, 2018. 8(11): p. 1390-1403.

60. O'Leary, B., et al., Early circulating tumor DNA dynamics and clonal selection with palbociclib and fulvestrant for breast cancer. Nat Commun, 2018. 9(1): p. 896.

61. Jeselsohn, R., et al., Allele-Specific Chromatin Recruitment and Therapeutic Vulnerabilities of ESR1 Activating Mutations. Cancer Cell, 2018. 33(2): p. 173-186.e5.

62. Carlson, K.E., et al., Altered ligand binding properties and enhanced stability of a constitutively active estrogen receptor: evidence that an open pocket conformation is required for ligand interaction. Biochemistry, 1997. 36(48): p. 14897-905.

63. Wardell, S.E., et al., Efficacy of SERD/SERM Hybrid-CDK4/6 Inhibitor Combinations in Models of Endocrine Therapy-Resistant Breast Cancer. Clin Cancer Res, 2015. 21(22): p. 5121-5130.

64. Harrod, A., et al., Genomic modelling of the ESR1 Y537S mutation for evaluating function and new therapeutic approaches for metastatic breast cancer. Oncogene, 2017. 36(16): p. 2286-2296.

65. Joseph, J.D., et al., The selective estrogen receptor downregulator GDC-0810 is efficacious in diverse models of ER+ breast cancer. Elife, 2016. 5.

144

66. Martin, L.A., et al., Discovery of naturally occurring ESR1 mutations in breast cancer cell lines modelling endocrine resistance. Nat Commun, 2017. 8(1): p. 1865.

67. Toy, W.e.a., Activating ESR1 Mutations Differentially Affect the Efficacy of ER Antagonists. 2017, AACR Publications: Cancer Discovery. p. 277-287.

68. Kuang, Y., et al., Unraveling the clinicopathological features driving the emergence of ESR1 mutations in metastatic breast cancer NPJ Breast Cancer, 2018. 4: p. 22.

69. Bielski, C.M., et al., Widespread Selection for Oncogenic Mutant Allele Imbalance in Cancer. Cancer Cell, 2018. 34(5): p. 852-862.e4.

70. Lai, A., et al., Identification of GDC-0810 (ARN-810), an Orally Bioavailable Selective Estrogen Receptor Degrader (SERD) that Demonstrates Robust Activity in Tamoxifen- Resistant Breast Cancer Xenografts. J Med Chem, 2015. 58(12): p. 4888-904.

71. Wardell, S.E., et al., Evaluation of the pharmacological activities of RAD1901, a selective estrogen receptor degrader. Endocr Relat Cancer, 2015. 22(5): p. 713-24.

72. Wittmann, B.M., A. Sherk, and D.P. McDonnell, Definition of functionally important mechanistic differences among selective estrogen receptor down-regulators. Cancer Res, 2007. 67(19): p. 9549-60.

73. Wijayaratne, A.L., et al., Comparative analyses of mechanistic differences among antiestrogens. Endocrinology, 1999. 140(12): p. 5828-40.

74. Willson, T.M., et al., Dissection of the molecular mechanism of action of GW5638, a novel estrogen receptor ligand, provides insights into the role of estrogen receptor in bone. Endocrinology, 1997. 138(9): p. 3901-11.

75. Connor, C.E., et al., Circumventing tamoxifen resistance in breast cancers using antiestrogens that induce unique conformational changes in the estrogen receptor. Cancer Res, 2001. 61(7): p. 2917-22.

145

76. Wu, Y.L., et al., Structural basis for an unexpected mode of SERM-mediated ER antagonism. Mol Cell, 2005. 18(4): p. 413-24.

77. De Savi, C., et al., Optimization of a Novel Binding Motif to (E)-3-(3,5-Difluoro-4- ((1R,3R)-2-(2-fluoro-2-methylpropyl)-3-methyl-2,3,4,9-tetrahydro-1H-pyrido[3,4- b]indol-1-yl)phenyl)acrylic Acid (AZD9496), a Potent and Orally Bioavailable Selective Estrogen Receptor Downregulator and Antagonist. J Med Chem, 2015. 58(20): p. 8128-40.

78. Weir, H.M., et al., AZD9496: An Oral Estrogen Receptor Inhibitor That Blocks the Growth of ER-Positive and ESR1-Mutant Breast Tumors in Preclinical Models. Cancer Res, 2016. 76(11): p. 3307-18.

79. Hamilton, E.P., et al., A First-in-Human Study of the New Oral Selective Estrogen Receptor Degrader AZD9496 for ER. Clin Cancer Res, 2018. 24(15): p. 3510-3518.

80. Cheung, K.W.K., et al., GDC-0810 Pharmacokinetics and Transporter-Mediated Drug Interaction Evaluation with an Endogenous Biomarker in the First-in-Human, Dose Escalation Study. Drug Metab Dispos, 2019. 47(9): p. 966-973.

81. Wardell, S.E., et al., Bazedoxifene exhibits antiestrogenic activity in animal models of tamoxifen-resistant breast cancer: implications for treatment of advanced disease. Clin Cancer Res, 2013. 19(9): p. 2420-31.

82. Ronkin, S., et al., Endometrial effects of bazedoxifene acetate, a novel selective estrogen receptor modulator, in postmenopausal women. Obstet Gynecol, 2005. 105(6): p. 1397- 404.

83. Bihani, T., et al., Elacestrant (RAD1901), a Selective Estrogen Receptor Degrader (SERD), Has Antitumor Activity in Multiple ER+ Breast Cancer Patient-derived Xenograft Models. . Clin Cancer Res, 2017. 23(16): p. 4793-4804.

146

84. Fanning, S.W., et al., The SERM/SERD bazedoxifene disrupts ESR1 helix 12 to overcome acquired hormone resistance in breast cancer cells. Elife, 2018. 7.

85. Wardell, S.E., J.R. Marks, and D.P. McDonnell, The turnover of estrogen receptor α by the selective estrogen receptor degrader (SERD) fulvestrant is a saturable process that is not required for antagonist efficacy. Biochem Pharmacol, 2011. 82(2): p. 122-30.

86. Wardell, S.E., et al., Pharmacokinetic and pharmacodynamic analysis of fulvestrant in preclinical models of breast cancer to assess the importance of its estrogen receptor-α degrader activity in antitumor efficacy. Breast Cancer Res Treat, 2019.

87. Agrawal, A., et al., Biological effects of fulvestrant on estrogen receptor positive human breast cancer: short, medium and long-term effects based on sequential biopsies. Int J Cancer, 2016. 138(1): p. 146-59.

88. Guan, J., et al., Therapeutic Ligands Antagonize Estrogen Receptor Function by Impairing Its Mobility. Cell, 2019. 178(4): p. 949-963.e18.

89. Wardell, S.E., E.R. Nelson, and D.P. McDonnell, From empirical to mechanism-based discovery of clinically useful Selective Estrogen Receptor Modulators (SERMs). Steroids, 2014. 90: p. 30-8.

90. Vogel, V.G., et al., Effects of tamoxifen vs raloxifene on the risk of developing invasive breast cancer and other disease outcomes: the NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 trial. JAMA, 2006. 295(23): p. 2727-41.

91. Cummings, S.R., et al., Lasofoxifene in postmenopausal women with osteoporosis. N Engl J Med, 2010. 362(8): p. 686-96.

92. Andreano, K.J.e.a., The dysregulated pharmacology of clinically relevant ESR1 mutants is normalized by ligand- activated WT receptor. Molecular Cancer Therapeutics

147

93. Brodie, A., Aromatase inhibitors in breast cancer. Trends Endocrinol Metab, 2002. 13(2): p. 61-5.

94. Dowsett, M. and A. Howell, Breast cancer: aromatase inhibitors take on tamoxifen. Nat Med, 2002. 8(12): p. 1341-4.

95. Mokbel, K., The evolving role of aromatase inhibitors in breast cancer. Int J Clin Oncol, 2002. 7(5): p. 279-83.

96. Palmieri, C., et al., Breast cancer: current and future endocrine therapies. Mol Cell Endocrinol, 2014. 382(1): p. 695-723.

97. van Kruchten, M., et al., Measuring residual estrogen receptor availability during fulvestrant therapy in patients with metastatic breast cancer. Cancer Discov, 2015. 5(1): p. 72-81.

98. Razavi, P., et al., The Genomic Landscape of Endocrine-Resistant Advanced Breast Cancers. Cancer Cell, 2018. 34(3): p. 427-438.e6.

99. Wang, T., et al., A multiplex allele-specific real-time PCR assay for screening of ESR1 mutations in metastatic breast cancer. Exp Mol Pathol, 2015. 98(2): p. 152-7.

100. Gaillard, S.L., et al., Constitutively active ESR1 mutations in gynecologic malignancies and clinical response to estrogen-receptor directed therapies. Gynecol Oncol, 2019. 154(1): p. 199-206.

101. Brzozowski, A.M., et al., Molecular basis of agonism and antagonism in the oestrogen receptor. Nature, 1997. 389(6652): p. 753-8.

102. Norris, J.D., et al., Peptide antagonists of the human estrogen receptor. Science, 1999. 285(5428): p. 744-6.

148

103. Norris, J.D., et al., Differential presentation of protein interaction surfaces on the androgen receptor defines the pharmacological actions of bound ligands. Chem Biol, 2009. 16(4): p. 452-60.

104. Gee, A.C., et al., Coactivator peptides have a differential stabilizing effect on the binding of estrogens and antiestrogens with the estrogen receptor. Mol Endocrinol, 1999. 13(11): p. 1912-23.

105. Huang, H.J., J.D. Norris, and D.P. McDonnell, Identification of a negative regulatory surface within estrogen receptor alpha provides evidence in support of a role for corepressors in regulating cellular responses to agonists and antagonists. Mol Endocrinol, 2002. 16(8): p. 1778-92.

106. Kobayashi, S., et al., Mechanisms of progesterone receptor inhibition of inflammatory responses in cellular models of breast cancer. Mol Endocrinol, 2010. 24(12): p. 2292- 302.

107. Yi, P., et al., Structure of a biologically active estrogen receptor-coactivator complex on DNA. Mol Cell, 2015. 57(6): p. 1047-1058.

108. Rioux, N., et al., Nonclinical pharmacokinetics and in vitro metabolism of H3B-6545, a novel selective ERα covalent antagonist (SERCA). Cancer Chemother Pharmacol, 2019. 83(1): p. 151-160.

109. Puyang, X., et al., Discovery of Selective Estrogen Receptor Covalent Antagonists for the Treatment of ERα. Cancer Discov, 2018. 8(9): p. 1176-1193.

110. Osborne, C.K., et al., Role of the estrogen receptor coactivator AIB1 (SRC-3) and HER- 2/neu in tamoxifen resistance in breast cancer. J Natl Cancer Inst, 2003. 95(5): p. 353- 61.

149

111. Tria, G.S., et al., Discovery of LSZ102, a Potent, Orally Bioavailable Selective Estrogen Receptor Degrader (SERD) for the Treatment of Estrogen Receptor Positive Breast Cancer. J Med Chem, 2018. 61(7): p. 2837-2864.

112. Chang, M.T., et al., Accelerating Discovery of Functional Mutant Alleles in Cancer. Cancer Discov, 2018. 8(2): p. 174-183.

113. Merenbakh-Lamin, K., et al., D538G mutation in estrogen receptor-α: A novel mechanism for acquired endocrine resistance in breast cancer. Cancer Res, 2013. 73(23): p. 6856-64.

114. Tomita, S., et al., Estrogen receptor alpha gene ESR1 amplification may predict endocrine therapy responsiveness in breast cancer patients. Cancer Sci, 2009. 100(6): p. 1012-7.

115. Yi, J.W., et al., Upregulation of the ESR1 Gene and ESR Ratio (ESR1/ESR2) is Associated with a Worse Prognosis in Papillary Thyroid Carcinoma: The Impact of the Estrogen Receptor α/β Expression on Clinical Outcomes in Papillary Thyroid Carcinoma Patients. Ann Surg Oncol, 2017. 24(12): p. 3754-3762.

116. Rahman, M.T., et al., ESR1 gene amplification in endometrial carcinomas: a clinicopathological analysis. Anticancer Res, 2013. 33(9): p. 3775-81.

117. Lebeau, A., et al., Oestrogen receptor gene (ESR1) amplification is frequent in endometrial carcinoma and its precursor lesions. J Pathol, 2008. 216(2): p. 151-7.

118. Tan, D.S., et al., ESR1 amplification in endometrial carcinomas: hope or hyperbole? J Pathol, 2008. 216(3): p. 271-4.

119. Holst, F., Estrogen receptor alpha gene amplification in breast cancer: 25 years of debate. World J Clin Oncol, 2016. 7(2): p. 160-73.

150

120. Backes, F.J., et al., Estrogen receptor-alpha as a predictive biomarker in endometrioid endometrial cancer. Gynecol Oncol, 2016. 141(2): p. 312-317.

121. McIntyre, J.B., et al., Molecular alterations in indolent, aggressive and recurrent ovarian low-grade serous carcinoma. Histopathology, 2017. 70(3): p. 347-358.

122. Jones, S., et al., Genomic analyses of gynaecologic carcinosarcomas reveal frequent mutations in chromatin remodelling genes. Nat Commun, 2014. 5: p. 5006.

123. Consortium, A.P.G., AACR Project GENIE: Powering Precision Medicine through an International Consortium. Cancer Discov, 2017. 7(8): p. 818-831.

124. Network, C.G.A.R., Integrated genomic analyses of ovarian carcinoma. Nature, 2011. 474(7353): p. 609-15.

125. Network, C.G.A.R., et al., Integrated genomic and molecular characterization of cervical cancer. Nature, 2017. 543(7645): p. 378-384.

126. Kandoth, C., et al., Integrated genomic characterization of endometrial carcinoma. Nature, 2013. 497(7447): p. 67-73.

127. Gao, J., et al., Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal, 2013. 6(269): p. pl1.

128. Cerami, E., et al., The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov, 2012. 2(5): p. 401-4.

129. Forbes, S.A., et al., COSMIC: exploring the world's knowledge of somatic mutations in human cancer. Nucleic Acids Res, 2015. 43(Database issue): p. D805-11.

130. Highlights of Prescribing Information: Evista, C.f.D.E.a. Research, Editor. 2007, Food and Drug Administration Silver Spring, MD

151

131. Summary of Product Characteristics: Conbriza 2009, European Medicines Agency London, England.

132. Assessment Report for Fablyn. 2009, European Medicines Agency London, England.

133. Highlights of Prescribing Information: Faslodex, C.f.D.E.a. Research, Editor. 2012, Food and Drug Administration: Silver Spring, MD.

134. Lien, E.A., et al., Serum concentrations of tamoxifen and its metabolites increase with age during steady-state treatment. Breast Cancer Res Treat, 2013. 141(2): p. 243-8.

135. Chung, J.H., et al., Hybrid capture-based genomic profiling of circulating tumor DNA from patients with estrogen receptor-positive metastatic breast cancer. Ann Oncol, 2017. 28(11): p. 2866-2873.

136. Stover, E.H., et al., Targeted Next-Generation Sequencing Reveals Clinically Actionable. JCO Precis Oncol, 2018. 2018.

137. Ma, C.X., et al., A Phase II Trial of Neoadjuvant MK-2206, an AKT Inhibitor, with Anastrozole in Clinical Stage II or III. Clin Cancer Res, 2017. 23(22): p. 6823-6832.

138. Fader, A.N., et al., Primary cytoreductive surgery and adjuvant hormonal monotherapy in women with advanced low-grade serous ovarian carcinoma: Reducing overtreatment without compromising survival? Gynecol Oncol, 2017. 147(1): p. 85-91.

139. Chu, D., et al., ESR1 Mutations in Circulating Plasma Tumor DNA from Metastatic Breast Cancer Patients. Clin Cancer Res, 2016. 22(4): p. 993-9.

140. Andreano, K.J., et al., G1T48, an oral selective estrogen receptor degrader, and the CDK4/6 inhibitor lerociclib inhibit tumor growth in animal models of endocrine-resistant breast cancer. Breast Cancer Res Treat, 2020. 180(3): p. 635-646.

152

141. Font de Mora, J. and M. Brown, AIB1 is a conduit for kinase-mediated growth factor signaling to the estrogen receptor. Mol Cell Biol, 2000. 20(14): p. 5041-7.

142. Campbell, R.A., et al., Phosphatidylinositol 3-kinase/AKT-mediated activation of estrogen receptor alpha: a new model for anti-estrogen resistance. J Biol Chem, 2001. 276(13): p. 9817-24.

143. Yue, W., et al., Activation of the MAPK pathway enhances sensitivity of MCF-7 breast cancer cells to the mitogenic effect of estradiol. Endocrinology, 2002. 143(9): p. 3221-9.

144. Medunjanin, S., et al., Glycogen synthase kinase-3 interacts with and phosphorylates estrogen receptor alpha and is involved in the regulation of receptor activity. J Biol Chem, 2005. 280(38): p. 33006-14.

145. Heidari, P., et al., Pharmacodynamic imaging guides dosing of a selective estrogen receptor degrader. Clin Cancer Res, 2015. 21(6): p. 1340-7.

146. Xiong, R., et al., Novel Selective Estrogen Receptor Downregulators (SERDs) Developed against Treatment-Resistant Breast Cancer. J Med Chem, 2017. 60(4): p. 1325-1342.

147. Bisi, J.E., et al., Preclinical development of G1T38: A novel, potent and selective inhibitor of cyclin dependent kinases 4/6 for use as an oral antineoplastic in patients with CDK4/6 sensitive tumors. Oncotarget, 2017. 8(26): p. 42343-42358.

148. Stice, J.P., et al., CDK4/6 Therapeutic Intervention and Viable Alternative to Taxanes in CRPC. Mol Cancer Res, 2017. 15(6): p. 660-669.

149. Kahraman, M., et al., Maximizing ER-α Degradation Maximizes Activity in a Tamoxifen-Resistant Breast Cancer Model: Identification of GDC-0927. ACS Med Chem Lett, 2019. 10(1): p. 50-55.

153

150. Draper, M.W., et al., A controlled trial of raloxifene (LY139481) HCl: impact on bone turnover and serum lipid profile in healthy postmenopausal women. J Bone Miner Res, 1996. 11(6): p. 835-42.

151. Wardell, S.E., D. Kazmin, and D.P. McDonnell, Research resource: Transcriptional profiling in a cellular model of breast cancer reveals functional and mechanistic differences between clinically relevant SERM and between SERM/estrogen complexes. Mol Endocrinol, 2012. 26(7): p. 1235-48.

152. Nagy, L. and J.W. Schwabe, Mechanism of the nuclear receptor molecular switch. Trends Biochem Sci, 2004. 29(6): p. 317-24.

153. Love, R.R., et al., Effects of tamoxifen on bone mineral density in postmenopausal women with breast cancer. N Engl J Med, 1992. 326(13): p. 852-6.

154. Love, R.R., et al., Effects of tamoxifen on cardiovascular risk factors in postmenopausal women. Ann Intern Med, 1991. 115(11): p. 860-4.

155. Mader, S., et al., Three amino acids of the oestrogen receptor are essential to its ability to distinguish an oestrogen from a glucocorticoid-responsive element. Nature, 1989. 338(6212): p. 271-4.

156. Fanning, S.W., et al., Estrogen receptor alpha somatic mutations Y537S and D538G confer breast cancer endocrine resistance by stabilizing the activating function-2 binding conformation. Elife, 2016. 5.

157. Li, Z., et al., Upregulation of IRS1 Enhances IGF1 Response in Y537S and D538G ESR1 Mutant Breast Cancer Cells. Endocrinology, 2018. 159(1): p. 285-296.

158. Gelsomino, L., et al., ESR1 mutations affect anti-proliferative responses to tamoxifen through enhanced cross-talk with IGF signaling. Breast Cancer Res Treat, 2016. 157(2): p. 253-265.

154

159. Pellegrino, M., et al., High-throughput single-cell DNA sequencing of acute myeloid leukemia tumors with droplet microfluidics. Genome Res, 2018. 28(9): p. 1345-1352.

160. Pellegrino, M., et al., RNA-Seq following PCR-based sorting reveals rare cell transcriptional signatures. BMC Genomics, 2016. 17: p. 361.

161. Eastburn, D.J., et al., Microfluidic droplet enrichment for targeted sequencing. Nucleic Acids Res, 2015. 43(13): p. e86.

162. Eastburn, D.J., A. Sciambi, and A.R. Abate, Identification and genetic analysis of cancer cells with PCR-activated cell sorting. Nucleic Acids Res, 2014. 42(16): p. e128.

163. Xu, L., et al., Clonal Evolution and Changes in Two AML Patients Detected with A Novel Single-Cell DNA Sequencing Platform. Sci Rep, 2019. 9(1): p. 11119.

164. McMahon, C.M., et al., Clonal Selection with RAS Pathway Activation Mediates Secondary Clinical Resistance to Selective FLT3 Inhibition in Acute Myeloid Leukemia. Cancer Discov, 2019. 9(8): p. 1050-1063.

165. Wei, A.H. and A.W. Roberts, Polyclonal Heterogeneity: The New Norm for Secondary Clinical Resistance to Targeted Monotherapy in Relapsed Leukemia? Cancer Discov, 2019. 9(8): p. 998-1000.

166. Gao, Y., et al., V211D Mutation in MEK1 Causes Resistance to MEK Inhibitors in Colon Cancer. Cancer Discov, 2019. 9(9): p. 1182-1191.

167. Blanchard, Z., et al., Estrogen-independent molecular actions of mutant estrogen receptor 1 in endometrial cancer. Genome Res, 2019. 29(9): p. 1429-1441.

168. Norris, J., et al., Identification of a new subclass of Alu DNA repeats which can function as estrogen receptor-dependent transcriptional enhancers. J Biol Chem, 1995. 270(39): p. 22777-82.

155

169. Nagel, S.C., J.L. Hagelbarger, and D.P. McDonnell, Development of an ER action indicator mouse for the study of estrogens, selective ER modulators (SERMs), and Xenobiotics. Endocrinology, 2001. 142(11): p. 4721-8.

170. Chang, C., et al., Dissection of the LXXLL nuclear receptor-coactivator interaction motif using combinatorial peptide libraries: discovery of peptide antagonists of estrogen receptors alpha and beta. Mol Cell Biol, 1999. 19(12): p. 8226-39.

171. He, J., et al., Integrated genomic DNA/RNA profiling of hematologic malignancies in the clinical setting. Blood, 2016. 127(24): p. 3004-14.

172. Frampton, G.M., et al., Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol, 2013. 31(11): p. 1023-31.

173. Livak, K.J. and T.D. Schmittgen, Analysis of relative gene expression data using real- time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods, 2001. 25(4): p. 402-8.

156

Biography

Kaitlyn Jo Andreano is the eldest daughter of Jeffrey and Kelly Andreano. She was born in Pennsylvania and raised in New York State. She earned her Bachelors of

Science in Biochemistry with minors in Chemistry and Leadership and Social Change from Virginia Polytechnic Institute and State University in 2014. She will earn her Ph.D. in Pharmacology from Duke University in spring 2020. During her graduate career, she was a co-author on seven primary research papers and one patent (listed below). She was a recipient of the Ruth L. Kirschstein Predoctroal Individual National Research

Service Award (F31). In Summer 2019, she spent the summer as an intern at Novartis

Institutes of Biomedical Research (NIBR) in Cambridge, Massachusetts as a part of the

Department of Pharmacology and Cancer Biology’s industry internship program.

Patent

Andreano, K.J., Chang, C., Gaillard S.L., McDonnell, D.P. Lasofoxifene Treatment of Breast Cancer. United States (Pub. No. US201903236S)

Publications

Gajadeeral, N., Mendes, O.P., Andreano, K.J. et. al Preliminary biologic evaluation of fluorescently labeled steroidal antiestrogens. In Preparation.

Andreano, K.J. et al. The dysregulated pharmacology of clinically relevant ESR1 mutants is normalized by ligand-activated WT receptor. Molecular Cancer Therapeutics. Accepted. (Chapter 2)

157

Ochsner, S.A., Abraham, D., Martin, K., Ding, W., McOwiti, A., Kankanamge, W., Wang, Z, Andreano K. et.al. The Signaling Pathways Project: an integrated 'omics knowledgebase for mammalian cellular signaling pathways. Scientific Data, 2019. 252 (6).

Wardell, S.E., Yllanes, A.P., Chao, C.A., Bae, Y., Andreano, K.J. et. al., Pharmacokinetic and pharmacodynamic analysis of fulvestrant in preclinical models of breast cancer to assess the importance of its estrogen receptor-α degrader activity in antitumor efficacy. Breast Cancer Res Treat, 2019.

Norris, J.D., Ellison, S.J., Baker, J.G., Stagg, D.B., Wardell, S.E., Park, S., Alley, H.M., Baldi, R.M., Yllanes, A., Andreano, K.J. et al., Androgen receptor antagonism drives cytochrome P450 17A1 inhibitor efficacy in prostate cancer. The Journal of Clinical Investigation, 2017. 127(6): p. 2326-2338

158