Defining the Mechanism of Action of Bromodomain And

DEFINING THE MECHANISM OF ACTION OF BROMODOMAIN AND

EXTRATERMINAL INHIBITORS IN TRIPLE-NEGATIVE BREAST CANCERS

JENNIFER MICHELE BRANCATO

Submitted in partial fulfillment of the requirements

for the degree of Doctor of Philosophy

Dissertation Advisor: Ruth A. Keri, Ph.D.

Department of Pharmacology

CASE WESTERN RESERVE UNIVERSITY

May, 2018 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Jennifer Michele Brancato

candidate for the degree of Doctor of Philosophy*.

Youwei Zhang, Ph.D. (Committee Chair)

Ruth Keri, Ph.D. (Committee Member)

Goutham Narla, M.D., Ph.D. (Committee Member)

Peter Scacheri, Ph.D. (Committee Member)

Analisa DiFeo, Ph.D. (Committee Member)

William Schiemann, Ph.D. (Committee Member)

Date of Defense

February 2, 2018

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

DEDICATION

I dedicate this work to my family. First, to my parents who have always pushed me

to dream big and work hard to achieve my goals. I would not be where I am today

without your love and support. Also, to my husband and my son. I know the last

several years were not always easy, especially when I had to focus on my work

instead of spending time with you. Your patience and encouragement helped to

carry me through this process. I love you.

TABLE OF CONTENTS

List of Tables ...... 6 List of Figures ...... 7 Acknowledgements ...... 10 ABSTRACT ...... 12 Chapter 1: Introduction ...... 14 1.1 Overview of breast cancer ...... 15 1.1.1 Breast cancer subtypes ...... 16 1.1.1.1 Receptor status ...... 17 1.1.1.2 Intrinsic subtypes of breast cancer ...... 21 1.1.1.3 Additional breast cancer classification systems ...... 26 1.1.2 Therapeutic options for the treatment of breast cancer...... 30 1.1.3 Resistance to cytotoxic chemotherapy ...... 37 1.1.3.1 Taxane-specific resistance mechanisms ...... 39 1.1.3.2 Anthracycline-specific resistance mechanisms ...... 40 1.2 Basics of transcriptional control of genes ...... 41 1.2.1 Transcriptional regulatory elements ...... 41 1.2.2 RNA polymerase ...... 45 1.2.3 Other transcription-related proteins...... 46 1.2.4 Transcription initiation, elongation, and termination ...... 50 1.3 Targeting the epigenome in cancer ...... 52 1.3.1 Chromatin structure ...... 53 1.3.2 DNA methylation ...... 54 1.3.3 Histone modifications ...... 57 1.4 Targeting Bromodomain and extraterminal proteins in breast cancer 61 1.4.1 BET protein structure and function ...... 61 1.4.1.1 BET protein structure ...... 61 1.4.1.2 BET proteins and transcription ...... 63

1.4.1.3 BET proteins and the cell cycle ...... 64 1.4.1.4 BET proteins and inflammation ...... 65 1.4.1.5 BET proteins and development ...... 66 1.4.1.6 BRDT and spermatogenesis ...... 66 1.4.2 BET proteins in cancer ...... 66 1.4.2.1 BET proteins and super-enhancers ...... 66 1.4.2.2 Expression of the BET family in breast tumors and cell lines ...... 68 1.4.2.3 Essentiality of BET proteins in breast cancer ...... 69 1.4.3 Targeting BET proteins in breast cancer ...... 70 1.4.3.1 BET inhibitor structure and selectivity ...... 70 1.4.3.2 Impact of breast cancer cell growth and tumor formation .. 72 1.4.3.3 BET inhibitors and hypoxia ...... 76 1.4.3.4 BET inhibitors and angiogenesis ...... 77 1.4.3.5 BET inhibitors and cancer stem cells ...... 79 1.4.3.6 BET inhibitors and metastasis ...... 81 1.4.3.7 BET inhibitors and metabolism ...... 85 1.4.3.8 Mechanism(s) of action of BET inhibitors in breast cancer 86 1.4.3.9 Resistance to BET inhibitors ...... 90 1.4.3.10 BET inhibitor adverse effects ...... 92 1.4.4 Alternative BET protein targeting mechanisms ...... 94 1.4.4.1 Combination therapies ...... 95 1.4.4.2 BET protein degraders ...... 106 1.4.4.3 Dual protein kinase/BET protein inhibitors ...... 108 1.4.4.4 CDK7 inhibitors ...... 109 1.4.5 Future perspectives and conclusions ...... 111 1.5 Mitosis ...... 113 1.5.1 The cell cycle and cell cycle checkpoints ...... 113 1.5.2 Cyclin-dependent kinases ...... 116

1.5.3 The mitotic kinases ...... 119 1.5.3.1 Polo-like kinases ...... 121 1.5.3.2 Aurora kinases ...... 122 1.5.4 Activity of mitotic kinases in M phase ...... 123 1.5.5 Targeting mitotic kinases in breast cancer ...... 128 1.5.5.1 Polo-like kinase 1 ...... 129 1.5.5.2 Aurora kinase A ...... 130 1.5.5.3 Aurora kinase B ...... 132 1.6 Statement of purpose ...... 133 Chapter 2: Bromodomain and extraterminal protein inhibition blocks growth of triple-negative breast cancers through the suppression of Aurora kinases ...... 158 2.1 Abstract ...... 159 2.2 Introduction ...... 159 2.3 Materials and methods ...... 161 2.4 Results ...... 167 2.4.1 BET inhibition blocks growth of diverse TNBC cells without consistently down-regulating MYC ...... 167 2.4.2 Sustained BET inhibition induces apoptosis and senescence in TNBC cells ...... 168 2.4.3 BET inhibition abrogates TNBC tumor growth ...... 170 2.4.4 Aurora kinases are downstream targets of BETi ...... 171 2.5 Discussion ...... 173 Chapter 3: Mitotic vulnerability in triple-negative breast cancer associated with LIN9 is targetable with BET inhibitors ...... 191 3.1 Abstract ...... 192 3.2 Introduction ...... 192 3.3 Materials and methods ...... 195 3.4 Results ...... 202 3.4.1 Sustained BET activity is necessary for normal progression through mitosis ...... 202

3.4.2 BET activity is necessary for sustained expression of cell-cycle- associated genes ...... 204 3.4.3 BET inhibitors fail to induce a luminal differentiation signature 205 3.4.4 BETi proteins directly modulate the mitotic transcriptional program ...... 206 3.4.5 BETi suppress mitosis transcription factors in the absence of SEs ...... 207 3.4.6 LIN9 is a key downstream effector of BET proteins ...... 209 3.5 Discussion ...... 212 Chapter 4: Discussion and future directions ...... 237 4.1 Summary ...... 238 4.2 The BET inhibitor-induced mitotic defect ...... 241 4.3 The role of Bcl-xL in the response of TNBC cells to BETi-induced mitotic catastrophe ...... 244 4.4 The mechanism of action of BETi in other subtypes of breast cancer ...... 245 4.5 The role of LIN9 in BET inhibitor-induced apoptosis and senescence ...... 248 4.6 The impact of LIN9 in BET inhibitor-induced mitotic catastrophe ..... 250 4.7 The LIN9-controlled transcriptome ...... 252 4.8 Alternative LIN9-containing complexes ...... 253 4.9 LIN9 and tumorigenesis ...... 256 4.10 Additional inhibitors of LIN9 ...... 257 4.11 Conclusions ...... 258 References ...... 265

LIST OF TABLES

Table 1.1: Examples of classification systems used to subtype breast tumors 135 Table 1.2: The characteristics of the intrinsic subtypes of breast cancer ...... 137 Table 1.3: FDA-approved therapies for breast cancer ...... 139 Table 1.4: The core histones and their variants ...... 141 Table 1.5: Breast cell lines used to study BET proteins and their inhibitors ..... 144 Table 1.6: Current and completed clinical trials evaluating AURKA inhibitors in breast cancer ...... 157 Table 2.1: TNBC subtypes and responses to BET inhibition ...... 180

LIST OF FIGURES

Figure 1.1: Receptor status-based subtypes are not interchangeable with the intrinsic subtypes of breast cancer ...... 136 Figure 1.2: The intrinsic subtypes of breast cancer subdivide into the pathway- derived subgroups of breast cancers ...... 138 Figure 1.3: Histone modification proteins add, remove, and read acetylated histone marks ...... 142 Figure 1.4: BET inhibitors suppress several oncogenic pathways in breast cancer ...... 146 Figure 1.5: Mechanisms of BET inhibitor resistance in breast cancer ...... 148 Figure 1.6: Combination treatments with BET inhibitors in breast cancer ...... 150 Figure 1.7: The response of the hallmarks of cancer to BET inhibitor treatment in breast cancer ...... 151 Figure 1.8: Activation and inhibition of cyclin-CDK complexes ...... 152 Figure 1.9: Cyclin-CDK activity in G1 phase ...... 153 Figure 1.10: Cyclin-CDK activity in S, G2, and M phases ...... 155 Figure 2.1: BET inhibition blocks growth of TNBC cells without consistently down- regulating MYC ...... 178 Figure 2.2: Sustained BET inhibition induces apoptosis and senescence in TNBC cells ...... 182 Figure 2.3: BET inhibition abrogates tumor growth ...... 184 Figure 2.4: BETi lack toxic side effects and do not impact normal adult mouse mammary gland morphology or proliferation ...... 186 Figure 2.5: Aurora kinases are downstream targets of BETi ...... 187 Figure 2.6: Aurora kinase inhibitors phenocopy BETi ...... 189 Figure 3.1: BET inhibition induces apoptosis and senescence in TNBC cells .. 218 Figure 3.2: Sustained BET activity is necessary for timely progression through mitosis ...... 219 Figure 3.3: BET inhibitors promote mitosis-associated death or prolonged interphase ...... 221 Figure 3.4: BET activity is necessary for sustained expression of cell cycle- associated genes ...... 223 Figure 3.5: BET proteins directly modulate the mitotic transcriptional program 225 Figure 3.6: BET inhibitors fail to induce a luminal differentiation signature ...... 226

Figure 3.7: I-BET762 suppresses expression of master regulators of mitosis .. 227 Figure 3.8: Simultaneous knockdown of FOXM1, E2F2, E2F8, LIN9, and MYBL2 phenocopies BET inhibition ...... 228 Figure 3.9: BETi suppress mitosis transcription factors in the absence of SEs 229 Figure 3.10: Simultaneous knockdown of FOXM1, LIN9, and MYBL2 phenocopies BET inhibition ...... 231 Figure 3.11: LIN9 mediates the effects of BET inhibition ...... 233 Figure 3.12: Individual knockdown of FOXM1, E2F2, E2F8, and MYBL2 does not phenocopy BET inhibition ...... 235 Figure 3.13: High expression of LIN9 is correlated with lower overall survival rates ...... 236 Figure 4.1: BETi induce mitotic catastrophe in TNBC by suppressing expression of LIN9 ...... 261 Figure 4.2: JQ1 may lead to aborted cleavage furrow ingression ...... 262 Figure 4.3: BETi alter growth patterns and cellular morphology of luminal breast cancer cells...... 263 Figure 4.4: LIN9 is amplified and overexpressed in cancer and particularly in TNBC ...... 264

ACKNOWLEDGEMENTS

I would first like to thank my advisor, Ruth Keri, for giving me the opportunity to

work in her lab. Even though she is very busy, Ruth has always taken the time to

help me with both work-related and personal issues. I have learned so much from

her, including how to think critically about my data, how to write compelling papers

and grants, and how to create eye-catching diagrams. She has helped me grow

not only as a scientist but also as a person. Even when my confidence has

wavered, she has had faith in me. Ruth has set an excellent example of what a

mentor should be, using my mistakes as teachable moments, encouraging me when I doubt myself, and celebrating my successes, both big and small. Everything

I have accomplished is a reflection of her excellent mentorship. I would not be where I am today without her, and for that I am truly grateful.

Secondly, thank you to the members of the Keri lab, both past (Jennifer Yori, Nicole

Restrepo, Valery Adorno-Cruz, Erika Ramos, Myrielis Rivera, and Victoria Osorio

Vasquez) and present (Darcie Seachrist, Kristen Weber Bonk, Sylvia Gayle, Elyse

Donaubauer, Lindsey Anstine, Leslie Cuellar Vite, Bryan Webb, and Melyssa

Shively). They have been amazing colleagues and friends and all contributed to my project in one way or another. I have learned so much from them, and I appreciate all of the support they have given me over the years. They have been with me every step of the way. I could not have done this without them.

I would also like to thank my thesis committee members for their constructive

guidance and support. Their comments and questions during my committee

meetings not only improved my project but also helped me to become a better

scientist. In addition, thank you to my collaborators: Vinay Vardan, Salendra Singh,

Gurkan Bebek, Matthew Summers, Steven Sizemore, James Bradner, Jenny

Chang, and Melissa Landis. I have been incredibly lucky to work with these

talented researchers, and my project greatly benefited from their expertise and

resources.

Thank you to the Cancer Biology Training Grant and the NCI Predoctoral to

Postdoctoral Fellow Transition Award (F99/K00) for supporting my growth as a scientist. I would also like to acknowledge those who assisted in the preparation of my F99/K00 application. In particular, thank you to Ruth Keri, Mark Jackson, and Damian Junk for helping me navigate the application process and for reviewing the various components of my application. Also, thank you to Darcie

Seachrist and Kristen Weber Bonk for proofreading all of the documents. Finally, thank you to Vida Tripodo for helping me submit the application and troubleshoot the issues that arose during the submission process. These six people worked so hard to ensure I submitted the best application possible, and I cannot thank them enough.

Lastly, I would like to thank my family and friends for their unconditional love and support during my time in graduate school. Even though I am sure they did not completely understand my project or the various requirements within the pharmacology department, they always listened when I was excited about a new piece of data or nervous about an upcoming challenge. Most importantly, they helped provide balance in my life and ensured I did not get lost within my work. I love you all and attribute much of my success to you.

Defining the Mechanism of Action of Bromodomain and Extraterminal Inhibitors

in Triple-Negative Breast Cancers

Abstract

JENNIFER M. BRANCATO

Inhibitors of the Bromodomain and Extraterminal (BET) family of epigenetic

readers have been extensively studied in a wide range of cancers, and several are

being assessed in clinical trials for their safety and efficacy in both hematologic

and solid tumors. However, we have a very limited understanding of the

mechanism(s) of action of these drugs in cancer. It is important to determine how

BET inhibitors (BETi) elicit their effects in order to identify patients who will benefit

most from the therapy and to develop effective combination treatments that will

provide durable responses and prevent resistance. Our studies centered on

identifying the mechanism of action of BETi in triple-negative breast cancer

(TNBC), an aggressive disease that lacks effective targeted therapies. We found

that BETi induced multinucleation in TNBC due in part to the suppression of the

critical mitosis regulators Aurora kinases A/B, and this was followed by apoptosis

and senescence. The appearance of multinucleated cells coincided with the

suppression of mitosis genes, increased mitotic timing, and the induction of

apoptosis during mitosis or immediately following mitotic exit, all of which indicate

that BETi induce mitotic catastrophe. We further discovered that LIN9, a member

of the MuvB transcriptional regulatory complex, is a key mediator of BETi activity

in TNBC. BETi disrupt binding of the BET protein BRD4 to the promoter of LIN9,

leading to LIN9 downregulation, which in turn suppresses expression of mitosis

genes and induces multinucleation. While the selectivity of BETi for cancer cells

has been attributed to the dismantling of super-enhancers (SE) by BETi, our

studies revealed that LIN9 and four other critical mitosis regulators lacked SEs,

indicating BETi-induced mitotic catastrophe does not rely on the disruption of SEs.

Finally, we discovered TNBCs may be particularly dependent on LIN9 expression.

LIN9 is amplified/overexpressed in two-thirds of TNBCs, and high LIN9 expression

is associated with poor survival. Together, these data reveal that tumors with

amplified/overexpressed LIN9 such as TNBCs may be particularly responsive to

BETi. They further suggest that an effective therapy to treat TNBC could be the

combination of BETi and drugs that increase sensitivity to mitotic catastrophe.

CHAPTER 1: INTRODUCTION

This chapter was originally published, in part, in Pharmacological Research. Sahni

JM and Keri RA. Targeting Bromodomain and Extraterminal Proteins in Breast

Cancer. Pharmacol Res 2017 (In Press). © Elsevier Ltd.

1.1 Overview of breast cancer

Breast cancer is the most commonly diagnosed non-skin cancer-related

malignancy in women regardless of race and ethnicity, with one in eight women

being diagnosed with breast cancer in their lifetime1. After lung cancer, breast

cancer is the second leading cause of cancer-related deaths among women.

Breast cancer also impacts men, although males comprise less than 1% of all

breast cancer diagnoses2. Women over 50 years of age account for the majority

of new breast cancer cases and breast cancer-related deaths each year, while

younger women are more likely to relapse3-5. Of all of the racial and ethnic groups,

breast cancer incidence and mortality is highest among non-Hispanic white and

non-Hispanic black women and is lowest for Asian Pacific Islanders6.

Risk factors for developing breast cancer include a family history of breast cancer,

previous breast cancer diagnosis, smoking, obesity, alcohol consumption, and

dense breasts7,8. Experiencing later menarche, having children earlier in life,

bearing a greater number of children, and breastfeeding are all linked to a lower

breast cancer risk9,10. About 10-30% of breast cancer patients have a family history

of breast cancer, and BRCA1/2 mutations contribute to about 30% of these hereditary breast cancer cases11-14. BRCA1 and BRCA2 are tumor suppressors

that play important roles in cell division, DNA damage repair, and apoptosis15.

These proteins are crucial players in the repair of double-strand breaks by

homologous recombination, and cells that lose expression of these proteins are

unable to undergo homologous recombination16. While the vast majority of

mutations in BRCA1/2 are hereditary, they can also occur spontaneously.

BRCA1/2 mutations not only dramatically increase the lifetime risk of developing

breast cancer (65-80% with a BRCA1 mutation, 45-85% with a BRCA2 mutation)

but also pose an increased risk for the development of ovarian, pancreatic, and

prostate cancers17,18. Suspicion of BRCA1/2 mutations is confirmed by genetic

testing which can inform cancer screening and prevention strategies as well as

therapeutic options.

Unlike other cancers, breast cancer as a whole has few oncogenic driver

mutations. While 184 mutational drivers are reported in the IntOGen-mutations

database19,20 from six breast carcinoma projects21-26, the only three genes that

have more than a 10% incidence rate in all breast cancers are TP53, PI3K, and

GATA321. Instead, copy number variants (CNVs) and somatic copy number

aberrations (CNAs) are very common in this disease, with one study finding about

40% of genes in 2000 primary breast tumors exhibit altered expression due to

CNVs, single nucleotide polymorphisms, and CNAs27. However, when breast

cancers were subdivided based on their gene expression patterns, a number of

subtype-specific mutations were identified21. This highlights the importance of

regarding breast cancer as a collection of diseases that require unique treatment

strategies instead of a single uniform disease.

1.1.1 Breast cancer subtypes

Breast cancer is a collection of diverse tumors that are driven by their unique gene

expression profiles, or transcriptomes. Various classification systems have been

developed to subdivide breast tumors into distinguishable subtypes (Table 1.1), although most are not yet used in the clinic. Some of these subtyping systems,

including those based on receptor status and several molecular profiling

techniques, are discussed below. It is important to appropriately subdivide breast

cancer into distinct subtypes, because this allows us to better understand the

behavior of these diverse tumors. These subtyping approaches also aid in the

development of novel therapeutic strategies and in the selection of patients for

existing anti-cancer therapies.

1.1.1.1 Receptor status

In the clinic, besides taking into account tumor size, histologic grade, and lymph

node involvement, breast tumors are routinely profiled based on expression of the

hormone receptors (HR) estrogen and progesterone receptors (ER and PR) and

amplification/overexpression of HER2. ER and PR expression is determined using immunohistochemistry (IHC), and IHC or fluorescence in situ hybridization (FISH) are used to assess HER2 amplification. Targeted therapies have been developed against these receptors. Thus, the status of these three receptors heavily influences therapeutic decisions.

Hormone receptor-positive breast cancers

ER is a member of the nuclear receptor family of ligand-activated transcription factors that regulates the expression of genes involved in several biological processes, including proliferation, cell cycle regulation, and survival28. There are

two forms of ER that have similar sequences: ERα (encoded by ESR1) and ERβ

(encoded by ESR2). In canonical estrogen signaling, binding of estrogen to ER induces a conformational change that allows ER to dissociate from heat shock

proteins and dimerize29. The ER dimer translocates into the nucleus where it binds estrogen response elements (EREs) that are located upstream of estrogen-

dependent genes. In general, ER downregulates genes that encode transcriptional

repressors, promoters of apoptosis, and proliferation suppressors and upregulates

genes that promote proliferation and cell cycle progression30. Upregulation is

mediated by two transactivation domains, AF1 and AF2. AF1 is located in the N

terminus of ER, is regulated by phosphorylation, and acts independently of ligand

binding. AF2 is located in the ligand-binding domain and its activity is thus ligand-

dependent31,32. AF1 and AF2 function in a cell-type dependent manner: in most

cases, these two domains work synergistically, but in certain contexts either AF1

or AF2 alone controls activation of transcription of the target gene. AF1 and AF2

interact with a large number of coactivators which is important for the localization

of the transcriptional machinery to the target gene. Other mechanisms of ER

activity exist. Estrogen-bound ER can elicit its effects by binding other transcription

factors, such as activating protein 1, Sp-1, and NF-κB, instead of binding DNA

directly33-35. ER can also act independently of estrogen. Signaling molecules and

RTKs, including EGFR, HER2, and IGF1-R, can phosphorylate ER which allows

ER to dimerize and localize to the DNA where it elicits its effects on the transcription of its target genes36.

HR+ tumors comprise 75-80% of all breast cancers. HR+/HER2- breast cancer

incidence is highest among non-Hispanic white women1. These tumors tend to be

of lower grade and make up the least aggressive breast cancer subtype1. They have the best overall survival and a decreased incidence of metastasis compared

to the other subtypes37,38. PI3K mutations are the most frequent mutations in ER+

breast tumors21. Patients with HR+ breast cancer are primarily treated with endocrine therapy, which has provided enormous clinical benefit for this set of

tumors.

HER2-positive breast cancers

HER2, also known as ErbB2 or neu (in rodents), is a 185 kDa protein that belongs

to the ErbB/HER family of transmembrane glycoproteins. There are three

additional members of this growth factor receptor family: EGFR/ErbB1/HER1,

ErbB3/HER3, and ErbB4/HER4. ErbB proteins have a high degree of sequence

homology and have a similar structure: an extracellular ligand binding domain, a

transmembrane domain, and an intracellular tyrosine kinase domain39. Typically,

binding of a ligand to an ErbB protein leads to homo- or hetero-dimerization with

other ErbB receptors. HER2, the preferred binding partner of all three of the other

ErbB proteins, is an exception as it has no endogenous ligand and is activated

following heterodimerization40. Once dimerization occurs, the ErbB receptors are

internalized and/or phosphorylated at tyrosine residues in the intracellular domain.

Phosphorylation of these proteins allows for interaction with adaptor proteins which

activates several signaling cascades, most notably the PI3K/AKT, Ras/MAPK, and

JAK/STAT pathways, impacting proliferation, adhesion, motility, survival,

angiogenesis, and invasion41.

While all four ErbB receptors as well as many of their ligands are overexpressed

in breast cancer, amplification of HER2 is the most common, occurring in 15-30%

of breast cancers42. HER2+/ER- tumors are more sensitive to cytotoxic

chemotherapies compared to ER+ breast cancers, as significantly more

HER2+/ER- tumors achieve a pathological complete response (pCR)43.

Overexpression of HER2 is associated with poor prognosis, although these tumors

are treatable using HER2-targeting agents42. HER2 dimerization with HER3, which

is the strongest ErbB activator of the PI3K signaling pathway, has been shown to

be particularly important for the development of breast cancer and maintenance of

breast cancer cell proliferation44-46. HER2 expression influences several additional oncogenic phenotypes. Aside from increased PI3K signaling, the anti-apoptotic

proteins NF-κB and survivin are also regulated by HER247,48. In addition, increased

HER2 activity is linked to elevated levels of HIF-1α and VEGFA, factors critical for

the hypoxic response and angiogenesis, respectively49-51.

Although all tumors within this subtype have high expression of HER2,

heterogeneity exists within this group of breast cancers. For example, they have

different patterns of gene expression, DNA methylation, and copy number

alterations52,53. In addition, two groups have found that the HER2+ subtype is

actually an amalgam of at least two types that can be further subdivided based on

gene expression and pathway signatures54-56. These subtypes have distinct clinical

outcomes and, because they have different gene expression patterns, may

respond differently to treatment.

Triple-negative breast cancers

TNBCs comprise about 10-20% of all breast cancers, and women who are young,

African American, or Hispanic are more likely to be diagnosed with TNBC

compared to all other age and racial/ethnic groups57-60. TNBC tumors respond very

well to cytotoxic chemotherapies (including taxanes and anthracyclines),

especially compared to ER+ tumors, and up to 85% of TNBC patients achieve

some level of initial clinical response43,61,62. However, TNBC rapidly recurs due to

the presence of residual disease, and, compared to the other subtypes of breast

cancer, TNBC has the worst prognosis and the lowest distance metastasis-free

rate during the first three years following diagnosis63-65. However, after three years, survival rates begin to improve and eventually are similar to survival rates of ER+ breast cancers64. Women diagnosed with metastatic TNBC have particularly poor

prognosis, and most succumb to their disease66.

TNBC tumors tend to have a high histological grade, are highly proliferative, are

larger, and are more likely to be aneuploid compared to HR+ and HER2+

tumors64,65. EGFR, the expression of which is linked with low HR levels, high

proliferation, large tumor size, poor differentiation, high relapse rates, and worse

prognosis67-69, is highly expressed in roughly half of all TNBC tumors65. Tumors

with BRCA1 or BRCA2 mutations are also significantly more likely to be triple-

negative70. Other frequent alterations that occur in TNBC include the activation of

the RAS-RAF-MEK and PI3K pathways, mutation of TP53, activation of MYC, loss

of retinoblastoma protein (RB1), expression of programmed death ligand 1 (PD-

L1), and enrichment of the androgen receptor (AR)21,71-76. Despite these

commonalities, TNBC tumors are quite diverse and can be further subdivided by

gene expression profiling, as is illustrated below.

1.1.1.2 Intrinsic subtypes of breast cancer

While not currently used to determine treatment options for patients, breast

cancers can be classified based on molecular profiling. This has led to the

definition of the intrinsic subtypes of breast cancer, which is one of the most

frequently used classification systems for defining breast cancers in the research

setting. Unsupervised clustering analysis of global gene expression patterns of

breast tumors identified five intrinsic breast cancer subtypes (luminal A, luminal B,

HER2-enriched, basal-like, and claudin-low) and a normal-like group21,77-81. A 50 gene set (PAM50) can identify these subtypes in breast cancer tissue samples82.

It is important to note that these subtypes are not interchangeable with the

receptor-based subtypes (Figure 1.1)78,83. For example, based on the UNC337 and

MDACC133 datasets, the majority of triple-negative tumors classified as either

basal-like or claudin-low, but 27% were divided among the remaining intrinsic subtypes79,84.

Luminal A and B breast cancers

Because the intrinsic breast cancer subtypes are controlled by unique

transcriptomes, they have different defining characteristics and clinical features, including incidence, prognosis, response to therapy, and patterns of metastasis

(Table 1.2). The luminal subtypes are the most commonly diagnosed breast cancers21. The majority of luminal A and B tumors express ER and low molecular

weight keratins, and some (6-8% of luminal A and 16-21% of luminal B tumors)

have HER2 amplification85. Luminal A and B tumors can be distinguished by

expression of proliferation/cell cycle-related genes and proteins, such as MKI67

and AURKA (higher in luminal B tumors), and luminal-related genes and proteins,

including PR and FOXA1 (lower in luminal B tumors)85. Reflecting differences in

proliferation rates, luminal A tumors tend to be of lower histological grade

compared to luminal B tumors85. Luminal A tumors also have lower overall

mutation rates (including TP53 mutation) and CNVs, although they have a higher

incidence of PIK3CA and MAP3K1 mutations compared to luminal B tumors21.

While luminal B tumors are more sensitive to cytotoxic chemotherapies than

luminal A cancers, both tumor types are less responsive to this line of treatment

compared to the HER2-enriched and basal-like subtypes62,86-89.

Luminal A and luminal B breast cancers also differ in terms of prognosis. The luminal A subtype has the best prognosis of all of the intrinsic breast cancer

subtypes and has better overall and distant recurrence-free survival compared to

luminal B breast cancers90-95. Luminal B tumors are more aggressive, have a higher metastatic rate, and are more likely to relapse despite treatment with

tamoxifen or the aromatase inhibitor anastrozole96-99. In addition, by ten years

post-diagnosis, overall survival of patients with luminal B tumors is similar to that

of basal-like breast cancers, which have the poorest prognosis especially within

the first five years90,94. It is therefore critical to develop approaches to accurately

identify luminal B tumors at the time of diagnosis in order to provide appropriate treatment to these patients and thus improve therapeutic outcomes. It has been

proposed that using PR activity by measuring the expression of PR-target genes may more accurately distinguish between luminal A and B subtypes compared to

IHC or RT-PCR measurements of PR expression, but this has yet to be used in

the clinic100.

HER2-enriched breast cancers

HER2-enriched breast cancers are defined by the overexpression of genes within

the HER2 amplicon at 17q22.24, including HER2 and GRB7. These tumors also

have a high number of mutations, including those in TP53, PIK3CA, and

APOBEC3B. They also have intermediate expression of luminal-related genes,

and express low levels of basal-like genes101,102. Tumors within this subtype are

highly proliferative, tend to be of high grade, are likely to have metastasized to the

lymph nodes by the time of diagnosis, and have poor prognosis. Despite their

classification, about 30% of HER2-enriched tumors are HER2-78. While these

tumors may not overexpress HER2, they could instead be driven by a mutation in

the HER2-encoding gene (ERBB2) or by amplification, mutation, or activation of a

downstream member of a HER2-regulated pathway.

Basal-like breast cancers

Basal-like breast cancers compose 11-23% of all breast cancers, express basal

cytokeratins 5/6, 14, and/or 17, and have high expression of proliferation genes,

including MKI6778,103. These tumors also express higher levels of PD-L1 compared

to other subtypes of breast cancer104. Basal-like tumors have poor prognosis and

frequently metastasize, often within the first five years following diagnosis105-107.

Certain mutations and signaling pathway alterations are characteristic of basal-like

breast cancers. For example, tumors carrying a BRCA1 mutation are much more

likely to categorize as basal-like compared to the other intrinsic subtypes105,108,109.

Another commonly mutated gene is TP53, which occurs in 62-80% of basal-like

tumors21,25,83. In addition, several members of the PI3K and RAS-RAF-MEK

pathway are frequently altered, including deletion of PTEN and INPP4B and

amplification of KRAS, PIK3CA, and EGFR21. Activation of MYC, loss of RB1, and enrichment of AR are also common21,110-112. Interestingly, basal-like breast cancers

are distinct from other breast cancers and have been found to be more similar to

squamous cell lung cancer, high grade serous ovarian carcinoma, and a subtype

of bladder cancer compared to other breast cancer subtypes21,113,114.

The terms triple-negative and basal-like are often used interchangeably in the literature. 75-85% of TNBC tumors classify as basal-like115. Similar to triple-

negative tumors, basal-like breast cancers tend to impact younger women and

have a high histological grade, central necrosis, a high mitotic index, and a pushing

border of invasion78,106,116. However, it is important to note that not all basal-like tumors are triple-negative, and 15-45% of basal-like tumors have been found to express ER and/or HER2101,115,117. It is not currently possible to distinguish

between TNBC and true basal-like breast cancer at the time of diagnosis. Cheang,

et al. has proposed a method to replace the currently used assessment of

ER/PR/HER2 expression with IHC analysis of ER, PR, HER2, EGFR, and

cytokeratin 5/6118. However, this is not yet utilized in the clinic.

Claudin-low breast cancers

The claudin-low subtype is the least prevalent subtype of breast cancer, occurring

in 7-14% of breast cancer patients78. Claudin-low tumors are characterized by low expression of luminal differentiation markers and cell-cell adhesion proteins (e.g. occludin, E-cadherin, and claudins 3, 4, and 7), enrichment of EMT genes, presence of infiltrating immune cells, and stem-cell-like features79,81,119-121. About

20% of claudin-low tumors possess mutated BRCA178,80. Claudin-low breast

cancers share some characteristics with basal-like tumors; the majority of claudin-

low breast cancers are high grade and triple-negative, although 15-25% express

ER and/or PR78. However, compared to basal-like cancers, claudin-low cells

express low levels of proliferation genes, including MKI67, suggesting these

tumors may be slow-cycling79. In addition, while basal-like tumors respond well to

chemotherapy, claudin-low tumors have an intermediate response and have poor

prognosis79.

Although the intrinsic subtypes of breast cancer are not currently used to classify

tumors in the clinic, they have provided important information regarding the

characteristics of diverse breast cancer tumors and potential therapeutic

vulnerabilities. Numerous other breast cancer subtypes have also been identified

by gene expression, copy number, or pathway signature profiling. While they are

not as widely used as receptor status or the intrinsic subtypes, these additional

breast cancer subgroups provide information that can improve our understanding

of this diverse collection of diseases. They also highlight the heterogeneity that

exists within the more established breast cancer subtypes. A few examples of

these additional molecular subtypes are given below.

1.1.1.3 Additional breast cancer classification systems

To stratify breast cancers into subgroups and identify their unique drivers, Curtis,

et al. combined copy number and gene expression profiles of 997 breast cancers27.

This led to the definition of 10 integrative clusters (IntClust1-10) that had distinct

clinical outcomes and gene expression profiles. For example, IntClust2 was

composed of ER+ tumors that expressed the 11q13/14 amplicon which includes

CCND1 and PAK1, genes known to play a role in breast and ovarian cancers. This subtype is associated with poor survival despite the expression of ER. IntClust5 was comprised of both HER2-enriched tumors and luminal tumors with HER2 amplification, indicating these tumors should be responsive to HER2-targeted therapies. Most basal-like tumors were members of IntClust10, and these tumors exhibited altered expression of several genes involved in cell cycle progression and mitosis, including AURKB, FOXM1, CDC20, and KIF2C. Despite poor initial survival, tumors within this subtype had relatively good overall prognosis. Thus,

integrating copy number data with transcriptome information can distinguish

between diverse breast tumors and identify unique drivers. Moreover, inhibitors

exist for some of the protein products of genes identified within these clusters,

suggesting this subtyping method can inform therapeutic decisions.

Another method used to identify breast cancer subtypes is based on expression of pathway signatures56. This technique resulted in the classification of 17 breast

cancer subtypes (subgroups 1-17). The number of subtypes identified highlights

the heterogeneity of not only breast cancer as a whole but also the diversity that

exists within the intrinsic subtypes. In general, each intrinsic subtype was

subdivided into several pathway-defined subgroups (Figure 1.2). The individual

subgroups had unique CNVs, biological features, and prognoses, and this was true

even for the subtypes that contained tumors belonging to the same intrinsic

subtype of breast cancer. For example, despite the association of the basal-like

subtype with poor prognosis, the subgroups that contained basal-like tumors

(subgroups 2, 5, and 8) had different survival rates: subgroup 8 was associated

with good overall survival while subgroups 2 and 5 had very poor outcomes. The

differences observed in clinical outcome reflected variations in pathway activities.

While subgroup 8 tumors expressed high levels of EGFR and had low SRC activity,

the opposite was true for subgroups 2 and 5, suggesting subgroup 8 tumors will

have different responses to EGFR- or SRC-targeting agents compared to tumors

within subgroups 2 and 5. Similar differences were seen among the subgroups

containing luminal tumors. Thus, distinct breast cancer subtypes can be defined

by pathway activity, and this method of tumor characterization can be used to choose an appropriate therapeutic strategy.

Molecular profiling has also produced subgroups specifically within the TNBC subtype, reflecting the diversity of tumors in this category of breast cancers.

Analysis of gene expression profiles of 587 TNBC tumors by Lehmann et al.

revealed six TNBC subtypes (TNBCtypes): basal-like 1 (BL1), basal-like 2 (BL2),

immunomodulatory (IM), mesenchymal (M), mesenchymal stem-like (MSL), and

luminal androgen receptor (LAR)122. These tumors had distinct features and

different responses to anthracycline and taxane neoadjuvant treatment, with BL1

tumors having the highest rate of pCR123. When compared to the intrinsic breast

cancer subtypes, most of the tumors within the BL1, BL2, IM, and M TNBCtype

subtypes classified as basal-like124. About half of MSL tumors categorized as

basal-like while the remaining were divided among the luminal B (14%) and

normal-like (28%) subgroups. Tumors within the LAR subtype mostly classified as

either HER2-enriched (74%) or luminal B (14%). This classification system has

since been updated due to the discovery that the IM and MSL subtype transcriptomes were heavily influenced by infiltrating leukocytes and adjacent stromal cells, respectively125.

Similar to the original TNBCtypes, the four remaining subtypes (BL1, BL2, M, and

LAR), referred to as TNBCtype-4, had unique clinical characteristics. The BL1

subtype, the largest TNBCtype-4 subgroup, contained tumors expressing high

levels of cell cycle and DNA damage response signatures. Patients with BL1

tumors had the best overall and relapse-free survival and had a higher incidence

of pCR compared to the other TNBCtype-4 subgroups. BL2 tumors were enriched

in growth factor signaling and myoepithelial markers, and patients with these tumors had a lower frequency of pCR. Tumors within the M subtype had high expression of differentiation and growth factor pathway genes and were sensitive to PI3K/mTOR inhibitors and the Src inhibitor dasatinib. 21% of patients with M tumors had lymph node metastasis, and these tumors had a higher propensity to metastasize to the lungs compared to other subtypes. Finally, patients diagnosed with LAR tumors tended to be older, and roughly half presented with lymph node metastasis. LAR tumors were much more likely to metastasize to the bone, with

46% of LAR patients experiencing metastasis to this site compared to 16% for the other subtypes. There was a higher incidence of PIK3CA mutation in LAR tumors compared to tumors within the other triple-negative subtypes126. LAR tumors

expressed AR and are thus sensitive to AR antagonists.

Similarly, Burstein et al. performed mRNA and DNA profiling on 198 TNBC tumors

and identified four subtypes: luminal-AR (LAR), mesenchymal (MES), basal-like

immune suppressed (BLIS), and basal-like immune-activated (BLIA)127. These subtypes had unique gene expression profiles, CNVs, and clinical outcomes.

Importantly, the authors were able to identify distinct expression patterns within

each subtype that are targetable by already approved therapies. This study and

the definition of TNBCtype-4 subtypes illustrate the heterogeneity within TNBC and

the potential utility of these subtypes in identifying beneficial therapeutic strategies

for these tumors.

1.1.2 Therapeutic options for the treatment of breast cancer

Receptor status is frequently used to inform breast cancer therapy, and commonly

prescribed agents for each subtype are outlined in Table 1.3. Patients whose

tumors have at least 1% ER expression are treated with endocrine therapy.

Targeting the oncogenic effects of ER can be accomplished by reducing estrogen levels (aromatase inhibitors: anastrozole, letrozole, and exemestane), suppressing

ER activity (selective estrogen-receptor modulators: tamoxifen and raloxifene), or degrading ER (selective estrogen receptor degraders: fulvestrant). Besides endocrine therapies, several other agents targeting pathways frequently dysregulated in HR+ breast cancers have been developed. For example, numerous drugs have been designed to target the PI3K/AKT/mTOR pathway which is involved in several cellular processes critical for breast cancer growth and survival and is commonly mutated in ER+ breast cancers128,129. Two ongoing

phase III trials are studying the combination of a pan-class I PI3K inhibitor

(buparlisib/BKM120) with fulvestrant in women with HR+/HER2- breast cancer that

progressed or recurred following treatment with an aromatase inhibitor

(NCT01610284) or mTOR inhibitor (NCT01633060). In addition, the mTOR inhibitor everolimus is approved to treat HR+/HER2- breast cancer in combination with exemestane, and additional combination therapies utilizing everolimus are currently being studied in several clinical trials. Other common therapeutic targets for ER+ breast cancers are proteins that control cell cycle progression, such as

CDK4 and CDK6. These proteins form a complex with cyclin D, and this complex phosphorylates Rb, which is involved in the G1 to S phase transition130. This pathway is frequently deregulated in breast cancer, which inactivates the G1-S phase checkpoint and drives uncontrolled proliferation. Three CDK4/6 inhibitors, palbociclib, abemaciclib, and ribociclib, are approved for use in advanced or metastatic HR+/HER2- breast cancers and are currently being evaluated in numerous clinical trials both as monotherapies and in combination with other agents.

Several therapeutics have been developed to target tumors dependent on HER2 amplification/overexpression. The first to be generated was trastuzumab, a humanized IgG1 monocloncal antibody that targets the extracellular domain of

HER2. Use of trastuzumab in the clinic has led to substantial improvements in both recurrence-free and overall survival of patients with early and advanced HER2+ breast cancer, and trastuzumab can be given either alone or in combination with other agents. Another HER2-targeted therapy is pertuzumab, a humanized IgG1 monoclonal antibody against the extracellular dimerization domain of HER2, which prevents dimerization with HER3 and subsequent activation of the receptor.

Pertuzumab has been approved for use with trastuzumab or docetaxel in the

neoadjuvant or metastatic setting. Small molecule inhibitors of HER2 have also

been developed. One example is lapatinib, which targets the kinase domain of

both EGFR and HER2, reversibly binding the ATP-binding pocket. This prevents

autophosphorylation of the receptors. Lapatinib is approved for use in combination

with the antimetabolite capecitabine or, in HR+/HER2+ breast cancers, with

letrozole. Lastly, the antibody-drug conjugate trastuzumab emtansine (T-DM1) is used to treat women with metastatic HER2+ breast cancer and is composed of trastuzumab covalently bound to the cytotoxic agent emtansine (DM1) via a non- reducible linker131. This design effectively targets a highly potent chemotherapy

drug to HER2+ tumors where the drug enters the cell by endocytosis and is

delivered to the lysosome. There, lysosomal processing releases active DM1

which binds tubulin and prevents microtubule polymerization, inhibiting mitosis and

inducing apoptosis132,133.

As previously stated, TNBC tumors lack expression of ER, PR, and HER2 and thus

are not affected by endocrine or anti-HER2 therapies. Instead, cytotoxic

chemotherapy, such as taxanes and anthracyclines, is the mainstay of treatment

for triple-negative tumors. TNBCs respond better to these drugs than tumors that

express ER89. The taxanes paclitaxel and docetaxel are widely used

chemotherapy agents that prevent microtubule depolymerization by stabilizing

GDP-bound tubulin, thus inhibiting mitosis. Anthracyclines (e.g. doxorubicin) are

some of the most effective anti-cancer drugs and have several mechanisms of

action, including 1) production of free radicals that induce DNA damage and disrupt

cell membranes via lipid peroxidation, 2) induction of DNA double-strand breaks

by inhibiting topoisomerase II, 3) eviction of histones which reduces the response

to DNA damage and disrupts the epigenome and transcriptome, and 4)

suppression of DNA synthesis by intercalating into the DNA134,135. No single

therapeutic strategy using cytotoxic chemotherapies has been identified as being superior for neoadjuvant or adjuvant treatment of TNBC, although TNBC patients particularly benefit from dose-dense and high-dose treatments136-138. Because these types of therapies target all rapidly dividing cells and are not selective for cancer cells, toxic side effects, which can be severe, are extremely common. This can lead to termination of treatment and even death. Another drawback of this line

of therapy is that less than half (about 30-40%) of TNBC patients achieve pCR,

leaving a large number of patients with residual disease, which leads to

recurrence62. It is thus important to identify treatment strategies that target

alterations unique to cancer cells to prevent disease recurrence in these patients.

The vast majority (90%) of residual tumors following neoadjuvant chemotherapy

contain altered pathways that are targetable by agents already in clinical trials,

such as inhibitors of PARP, PI3K, MEK, heat shock protein 90 (HSP 90), VEGF,

HIF-1α, and histone deacetylase (HDAC)139,140. This makes it possible for the rapid

assessment of these agents in patients with TNBC, and a wide variety of therapies

that capitalize on altered expression of targetable proteins in TNBC tumors are

under investigation. For example, TNBC tumors express higher levels of PD-L1, making them a prime candidate for treatment with anti-PD-1 and anti-PD-L1 antibodies71,72,104. This treatment strategy is currently being assessed in clinical trials in combination with chemotherapy (NCT02489448, NCT02530489, and

NCT02620280). Some of these treatment options are discussed here. However,

their utility in TNBC will be dependent on the proper identification of patients who

will benefit from their use.

Inhibitors of poly(ADP-ribose) polymerase (PARP), a nuclear enzyme that

activates DNA damage repair, are being explored as a therapeutic option for breast

tumors with BRCA1/2 mutations, the majority of which classify as basal-like or

claudin-low. PARP binds DNA in response to single-strand breaks and initiates

DNA damage repair by base excision repair141. Inhibition of PARP causes single- strand breaks to become double-strand breaks due to collapsed replication forks during DNA synthesis, and these breaks are then repaired via homologous recombination142,143. The use of PARP inhibitors against tumors with BRCA1/2

mutations induces synthetic lethality. Tumor cells with BRCA1/2 mutations are unable to perform DNA damage repair by homologous recombination, forcing cells

to use the more error-prone non-homologous end joining (NHEJ) instead144. The treatment of homologous recombination-deficient tumors with PARP inhibitors that suppress DNA damage repair pathways induces genomic instability, growth inhibition, and cell death143-145. The PARP inhibitors olaparib, rucaparib, and

niraparib have been approved for use in ovarian, fallopian tube, and peritoneal

cancers. These and other PARP inhibitors have performed well when used in breast cancer patients with BRCA1/2 mutations as a monotherapy, and several

clinical trials determining the efficacy of PARP inhibitors in breast cancer are

ongoing146-148.

A subset of TNBC tumors express AR, providing another potential therapeutic

target for this disease. Preclinical studies have revealed that targeting AR

suppresses growth of LAR cells in vitro and in vivo122, and clinical trials have shown

some clinical benefit140,149. LAR breast cancer cells are also sensitive to PI3K-

targeting agents, due to the frequent rate of activating mutations in PIK3CA in

these cells122,126,150. The efficacy and toxicity of the combination of taselib (PI3K

inhibitor) and enzalutamide (AR inhibitor) is currently being assessed in AR+

TNBC patients (NCT02457910). In addition, while most TNBC tumors are

unresponsive to CDK4/6 inhibitors, LAR cells are highly sensitive to this line of

therapy151,152. An ongoing phase II clinical trial is evaluating combining the antiandrogen bicalutamide with ribociclib in AR+ TNBC patients (NCT03090165).

EGFR overexpression is observed in all subtypes of breast cancer but most frequently occurs in TNBC and inflammatory breast cancer153-155. It is thus a prime therapeutic target for these diseases. Inhibition of EGFR activity can be achieved via small molecule tyrosine kinase inhibitors or monoclonal antibodies. Due to mostly poor response rates in numerous clinical trials, EGFR-targeting agents have not been approved for use in TNBC156. However, a subset of patients did

respond to EGFR inhibition, highlighting the need to develop a method to identify

probable responders and to create more effective therapeutic strategies for these

patients.

It has been suggested that therapies that are effective in high grade serous ovarian

carcinoma (HGSOC) may also be effective in TNBC, because these diseases are

molecularly similar21. Platinum-containing agents, such as cisplatin and

carboplatin, are commonly used to treat HGSOC and have been approved for use

in several cancers including ovarian, testicular, and bladder cancers. These drugs

crosslink DNA, inducing DNA damage followed by apoptosis. The efficacy of

platinum-containing agents in TNBC has been evaluated in several clinical trials.

In the neoadjuvant setting, platinum salts alone and in combination with paclitaxel

or anthracycline/taxane/bevacizumab (anti-VEGF antibody) were effective,

achieving good clinical response rates157-159. However, increased toxicity was

observed when combining platinum agents with other chemotherapies, with as many as 48% of patients having to discontinue treatment158,159. Functional

BRCA1/2 is necessary for proper DNA damage repair following treatment with

platinum salts, indicating BRCA1/2 mutations can confer additional sensitivity to

these drugs160,161. Indeed, evidence from clinical trials in both the metastatic and

neoadjuvant settings supports these findings157,162-164. TNBC patients with

BRCA1/2 mutations therefore may particularly benefit from therapies containing

platinum-based drugs.

An alternative strategy to identifying targetable pathways that are altered in

individual tumors is to develop therapies that can target multiple critical pathways

at once. This can be accomplished by using epigenome-modulating drugs.

Epigenetic modification proteins represent attractive therapeutic targets, because

during carcinogenesis, extensive restructuring of the epigenome occurs, including

aberrant acetylation, alteration of methylation patterns, and accumulation of

epigenetic readers at oncogenes. As epigenetic alterations are reversible, epigenome-targeting agents could provide a mechanism to silence numerous

oncogenes simultaneously. The focus of this thesis is the use of inhibitors of the

Bromodomain and Extraterminal (BET) family of epigenetic readers in TNBC. As

discussed in the subsequent chapters, I found BET inhibitors suppress growth and

induce apoptosis and senescence in diverse models of TNBC in vitro and in vivo.

These findings suggest BET inhibitors could be effective across multiple TNBC

subtypes, as opposed to other therapies that are only effective in tumors with

specific genetic alterations.

1.1.3 Resistance to cytotoxic chemotherapy

Due to advances in radiotherapy and surgical procedures and the development

and widespread use of targeted therapies, there has been a decrease in local

recurrence in breast cancer during the last several decades. Recurrence rates vary by subtype. For example, unlike HR- breast cancers which tend to relapse within the first five years, a large number of recurrences of HR+ breast cancers occur later, with the risk of relapse remaining unchanged even after 10 years post- diagnosis165. A major issue impacting all subtypes of breast cancer is the

development of therapeutic resistance, which limits clinical response. It is thus

necessary to develop treatment strategies that can prevent resistance as well as

targeted therapies for drug resistant breast cancers to improve patient outcomes.

There are two types of therapeutic resistance. First, cells within a tumor can be

intrinsically resistant to the therapy. Thus, while the bulk of the tumor may respond, residual disease will lead to recurrence. Second, tumor cells can acquire resistance which decreases sensitivity to the therapeutic agent. It is thought that using combination therapies or dual targeting agents as opposed to a monotherapy

can more effectively silence oncogenic signaling and inhibit activation of

compensatory pathways, thus preventing acquired resistance.

A variety of molecular mechanisms have been identified that contribute to drug

resistance. Besides alterations in drug uptake and export by tumor cells, resistance

mechanisms generally fall into three categories: alterations of the drug target,

mutations or changes in expression of factors upstream or downstream of the drug

target, and activation of compensatory signaling pathways. As an example, some

of the common resistance mechanisms specific to traditional chemotherapy are

reviewed here.

Chemotherapy resistance is a major obstacle in the treatment of breast cancer.

Once resistance to taxanes and anthracyclines occurs, few additional treatment

options exist, especially for TNBC patients. It is thus critical to understand the

mechanisms of resistance to these agents in order to define drivers of resistance.

This will aid in the identification of patients who are likely to develop resistance and

the creation of targeted therapies to inhibit or overcome resistance.

Compared to normal breast tissue, breast cancers express higher levels of

P-glycoprotein (Pgp), an ATP binding cassette (ABC) transporter that functions as

a drug efflux pump. Pgp expression is also elevated in treated versus non-treated

breast tumors166. Both taxanes and anthracyclines are Pgp substrates. Thus,

overexpression of Pgp reduces the concentration of these drugs within the cell,

limiting their effects167. Resistance also occurs when cells fail to undergo apoptosis

following the addition of cytotoxic agents. For example, p53 mutations, which are common in breast cancer, are linked to intrinsic resistance to doxorubicin, and

acquired doxorubicin resistance is associated with the downregulation of the CDK

inhibitor p27168,169. In addition, taxane-mediated cytotoxicity depends on the

balance between anti- and pro-apoptotic proteins. Overexpression of the anti-

apoptotic proteins Bcl-xL and Bcl-2 has been linked to taxane resistance, while

sensitivity to taxanes is associated with overexpression of the pro-apoptotic

proteins Bax and Bad170-173. This suggests that drugs that target the Bcl-2 family may reverse resistance to taxanes.

1.1.3.1 Taxane-specific resistance mechanisms

Several mechanisms of taxane resistance involve microtubule alterations, including changes in the expression patterns of α- and β-tubulin, increased expression of α-tubulin, expression of α-tubulin variants, high expression of microtubule associated proteins (MAPs), and β-tubulin mutations that prevent taxane binding167,174-176. Disruption of the spindle assembly checkpoint (SAC)

during mitosis can also confer taxane resistance. In actively dividing cells, taxanes

stabilize microtubules, and signaling from resulting unattached kinetochores

activates the SAC. This prevents chromosome separation and induces mitotic

arrest177. Cells that cannot adapt undergo apoptosis in mitosis or soon after mitotic

exit. Several defects in the SAC that correlate with resistance to taxanes have

been identified in preclinical studies. For example, MAD2 and BUBR1 are

checkpoint proteins that prevent the metaphase-anaphase transition by inhibiting

the degradation of cyclin B. Downregulation of the genes encoding MAD2 and

BUBR1, either directly or through the suppression of BRCA1, induces premature

transition into anaphase and taxane resistance178,179. In addition, stathmin, a

protein that destabilizes microtubules, is overexpressed in breast cancer180.

Increased stathmin expression induces taxane resistance by decreasing the

binding of taxanes to the microtubules and inhibiting the transition from G2 to M

phase, which prevents the mitosis-associated cytotoxic effects of drugs within this

class181. Finally, CDK2 activity is necessary for progression into mitosis and proper

function of the SAC, both of which are essential for sensitivity to taxanes182. This suggests that co-treatment with a taxane and a CDK2 inhibitor should be avoided because the CDK2 inhibitor could augment taxane resistance.

Besides proteins involved in the SAC, other critical mediators of mitosis, including

CDK1 and Aurora kinase A, are dysregulated in cancer and can lead to taxane resistance by suppressing apoptosis and promoting mitotic exit183,184. Defining the

role of mitosis regulators in the acquisition of taxane resistance should provide

potential drug targets to prevent or overcome resistance. Indeed, inhibitors have

been developed against Aurora kinase A, which is overexpressed in breast tumors,

and these agents synergize with taxanes in several types of cancer, including

TNBC185-188. It will be important in the future to determine if this drug combination

can also delay or prevent taxane resistance.

1.1.3.2 Anthracycline-specific resistance mechanisms

Resistance to anthracyclines can occur via several mechanisms, such as

alterations in DNA damage repair pathways. For example decreased expression

levels or activity of topoisomerase II can confer resistance to this drug class189. In

addition, anthracycline-induced DNA damage activates the DNA mismatch repair

pathway. Reduced activity of this pathway through the loss of function of pathway

proteins prevents the signaling cascades that promote apoptosis, causing

resistance190,191. Anthracyclines also produce reactive oxygen species (ROS), and

increased expression of the antioxidant enzymes glutathione S-transferase P1,

superoxide dismutase, and catalase reduce the cytotoxicity of

anthracyclines192,193. Thus, inhibition of these enzymes may improve clinical

outcomes to anthracycline treatment.

1.2 Basics of transcriptional control of genes

In response to an outside or internal stimulus, a number of factors, including

transcription factors, co-activator complexes, chromatin remodelers, and histone

modifying proteins, rapidly localize to target genes and stimulate the initiation of

transcription. The binding and disassociation of transcription-related proteins to

regulatory elements of target genes is highly coordinated, and these mechanisms

are frequently altered in cancer. As such, transcription modulators are attractive

targets for anti-cancer therapy. As mentioned previously, breast cancer subtypes

are defined by unique sets of gene expression patterns. These transcriptomes are

maintained by the epigenome, the components of which are described in the next

section. Here, I provide an overview of mRNA transcription and the proteins and

regulatory elements involved in this tightly-controlled process.

1.2.1 Transcriptional regulatory elements

Cis-acting transcriptional regulatory elements can be found upstream and

downstream of the open reading frame. These elements are responsible for the

proper expression of their associated gene and can be divided into two groups: 1)

core promoter and promoter-proximal elements and 2) distal elements including

enhancers, silencers, insulators, and locus control regions194.

The core promoter is located at the beginning of the gene and is the region where general transcription factors (GTFs) and RNA polymerase II bind to form the preinitiation complex (PIC). It also determines the direction of transcription. The

core promoter contains the transcriptional start site (TSS) and extends upstream

or downstream of this site by about 35 nucleotides. Several core promoter

elements have been identified, the most common of which are the TATA box, the

initiator element, the TFIIB recognition element, and the downstream promoter

element194. Despite what was initially thought, these elements are not present in

all core promoters. For example, the TATA box is found in only about 24% of

human genes195. It is possible that the diversity in core promoter composition plays

a role in transcription regulation, as transcription factors and enhancers display

specificity for certain core promoters196.

Promoter-proximal elements are found within a few hundred nucleotides of the

TSS and contain activator binding sites194,196. In addition, CpG islands, stretches

of DNA with high GC content, commonly occur near promoters197. CpG dinucleotides can be methylated, and this is associated with transcriptional repression. However, CpG islands are usually unmethylated, suggesting the presence of these elements prevent gene silencing198.

Enhancers can be located upstream, downstream, or within a gene and can be

between hundreds and millions of base pairs away from the promoter199. These

regions contain binding sites for transcription factors that work cooperatively to

influence the rate of transcription initiation. DNA looping helps to bring enhancers

close to promoters and delivers proteins bound to enhancers to the promoter

where they can influence multiple steps of transcription initiation200. The protein

complex Mediator can also form a bridge between enhancer-bound proteins and

proteins at the promoter and recruit transcription factors to these loci201.

Enhancers exist in four states which can be identified by different epigenetic

marks202. The repressive H3K27me3 mark is found at inactive enhancers, which

lack transcription factor binding and are found within areas of condensed

chromatin. Primed enhancers lie within areas of open, nucleosome-free chromatin

and are bound by transcription factors but require more steps, such as the

recruitment of additional transcription factors and co-activators, to become fully

active. These enhancers are marked with H3K4me1 and H3K4me2. A subset of

primed enhancers is poised enhancers, which are common features of genes

during embryonic development and are acquired by progenitor cells203

Similar to active enhancers, poised enhancers are enriched for H3K4me1, have

low nucleosomal density, and are bound by transcription factors and co-

activators204,205. However, unlike active enhancers, poised enhancers cannot drive transcription. Active and poised enhancers can be distinguished by their histone modifications: poised enhancers are characterized by H3K9me3 and H3K27me3 while active enhancers are marked by H3K27ac206-208. Poised enhancers are

common features of genes during embryonic development and are acquired by

progenitor cells203. The shared properties of poised and active enhancers allow the

quick conversion of a poised enhancer to an active enhancer to initiate

transcription of the target gene in response to a stimulus.

Silencers share similar properties with enhancers but function to repress

transcription of the target gene194. They can exist on their own or as parts of

enhancers or proximal promoters and contain binding sites for repressor proteins.

Another type of regulatory element, the insulator, prevents its target gene from

being influenced by the transcriptional regulatory elements of nearby genes209.

This is accomplished by preventing the interaction between enhancers and

promoters and by preventing the spread of heterochromatin. In vertebrates,

insulators contain a binding site for CTCF, suggesting this protein plays a role in

directional enhancer blocking210.

Finally, locus control regions (LCRs) contain multiple cis-acting regulatory

elements, including enhancers, silencers, and insulators, that strongly enhance

transcription of a gene cluster211. They are usually located near DNase I

hypersensitivity sites which provides areas of open chromatin for the target genes.

Transcription factors, activators/repressors, co-activators/co-repressors, and

chromatin modifying enzymes localize to an LCR, and, while each of these proteins

influences gene expression, it is their combined effect that defines the activity of

the LCR and the spatial/temporal expression of the target genes. Similar to

enhancers and silencers, LCRs can be located far away from their target genes.

DNA looping has been proposed as a potential mechanism that could explain how

LCRs can elicit long-range control of transcription of their target genes194.

A wide variety of proteins recognize specific DNA sequences within the regulatory

elements described above. The binding and dissociation of these proteins control

the timing and execution of events during transcription. Some of these

transcription-related proteins are described in the following two sections.

1.2.2 RNA polymerase

RNA polymerases are enzymes that transcribe the DNA by catalyzing the

formation of phosphodiester bonds that link ribonucleotides together. Eukaryotes

express three RNA polymerases which are responsible for the transcription of

different types of RNA: RNA polymerase I (5.8S, 18S, and 28S rRNA genes), RNA

polymerase II (protein-coding, snoRNA, miRNA, siRNA, and most snRNA genes),

and RNA polymerase III (tRNA, 5S RNA, and some snRNA genes)212. RNA

polymerase II (RNAPII) is composed of two large subunits and several smaller

subunits. In humans, the C terminus of the largest subunit has 52 heptad repeats

of the consensus sequence Tyr-Ser-Pro-Thr-Ser-Pro-Ser213. This C-terminal

domain (CTD) is unique to RNAPII, and its phosphorylation, which mostly occurs

at the Ser residues, regulates RNAPII activity at multiple stages of transcription214-

217. The phosphorylation status of the CTD changes throughout the transcription

cycle: pSer5 levels are highest at the promoter and TSS of genes, while pSer2 levels increase as RNAPII moves along the gene body218,219. This is thought to

influence which proteins bind to the CTD and the timing when this association

occurs.

The CTD acts as a scaffold for various proteins220. For example, pre-mRNA

processing is highly dependent on the phosphorylation status of the CTD. In order

for the 5ʹcap to be added to the RNA transcript, the capping enzymes must interact

with the CTD that is phosphorylated at Ser5218,219,221,222. Similarly, proteins

involved in splicing and the two steps of 3ʹ-end processing (cleavage and

polyadenylation) also bind phosphorylated CTD223-226. Truncation of the CTD

inhibits all of these process by preventing pre-mRNA processing factors from

interacting with RNAPII221,223. Besides mRNA processing proteins that associate with RNAPII at the beginning and end of a gene, other proteins bind the

phosphorylated CTD within the gene body. For example, SETD2, a histone

methyltransferase, binds the phosphorylated CTD and adds methyl groups to

histone H3 at K36, a histone mark associated with active transcription227,228. The

CTD is thus important for linking several processes to transcription.

1.2.3 Other transcription-related proteins

A wide variety of proteins besides RNAPII are involved in the production of RNA transcripts. Transcription factors that regulate the formation of the PIC and transcription initiation are called activators and repressors. Activators are sequence-specific DNA binding proteins that activate gene transcription when bound to the DNA. Many different transcription activating proteins exist and tend to contain at least two domains: a DNA binding domain and an activation domain.

Through several mechanisms, activators function to recruit GTFs, Mediator, and

RNAPII to the promoter and can modify these proteins229. Only after these critical

proteins localize to the promoter can transcription be initiated. Some activators

directly bind the core transcriptional machinery while others alter the structure of chromatin at the promoter to allow GTFs, RNAPII, and other gene regulatory

proteins access to binding sites within the DNA230. Activators can recruit histone

modification enzymes, ATP-dependent chromatin remodeling complexes, and histone chaperones which lead to histone modifications, nucleosome remodeling, nucleosome eviction, and nucleosome replacement231-233. These events create

open areas of chromatin, a step that is necessary for transcription to occur.

Gene repressor proteins also influence gene transcription and are important for

the prevention of gene mis-expression. Repressors can suppress gene

transcription through several mechanisms: 1) competing with activators for binding

to the DNA, 2) binding to and masking the activation domain of DNA-bound

activators, 3) blocking assembly of GTFs at the promoter, 4) recruiting chromatin

remodeling complexes that cause chromatin to adopt a closed conformation, 5)

bringing HDACs to the DNA, and 6) recruiting histone methyltransferases234-237.

Both gene activator and repressor proteins can exist in complexes with other

activators/repressors as well as co-activators/co-repressors, proteins that do not

bind directly to the DNA but instead interact with other DNA-bound regulatory

proteins194.

Another group of proteins involved in transcription initiation and elongation are the

GTFs TFIIA, TFIIB, TFIID, TFIIE, TFIIF, and TFIIH. They are recruited to the core

promoter by interacting with activators and are critical for the activity of RNAPII238.

The GTFs have different functions. TFIIA, TFIIB, TFIID, and TFIIF recruit RNAPII

to the DNA to form the PIC239,240. TFIIH has DNA helicase activity and, together with TFIIE, is responsible for opening promoter DNA which is required for RNAPII

promoter clearance241,242. GTFs are required for transcription, and transcription

initiation in vitro does not occur unless they are present243.

Mediator is also required for transcription. This very large co-activator protein

complex serves as a docking site for transcription factors and co-activators and

plays a role in numerous cellular processes, including transcription, chromatin

looping, mRNA processing and export, transcriptional memory, and DNA repair244.

It is composed of 30 subunits organized into four modules: Head, Middle, Tail, and

Kinase. The Head and Middle modules interact with RNAPII and the GTFs while

the Tail module binds sequence-specific transcription factors. Through DNA

looping, Mediator can link transcription factors bound to enhancer regions with the

transcriptional machinery at the promoter245. Prior to transcription initiation,

activators recruit Mediator to the PIC, and Mediator in turn can recruit GTFs246,247.

Mediator interacts with the unphosphorylated CTD and body of RNAPII and can

bring RNAPII to the PIC as well247. It is also important for stabilizing the PIC and

catalyzes its dissolution244.

Elongation factors control the rate of transcriptional elongation, alleviate pausing,

and aid in the movement of RNAPII along the DNA248. One example is the positive

transcription elongation factor b (P-TEFb). A rate-limiting step of transcription is

RNAPII pausing following initiation. P-TEFb phosphorylates RNAPII which then

initiates elongation249,250. P-TEFb has two subunits, CDK9 and the regulatory

subunit cyclin T1251. Several proteins can interact with P-TEFb to bring it to the

chromatin. The primary recruitment factors are BRD4 and the super elongation

complex (SEC), although other proteins have also been found to recruit P-TEFb,

including Mediator, Myc, ERα, and RelA244,252-257. Besides P-TEFb, other

elongation factors, histone chaperones, histone modification enzymes, and

nucleosome remodelers assist in the movement of RNAPII along the DNA. For

example, the histone methyltransferases SETD1A, SETD2, and DOTL1 add the

histone marks H3K4me3, H3K36me3, and H3K79me3, respectively, all of which

are marks of active transcription. These histone modifications open the chromatin

by generating binding sites for chromatin remodeling factors258-260.

Chromatin organization plays an important role in the regulation of transcription.

For example, nucleosomes located within promoters and gene bodies can block the binding of transcription factors and the transcriptional machinery and can impede RNAPII from moving along the DNA during elongation. Chromatin remodeling enzymes, multi-subunit proteins that have ATPase activity, alleviate this issue by sliding nucleosomes along the DNA, ejecting nucleosomes, or altering the histones that compose the nucleosomes261,262. They can also assist in

unwrapping DNA from around histones which exposes transcription factor binding

sites263. Finally, chromatin remodelers can replace nucleosomes that were

removed during transcription264.

While the diverse proteins described here each have their own unique

transcription-related functions, they work together to ensure the proper timing and

level of gene expression. Disruption of any of these proteins can lead to

transformation and cancer progression. It is thus important for cells to tightly

regulate the expression and activity of these proteins to prevent oncogenesis.

1.2.4 Transcription initiation, elongation, and termination

Activation of transcription begins with the removal of transcriptional repressors265.

In addition, the DNA within the gene’s promoter region must be unwound from the

nucleosome. This occurs when a sequence-specific activator binds the DNA and

interacts with chromatin remodelers and nucleosome-modifying enzymes to

remove nucleosomes, opening the DNA and revealing binding sites for the

transcriptional machinery231-233. Next, the TATA-binding protein (TBP) subunit of

the GTF TFIID binds the TATA box266. Several other GTFs, coactivators, and

hypophosphorylated RNAPII bind the open region of DNA and form the PIC at the

gene promoter216,267,268. Activators recruit histone modifying enzymes, resulting in the addition of epigenetic marks to the histones that are bound by the GTFs, helping to stabilize the PIC269,270. Following formation of the PIC, promoter DNA is unwound, at least in part by TFIIE and TFIIH, and RNAPII can then initiate

transcription271. Subsequent phosphorylation of RNAPII at Ser5 by TFIIH (a

multimeric complex that includes CDK7 and cyclin H), facilitated by Mediator, disrupts the interaction between RNAPII and promoter-bound factors. This destabilizes the PIC, allowing RNAPII to escape the promoter246,247,272-275. Soon after, RNAPII pauses 30-60 nucleotides downstream of the TSS following its interaction with the pausing factors DRB sensitivity-inducing factor (DSIF) and negative elongation factor (NELF)276-278. While paused, RNAPII creates short,

nascent RNA species but cannot continue with transcriptional elongation until it is

phosphorylated at a second site along with NELF and DSIF. P-TEFb is responsible

for unpausing RNAPII279. Once PTEF-b localizes to the DNA, it phosphorylates the

CTD of RNAPII at Ser2249. It also phosphorylates DSIF and NELF, which releases

NELF from the elongation complex280,281. DSIF remains bound to RNAPII but

transitions into a positive elongation factor276,280. Transcriptional elongation can

then continue.

During elongation, RNAPII exists in a complex with the nascent RNA, elongation factors, RNA processing factors, and other proteins such as those involved in DNA repair, chromatin modification, and gene silencing282. When RNAPII reaches the

3ʹ-end of the gene, it is highly phosphorylated at Ser2 of the heptad repeats in the

CTD, which is important for the interaction between RNA processing proteins and

the CTD. Recognition of the poly(A) site in the nascent RNA by RNA-binding

factors recruits a set of proteins that cleave the RNA product283. This pre-mRNA is

then further processed into a fully functional mRNA, and successfully processed

mRNAs are exported from the nucleus via nuclear pore complexes into the cytosol

where they are translated.

There are two models of transcription termination, the torpedo model and the

allosteric model, both of which depend on recognition of the poly(A) site. In the

torpedo model, cleavage of the pre-mRNA transcript exposes the remaining

uncapped RNA transcript to the 5ʹ-3ʹ exonuclease Rat1/Xrn2 that degrades the

transcript284,285. When the exonuclease reaches RNAPII, it destabilizes the

complex of RNAPII and elongation factors, terminating transcription. The allosteric

model involves the release of anti-termination factors and/or the recruitment of

negative elongation factors which causes a conformational change in the RNAPII-

elongation factor complex, destabilizing the complex and causing it to dissociate

from the DNA286.

This tightly controlled process of mRNA production is not only regulated by

transcription-related proteins but also by epigenetic mechanisms. Epigenetic

marks and proteins dictate which genes are activated or repressed and are thus

responsible for establishing, maintaining, and editing the transcriptome. It is

important to understand these epigenetic modifications, as these alterations are

frequently disrupted in cancer and represent potential therapeutic targets.

1.3 Targeting the epigenome in cancer

The transcriptome is controlled by epigenetic mechanisms: reversible alterations in chromatin structure that do not involve changes to the DNA sequence. These mechanisms are critical for normal development and for maintaining cell identity.

Dysregulation of the epigenome induces changes in the activity of signaling pathways involved in many diseases. Extensive epigenetic reprogramming is common in cancer and, together with genetic alterations, induces vast changes in the transcriptome that drive transformation and tumor progression and maintain cancer phenotypes. Mutations in the epigenome and changes in expression patterns of proteins that regulate the epigenome can both silence tumor suppressors and activate oncogenes. Unlike genetic mutations, epigenetic aberrations are reversible, providing an opportunity for therapeutic intervention. As a result, several drug classes have been developed to target components and regulators of the epigenome. Many have entered clinical trials and are being assessed in various tumor types, including breast cancer.

Understanding epigenetic mechanisms that regulate gene expression in normal

cells is crucial for anticipating how they can be manipulated by cancer cells and

for developing novel therapeutics. There are two major categories of epigenetic

alterations: DNA methylation and histone modification. These mechanisms and

how they are manipulated in cancer are described here following a brief overview

of chromatin structure.

1.3.1 Chromatin structure

The most basic unit of chromatin packing is the nucleosome. Each nucleosome

core particle is composed of 145-147 nucleotide pair-long DNA wound around the

histone octamer (two of each of histones H2A, H2B, H3, and H4)287. Each histone

has a histone tail at its N-terminus and can be covalently modified. The

nucleosome itself is the nucleosome core particle and one of the adjacent linker

DNA segments. Nucleosomes tend to occur every 200 nucleotide pairs288. The linker histone H1, which is usually present but is not part of the core complex, helps to organize nucleosomes into higher-order structures288.

Nucleosome positioning is influenced by DNA binding proteins289. For example,

some DNA-bound proteins stabilize nearby nucleosomes whereas others make it

more difficult for a nucleosome to maintain its position, increasing the likelihood

that the nucleosome will be ejected or moved elsewhere. The DNA sequence itself

can also affect the position of nucleosomes. A/T rich sequences, such as those

found within a promoter, as well as G/C rich sequences are conformationally rigid

and therefore are resistant to being wound around histones290-292. As mentioned

above, nucleosomes can be moved along the DNA, removed from the DNA, or

have their composition altered by chromatin remodeling complexes261. The

position of nucleosomes within chromatin is important because it serves as a

regulatory mechanism that impacts transcription, DNA repair, and other DNA-

dependent processes.

The composition of nucleosomes as well as the affinity of histones for the DNA can

be manipulated by epigenetic modifications. Certain alterations are associated

with specific cellular functions. In addition, changes in the epigenome can cause

the chromatin to adopt a more open or closed conformation, which influences the

activity of associated genes. Finally, epigenetic modifications serve as binding

sites for proteins that either enhance or suppress transcription. One such

modification is DNA methylation.

1.3.2 DNA methylation

Methylation of cytosine residues in CpG dinucleotides, forming 5-methylcytosine

(5mC), regulates gene architecture and gene expression. CpG dinucleotides tend

to exist within two types of sites: CpG islands and large repetitive sequences (e.g.

centromeres, telomeres, and retrotransposon elements)293,294. CpG dinucleotides

within large repetitive sequences are methylated while CpG islands are primarily unmethylated. Methylation of CpG islands can lead to long-term transcriptional

silencing such as that which occurs during X-chromosome inactivation295.

Methylated DNA functions as a gene silencing mechanism by blocking the binding

of proteins required for transcription and by recruiting proteins that suppress

transcription296-299. DNA methylation and the modification of histones are

intrinsically linked. Proteins that bind 5mC, such as MECP2 and MBD2, and DNA

methyltransferases (DNMTs) can recruit HDACs to chromatin which remove acetyl

groups from histone tails and further suppress transcription298-301.

Three DNMTs are responsible for DNA methylation in humans. DNMT1 maintains patterns of DNA methylation during DNA replication, although it also participates

in de novo methylation302. DNMT3A and DNMT3B, on the other hand, are primarily

responsible for establishing DNA methylation patterns303. DNA demethylation is

performed by the TET family of DNA dioxygenases (TET1-3) which convert 5mC

to 5-hydroxymethylcytosine (5hmC)304-306. They then further convert 5hmC to 5-

formylcytosine (5fC) and 5-carboxycytosine (5caC) which are excised by thymine-

DNA glycosylase307-310. This creates an abasic site that is removed and replaced

with an unmodified deoxycytodine triphosphate through the base excision repair pathway. Passive demethylation can also occur through the loss of 5hmC during

DNA replication311.

Aberrant DNA methylation is linked to cancer development and progression312,313.

In cancer cells, DNA is globally hypomethylated, which is associated with gene

activation and increased chromosomal instability, while tumor suppressor genes

are hypermethylated314-319. Disruption of several proteins responsible for DNA

methylation patterns have been identified in cancer. For example, TET protein

degradation initiated by IDAX and TET3 itself as well as negative regulation by

microRNAs, including the oncogenic miR-22, are associated with cancer

development and metastasis in several cancers, including breast cancer320-322.

TET and DNMT mutations also contribute to cancer development and, in some

cancers such as acute myeloid leukemia (AML), confer poor outcomes323-325. In

addition, normal methylation patterns within a cell can be disrupted by

carcinogens, such as cadmium and arsenic, that alter DNMT activity326,327.

Silencing of DNMTs in cancer cells induces global demethylation, re-expression of

tumor suppressor genes, and growth suppression328.

Inhibitors of DNMTs have been developed and are currently under investigation as

anti-cancer therapies. At low doses, DNMT inhibitors lead to epigenetic

reprogramming by reducing methylation at gene promoters, inhibiting cancer stem

cells without inducing cytotoxicity329,330. The DNMT inhibitors azacitidine and

decitabine have been approved for use in myelodysplastic syndromes and are

currently being studied extensively in combination with HDAC inhibitors. In

preclinical studies, this combination induced the re-expression of many genes

silenced in cancer cells, including known tumor suppressors331. However,

resistance to DNMT inhibitors is common332,333. Guadecitabine is a second

generation hypomethylating agent that is well tolerated and performed well in

patients with myelodysplatic syndrome or AML in a phase I clinical trial334. It is currently being assessed in hematologic and solid tumors in at least 20 clinical trials both alone and in combination with other agents including cisplatin and the monoclonal antibody pembrolizumab that targets PD-1. Together, these studies illustrate the ability of DNMT inhibitors to lead to the activation of important tumor suppressor genes. Improvement in the design of these agents as well as the discovery of efficacious combination therapies should improve the utility of these agents in the clinic.

1.3.3 Histone modifications

Histone modifications can be non-covalent or covalent. Non-covalent modifications

refer to such mechanisms as nucleosome repositioning and histone variants. The

location of nucleosomes mark specific regulatory sites within a gene and reveal or conceal transcription factor binding sites335. As a result, transcription regulation

and nucleosome positioning are tightly linked. The 5ʹ- and 3ʹ-ends of genes tend

to be nucleosome-free regions (NFRs), which allows the transcription machinery to assemble and disassemble at these loci336. Nucleosome eviction near the TSS

of a gene can lead to transcription initiation while the addition of a nucleosome at

this locus represses transcription337-339. As discussed previously, nucleosomes are repositioned by ATP-dependent chromatin remodeling complexes.

Variant histones that replace canonical histones also influence gene activity. While the core histones are incorporated into chromatin during S phase, variant histones can be deposited onto chromatin throughout the cell cycle340,341. Histone variants

have been identified for all of the core histones except histone H4 and have specific

functions (Table 1.4)342. For example, histone H3 can be replaced by several

variants, including H3.3 and CENPA. While H3.3 is involved in transcription

activation, CENPA localizes to the centromere and is important for the assembly

of the kinetochore and chromosome segregation340,343,344. Posttranslational

modifications can alter the functions of variant histones. As an example, if H2A.Z,

one of the histone variants that can replace histone H2A, is acetylated, it localizes

to the 5ʹ-end of genes and activates transcription345. On the other hand,

ubiquitylated H2A.Z silences transcription and associates with facultative

heterochromatin346. ATP-dependent chromatin remodeling complexes and histone

chaperones are responsible for exchanging core histones for histone variants at

specific sites on chromosomes347.

As with other epigenetic modifications, variant histones are deregulated in cancer.

Overexpression and downregulation of H2A.Z, which promote cell cycle

progression and the spreading of heterochromatin, respectively, have been found in several types of cancer348,349. Similarly, mutations in histones H3 and H3.3 have

been identified in glioblastoma and diffuse intrinsic pontine glioma (DIPG) and are

thought to be involved in the development of these diseases350,351. ATP-dependent

chromatin remodeling complexes and histone chaperones are also implicated in

cancer development. The H3.3 histone chaperone Daxx and chromatin remodeling

factor ATRX, which form a complex, are mutated in pancreatic neuroendocrine and

glioblastoma tumors, leading to altered gene expression patterns350,352. In addition,

alterations in the subunits of the SWI/SNF chromatin remodeling complex, such as

the silencing of the catalytic subunits BRG1 and BRM, are linked to oncogenesis

in several types of cancer353,354. Administration of HDAC inhibitors can stimulate

the re-expression of BRM, suggesting the utility of these drugs in tumors with

silenced BRM354.

Similar to non-covalent histone alterations, covalent modifications of histone tails

influence chromatin structure and gene expression and include acetylation,

methylation, phosphorylation, and ubiquitylation355. These modifications can

activate or repress transcription depending on the type of modification added and

the residue that is modified. For example, the addition of an acetyl group to a lysine

removes the lysine’s positive charge, disrupting the interaction between the

histone tail and the DNA and causing chromatin to adopt a more open

conformation355,356. Lysine methylation, on the other hand can either mark active

or repressed genes. The collection of histone tail modifications make up a histone

code that determines chromatin structure and gene activity357.

The histone code is dynamic, and specialized proteins referred to as writers,

erasers, and readers are required to modify and translate it (Figure 1.3)358. Writers, such as histone acetyltransferases (HATs) and histone methyltransferases

(HMTs), add modifications to the histone tail, while erasers, such as HDACs and histone demethylases (HDMs), remove them. Proteins called readers bind to histone modifications and alter gene expression by recruiting transcriptional machinery and chromatin remodelers. Each histone modification is recognized by unique domains within reader proteins. For example, the reader motifs responsible for binding acetylated lysines are tandem bromodomains and PHD fingers while methylated lysines are recognized by chromodomains, Tudor domains, MBT domains, PWWP domains, PHD fingers, and WD40/β propellers359. Later in this

chapter, I will discuss a family of epigenetic readers, called Bromodomain and

Extraterminal (BET) proteins, in more detail.

Due to their impact on chromatin structure and gene expression, it is important that

writers, erasers, and readers localize to chromatin at the appropriate time, and

changes in the expression of these proteins can lead to a reorganization of the

histone code. In cancer, for example, there is a global loss of specific acetylation

(e.g. H3K18ac, H3K9ac, and H4K16ac) and methylation (e.g. H3K4me2,

H3K27me3, and H4K20me3) marks, and these are associated with poor prognoses360-364. Tumors also gain methylation marks, such as H3K9me2, which

aberrantly silence genes, and acetyl groups are deposited to newly active

enhancers and promoters and are enriched at super-enhancers that drive the

expression of oncogenes365,366. Mutations and alterations in the expression of

histone modification proteins are also frequently observed in tumors294,367-369.

HDACs, such as HDAC1 and HDAC2, and HMTs, such as EZH2 and G9a, are

often overexpressed in tumors370-372. Histone modification proteins are thus

attractive therapeutic targets for the treatment of cancer.

Many agents targeting various HDACs, HATs, HDMs, HMTs, and reader proteins have been developed. For example, numerous studies in the 1990s and early

2000s found that HDAC inhibitors effectively suppress tumor growth while inducing minimal toxicity, and these inhibitors, four of which are already FDA-approved, are currently being assessed in clinical trials for the treatment of a variety of cancers373,374. HDAC inhibitors are also being studied in combination with other

drugs targeting the epigenome. Preclinical studies have shown DNA demethylating agents and HDAC inhibitors synergize to induce the re-expression of aberrantly silenced genes in cancer cells375,376. Similarly, HDAC inhibitors

synergize with BET protein inhibitors in several types of cancer377-381. In addition,

several epigenetic drugs are effective in combination with established anti-cancer

agents. Some, such as BET inhibitors and HDAC inhibitors, reverse acquired drug

resistance in ovarian and breast cancers382,383. These findings indicate that drugs

that inhibit histone modification proteins may provide immense clinical benefit and

potentiate the effects of currently available therapies. However, more work is

necessary to identify biomarkers that can predict which patients will respond to these therapies and to develop effective combination therapies with limited toxicities and durable responses. With regard to the current thesis, I have studied the effects of BET protein inhibitors in breast cancer. Thus, I will now describe the effects of drugs targeting BET proteins in breast cancer and efforts focused on optimizing their use in this disease.

1.4 Targeting Bromodomain and extraterminal proteins in breast cancer

Breast cancers are driven by numerous oncogenic pathways which can be subtype-specific. As such, the various subtypes must be treated with different agents. To target a diverse array of breast tumors and to prevent recurrence, it should also be useful to develop therapies that can target multiple pathways simultaneously and have broad implications for this group of diseases as a whole.

Here, I discuss targeting BET proteins, an approach already in clinical trials that has the potential to benefit all subtypes of breast cancer.

1.4.1 BET protein structure and function

1.4.1.1 BET protein structure

As mentioned above, various posttranslational modifications are added to nucleosomes that impact their association with chromatin and the recruitment of proteins to DNA. One such modification is lysine acetylation, which marks areas of chromatin for active transcription and is recognized by bromodomains (BRDs) in various proteins355. The BRD is a conserved 110 amino acid structural motif

composed of four α-helices (αZ, αA, αB, and αC) that comprise a left-handed

bundle384. Two loop regions (ZA and BC) connect the α-helices and form a surface

that interacts with acetylated lysines in nucleosomal histones385. In humans, there

are 61 BRDs found within 42 multi-domain proteins that regulate transcription,

including ATP-dependent chromatin remodeling complexes, transcriptional co-

activators, HATs, and BET proteins386.

The BET protein family consists of four members (BRD2, BRD3, BRD4, and

BRDT) that reside in the nucleus and play critical roles in transcription387. BET proteins act as epigenetic readers and are characterized by two tandem N-terminal

BRD regions followed by an extraterminal domain. The BRD regions recognize and bind acetylated lysines in histone tails (histones H3 and H4) and transcription factors. The extraterminal domain is involved in protein–protein interactions with proteins such as E2Fs and latent nuclear antigen of Kaposi’s sarcoma-associated herpes virus388,389. BRD4 and BRDT have an additional C-terminal motif that links

their reader function to transcriptional elongation: following the binding of BRD4/T

to acetylated histones, the C-terminal motif interacts with P-TEFb390. This localizes

P-TEFb to target promoters where it phosphorylates RNAPII and releases it from

pausing.

Alternative splicing generates three isoforms of BRD4: the long-form isoform A

(13622 aa), isoform B (796 aa), and isoform C (722 aa)391. Isoforms B and C lack

the C-terminal domain and are distinguishable from each other by the presence of

an additional 76 amino acid peptide at the C terminus of isoform B. Isoform B has

only been identified in U2OS cells, and its activity is not well characterized391. A

separate set of studies identified two isoforms, a long-form (BRD4-LF) that

corresponds to isoform A and a short-form (BRD4-SF) that most likely corresponds

to isoform C392,393. BRD4-LF and BRD4-SF have differing effects on metastasis:

BRD4-LF reduces metastasis while BRD4-SF promotes metastasis of the mouse

mammary tumor Mvt-1 model of breast cancer392,394. It is possible that the ratio

between the long and short isoforms of BRD4 dictate the oncogenic potential of

BRD4, and differences in the ratio of these isoforms could explain why only a few

studies have generated data suggesting that BRD4 acts as a breast tumor and

metastasis suppressor394-396 while the majority of studies demonstrate that BRD4

is an oncogenic driver.

Binding of BET proteins to acetylated histones recruits BET proteins to the

enhancer and promoter regions of genes marked for active transcription. Here,

they interact with co-activators/repressors, transcription factors, and the

transcriptional machinery, forming protein complexes that influence target gene

transcription397. While they have a similar structure and usually enhance

transcription, BET proteins regulate different processes based on their binding

partners, which are often tissue-specific.

1.4.1.2 BET proteins and transcription

When BET proteins bind acetylated histones, they recruit several regulatory

complexes that influence various aspects of chromatin structure and transcription.

For example, BRD4 is important for transcription initiation. BRD4 regulates

monoubiquitination of histone H2B (H2Bub1) and interacts with histone modifying

proteins, such as the arginine demethylase JMJD6, and chromatin remodeling

enzymes, such as CHD8398-401. This leads to a more open conformation of the

chromatin and allows BRD4 to recruit key transcriptional regulators, such as

Mediator and transcription elongation factors, to the DNA402-405. BRD4 additionally

enhances transcription of genes with ERα-bound estrogen response elements

(EREs) that are also enriched for FOXA1 binding. BRD4 binds acetylated histones,

including H4K12ac, at these elements and recruits RNAPII which initiates the

transcription of eRNAs398,406.

In addition, BET proteins are involved in transcriptional elongation. BRD4 directly

promotes elongation by phosphorylating RNAPII at Ser2 of the CTD407 and

recruiting P‐TEFb255,400,408. P-TEFb in turn directly phosphorylates paused RNAPII

and disrupts the interaction between RNAPII and the regulatory complexes DSIF

and NELF249,277. Both events release RNAPII from the paused state at the

promoter and thereby induce transcriptional elongation. During elongation, both

BRD2 and BRD3 act as histone chaperones, remodeling histones in order for

RNAPII to move along the DNA409. BRD3 also recruits the polymerase-associated

factor complex (PAFc), which coordinates several events during transcription, and the super elongation complex, which also regulates elongation410. Based on the

activity of BET proteins during transcription initiation and elongation, the disruption

of BET protein function should have an enormous impact on the production of RNA

transcripts in varied cell contexts.

1.4.1.3 BET proteins and the cell cycle

BET proteins play diverse roles in multiple phases of the cell cycle and control

expression of cell cycle and proliferation genes389,409,411-413. During mitosis, BRD4

ensures proper chromosomal segregation and cytokinesis by regulating

expression of Aurora kinase B414. Both BRD2 and BRD4 facilitate mitotic memory

by remaining bound to M/G1 genes during mitosis and recruit P‐TEFb to the DNA

late in mitosis. This marks G1 genes for immediate transcription following mitotic

exit, ensuring cell cycle progression415-417. BRD4 also promotes G1-S and G2-M

phase transitions418,419. As a result of their role in the regulation of the cell cycle,

loss of BET expression induces cell cycle arrest420-426.

1.4.1.4 BET proteins and inflammation

BET proteins also control inflammation. BRD2 binds to genes regulated by STAT5, an important mediator of cytokine signaling, and pan-BET inhibition suppresses expression of STAT5-target genes427. Furthermore, mice that express half the

amount of Brd2 as wild-type mice (brd2 lo) develop severe obesity but are

protected against insulin resistance and the obesity-induced inflammatory

response428. siRNA-mediated knockdown of BRD2 suppresses NF-κB

transcriptional activity429. In addition to regulating pro-inflammatory gene

expression though NF-κB, BRD2 directly binds the promoter regions of pro-

inflammatory genes, especially following LPS stimulation, and macrophages from brd2 lo mice produce significantly less pro-inflammatory cytokines such as

TNF-α, IL1β, and IL6429. NF-κB activity is also regulated by BRD4; BRD4 binds

acetylated RelA, a subunit of NF-κB, and enhances transcription of NF-κB-

dependent inflammatory genes430. Suppression of BET proteins induces anti-

inflammatory responses429,431, and BET inhibitors are currently being investigated

as potential therapeutic options for the treatment of inflammatory diseases432.

1.4.1.5 BET proteins and development

BET proteins are crucial during development, and loss of either BRD2 or BRD4

results in embryonic lethality. BRD2 controls neuronal differentiation433,

with Brd2 null embryos displaying deficient neural tube formation and dying during

mid-gestation434. BRD4 maintains the self-renewal capability of stem cells by

stimulating expression of genes involved in pluripotency. Over 20% of pluripotency

genes, including NANOG, OCT4, SOX2, and PRDM14, are bound by BRD4 in

embryonic stem cells435-439. BRD4 also localizes to a number of stem cell genes in

preimplantation embryos and maintains the inner cell mass, with BRD4 null

embryos dying shortly after implantation437,440.

1.4.1.6 BRDT and spermatogenesis

Unlike the other three BET proteins which are ubiquitously expressed, BRDT

expression is normally testis-specific441. It is expressed in pachytene and diplotene

spermatocytes and round spermatids, and its expression decreases during

spermatid differentiation442. BRDT is essential for spermatogenesis,

and BRDT knockdown leads to sterility in male mice442. Pharmacological inhibition

of BRDT also confers reversible infertility in male mice without impacting

testosterone levels, and offspring produced by these mice once treatment is

removed are normal443. As a result, BRDT is seen as a potential target for

reversible male contraception.

1.4.2 BET proteins in cancer

1.4.2.1 BET proteins and super-enhancers

BET proteins are involved in various diseases, including inflammation, viral

infection, heart failure, and cancer444,445. It is thought that BET proteins primarily

mediate their effects in disease pathogenesis and progression primarily by

localizing to super-enhancers (SEs) at pathology-associated genes and driving

their expression366,410,446. An SE is a large contiguous cluster of enhancers within

a locus that is associated with increased gene expression, DNase I sensitivity,

histone tail acetylation, and transcription factor and co-activator binding366,447. SEs are frequently identified by ChIP-seq analysis using antibodies against histone marks such as H3K27ac or the transcriptional regulators Mediator and BRD4, as these proteins are enriched at SEs366,447,448. SEs vary by cell type and the

different binding sites in distinct cells underlie the expression of cell identity genes

that specify that cell type447,448. In addition, SEs drive expression of disease- associated genes in numerous diseases, including Alzheimer’s disease, type 1 diabetes, and cancer448,449. In cancer, SEs are enriched at oncogenes known to play a role in specific cancer types, including MYC and IRF4 in multiple

myeloma, RUNX1 and FOSL2 in glioblastoma, and CD79 B in diffuse large B cell

lymphoma366,446. They are also associated with many oncogenes that are linked to

general cancer pathogenesis, including CCND1, MCL1, and BCL2L1366.

Only a fraction of enhancer regions are classified as SEs. For example, in multiple

myeloma there are 308 putative SEs compared to nearly 8000 typical

enhancers366. Despite the relatively small number of SEs, BRD4 disproportionately

accumulates at these regions, with up to 40% of all bound BRD4 being localized

to SEs366,446. BRD4, like other co-activators, exhibits cooperative binding. Thus,

loss of BRD4 leads to a greater disruption of transcription at SE-associated genes compared to typical-enhancer associated genes, providing a mechanism to preferentially silence multiple SE-associated oncogenes at once366,446,450.

1.4.2.2 Expression of the BET family in breast tumors and cell lines

BRD2, BRD3, and BRD4 are expressed in breast tumors while BRDT is rarely expressed. When examining all breast cancers, regardless of subtype, the genes encoding the BET proteins are amplified and/or overexpressed in less than 10% of tumors in the TCGA451 and METABRIC27,452 datasets. However, when focusing

on breast cancer subtypes, BRD2 and BRD4 are amplified and/or overexpressed

in 12.1% and 20.6% of basal-like breast cancers, respectively451. Analysis of CNAs

of 9445 tumors representing 20 cancer types in the TCGA dataset confirmed

that BRD4 is more commonly amplified in breast cancer as well as ovarian, liver,

and endometrial cancers compared to cancers from other organs453. In

addition, BRD4 was found to be amplified in a study evaluating focal amplification

events in a panel of 10 DCIS and 151 invasive breast tumors representing all

subtypes of breast cancer454. Lastly, BRD4 mRNA was more highly expressed in

tumors compared to normal breast tissue454. Together, these data suggest an

association between BRD4 levels and breast tumorigenesis.

BRD4 is expressed in non-transformed breast epithelial, luminal breast cancer,

and TNBC cell lines455. By interrogating a microarray dataset of 477 breast cancer

samples, Shi et al. found there was no difference in BRD4 mRNA expression between ER+ and ER- tumors455. However, BET expression is less consistent at

the protein level. When comparing BRD4 protein expression in two TNBC (MDA‐

MB‐231 and BT549) and two ER+ (MCF7 and T47D) cell lines, T47D cells clearly

express higher levels of BRD4 compared to the other three lines456. In TNBC cell

lines (MDA-MB-231, MDA-MB-468, and HCC1937), BRD4 is expressed at similar

levels in all three lines while BRD2 and BRD3 expression was higher in HCC1937

cells421. In addition, a larger study examining basal protein expression of BRD2,

BRD3, and BRD4 in an 18 cell line panel representing non-transformed mammary,

luminal, HER2+, and TNBC cell lines found that expression of BET proteins was

variable and did not correlate with breast cancer subtype420. Thus, despite

relatively constant expression at the RNA level, BET protein expression varies by

cell line, suggesting post-transcriptional or post-translational mechanisms are

responsible for the observed differences in BET protein expression.

1.4.2.3 Essentiality of BET proteins in breast cancer

In an effort to identify subtype-specific and pan-disease essential genes in breast

cancer, Marcotte, et al. utilized pooled lentiviral shRNA dropout screens in 78

breast cancer and four non-transformed mammary epithelial cell lines and

analyzed the results using an algorithm they developed termed “the siRNA/shRNA

mixed effect model” (siMEM)457. This process successfully detected known drivers

of breast cancer and breast cancer subtypes, affirming its utility. In addition, this

study revealed new candidates for essential breast cancer genes, including BRD4.

Silencing of BRD4 gene expression using two shRNAs in SUM159, BT474, and

T47D cells resulted in decreased proliferation, and this response was prevented

by restoring BRD4 expression using an shRNA-resistant BRD4 cDNA. However,

sensitivity to BRD4 depletion does not always translate into sensitivity to BET

inhibitors, suggesting that BRD4 has both BRD-dependent and BRD-independent

roles in breast cancer. This concept is supported by the finding that SUM159 cells

that are resistant to BETi are still sensitive to genetic silencing of BRD4, indicating

BRD4 may be recruited to the chromatin through a BRD-independent mechanism

or that BRD4 has a role that is independent of chromatin binding420.

Evidence that BRD4 may be essential for growth of estrogen-dependent breast cancer cells was first provided by a group that applied a novel triclustering algorithm to a publicly available microarray dataset corresponding to a time course of estrogen response of MCF7 cells458. This approach revealed that BRD4 may be

a hub-gene in ER-driven breast cancer459. Later functional studies in four TNBC

cell lines (SUM159, MDA-MB-231, MDA-MB-468, and MDA‐MB-436) and one luminal line (ZR-75-1) using RNAi-mediated silencing of BET proteins demonstrated that suppression of either BRD2 or BRD4 reduces the growth of four of the cell lines with only siBRD4 inhibits growth of MDA‐MB‐436 cells420.

Subsequently, we reported that simultaneously

silencing BRD2 and BRD4 reduces expression of BRD3 as well, suggesting a

complex interplay in the regulation of these factors. We further showed that

combined BRD2/4 silencing reduces the expression of key mitotic regulators that

play a crucial role in the BETi response of TNBC cells411. Together, these findings

further revealed a role for BET proteins in breast cancer growth and pathogenesis.

1.4.3 Targeting BET proteins in breast cancer

1.4.3.1 BET inhibitor structure and selectivity

The first BETi to be developed were JQ1460 and I-BET431. Since then, multiple

derivatives have been utilized in preclinical settings, including I-BET151410, I-

BET762461, MS417462, and OTX015463, and at least 11 have moved on to be

assessed in early phase clinical trials for a wide variety of hematologic cancers

and solid tumors. BETi belong to varied chemical classes based on their core

scaffolds, such as azepines (JQ1, OTX-015, CPI-0160), 3,5-dimethyl isoxazoles

(I-BET151), pyridones (ABBV-075), and tetrahydroquinolones (I-BET762)464.

While there are 42 BRD-containing proteins in humans386, BETi selectively target

the BET family of proteins431,460,465. Most BETi inhibit all four members of this

family, although BETi that selectively bind a specific BET protein, particularly

BRD4, are currently being developed. BETi inhibit BET proteins by competing with

acetylated lysines for binding to both BRD regions, preventing BET proteins from

binding histones and thus from localizing to the chromatin460. The specific BET

protein(s) that must be suppressed for BETi to elicit their effects differs depending

on the context. For example, in breast cancer models, we found that loss of BRD2

and BRD4 expression together was necessary for the induction of mitotic

catastrophe in response to BETi in TNBC411 while other studies have identified

BRD4 as the sole critical target of BETi for controlling other phenotypic

responses383,420,422.

Multiple studies have examined the utility of BETi in various models of breast

cancer (Table 1.5). Depending on the subtype studied and cell lines used, BETi

can impact tumor formation, proliferation, the response to hypoxia, angiogenesis,

cancer stem cells, metastasis, and metabolism by repressing the expression of

genes that drive these critical oncogenic pathways (Figure 1.4). The effects of BETi

in breast cancer reported thus far are reviewed below. It is important to keep in

mind that the activity of BETi in vivo most likely depends on a convergence of

several of these BETi-induced outcomes as opposed to a singular response.

1.4.3.2 Impact of breast cancer cell growth and tumor formation

Many cell lines representing the luminal (ER+), HER2+, and TNBC subtypes of

breast cancer undergo growth inhibition in response to BETi treatment. Marcotte,

et al. found luminal and HER2+ cell lines were more dependent on BRD4

expression than TNBC cells, as siRNA-mediated knockdown of BRD4 led to a

greater suppression of growth in these lines compared to basal cell lines457.

However, when comparing the IC50 of five BETi in a panel of 41 cell lines that

represent luminal, HER2+, and TNBCs as well as non-transformed mammary

epithelial cells, Polyak and colleagues found that TNBC cell lines were generally

more sensitive to BETi than HER2+ and luminal cell lines420. The discrepancy

between these two studies could be due to the genetic silencing of a single BET

protein compared to inhibition of the entire family with a small molecule.

Several studies have specifically assessed BETi activity in luminal breast cancer.

BETi suppressed growth of MCF7, ZR75-1, and T47D cells with or

without estrogen stimulation in 2D and 3D culture, and JQ1 inhibited expression of

canonical estrogen-target genes398,423,456,466,467. Tamoxifen-resistant (Tam-R) and

estrogen-deprivation-resistant versions of ER+ cell lines were more sensitive to

BETi than the parental lines468. In addition, within two days, BETi treatment

induced apoptosis of Tam-R MCF7 cells but not of parental MCF7 cells468,

suggesting BETi may be an effective treatment option for hormone therapy-

resistant tumors. The difference in sensitivity of parental and Tam-R breast cancer cells to BETi likely stems from differences in the expression pattern of critical transcription factors. For example, parental MCF7 cells had higher expression of

GATA3, which is necessary for sustained ESR1 (ERα) gene expression,

compared to Tam-R cells469. Supporting this potential mechanism of BETi sensitization, silencing GATA3 in parental cells rendered them more sensitive to

BETi treatment468.

Affirming the relative resistance of parental MCF-7 cells to BETi in vitro, JQ1 failed

to impact MCF7-derived tumor growth in mice466. MCF-7 cells are a model of

luminal A breast cancer. In contrast to studies with these xenografts, JQ1 was reported to be effective in the MMTV-PyMT mouse model of luminal B breast cancer470. Precursor lesions in this model are ER+ while established tumors are

ER- but maintain a luminal gene expression signature80,471. Pérez-Salvia, et al.

discovered that not only could JQ1 suppress growth of established tumors in

MMTV-PyMT mice, but the drug could also slow the development of spontaneous

mammary tumors when administered to four week old mice prior to the detection

of palpable tumors467. In addition, JQ1 improved overall survival in MMTV-PyMT

mice. Based on these data, the authors suggested that BETi could be utilized as

a preventative agent in women who have a high risk of developing breast cancer.

BETi are quickly cleared from the target tissue, hence it would be necessary to

treat these patients daily. Importantly, it is not yet known whether long-term

exposure to BETi will lead to intolerable toxic side effects and increase the chances

of developing BETi resistance.

Information on BETi treatment alone in HER2+ breast cancer growth is limited. In

one study, both JQ1 and I-BET762 inhibited the in vitro growth of four HER2+ cell

lines within five days in a dose-dependent manner383. However, in a four-week

clonogenic assay, BETi-resistant colonies still formed. Addition of lapatinib to BETi

dramatically reduced the number of colonies formed, thereby inhibiting acquired

BETi resistance.

The response of TNBC to BETi has been more thoroughly documented. When

Shu, et al. treated 26 TNBC cell lines with BETi, the majority were highly sensitive

to this drug class420. Similarly, we found JQ1, I-BET151, and I‐BET762 treatment

suppressed growth of a panel of seven TNBC cell lines representing five of the six

TNBC subtypes described by Lehmann, et al.122 as well as both the claudin-low and basal subtypes in a dose-dependent manner472. Multiple other studies have

also shown that BETi suppressed 2D and 3D growth, wound-healing capacity, and

colony formation of diverse TNBC cell lines420,421,424,425,456,473, indicating BETi could

be an effective therapy across diverse TNBC tumor types. Our studies and others

also revealed that sustained inhibition of BET proteins induced two terminal

responses, apoptosis and senescence, and these responses did not correlate with

the extent of BETi-induced growth inhibition, impact on c-Myc expression, or TNBC

subtype472. These effects were recapitulated in vivo. Tumors derived from MDA-

MB-231 cells, which senesced in vitro, grew significantly slower when treated with

JQ1, while MDA-MB-468 tumors, which died in vitro, partially regressed472. In

addition, JQ1 suppressed growth of tumors formed from a TNBC patient-derived

xenograft (PDX). Three other studies confirmed the in vivo efficacy of BETi in

TNBC: JQ1 and MS417 inhibited growth of MDA-MB-231, SUM1315, and SUM159

xenografted tumors as well as two PDX models420,424,455. Together, these studies

indicate that models of TNBC are highly responsive to BET inhibition both in

vitro and in vivo.

It has been suggested that BETi induce subtype switching in TNBC, where TNBC

cells lose basal markers and gain luminal markers, due to differential expression

of luminal and basal cytokeratins following JQ1 treatment of the MDA‐MB‐231 and

SUM159 cell lines420. Differentiation of TNBC models following BETi treatment has also been assessed in vivo. Treatment of a PDX model with vehicle or JQ1 and

staining for low molecular weight (luminal) and high molecular weight (basal)

cytokeratins revealed that vehicle-treated tumors had very little expression of low molecular weight cytokeratins while their expression increased significantly with

JQ1 treatment. These data, coupled with the in vitro analysis, suggested that BETi

may induce differentiation of basal tumors to a more luminal phenotype. However,

this conclusion was based on the restricted analysis of cytokeratin gene

expression. In contrast, when vehicle- and JQ1-treated SUM159 xenografts were

stained for diverse luminal (luminal cytokeratin, CK18, and CD24) and basal (basal

cytokeratin, CK17, pSTAT, and CD44) markers, JQ1-treated tumors had variable

responses, and there was no consistent loss of basal and simultaneous gain of

luminal markers. Similarly, we performed GSEA using Neve474 and Charafe-

Jauffret475 breast cancer subtype classifiers on gene expression array data from

MDA-MB-231 and HCC70 cells treated with vehicle or JQ1411. While both cell lines

lost expression of a subset of basal-signature genes, there was no consistent gain

of expression of luminal genes, indicating TNBCs do not undergo extensive BETi-

mediated differentiation to a characteristic luminal expression signature.

1.4.3.3 BET inhibitors and hypoxia

Severe intratumoral hypoxia is common in breast cancer476. Hypoxia in solid tumors occurs due to increased metabolism and proliferation as well as poor vascular structure. It is linked to metastatic progression, resistance to radiation and chemotherapy, and poor prognosis477. As in many other types of cancer,

regions of hypoxia in breast tumors are associated with EMT and the acquisition

of cancer stem cell properties via signaling through hypoxia-inducible factors

(HIFs) which stimulate EMT, promote self-renewal, and inhibit differentiation478.

The phenotypes associated with hypoxia are regulated by HIF-1α and HIF-2α

which heterodimerize with HIF-1β in low oxygen conditions. This complex then

localizes to hypoxia response elements in the promoters of target genes to induce

transcription. Compared to the other subtypes of breast cancer, TNBC is

particularly associated with hypoxia, and HIF target genes are upregulated in

TNBC patient tumors479.

OTX015 suppressed growth of three TNBC cell lines in both normoxic and hypoxic

conditions, and GSEA following gene expression profiling revealed this drug

downregulated hypoxia-responsive genes421. In a second gene expression

analysis study, when MCF7 and MDA‐MB‐231 cells were treated with JQ1 in normoxic and hypoxic conditions, JQ1 also altered expression of hypoxia-related

genes and prevented the hypoxia-mediated upregulation of several gene sets,

including those involved in angiogenesis and the hypoxic pathway473. In MDA‐MB‐

231 cells in particular, JQ1 altered expression of 44% of hypoxia-responsive

genes, the majority of which were suppressed with drug exposure. While

expression of HIF‐1α and HIF-2α remained unchanged, JQ1 reduced expression

of carbonic anhydrase 9 (CA9), a known hypoxia-responsive gene that helps to

maintain a neutral intracellular pH480,481, in MCF7 cells, two TNBC cell lines (MDA-

MB-231 and HCC1806), and HCC1806 xenografts. Notably, high expression of CA9 has been associated with poor overall survival and a higher rate of distant metastases in a cohort of over 3600 breast cancer patients, and inhibition of CA9

suppresses metastasis482. Thus, BETi inhibition of CA9 expression may provide

an approach to limit metastatic progression. Mechanistically, the JQ1-induced

reduction in CA9 expression was accompanied by the loss of HIF-1β binding at

the CA9 promoter following JQ1 treatment in hypoxic conditions473. Thus, BETi prevented the localization of the HIF heterodimer to HIF target genes. Exposure to hypoxia can induce radio- and chemo-resistance, and inhibiting CA9 in combination with radiotherapy or chemotherapy is effective in preclinical models483,484. Together, these data suggest that BETi may be useful for sensitizing

cancers to radiotherapy and/or chemotherapy, making it an effective approach for

the treatment of solid tumors, including breast tumors.

1.4.3.4 BET inhibitors and angiogenesis

In addition to disrupting hypoxia-regulated pathways, BETi also appear to

suppress angiogenesis, one of the hallmarks of cancer485. Hypoxia and

angiogenesis are inherently linked, with hypoxia inducing expression of VEGF that

then stimulates the production of new blood vessels486. Angiogenesis is critical for

tumor growth and metastasis, making it a useful therapeutic target, particularly in

renal cell carcinoma. Monoclonal antibodies against VEGFA or the VEGF receptor

(VEGFR) and tyrosine kinase inhibitors that target VEGFR have been developed,

but clinical trials in breast cancer have yielded mixed results, leading to the

revocation of FDA approval of the anti-VEGFA antibody, bevacizumab, in breast

cancer patients in 2011487. Therefore, it is essential to develop additional strategies

to effectively disrupt angiogenesis. BETi have been shown to suppress

angiogenesis in rhabdomyosarcoma, Ewing sarcoma, and testicular germ cell

tumors488,489, suggesting they may also display anti-angiogenic activity in other

tumor types, including breast cancer.

Only one study has directly assessed the impact of BETi on angiogenesis in breast

cancer. In MCF7 and MDA-MB-231 cells, JQ1 prevented the upregulation of

angiogenic signature genes under hypoxic conditions473. BRD4 bound the

promoter of VEGFA in MDA-MB-231 cells, and this binding increased in hypoxia.

Both treatment with JQ1 and gene silencing of BRD4suppressed expression of

VEGFA in hypoxia in MCF7, MDA-MB-231, and HCC1806 cells. In HCC1806

xenografted tumors, JQ1 also suppressed expression of VEGFA, as well as

the TIE2 and NRP genes that are critical for angiogenesis. Immunostaining

revealed these tumors had lower expression of the blood vessel marker CD31.

These data indicate BETi may impair angiogenesis. This could be due to a double

hit: direct loss of BRD4 at the promoter regions of genes involved in angiogenesis

and the suppression of the hypoxic response leading to the inhibition of hypoxia-

induced angiogenic pathways.

Anti-angiogenic therapies are already used to treat numerous solid tumor types.

However, drug-induced hypoxia occurs in the tumors of about half of these

patients, leading to therapeutic resistance477. Combining bevacizumab with CA9

knockdown suppressed colon cancer and glioblastoma xenografts growth better

than bevacizumab alone490. As mentioned above, BETi downregulated CA9 in

breast cancer cell lines and xenografts, suggesting that combining BETi with anti-

angiogenic agents could be a beneficial treatment strategy and revive the use of

drugs such as bevacizumab in breast cancer.

1.4.3.5 BET inhibitors and cancer stem cells

Cancer stem cells (CSCs) are involved in numerous processes during tumor

initiation and progression, are resistant to traditional cytotoxic chemotherapies,

and play a role in metastasis and recurrence491, making them a desirable target

for anti-cancer therapies. A role for BET proteins in the maintenance of stem genes

is now well established435-438. In embryonic stem cells, inhibition of BET proteins

suppress expression of critical stem cell factors and induce differentiation438,439.

Extending to cancer, BETi induce apoptosis in progenitor and stem cells in AML

and glioblastoma492,493. In MYC-driven medullablastoma, BETi reduce stem cell

signaling and the self-renewal capacity of tumor cells494.

The only subtype of breast cancer that has been investigated thus far for the impact

of BETi on CSCs is TNBC. Expression of WNT5A, which plays crucial roles in

maintaining stem cell pluripotency, was suppressed by JQ1 treatment due to

reduced binding of BRD4 at its promoter455. JQ1 also inhibited activity of the

JAK/STAT pathway that promotes stem cell renewal as well as the pro-

inflammatory response and EMT420. Similarly, a more extensive study utilizing

OTX015 found that BETi altered stem cell-related gene expression patterns. GSEA of OTX015-treated TNBC cells showed an overall loss of expression of CSC genes421. In general, OTX015 downregulated CSC genes in three TNBC cell lines within 24 h, although some of the genes that changed and the direction in which they were altered were cell line-specific. NANOG and OCT4, transcription factors that promote stemness, were suppressed as were two additional stem cell markers, CD133 and Musashi-1. Breast CSCs are often defined by high expression of CD44 and low expression of CD24. OTX015 treatment reduced

CD44 expression in three cell lines while the impact on CD24 was variable. The suppression of stem cell markers by OTX015 was confirmed in vivo in mice bearing MDA-MB-231 tumors. However, expression of the epithelial marker

EpCAM did not increase in any of the cell lines assessed. No other epithelial markers were examined, so it is unclear if OTX015 is capable of initiating differentiation in CSCs. As mentioned above, our studies indicated that TNBC cells do not undergo a basal to luminal transdifferentiation in response to JQ1411. Thus,

while the loss of stem cell markers that occurs with BETi indicates a loss of the

stem cell phenotype in TNBC, it is not accompanied by the acquisition of a luminal

breast cancer profile.

In a separate study, JQ1 did not alter expression of three stem cell markers (CD44,

CD49, and CD133) in MDA-MB-231 cells, yet it did significantly decrease the formation of primary and secondary tumorspheres426. It will be important in the

future to perform additional functional tests in vitro and in vivo to determine if and how BETi directly impact the population of stem cells in breast cancers.

1.4.3.6 BET inhibitors and metastasis

The vast majority of breast cancer patients do not die from their primary tumor.

Instead, they succumb to metastatic lesions that develop in vital organs.

Metastasis is a multi-step process, which includes invasion of surrounding tissue, intravasation and survival within the bloodstream, extravasation, and colonization of a distant organ495. An early event during this cascade is EMT, a process that

enables epithelial cells to adopt a more mesenchymal, motile phenotype496. Not only do cells that undergo EMT become more migratory but they also acquire stem cell characteristics. BRD4 has been shown to regulate EMT in cancer. In several types of cancer, overexpression of BRD4 triggered metastasis while BET inhibition altered expression of key EMT genes, thereby preventing metastasis497-500.

Additionally, high expression of BRD4 correlated with lymph node metastasis in

non-small cell lung cancer and renal cell carcinoma501.

In breast cancer, changes in BRD4 expression or treatment with BETi impact

expression of genes linked to EMT and metastasis. Another set of proteins, those

belonging to the extracellular matrix (ECM), can also modulate the EMT response.

This has been demonstrated in mammary epithelial cells, with laminin inhibiting

EMT and fibronectin promoting EMT following the addition of the EMT-inducing

enzyme matrix metalloproteinase-3502. In addition, multiple mouse and human studies have revealed that expression of ECM genes are frequently dysregulated in tumors that are likely to metastasize503-507. Modulating BRD4 levels by ectopic

overexpression or gene silencing altered expression of ECM regulatory genes in

breast cancer cell lines394,422. Overexpression of BRD4 also changed the activity

of genes involved in other processes important for EMT and metastasis, including cytoskeletal remodeling and cellular adhesion394. In addition, BRD4 regulates the

expression of the HOX transcript anti-sense RNA (HOTAIR), a long non-coding

RNA that promotes metastasis, regulates breast CSC properties, and is a

biomarker for breast cancer diagnosis and metastasis508-511. When claudin-low

cells were grown in laminin rich ECM 3D cultures, BRD4

maintained HOTAIR expression by binding its promoter512. As expected, treatment

with JQ1 or gene silencing of BRD4 decreased HOTAIR expression. Together,

these data indicate that BET proteins, and specifically BRD4, control the

production and sensing of the extracellular matrix, a key regulator of cellular

motility.

In addition to modulating the ECM, BET proteins directly regulate EMT-modulating

transcription factors. Inhibiting BET proteins with JQ1 reduces the binding of the

transcription factor, activating enhancer binding protein 4 (AP4), to

the MYC promoter, leading to the downregulation of MYC425. AP4 expression is

linked to EMT, metastasis, and poor prognosis in several cancers, including

breast513-515, suggesting that BETi may suppress metastatic progression, at least

in part, by modulating AP4 activity.

In another study by Andrieu and colleagues, JQ1 suppressed migration and

invasion in MDA‐MB‐231 and SUM149T cells422. In this case, JAG1 was identified

as a key target of BRD4. This gene encodes Jagged, a Notch receptor family ligand

involved in EMT, metastasis, proliferation, survival, and resistance to therapy516.

The clinical relevance of this finding is demonstrated by Kaplan-Meier analysis of

664 breast cancer patients517 revealing that high co-expression

of BRD4 and JAG1 is associated with shorter distant metastasis-free survival422.

BRD4 binding to the promoter of JAG1 is enhanced by the pro-inflammatory

cytokine interleukin 6 (IL6), which is linked to EMT, migration, invasion, and

metastasis and is secreted by the tumor microenvironment518-520 and leads to

increased Jagged1 protein levels. Increased Jagged1 in turn activates Notch1 to

promote migration and invasion. Consistently, treatment with JQ1 prevents the

recruitment of BRD4 to the JAG1 promoter, reducing Notch1 activation. BETi have previously been shown to have anti-inflammatory activity429,431, and these findings are further enhanced by the discovery that BETi can combat the pro-metastatic effects of IL6 from the tumor microenvironment by modulating Jagged1/Notch signaling. These data further support the notion that BETi can inhibit the secretion of pro-metastatic factors into the tumor microenvironment and may improve patient outcomes. However, the specific impact of BETi on metastatic spread was not assessed.

Another mechanism by which BRD4 can directly modulate motility and invasion of breast cancer cells involves its interaction with Twist, a transcription factor that plays an essential role in the activation of EMT521. When Twist is di-acetylated, it

interacts with the second BRD of BRD4455. This Twist-BRD4 complex drives

expression of a set of EMT genes and thus maintains mesenchymal characteristics

of TNBC cells. One of the gene targets of this complex is WNT5A. As mentioned

above, WNT5A is a secreted factor that regulates various aspects of cancer cell properties, including proliferation, self-renewal, migration, and invasion522.

The WNT5A gene has a putative super-enhancer, and binding of Twist to this

locus is important for the recruitment of BRD4, P-TEFb, and RNAPII to

the WNT5A enhancer and promoter455. BETi disrupts the interaction between

BRD4 and Twist, leading to the suppression of invasiveness in TNBC cells. Both

JQ1 treatment and gene silencing of BRD4 in five basal-like breast cancer cell

lines suppressed the expression of WNT5A as well as invasion and tumorsphere

formation. These data suggested that BRD4 may modulate metastatic outgrowth.

However, similar to the analyses by Andrieu and colleagues, this was not directly

tested. Rather, the authors reported that two BETi, JQ1 and MS417, inhibited

growth of primary SUM1315 tumors partially via the suppression

of WNT5A expression.

Only two studies have directly assessed the ability of BETi to impact the breast

cancer metastatic cascade in vivo. We found that JQ1 treatment reduces the

number of liver macrometastases in mice with tumors derived from the metastatic

TNBC cell line, MDA-MB-231472. However, a second study using two highly

metastatic cell lines (Mvt-1 and 6DT1) revealed that while I-BET151 lowered

primary tumor weight, it did not suppress the incidence of pulmonary metastasis393.

Several differences between the models used may explain this discrepancy. MDA-

MB-231 cells are a claudin-low human breast cancer cell line that was examined

in immune-compromised mice whereas Mvt-1 and 6DT1 murine mammary cancer

cells were studied as allografts in immune-competent mice. In addition, the

molecular classification of Mvt-1 and 6DT1 cell lines is mixed with both having

elements of luminal and claudin-low gene expression signatures523. The use of cell

lines that have their own unique transcriptomes could account for the apparent

differences in the impact of BETi on metastasis in these two studies. Lastly, these

studies interrogated metastatic potential to different organs. The impact of I-

BET151 on metastasis of Mvt-1 or 6DT1 to the liver or organs other than the lungs

was not assessed, and it is possible that BETi modulate microenvironmental

sensing in one tissue context but not another. Additional analysis of a broader

spectrum of tumor models will be necessary to elucidate the molecular drivers that

define the impact of BETi on metastasis.

1.4.3.7 BET inhibitors and metabolism

Deregulated cellular metabolism is a hallmark of cancer, with cancer cells having

different energy needs than normal cells due to increased cell division and

proliferation as well as altered access to nutrients485. Very little is known regarding

the impact of BETi on cellular metabolism in breast cancer. One study found the

BETi, XD14, significantly altered the expression of 67 metabolites in the ER+

MCF7 cell line524. These metabolites included amino acids, fatty acids, lipids, and

phospholipids and could be grouped into 12 pathways including those that regulate

amino acid levels. Eight amino acids as well as glucose were elevated following

XD14 treatment, suggesting that XD14-treated MFC7 cells consumed less energy

than vehicle-treated cells. Similarly, we found that “metabolism” was one of the top

Reactome pathways altered in HCC70 cells after treatment with JQ1411. To

complement these descriptive reports, mechanistic and functional studies are

necessary to clarify the effect of BETi on metabolism and how this impacts breast

cancer pathogenesis.

1.4.3.8 Mechanism(s) of action of BET inhibitors in breast cancer

BET proteins act as co-activators or co-repressors depending on their binding partners and cellular context. Considering their critical roles in modulating transcription, it is not surprising that gene expression analyses have revealed that

BETi alter expression of hundreds of genes in breast cancer

cells383,398,411,421,424,456,467,468,525. The genes impacted by BETi vary depending on

breast cancer subtype and the cell lines used. However, one of the most consistent

findings among many of these reports is that BET inhibition induces cell

cycle arrest. The majority of studies that performed cell cycle analysis on BETi-

treated breast cancer cells found that multiple ER+ and TNBC cell lines arrest in

the G1 phase as early as 24 h after drug addition420-426. This occurred following the

suppression of cell cycle genes411,468. Nonetheless, the mechanism by which cell

cycle arrest occurs seems to be subtype, and even cell line, dependent.

A major regulator of proliferation in varied cell types is MYC. In numerous non-

breast cancer models, BETi dramatically suppress expression of c-MYC, and overexpression of c-MYC can reduce sensitivity to BETi526,527. Thus, BETi have

been touted as “Myc inhibitors.” However, multiple studies have found that

neither MYC amplification nor BETi-mediated suppression of c-Myc were involved

in the response of TNBC to BETi, because BETi had an inconsistent impact on c-

Myc expression in various cell lines representing this subtype of breast

cancer420,421,425,472,473. Furthermore, overexpression of MYC in SUM149, HCC38,

and EVSAT cells did not induce JQ1 resistance457, indicating that modulation of

MYC likely contributes only modestly to the BETi responsiveness of TNBC cells.

In contrast, c-Myc suppression may play a more substantial role in the BETi

response of ER+ and HER2+ breast cancers. Several studies reported

downregulation of c-Myc in cell lines representing both of these subtypes, including

those with acquired drug resistance383,425,456,466-468. Furthermore, MYC-

overexpressing BT474 cells were less sensitive to JQ1 treatment383. It is possible

these results could be partially driven by the time point selected, as at least one

study showed c-Myc expression initially decreased but then rebounded in BETi-

treated MCF7 cells468. Another caveat of these studies is that they utilized a limited

number of cell lines which may not fully represent the response of the luminal and

HER2+ subtypes as a whole. Indeed, RT-PCR analysis of 24 breast cancer cell

lines representing the luminal, HER2+, and TNBC subtypes treated with JQ1

revealed variable regulation of MYC mRNA, and the suppression or upregulation

of MYC did not correlate with growth sensitivity to JQ1457. Interestingly, GSEA of

MDA-MB-231 and MDA-MB-468 cells treated with vehicle or OTX015 revealed that

MYC target genes were downregulated following OTX015 treatment even

though MYC itself was not suppressed421. This could indicate that BETi can

regulate MYC-responsive genes through a mechanism other than by the direct

transcriptional repression of the MYC gene itself.

In contrast to MYC, we found that BETi profoundly disrupted mitosis in TNBC,

partially due to the downregulation of Aurora kinases A and B (AURKA/B)472.

AURKA and AURKB play critical roles in multiple steps of mitosis, and altered expression of these genes induces polyploidy528-531. Indeed, treatment with JQ1

suppresses expression of AURKA and AURKB, leading to polyploidy in MDA-MB-

231 cells and multi-nucleation in several TNBC cell lines. The impact of BETi

on AURKA and AURKB gene expression is direct. BRD4 binds to the promoter

regions of both genes and JQ1 reduces BRD4 recruitment to these loci. In addition,

treatment of TNBC cells with an AURKA-selective (MLN8237) or an AURKB-

selective (AZD1152) inhibitor phenocopies JQ1; both induced multi-nucleation

followed by either senescence or apoptosis. These data suggest Aurora kinases

could potentially be used as biomarkers to predict response to BETi therapy.

We further reported that mitotic dysfunction induced by BETi results in mitotic

catastrophe, as indicated by the suppressed expression of mitosis/cytokinesis-

associated genes, increased mitotic timing, and acquisition of multi-nucleation or

induction of apoptosis in or immediately following mitotic exit411. The global

downregulation of genes involved in mitosis/cytokinesis was mediated, at least in

part, by the suppression of the critical mitosis regulator LIN9 by JQ1.

Silencing LIN9 alone suppressed similar cell cycle genes as JQ1 and led to multi-

nucleation, indicating that LIN9 is an important regulator of mitotic progression in

TNBC that is also a key target of BETi. In addition, genes that were highly

correlated with LIN9 in breast cancer or had a LIN9 binding site in HeLa cells were

more likely to be downregulated by JQ1 than those that were not correlated.

Interrogation of publically available datasets revealed that LIN9 is amplified and/or

overexpressed in two-thirds of basal breast cancers, and its expression is linked

to poor prognosis, underscoring its clinical impact on this disease and its potential

utility as a biomarker for predicting BETi responsiveness of TNBCs.

As in other cancers, SEs have been detected in several oncogenes in breast

cancer cell lines411,420,532-534. Shu, et al. identified 219 SEs in SUM159 cells and

159 SEs in SUM149 cells, and SEs were found at known gene drivers of TNBC,

including MYC and HIF1A420. Treatment of SUM159 and SUM149 cells with JQ1

led to a rapid (within 12 h) suppression of transcription of SE-associated genes,

and the number of SEs increased when cells developed BETi resistance. These

studies suggested that SE modulation may play a key role in mediating the effects

of BETi in breast cancers. However, we found that the impact of BETi on mitosis

in TNBC was not related to the suppression of SE-associated genes, as LIN9 and

four other JQ1-regulated master mitosis transcription factors were not associated

with SEs411. Cancer cells are particularly sensitive not only to SE disruption but

also to mitotic catastrophe535. We thus hypothesize that the selectivity of BETi for

cancer cells, at least in TNBC, is due to a combination of loss of SEs and mitotic

dysregulation via the induction of mitotic catastrophe.

The observed downregulation of critical genes in breast cancer cells following BETi

treatment can also result from the prevention of transcriptional elongation rather than disruption of SEs. In a study using ER+ breast cancer cells, Sengupta, et al.

found that while BETi did not impact ERα expression or recruitment to promoters

and EREs, it does alter transcription of estrogen-target genes423. JQ1 only slightly

suppressed the recruitment of RNAPII to the promoters of the estrogen-target

genes TFF1, GREB1, and XBP1. In contrast, there was a large reduction in

RNAPII binding to the bodies of these genes following JQ1 treatment. This

indicates that the transcription complex was able to form and initiate transcription

in the presence of BETi but transcription elongation was inhibited. BRD4 recruits

P-TEFb to the chromatin which releases RNAPII from its paused state at

promoters to stimulate elongation255. It is likely that the impact of BETi on elongation is due to the loss of P-TEFb presence at gene promoters, thus restraining RNAPII at the promoter and preventing its progression to the gene body.

1.4.3.9 Resistance to BET inhibitors

Thus far, two mechanisms of resistance to BETi have been identified in breast

cancer. One of these is activation of the PI3K pathway (Figure 1.5A). JQ1 did not

inhibit proliferation of cell lines derived from mouse mammary tumors with

amplified Myc in conjunction with either an activating mutation in PI3K (MCCL-

278) or deletion of Pten (MCCL-357), despite suppressing MYC protein expression at higher (≥2 μM) doses536. A second BETi, MS417, did suppress

growth in both cell lines by at least 50% but only at high (≥4 μM) doses. Similar

observations were made using the TNBC cell line SUM159 which has an activating

mutation in PIK3CA and amplified MYC. These results indicate that tumors with activated PI3K pathway and high MYC expression could be intrinsically resistant

to BETi. Supporting this conclusion, analysis of shRNA dropout screens in 82

breast cancer and mammary epithelial cell lines using the algorithm siMEM also

identified PIK3CA mutation as a potential BETi resistance mechanism457. Cell lines that did not die in response to BETi were more likely to have a mutated form of PIK3CA, and SKBR3 cells overexpressing wild type or mutated PIK3CA were

less susceptible to JQ1-induced growth inhibition. Although tumors with altered

PIK3CA are less sensitive than their wild type counterparts, they can still respond

to BETi to some degree. Indeed, the BETi, MS417, suppressed growth of both

MCCL-278 and MCCL-357 tumors by about 50%536, and a separate study found

growth of SUM159 xenografts was significantly reduced in response to JQ1

treatment420. The partial responsiveness of PIK3CA mutant cancers to BETi

suggests that combining AKT pathway inhibitors with BETi should provide an

effective therapeutic approach for these cancers. This is supported by preclinical in

vivo data that will be discussed below in the section focused on combination

therapies.

A second mode of BETi resistance was discovered by Shu, et al.420. In this case,

the authors developed BETi-resistant SUM159 (SUM159R) and SUM149

(SUM149R) cells via long-term culture with gradually increasing doses of JQ1.

BRD2, BRD3, and BRD4 still localized to the chromatin despite the presence of

JQ1, indicating that a factor that controls BET protein binding to other chromatin-

associated proteins may be altered. SUM149R and SUM159R cells had higher

levels of phosphorylated BRD4 (pBRD4), and there was enhanced binding of

pBRD4 to MED1 compared to unphosphorylated BRD4, allowing pBRD4 to

localize to the chromatin even in the presence of BETi (Figure 1.5B). PP2A was

identified as a BRD4 phosphatase, and PP2A activity was diminished in BETi-

resistant cells. These data indicate phosphorylation of BRD4 can confer BRD-

independent functions and maintain activity of BRD4, thus providing a mechanism

to overcome the effects of inhibitors targeted to the BET BRDs. They further

suggest that activators of PP2A may prevent BETi resistance. Lastly, these studies support the development of alternative approaches for inhibiting BET protein function that do not rely on the BRDs. Inroads in this last effort will be discussed later in this review.

1.4.3.10 BET inhibitor adverse effects

BETi are generally non-toxic in mice. Mice treated with BETi did not lose weight424,453,467,468,472,537, and the only adverse effects observed were reversible

male infertility and reduced long-term memory formation443,538. BETi can also

decrease uterine size and weight in female mice due to the suppression of

estrogen-target genes398, but the potential impact of BETi on fertility in female

patients is unknown. In the normal adult mammary gland, BETi did not alter rates

of proliferation or apoptosis and had no impact on ductal branching following one

week of treatment, indicating BETi, at least in the short-term, do not negatively impact the virgin adult gland472. However, it is unclear if and how long-term BETi

treatment will affect the normal breast architecture or function in women, especially

those who are pre-menopausal. Estrogen signaling is critical for terminal end bud

formation during puberty in mice, and estrogen is involved in the formation of milk

ducts in women during the menstrual cycle539. Because BETi suppress expression

of estrogen-target genes423, it will be important to determine if BETi alter the

expression of such genes in the normal adult breast in order to predict the potential impact of BETi on this tissue, especially in pre-menopausal women.

One study has also utilized BRD4 silencing to predict untoward effects of BETi. In mice, loss of BRD4 reduced the T lymphoid and hematopoetic stem cell populations, indicating BETi may prevent normal hematopoiesis540. Furthermore,

after five weeks of silencing, the mice experienced alopecia and skin hyperplasia.

These animals also had reduced cellular diversity in the intestine, including a loss

of several cell types that compose the secretory lineage as well as stem cells in

the crypts. While the intestines of shBRD4 mice seemed to function normally, they

displayed impaired regenerative capability following insult by radiation

or chemotherapy agents that are known to damage the intestine. This suggests

that combining BETi with chemotherapies may result in intolerable gastrointestinal

and skin effects in patients.

Following these preclinical studies, BETi are currently being assessed in clinical

trials in a wide variety of hematologic and solid tumors, and results from three

phase 1 studies assessing OTX015 in acute leukemia (41 patients), NUT midline

carcinoma (NMC; 4 patients), lymphoma (33 patients), and multiple myeloma (12

patients) have been published541-543. Across these three studies, the most frequent

grade 3–4 adverse event was thrombocytopenia, with other common side effects

including diarrhea, anemia, neutropenia, fatigue, hyperglycemia, and nausea. The

NMC study is the only trial thus far to assess any clinical activity of BETi in solid

tumors. NMC is a rare but highly aggressive disease with a median survival of less

than one year. It is driven by the translocation of BRD3 or BRD4 which results in

the fusion of BRD3/BRD4 and NUT genes544,545. The encoded BRD3/4-NUT fusion

proteins have oncogenic activity and prevent epithelial differentiation, thereby

contributing to the poorly differentiated phenotype of cells in NMC tumors545. Of

the four NMC patients treated with OTX015 in the published phase 1 trial, one

patient achieved disease stabilization while two experienced tumor regression541.

Although these three patients initially rapidly responded to OTX015, they all

relapsed, highlighting the importance of developing effective combination

therapies with BETi. However, two patients survived significantly longer than the

median 6.7 month survival time for NMC (18 and 19 months from initial diagnosis).

Given the phase I nature of this study and the small cohort size, it is not possible

to make broad conclusions regarding BETi efficacy in this disease. Nevertheless,

it is promising that BETi more than doubled the survival time for two of four

patients.

1.4.4 Alternative BET protein targeting mechanisms

As described above, small molecule competitive inhibitors of BET proteins have

profound anti-tumor effects in breast cancer models via suppression of multiple

oncogenic pathways. However, the utility of BETi in the clinic could be hampered

by toxicity associated with long-term treatment and/or the development of drug

resistance. Additional therapeutic approaches targeting BET proteins are currently

being developed, including combining BETi with other therapies (Figure 1.6),

degrading BET proteins, and using dual kinase/BET protein inhibitors. These

strategies and their performance in breast cancer models are reviewed below. We

also discuss CDK7 inhibitors which may function in a similar manner to BETi and

thus may provide similar therapeutic benefits to breast cancer patients.

1.4.4.1 Combination therapies

BETi synergize with anti-mitotic (docetaxel and vinorelbine) and DNA-damaging

(cisplatin and carboplatin) agents424,453. Treatment of the TNBC cell lines MDA-

MB-231 and HS578T with any of these drugs in combination with BETi induces

greater apoptosis and growth inhibition than the single agents. It has also recently

been reported that BETi synergize with platinum-containing agents in HGSOC546,

which has a high degree of genetic and transcriptomic similarity with TNBC21.

Therapies that are effective in TNBC or ovarian cancer may also be efficacious in

the other21, further supporting the continued study of BETi

and chemotherapy combinations in TNBC, with a particular need to assess its

synergy with platinum-containing agents.

Another strategy that could be effective in TNBC is combining BETi with PARP

inhibitors (PARPi). Three separate BETi (JQ1, I-BET762, and OTX015) have been

shown to synergize with two PARPi (olaparib and veliparib) in MDA-MB-231 cells,

with the combination of BETi and PARPi also being effective in ovarian and

prostate cancers453. In mice, combined BET protein and PARP inhibition

significantly reduced growth of orthotopic MDA-MB-231 tumors compared to either

agent alone, and this treatment strategy did not induce obvious toxicity. In addition

to increasing growth inhibition in anchorage-dependent and anchorage-

independent conditions, adding BETi to PARPi treatment enhanced DNA damage.

JQ1 decreased homologous recombination in response to ionizing radiation and

instead increased repair by non-homologous end joining, supporting a mechanism

for the synergy between BETi and PARPi. This resulted, at least in part, from the

JQ1-mediated suppression of BRCA1 and RAD51, genes that encode proteins that repair DNA damage via homologous recombination. Reduced binding of

BRD2/3/4 at the promoter regions of these genes was observed in response to

BETi. Interestingly, cancer cell lines with BRD4 amplification were less sensitive to PARPi, suggesting that adding BETi to PARPi treatment could increase PARPi sensitivity of tumors with amplification of BRD4. PARPi are particularly effective in tumors that are unable to perform repair of double strand breaks by homologous recombination, such as those with BRCA1/2 mutations143,547. However,

while BRCA1/2mutations are frequent in TNBC, a large number of breast cancers

maintain the ability to perform homologous recombination548. Data from this study

indicate that combining BETi with PARPi can increase the efficacy of PARPi in

tumors that lack BRCA1/2 mutations and those that acquire homologous repair

proficiency.

HDAC inhibitors (HDACi) have anti-tumor effects, are being evaluated in clinical

trials, and several have been approved for use in specific types of cancer549.

Because BETi inhibit the recognition of acetylated histones and HDACi block

histone acetylation, it was logical to expect that these drug classes would

demonstrate synergism. Indeed, BETi synergize with HDACi in AML, pancreatic

ductal carcinoma, melanoma, murine lymphoma, and T-cell acute lymphoblastic

leukemia (T-ALL)377-381. A few studies have found that this drug combination could

also be a plausible treatment strategy in breast cancer. Comparing drug signatures

of four HDACi (trichostatin A, vorinostat, scriptaid, and CP‐690334‐01) from the

Connectivity Map database (CMAP)550 with GSEA results from two TNBC cell lines

(MDA‐MB‐231 and MDA-MB-468) treated with OTX015 revealed that HDACi and

OTX015 regulated similar genes, further supporting the potential for synergy

between these two drug classes in TNBC421. Indeed, combination treatment of

HDACi with BETi suppressed the growth of MCF7, MDA-MB-231, and BT549 cells in vitro and inhibited growth of two PDX TNBC tumor models in vivo to a greater extent than either the HDACi or BETi alone456,537. To identify mechanisms

of synergy, Borbely and colleagues performed gene expression profiling in MDA‐

MB-231 cells. This revealed that several members of the USP17 family of

deubiquitinating enzymes were upregulated following co-treatment with HDACi

and BETi456. The upregulation of USP17 and USP17L5 was confirmed in MCF7,

T47D, MDA-MB-231, and BT549 cells, indicating activity in both luminal and TNBC

cells. USP17 suppresses proliferation, and part of its anti-proliferative effect stems

from its ability to inhibit Ras, thereby decreasing the activity of the Ras/MAPK

pathway551,552. Combining BETi and HDACi repressed Ras at the protein level as

well as phosphorylation of MEK and ERK1/2, indicating that changes in Ras

function may underlie the synergy between the two agents. Recently, a dual

HDAC/BRD4 inhibitor has been developed that inhibits growth of AML and chronic

myelogenous leukemia (CML), in vitro553-555. These studies suggest that either

combining HDACi and BETi or use of a dual HDAC/BRD4 inhibitor may have

efficacy in at least two subtypes of breast cancer.

BETi have also been shown to combat resistance to standard therapies in various

breast cancer models. Tamoxifen resistance is a major clinical hurdle in the

treatment of ER+ breast cancers. Roughly half of ER+ cancers do not respond to

tamoxifen and as many as 40% of patients whose tumors initially respond relapse

and succumb to their disease556. This highlights the need to develop new treatment

options for women with de novo or acquired tamoxifen resistance. Of the patients

whose tumors acquire resistance to tamoxifen, 20% respond to second-line

treatment with aromatase inhibitors or the ERα degrader fulvestrant. In tamoxifen-

resistant (Tam-R) MCF-7 xenografts, BETi can restore fulvestrant sensitivity468.

Analysis of these tumors revealed reduced ERα protein expression and decreased

Ki67 and histone H3 p‐Ser10 staining in tumors treated with the combination

therapy compared to either agent alone. These results suggest that combining

BETi with fulvestrant could provide a novel therapeutic option for women with Tam-

R breast cancer. Indeed, the BETi GS-5892 is currently being investigated in a

phase 1 trial in combination with fulvestrant or the aromatase

inhibitor exemestane in post-menopausal women with advanced ER+ breast

cancer (NCT02392611).

A second study found that estrogen regulates the binding of BRD4 to both the

transcription start site of estrogen-dependent genes as well as distal ERα-

regulated enhancers406. The presence of H4K12ac was highly correlated with

BRD4 binding at these loci. When comparing global H4K12ac in MCF7 (ER+

breast cancer) to MCF10A (ER- non-transformed mammary epithelium) cells,

MCF7 cells had higher levels of this histone mark. In addition, treatment of MCF7

cells with fulvestrant globally reduced H4K12ac levels, indicating that BET proteins

work with ERα to position these marks. These results support the finding by Feng,

et al. that combining BETi with fulvestrant or other anti-estrogens could be a therapeutic strategy in Tam-R ER+ breast cancers468. These findings also suggest

the utilization of agents that target HATs and HDACs (described above) that modulate H4K12 acetylation could prevent BRD4 localization to the chromatin and thus have anti-tumor effects in ER+ breast cancers.

Numerous kinase inhibitors have been approved for use in patients with breast cancer. These agents target receptor tyrosine kinases (RTKs) as well as various signaling intermediates. However, acquired resistance to kinase inhibitors is common and can occur via multiple mechanisms, including mutation of the targeted protein(s) or the activation of alternative pathways or pathway members, such as RTKs, that bypass the suppression of the targeted kinase and maintain activity of an oncogenic pathway557,558. It is critical to develop new therapeutic

options that will prevent or overcome resistance to kinase inhibitors to extend their

clinical utility and prevent disease recurrence. In this regard, BETi have been

shown to synergize with a wide variety of kinase inhibitors. Nieto-Jiménez, et al.

performed gene expression analysis of a publically available

dataset559,560 comparing normal epithelial cells and basal-like breast tumors to

identify differentially expressed kinases426. Several druggable mitotic kinases,

including AURKA/B and polo-like kinase 1 (PLK1), were expressed at higher levels

in breast tumors compared to the normal epithelium. The IC50s for drugs targeting

these kinases were determined in four basal-like cell lines (MDA-MB-231, HS578T,

BT549, and HCC3153) with volasertib, a PLK1 inhibitor, having the lowest IC50 in

all four cell lines. Moreover, this drug synergized with JQ1 to repress growth and

stimulate apoptosis. This is particularly notable given our studies that have

demonstrated that BETi induce profound mitotic dysfunction in TNBC cells and

underscores the potential utility of targeting mitotic vulnerabilities through varied

avenues411.

Using the connectivity maps (CMAP) described above to compare drug signatures

to the results of gene expression profiling of two TNBC cell lines treated with

OTX015, Vázquez, et al. found that OTX015-target genes overlapped with those that are modulated by a number of agents, including the PI3K inhibitor

LY294002421. This highlights the potential utility of combining BETi with agents

targeting the PI3K/AKT/mTOR pathway. Treatment of two TNBC cell lines (MDA‐

MB‐231 and HCC1937) with a combination of OTX015 and the mTOR inhibitor

everolimus for 72 h had an additive effect, suppressing growth in both lines. In

contrast, the combination was antagonistic in MDA-MB-468 cells. It is unclear what

determines the outcome of combining BETi and inhibitors of the PI3K/AKT/mTOR

pathway, and this warrants future study. The in vivo efficacy of combining OTX015

and everolimus was also assessed. Mice with xenografted tumors derived from

MDA-MB-231 cells were treated with OTX015, everolimus, or the combination for

four weeks. As previously reported561, the MDA-MB‐231 model was resistant to

mTOR inhibition. In contrast, OTX015-treated tumors were significantly smaller

than vehicle-treated tumors. More importantly, the combination of these two drugs

significantly reduced tumor size and was more effective than either single agent.

100

In addition, tumors treated with the combination remained smaller than tumors

from the other treatment groups 20 days after the end of treatment, indicating a

durable response to BETi/everolimus in this model. Notably, the combination of

OTX015 and everolimus was also more effective than treatment with paclitaxel,

the current standard of care for TNBC562.

In contrast to TNBC, mTOR inhibitors are much more efficacious in luminal breast

cancer models, and everolimus is currently FDA-approved for treating patients with

ER+ breast cancers when used in combination with exemestane563. However,

everolimus resistance is common and understanding the mechanisms underlying

resistance may reveal new therapeutic approaches. Relevant to BETi, enhanced

BRD4 binding at the MYC locus can confer resistance to everolimus in parental

ER+ breast cancer cells and in long-term estrogen deprived (LTED) ER+ breast

cancer cells that mimic tumors with acquired resistance to tamoxifen or aromatase

inhibitors466,564. Treatment of everolimus-resistant (Eve-R) parental and Eve-R

LTED MCF7 cells with JQ1 reduced BRD4 binding to the MYC gene and suppressed MYC expression466. JQ1 also resensitized Eve-R parental and LTED

cells to everolimus, as the combination treatment had a greater growth

suppressive effect and induced cell death. To assess the in vivo efficacy of

combining BETi and everolimus, MCF7 xenografts were treated with vehicle, JQ1,

everolimus, or JQ1 + everolimus for three weeks. Everolimus suppressed tumor

growth by approximately 50%, while JQ1 alone had no impact. However,

combining JQ1 and everolimus synergized to significantly reduce tumor size

compared to either single agent. In addition, when treatment was removed and

101

tumors were allowed to regrow for 25 days, tumors that had been treated with the

combination therapy were significantly smaller than everolimus-treated tumors. A

second study confirmed these findings. In this case, JQ1 synergized with the

mTOR inhibitors rapamycin or Torin to suppress growth of MCF7 and T47D

cells457. MCF7 xenografts began to regress within 15 days of JQ1 and everolimus

co-treatment, while JQ1 alone did not impact tumor growth and everolimus alone

suppressed tumor growth but did not induce regression. These results further

affirm the potential addition of BETi to everolimus to provide greater therapeutic

benefit to patients with ER+ breast cancer and potentially overcome resistance to

mTOR inhibitors. It will be necessary in the future to assess if these findings are

applicable to other mTOR-targeting agents.

Lapatinib, a reversible competitive inhibitor of EGFR and HER2, is FDA-approved

for use in patients with HER2+ breast cancer who have already been treated with

an anthracycline, a taxane, or trastuzumab. While tumors initially respond to this

small molecule, resistance is common565. During the development of resistance,

cancer cells undergo kinome reprogramming wherein a variety of protein kinases

are upregulated to overcome loss of HER2-mediated signaling. Stuhlmiller, et al.

found that an entire network of kinases are altered within 48 h of lapatinib treatment in vitro, and the kinases responsible for lapatinib resistance were cell- line dependent383. The authors reasoned that utilizing a single kinase inhibitor

would be unable to address the heterogeneity of response to lapatinib due to the

variety of kinases that drive lapatinib resistance. Instead, they focused on

preventing the adaptive kinome response by targeting epigenetic modulating

102

enzymes and found BETi could not only resensitize resistant HER2+ cells to

lapatinib but also prevent the onset of lapatinib resistance. Combination treatment

of lapatinib with JQ1 increased the extent of apoptosis and was more effective than

combinations of lapatinib with dasatinib (Bcr-Abl and Src family kinase inhibitor) or

other kinase inhibitors targeting IGF1R/INSR (BMS754), FAK (PF228), or FGFR

(BGJ298) in parental and lapatinib resistant (Lap-R) cells. In the presence of

lapatinib, BETi inhibited expression of the kinases involved in lapatinib resistance

in both parental and Lap-R HER2+ cells. BETi prevented BRD4 localization to the

promoters and enhancers of lapatinib-responsive genes and decreased the

accumulation of p-Ser2 RNAPII at their promoters and gene bodies. Addition of

BETi to lapatinib further inhibited the recruitment of BRD4 and RNAPII to the

chromatin. This suggests that the mechanism by which BETi prevent lapatinib-

induced kinome reprogramming is by inhibiting the recruitment and formation of

the P-TEFb complex with BRD4 and RNAPII, thus leaving RNAPII paused at the

promoters of lapatinib-responsive genes. This mechanism is consistent with the

model proposed by Sengupta, et al. for ER+ breast cancer where BETi prevent

transcription elongation at estrogen-target genes423 and suggests that a common

path by which BETi prevents therapeutic resistance is through blocking the reprogramming of the active genome in response to various growth-inhibitory drugs.

The MEK inhibitor trametinib is approved for use in patients with melanoma and non-small cell lung cancer with activated BRAF and is currently being evaluated for efficacy in TNBC. While tumors initially respond, resistance occurs due to the

103

upregulation and activation of other kinase pathways that can overcome

dependence on the MEK-ERK signaling pathway566. In TNBC cell culture, mouse

xenografts, and biopsied TNBC patient tumors, trametinib treatment quickly

induced the upregulation of several tyrosine kinases with dramatic restructuring of

the enhancer landscape and acquisition of over 1000 enhancers, including

numerous SEs532. This is very similar to the kinome reprogramming observed in lapatinib-treated breast cancer cells discussed above, and, consistent with that study, BETi can collaborate with trametinib in SUM159PT and MDA-MB-231 cells to inhibit proliferation better than either agent alone. Adding a BETi also prevents the de novo formation of typical enhancer and SE regions observed with trametinib treatment alone. As anticipated, the combination treatment also blocked the upregulation of several genes involved in the adaptive response. While the authors did not show evidence of apoptosis, the combination also increased the expression of BIM, suggesting an increase in BAX/BAK priming. Lastly, BETi were able to overcome acquired resistance to trametinib in SUM159 cells, similar to what was observed in lapatinib-resistant HER2+ cells383. In vivo, the combination of

trametinib and I-BET151 significantly inhibited tumor growth compared to either

agent alone in SUM159PT xenografts as well as in two TNBC orthotopic syngeneic

mouse tumor transplant models. This study provides foundational support for

clinically assessing the utility of adding BETi to kinase inhibitors to prevent

acquired resistance that occurs via adaptive bypass signaling.

In addition to using BETi to overcoming resistance to various therapies,

approaches for blocking resistance to BETi themselves have also been identified.

104

As mentioned above, Shu, et al. discovered that hyperphosphorylated BRD4 can confer resistance to BETi in TNBC420. Either targeting CK2, the kinase responsible

for phosphorylating BRD4, using the CK2 inhibitor CX-4945 or activating the

phosphatase PP2A using perphenazine suppressed growth of both parental and

BETi-resistant TNBC cell lines when used in concert with JQ1. In addition to

excessive phosphorylation of BRD4 in BETi resistant cells, these cells acquire an

SE at the gene encoding Bcl-xL. ABT737, a pan-Bcl-2 family inhibitor, synergized

with JQ1 in both parental and resistant cell lines.

As previously mentioned, human breast cancer and mouse mammary epithelial

cell lines with either mutated PI3K alone or mutated PI3K plus amplified MYC are

resistant to BETi457,536. Co-inhibition of BET proteins and PI3K effectively

suppressed growth and induced apoptosis of breast cancer cells. T47D and MCF7 cells treated with JQ1 and A66 (PIK3CA-selective inhibitor) underwent greater growth inhibition compared to JQ1 alone or JQ1 plus TGX221, a PIK3CB-selective inhibitor457. The combination of JQ1 and PIK3CAi was also effective in the basal

cell line SUM159 which has an activating PIK3CA mutation. In another study, the

combination of BETi and GDC-0941, a PI3K inhibitor, suppressed growth and induced cell death more than either single agent in vitro and in vivo536. Treatment

with GDC-0941 alone induced rebound activation of AKT as well as enhanced

expression of a number of RTKs, both of which indicate reactivation of the PI3K

signaling pathway, an established mechanism of resistance. However, BETi

prevented these compensatory changes by reducing BRD4 binding to GDC-0941- responsive genes. Adding BETi to GDC-0941 blocked the reactivation of the PI3K

105

pathway in cell lines representing numerous cancer types, including seven breast

cancer cell lines, six of which had activating mutations in the PI3K pathway but

lacked MYC amplification. Hence, dual inhibition of BET proteins and PI3K could

impact a broad spectrum of cancers by preventing feedback reactivation of the

PI3K pathway. These data also suggest combining BETi with inhibitors that target

the PI3K signaling pathway will be efficacious in patients with PIK3CA mutations

and may combat PIK3CA-mediated BETi resistance, underscoring the need for

clinical trials that assess the efficacy of combining BETi with PI3K/AKT/mTOR-

targeting agents.

1.4.4.2 BET protein degraders

Generally speaking, small-molecule inhibitors are ideal agents for cancer therapy

and have provided immense clinical benefit. However, they do have notable

disadvantages567, including the typical requirement for high concentrations to

provide efficacy, increasing the risk for off-target effects. Small molecules that act

by competitive inhibition such as ATP analogs also leave the targeted pathogenic

protein(s) intact, providing an opportunity for resistance to develop. Small-

molecule-based proteolysis-targeting chimeras (PROTACs) present an improved small-molecule approach that reduces the expression of target proteins by marking them for proteasomal degradation. A PROTAC is a heterobifunctional compound that combines two moieties: one that targets the protein of interest while the other binds a specific E3 ligase, effectively recruiting the target protein to the E3 ligase for ubiquitination and subsequent degradation by the proteasome567. Several BET

PROTACS have been developed, including dBET1, ARV-825, and ARV-771, all of

106

which degrade BET proteins. These inhibitors link JQ1 and thalidomide (engaging

cereblon E3 ligase), OTX-015 and pomalidomide (also engaging cereblon E3 ligase), or OTX-015 and a ligand that targets the von Hippel-Landau E3 ligase, and have shown greater efficacy against AML, Burkitt’s lymphoma, mantle cell lymphoma, and castration resistant prostate cancer cell lines and PDX models compared to standard BETi treatment568-571.

Thus far, only one report has examined the utility of PROTAC technology in breast cancer. Bai, et al. developed a PROTAC molecule (BETd-246) that tethers the

potent BETi, BETi-211, to thalidomide525. Treatment of several TNBC cell lines

with BETd-246 effectively and selectively degraded the majority of BRD2, BRD3,

and BRD4 within three hours. BETd-246 suppressed growth of 13 TNBC cell lines

and was more potent than the parent BETi-211 compound. In the majority of TNBC

cell lines tested, BETd-246 rapidly induced apoptosis, activating multiple apoptotic

pathways within as little as five hours in some cell lines. Interestingly, RNA-seq

analysis revealed BETd-246 evoked a unique transcriptional response compared

to BETi-211 in TNBC cells, suggesting that the mechanism(s) of action of BETi

may not be the same as BET PROTACs. BETd-246 also synergized with two dual

Bcl-2/Bcl-xL inhibitors and a selective Bcl‐xL inhibitor in six TNBC cell lines to

produce strong anti-proliferative and pro-apoptotic responses. When examined in

vivo, no toxicity was observed in several models representing immunocompetent

or immunodeficient mice. Treatment of the Washington Human in Mouse (WHIM)

PDX model and MDA-MB-453 xenografts with BETd-246 suppressed tumor

growth and, at higher doses (10 mg/kg vs. 3 mg/kg), induced partial tumor

107

regression. In these tumors, BET protein levels were significantly reduced within

one hour after the first dose but rebounded once BETd-246 was cleared from the

tissue (around 12 h). In contrast to these results, BETd‐246 was not effective in

MDA‐MB-231 or MDA-MB-468 xenografts due to poor delivery of the drug to the tumor tissue. This led to the optimization of BETd-246 to generate BETd‐260. This derivative had anti-proliferative activity in TNBC cells, in vitro, but more importantly, suppressed growth of MDA‐MB‐231 and MDA-MB-468 xenografts without inducing observable toxic effects. Overall, these data suggest that

PROTAC approaches may be more efficacious for breast cancer than BET inhibition and may prevent some of the acquired resistance mechanisms identified for BETi treatment due to the ability of PROTACs to selectively degrade their protein targets.

1.4.4.3 Dual protein kinase/BET protein inhibitors

One of the hallmarks of cancer is sustained proliferative signaling via the dysregulation of signaling pathways and aberrant regulation of signaling kinases485. While oncogenic kinases are attractive therapeutic targets,

compensatory pathways that cause acquired resistance have limited their long-

term therapeutic impact572,573. Combination therapies have been proposed to

subvert these issues. However, this approach has its own complications, including

toxicities, drug–drug interactions, and increased costs. Ciceri, et al. proposed an

alternative approach to combination therapy that could provide similar benefits

while avoiding the same clinical development challenges by targeting multiple proteins with a single small molecule574. They screened 628 kinase inhibitors for

108

evidence of binding to BRD4 and found that nine directly interacted with BRD4 and prevented its binding to acetylated H4. These kinase inhibitors docked into the first

BRD of BRD4 and included the PLK inhibitors BI-2536 and volasertib, the JAK inhibitor TG-101348, and the PI3K-mTOR inhibitors GSK2636771 and PP-242.

Using BI‐2536 and TG‐101348 as exemplars, they further confirmed that these kinase inhibitors prevented the recruitment of BRD4 to chromatin and suppressed expression of c-Myc in the multiple myeloma cell line MM.1S. In addition, TG-

101348 inhibited growth of MM.1S cells. Growth inhibition was restricted to these two inhibitors, as other selective PLK and JAK inhibitors did not interact with BRD4, repress c-Myc, or inhibit growth in MM.1S cells. Following this preliminary study, additional dual kinase/BET inhibitors have since been identified, and their efficacy has been assessed in a limited number of cancer models575-581. While these

preliminary studies are promising, more work is necessary to fully characterize the

efficacy and mechanism(s) of action of dual kinase/BET inhibitors in different

cancer types, including breast cancer.

1.4.4.4 CDK7 inhibitors

One of the mechanisms of action of BETi is the dismantling of SEs. Thus, agents

that disrupt SE architecture through different mechanisms should also be useful

either as single agents or in combination with BETi. In this regard, CDK7 inhibitors

(CDK7i) have recently been shown to alter the composition of SEs in many cell

types. Similar to BRD4, inhibition of CDK7 results in the preferential

downregulation of SE-associated genes, including RUNX1 in T-

ALL, MYCN and PHOX2B in neuroblastoma, SOX2 and SOX4 in small cell lung

109

cancer, and PAK4 and YAP1 in esophageal squamous cell carcinoma582-585.

CDK7 is responsible for phosphorylating, and thus activating, cell cycle CDKs as

well as RNAPII at transcriptional start sites586. CDK7, cyclin H, and MAT1 form a

subcomplex of TFIIH which plays roles in both transcription and DNA nucleotide

excision repair. Before transcription can begin, TFIIH is recruited to the pre-

initiation complex and unwinds a small portion of the DNA, giving RNAPII access

to the transcriptional start site of the target gene. CDK7 phosphorylates RNAPII at

Ser5, the first of two phosphorylation events required before transcription

elongation can occur.

CDK7, cyclin H, and MAT1 expression levels are elevated in breast cancers,

particularly in ER+ tumors, compared to normal tissue and are associated with

better clinical outcome587. However, analyses of publically available datasets and

tissue microarrays have revealed high CDK7 protein expression is associated with

poor outcome of TNBC patients588, suggesting the potential utility of inhibiting

CDK7 in TNBC. Several small molecules have been developed that target CDK7,

and they have shown preclinical efficacy in breast cancer. CDK7i prevent

phosphorylation of RNAPII, thus inhibiting transcription589-591. This has a profound

effect on short-lived proteins, including the depletion of c-Myc and the anti- apoptotic proteins Mcl-1 and X-linked inhibitor of apoptosis protein (XIAP)591,592. In

ER+ breast cancer cells, CDK7i (BS‐181, roscovitine, ICEC-0782, and SNS-032) suppressed proliferation and induced cell cycle arrest and apoptosis589-593. These drugs also inhibited growth of MCF7 and MDA‐MB-453 xenografts without inducing overt toxicity or impacting the normal mammary gland589,591,592. In TNBC, shRNA-

110

mediated knockdown of CDK7 reduced proliferation and migration and increased sensitivity to doxorubicin supporting the potential utility of CDK7i in TNBC588.

Indeed, the CDK7i THZ1 and BS-181 suppressed proliferation, induced apoptosis,

reduced phosphorylation of CDK1 and RNAPII, and lowered expression of c-Myc

and Mcl-1 in TNBC models. THZ1 additionally synergized with the BH3 mimetic

ABT-263 to suppress growth and induce apoptosis.

While BETi and CDK7i produce similar outcomes in cancer cells, BRD4 and CDK7

control different sets of genes. For example, in diffuse intrinsic pontine glioma

(DIPG), JQ1 altered expression of genes involved in nervous system development

while THZ1 disrupted transcription-associated genes594. Inhibition of CDK7 may

thus provide an alternative to BETi, and it will be important in the future to

determine which inhibitor class is more effective. It may also be possible to use co-

inhibition of BET proteins and CDK7 as a therapeutic strategy, as JQ1 and THZ1

synergized in DIPG594.

1.4.5 Future perspectives and conclusions

BET proteins are epigenetic readers that bind acetylated lysines in histones and

transcription factors and tether transcriptional regulators to chromatin, thus

impacting gene expression. The functions of BET proteins are dictated by their

binding partners, which are context dependent. Several diseases, including cancer, have been linked to excessive BET activity, highlighting the clinical significance of developing drugs that target this protein family. In breast cancer,

BETi impact the expression of genes that drive multiple hallmarks of cancer485 such as inducing angiogenesis, resisting programmed cell death,

111

activating invasion and metastasis, and deregulating cellular energetics (Figure

1.7). Furthermore, multiple mechanisms of action of BETi have been identified in breast cancer, including the suppression of LIN9 expression and the disruption of

SEs. It is unlikely that the efficacy of BETi is due to the suppression of a single gene or signaling pathway. Rather, BETi likely alter several key oncogenic pathways that work in concert to suppress cancer-associated pathologies.

Breast cancer is a highly heterogeneous disease comprised of several distinct subtypes, each of which is driven by a unique transcriptome. Despite the diversity within this collection of diseases, BETi elicit anti-tumor responses in several subtypes of breast cancer. However, to date there are no studies evaluating the response of inflammatory breast cancer (IBC) to BET-targeting agents. IBC is a rare but very aggressive subtype of breast cancer that is associated with poor prognosis. The role of BET proteins in inflammation and the observation that suppression of BET proteins reduces the expression of pro-inflammatory cytokines420,422,427-429,465 suggests BETi may provide a new treatment avenue for

this poorly understood disease, and it will be important to test this possibility.

While BETi have a profound effect on hematologic cancers, many studies of BETi

in solid tumors, including breast cancer, have concluded that BETi are more likely

to induce tumor stasis rather than regression. Understanding the mechanism(s) of

action of BETi in the individual subtypes of breast cancer should aid in the selection

of appropriate patients for BETi therapy. For example, we discovered that LIN9 is

a critical mediator of the BETi response in TNBC411. This suggests tumors with

high LIN9 expression may be particularly responsive to treatment with BETi, and

112

LIN9 could potentially be used as a predictor of response to BETi in the clinic. In

addition, single-agent therapies are risky due to the likelihood of acquired

resistance. Therefore, it is important to identify alternative strategies to single-

agent BETi treatment to improve their clinical utility. Numerous preclinical

studies have revealed that BETi can be combined with agents targeting various

oncogenic pathways to induce more durable, improved responses over single-

agent treatment and to overcome acquired resistance. Many of the combination

strategies involved agents that are already approved for use in various cancers. In

addition, several FDA-approved kinase inhibitors also target BET proteins. Thus,

alternatives to single-agent BETi treatment should be rapidly assessed in clinical

trials following the conclusion of phase I studies examining the safety of these

drugs. Overall, the impact of BETi in preclinical models of breast cancer is

promising and further studies into their mechanisms of action should facilitate identification of the most effective combinations to accelerate their use in this heterogeneous disease.

1.5 Mitosis

1.5.1 The cell cycle and cell cycle checkpoints

As described above, BET proteins are involved in multiple aspects of the cell cycle, which consists of five phases. The first four (G0, G1, S, and G2) comprise interphase. G0 is a specialized quiescent state where a cell is metabolically active but not dividing595. This can be a temporary pause in proliferation or a more

permanent arrest. When a cell in G0 receives an appropriate external stimulus, it

will enter G1 where it grows and synthesizes the mRNAs and proteins that will be

113

required for DNA replication. Also during this phase, the cell receives numerous

signals that together determine if the cell will continue through the cell cycle596. If

the conditions are appropriate, the cell moves into S phase where it replicates its

DNA, forming sister chromatids that are physically connected. Sister chromatid

cohesion is mediated by the multi-subunit complex cohesin597. The centrosome,

which nucleates the cell’s cytoplasmic microtubules, also duplicates as the cell

enters S phase598. In G2 phase, the cell continues to grow and prepares for entry

into the final phase, M phase, which encompasses mitosis (nuclear division) and

cytokinesis (cytoplasmic division).

Mitosis is divided into five phases: prophase, prometaphase, metaphase,

anaphase, and telophase. During prophase, chromosomes, which consist of two

sister chromatids, condense. In addition, the duplicated centrosomes move to

opposite ends of the cell and mature. The mitotic spindle forms between the

centrosomes. In prometaphase, the nuclear envelope breaks down. Once the

nuclear envelope is removed, spindle microtubules can attach to the kinetochores

of chromosomes. In metaphase, chromosomes align at the metaphase plate, and

the sister chromatids are pulled towards opposite spindle poles during anaphase.

In telophase, the mitotic spindle disassembles. Nuclear envelope reformation,

which begins in late anaphase, is completed during this phase, creating two nuclei

around the daughter chromosomes. Daughter chromosomes then decondense

and transcription to resumes. The contractile ring, assembled during anaphase,

begins to contract, and this will continue during cytokinesis. Cytokinesis, the

process by which the cytoplasm divides, is initiated during anaphase and ends

114

shortly after telophase. Once cytokinesis is completed, the two resulting daughter

cells are able to reenter the cell cycle.

There are three main cell cycle checkpoints that ensure faithful progression

through the cell cycle595. The first and most stringent checkpoint occurs near the

end of G1, and passage through this checkpoint commits the cell to duplicating its

chromosomes. If conditions are not adequate for a cell to divide or if DNA damage

is detected, the cell will arrest here until conditions are more favorable and the

damage is repaired. There is also a G2/M checkpoint which ensures a cell enters

mitosis only if it has duplicated all of its chromosomes and that there is no DNA

damage. Similar to the G1 checkpoint, the cell will also arrest at the G2/M transition

if the outside environment is not favorable. The third major cell cycle checkpoint,

the SAC, occurs during mitosis and ensures the proper segregation of

chromosomes. While not traditionally included as one of the main cell cycle

checkpoints that prevent transition through the cell cycle when defects are

detected, a fourth checkpoint, the intra-S phase checkpoint, prevents replication

fork collapse and origin firing while DNA damage is being repaired during DNA

replication599-602. ATM and ATR mediate this checkpoint in response to DNA

double-strand breaks and accumulating single-stranded DNA at replication forks,

respectively603,604.

Normally, cell cycle checkpoints tightly regulate the cell cycle so that cells divide

only when conditions are acceptable and the timing is appropriate. This is

important for maintaining proper tissue function and structure. However, in cancer, sustained proliferative signaling allows cancer cells to break through cell cycle

115

checkpoints and traverse mitosis unimpeded. Many drugs have been developed

that inhibit components of the cell cycle, and they have achieved varying degrees

of success in clinical trials. It has been suggested that agents that target proteins

that primarily function during interphase, such as the cyclin-dependent kinases

CDK2, CDK4, and CDK6, will mainly lead to cell cycle arrest and tumor stasis605.

On the other hand, drugs that disrupt cells undergoing mitosis have a far greater

likelihood of inducing cell death pathways. However, drugs that do not selectively

target cancer cells but instead destroy all rapidly dividing cells, such as those that

alter microtubule dynamics, are associated with severe toxicities. Thus it is

important to develop anti-mitotic agents that can selectively target tumors while

sparing normal tissues.

Mitotic kinases and their associated proteins are frequently deregulated in cancer,

and many agents have been developed to target them. Below, I discuss three of

the kinase families that include mitotic kinases, highlight their activity during normal

progression through the cell cycle, and review efforts to inhibit them in breast

cancer.

1.5.2 Cyclin-dependent kinases

Progression through the cell cycle and its checkpoints is regulated by sequential

activation of cyclin-dependent kinases (CDKs)606. CDKs phosphorylate proteins

involved in multiple aspects of the cell cycle and carry out their functions only when

they interact with the appropriate cyclin subunits and are phosphorylated by Cdk-

activating kinase (CAK) (Figure 1.8A)607-609. Another important mechanism that

controls the activity of CDKs is negative inhibition by CDK inhibitors (CKIs). These

116

small inhibitory proteins are divided into two families: the INK4/ARF family which

includes p16INK4a, p15INK4b, p18INK4c, and p19INK4d; and the Cip/Kip family

composed of p21Cip1, p27Kip1, and p57Kip2 610. These two families differ in the

mechanisms they use to inhibit CDKs. Proteins within the INK4/ARF family bind

CDK4/6 and prevent their association with cyclin D (Figure 1.8B). In contrast,

Cip/Kip proteins bind cyclin-CDK complexes which suppresses the kinase activity

of CDKs (Figure 1.8C). Both processes induce cell cycle arrest. The expression of

CKIs is induced by antiproliferative signals such as those elicited at cell cycle

checkpoints and during senescence611,612.

Different cyclin-CDK complexes control the various transitions through the cell cycle, and the timing of their activity is mediated by fluctuations in the expression of cyclins that undergo synthesis and degradation at specific time points in the cell cycle. Cyclin E drives the transition from G1 into S phase613. To prevent premature

entry into S phase, post-mitotic cells express low levels of cyclin E. When a cell

exits mitosis and enters G1, E2F transcription factors that regulate the expression

of cyclin E, cyclin A, and proteins required for DNA replication are bound to

retinoblastoma protein (Rb) which inactivates them614,615. During G1, in response

to growth factors, CDK4 and CDK6 form complexes with cyclin D. The binding of

cyclin D activates CDK4/6, and they subsequently phosphorylate Rb, a suppressor

of E2F activity606,616,617. This releases E2Fs which, in turn, activate transcription of

cyclin E (Figure 1.9). Cyclin E then binds and activates CDK2 which

phosphorylates and fully inactivates Rb618,619. Cyclin E-CDK2 is responsible for the

assembly of the pre-replication complex and initiates centrosome duplication620,621.

117

Once sufficient CDK2 has been activated, the cell enters S phase. Following this

transition, cyclin E levels decrease while cyclin A levels rise. Cyclin A also

associates with and activates CDK2. However, in this case, the resulting complex

initiates DNA replication and prevents assembly of additional pre-replication

complexes (Figure 1.10)621. Towards the end of S phase, CDK1 is activated by cyclin A. This complex pushes the cell into G2622. Cyclin A-CDK1 then stimulates

transcription of mitotic regulators, such as FOXM1c, preparing the cell for entry

into mitosis623.

Expression of another cyclin, cyclin B, increases during G2 and M, and there is a

corresponding increase in the formation of the cyclin B-CDK1 complex. To prevent

premature entry into mitosis, cyclin B is maintained in the cytoplasm until its

cytoplasmic retention signal is phosphorylated by PLK1624-626. CAK

phosphorylates CDK1 at an activating site which stabilizes its interaction with

cyclin B, but Myt1 and Wee1 also phosphorylate CDK1 at other sites that block its

active site607,627-631. Primed cyclin B-CDK1 complexes accumulate during G2 until

the protein phosphatase Cdc25 is activated by PLK1632,633. Cdc25 removes an

inhibitory phosphate from CDK1634. This coincides with the inactivation of Wee1, preventing the re-phosphorylation of the inhibitory site of CDK1. The accumulation of active cyclin B-CDK1 complexes drives entry into mitosis. CDK1 phosphorylates proteins involved in several processes early in mitosis, including the breakdown of the nuclear envelope, actin cytoskeleton rearrangement, chromosome condensation, mitotic spindle assembly, and bipolar spindle attachment at the kinetochore635. Cyclins A and B are marked for proteasome degradation at the

118

metaphase-to-anaphase transition, leading to the inactivation of CDK1, a step that

is necessary for mitotic exit636.

Cell cycle regulators are frequently altered in cancer. For example, the CKI p16 is

a tumor suppressor that is mutated in diverse tumor types637-639. High expression

of cyclin E and low levels of the CKI p27 in breast tumors is linked to poor

survival640. Cyclin D is overexpressed in breast cancer, especially in ER+ tumors,

and CNND1 amplification is associated with increased risk of recurrence in

patients treated with anastrozole or tamoxifen641,642. Cyclin B1 is also

overexpressed in breast cancers and correlates with several aggressive

phenotypes, including high tumor grade, nodal status, Ki67 expression, ER/PR

negativity, amplification of HER2, and shorter overall and metastasis-free

survival643-645. Nuclear staining of cyclin B1 in human breast carcinomas is

associated with poor outcomes and therapy resistance646. Overexpression of both

CDK1 and CDK4 has been found in colorectal tumors, and this is linked to poor

prognosis647,648. Together, these findings indicate that CDKs, cyclins, and CKIs as

well as the pathways that regulate these proteins could be potential therapeutic

targets in the treatment of cancer. Indeed, several inhibitors of CDKs are being

assessed in clinical trials, and the CDK4/6 inhibitor palbociclib has been approved

for use in breast cancer649,650.

1.5.3 The mitotic kinases

Besides CDK1, the activity of other kinases drive key events during mitosis. Two

of the main mitotic kinase families are the polo-like kinases (PLK) and Aurora

kinases (AURK). Members of these serine/threonine kinase families have

119

overlapping functions and sometimes work together to regulate mitotic events. In

addition, their functions are often interwoven with the activity of CDK1.

Expression of three mitotic kinases within these families (PLK1, AURKA, and

AURKB) is regulated by the transcription factor FOXM1 and the MuvB complex651.

MuvB consists of five proteins (LIN9, LIN37, LIN52, LIN54, and RBBP4) and

influences transcription at several phases of the cell cycle. During G0, MuvB

interacts with p107/p130, DP1/2/3, and the repressive E2Fs (E2F4 and E2F5) to

form the DREAM complex, which suppresses expression of cell cycle-dependent

genes and maintains the cell in a quiescent state652. When the cell enters the cell

cycle, MuvB dissociates from the DREAM complex which removes repressive

transcriptional regulators from the promoters of cell cycle genes, allowing G1/S

phase genes to be transcribed653. In S phase, MuvB forms a transcriptional

complex with B-MYB, a protein that is activated by cyclin A-CDK2 and associated

with cell cycle progression653,654. Together, MuvB and B-MYB localize to the

promoter regions of late S phase genes, driving their expression653. MuvB and B-

MYB also recruit FOXM1 to the DNA. B-MYB is then phosphorylated and

degraded, while FOXM1 is phosphorylated and activated. In G2, FOXM1 and

MuvB activate transcription of G2/M phase genes, including cylcin B, survivin, and, as previously stated, PLK1 and Aurora kinases. Despite its different interacting

partners, MuvB is the component responsible for localizing these various transcriptional regulatory complexes to cell cycle-related gene promoters655. As I discuss in Chapter 3, silencing one of the MuvB genes, LIN9, downregulated

120

expression of mitosis genes and induced mitotic defects411, highlighting the

importance of the MuvB complex for proper progression through mitosis.

1.5.3.1 Polo-like kinases

There are five members of the PLK family in humans, PLK1-5. The most studied

PLK is PLK1, which phosphorylates proteins involved in early mitosis. PLK1 is one

of the kinases that promotes centrosome maturation, as it is necessary for the

recruitment of γ-tubulin ring complexes to the centrosome and regulates

microtubule dynamics656-658. By activating Cdc25 and downregulating Wee1, PLK1

helps to push the cell into mitosis where it is involved in the spindle assembly

checkpoint (SAC) and regulates chromosome segregation659-663. In anaphase,

PLK1 moves from the centrosome and kinetochores to the central spindle and

promotes cleavage furrow formation and contractile ring assembly664. Lastly, PLK1

regulates abscission timing665. Cells treated with a PLK1 inhibitor are unable to complete mitosis, have monopolar or disorganized spindles, and undergo cell

death following mitotic arrest, highlighting its importance during cell division666.

Most of the other PLKs have cell cycle-related activities as well. Plk2 and Plk4 are required for centriole duplication which begins at the G1/S transition667,668. Plk3 regulates DNA replication during S phase and in G2 is involved in Golgi fragmentation, a process that is necessary for the Golgi apparatus to be distributed to daughter cells669,670. Plk5, which lacks the kinase domain possessed by the

other PLK proteins, is the exception. It is expressed predominantly in the brain and

is involved in neuronal differentiation671.

121

1.5.3.2 Aurora Kinases

There are three members of the AURK family: Aurora kinase A (AURKA), Aurora

kinase B (AURKB), and Aurora kinase C (AURKC). AURKC functions mainly

during meiosis, while AURKA and AURKB play several key roles throughout

mitosis672. Each AURK is activated following auto-phosphorylation of a specific

threonine residue in the catalytic T-loop of its C-terminus, and protein phosphatase

1 (PP1) deactivates AURKs673. AURKs, as well as PLK1, are marked for

degradation by the anaphase-promoting complex/cyclosome (APC/C) in late

mitosis which leads to their destruction, and this is essential for mitotic exit674-677.

AURKA was initially thought to only function in G2 and early mitosis, although it is

now believed to also play a role in late mitosis and cytokinesis. In S, G2, and early

mitosis, AURKA localizes to the duplicated centrosomes and mitotic spindle poles.

AURKA then moves to the central spindle and midbody during late

mitosis/cytokinesis678,679. Its location influences its function. AURKA stimulates

mitotic entry by phosphorylating and activating Cdc25 and PLK1680-682. Detection

of DNA damage in G2 inactivates AURKA, preventing the G2-to-M transition until

the damage is repaired683,684. AURKA is also involved in centrosome separation

and maturation, bipolar spindle assembly, and spindle function685-688. In late mitosis, AURKA aids in chromosome segregation as well as the formation of the central spindle and stabilization of astral microtubules, both of which influence cytokinesis674,689,690. Inhibition of AURKA after chromosome segregation leads to

aborted cytokinesis and the formation of multinucleated cells, while inhibition in

122

interphase and early mitosis induces mitotic delay and monopolar spindles,

causing chromosome segregation errors and aneuploidy690-693.

AURKB is the enzymatic component of the Chromosomal Passenger Complex

(CPC) which also includes Borealin, Survivin, and Inner Centromeric Protein

(INCENP)694. Like AURKA, AURKB localization changes throughout mitosis, and

this impacts its function. In prophase, AURKB is found along chromosome arms

and at centromeres. It then moves to inner centromeres in prometaphase and

metaphase. Here, it is involved in chromosome condensation, the proper

attachment of microtubules to kinetochores, the SAC, and chromosome alignment

and segregation673. In anaphase, AURKB localizes to the central spindle and

cortex where it stabilizes the central spindle and promotes the formation of the cleavage furrow, respectively695,696. AURKB also helps form the midbody, where it

localizes in telophase and regulates abscission at the end of cytokinesis697,698.

AURKB inhibition produces maloriented chromosomes and disrupts chromosome

segregation and cytokinesis, leading to the formation of polyploid cells followed by apoptosis528,699,700.

1.5.4. Activity of mitotic kinases in M phase

Cyclin B-CDK1, PLK1, AURKA, and AURKB are essential for the proper execution

of mitosis. In prophase, AURKB phosphorylates histone H3 at Ser10 and Ser28,

marks that are required for the completion and maintenance of chromosome

condensation701-703. Condensin, a five subunit protein that is activated following

phosphorylation by cyclin B-CDK1, also plays a crucial role in chromosome

condensation and resolution704,705. In addition to modulating condensing activity,

123

the cyclin B-CDK1 complex initiates spindle assembly, stability, and elongation

during this phase635. CDK1, PLK1, and AURKA work together to phosphorylate

kinesin-5 motors, stimulating centrosome separation and movement to opposite

ends of the cell706-708. Centrosome maturation, regulated by PLK1 and AURKA, also occurs at this phase, and this is marked by an increase in the amount of γ- tubulin ring complexes, multi-subunit protein complexes that are required for microtubule nucleation656,685. The accumulation of γ-tubulin ring complexes in

maturing centrosomes thus confers a greater ability to nucleate new microtubules,

a requirement for the proper separation of chromosomes.

The breakdown of the nuclear envelope in prometaphase is initiated by cyclin B-

CDK1 and PLK1 phosphorylation of components of the nuclear pore complexes

that reside within the nuclear envelope as well the nuclear lamina709,710. These

events remove nuclear pore complexes from the envelope and disassemble the

nuclear lamina711. Removal of the nuclear envelope allows spindle microtubules to

attach to the kinetochores of chromosomes. Spindle dynamics and stability are

now controlled by PLK1 and AURKA689,712-716.

The mitotic spindle is composed of three classes of microtubules: astral, interpolar,

and kinetochore717. Regardless of class, the minus ends of the microtubules are

anchored at the spindle poles emanating from the centrosomes. It is the plus-end

interaction partners that determine microtubule class. Astral microtubules interact

with the cell cortex which helps to anchor the spindle poles718. The plus ends of

interpolar microtubules from one spindle pole interact with the plus ends of

interpolar microtubules from the other pole at the spindle midzone. Lastly,

124

kinetochore microtubules attach to the kinetochores, located within the

centromeres of sister chromatids.

Before chromosomes can be separated, the SAC ensures that all chromosomes

are attached through their kinetochores to spindles originating from opposite

spindle poles719,720. If mitotic spindles are not properly attached to the

kinetochores, SAC components are brought to the kinetochores of maloriented chromosomes. PLK1 is involved in the activation and localization of AURKB to the kinetochore, and these two kinases work together to maintain SAC signaling721,722.

Both recruit components of the SAC to the kinetochore which leads to the

formation of the mitotic checkpoint complex, composed of BubR1, Bub3, Mad2,

and Cdc20723-728. This complex sequesters Cdc20 and prevents it from activating

the APC/C, effectively blocking the metaphase-to-anaphase transition720. AURKB links the SAC to the error correction mechanism for the maloriented chromosomes by phosphorylating several targets involved in the mitotic spindle-kinetochore interaction. This includes KIF2C, the NDC80 complex, and the spindle and kinetochore associated complex729-733. Phosphorylation of these proteins

destabilizes microtubules at the kinetochore, allowing them to break and re- establish proper connections with the kinetochore. Once all sister chromatids achieve accurate bipolar attachment, anaphase can begin.

Anaphase is initiated by the APC/C, an E3 ubiquitin ligase that tags several mitotic regulatory proteins for proteasomal degradation734. Activation of the APC/C occurs

following 1) its phosphorylation by cyclin B-CDK1 and subsequent binding of

Cdc20 and 2) the destruction of its inhibitor, EMI1, which is initiated by the

125

phosphorylation of EMI1 by PLK1735-739. To initiate sister chromatid segregation,

the APC/C ubiquitylates securin, a protein that binds to separase and inhibits its

protease activity740-743. Elimination of securin restores separase activity, and

separase subsequently cleaves cohesin, the protein responsible for holding sister

chromatids together744,745. Release of cohesin allows sister chromatids to

separate. The APC/C also ubiquitylates cyclins A and B746-748. This inactivates

CDKs, and CDK targets become dephosphorylated.

Two processes are involved in chromosome separation: anaphase A and

anaphase B. Anaphase A is primarily driven by activity at the kinetochore while

motor proteins play a critical role in anaphase B. In anaphase A, kinetochore

microtubules depolymerize, and chromosomes move towards the spindle poles749.

In anaphase B, motor proteins pull the spindle poles themselves apart750.

Movement of kinesin-5, which connects the plus ends of interpolar microtubules,

extends interpolar microtubules. With the help of AURKA, the interpolar

microtubules are remodeled to form the central spindle, which will eventually

become part of the midbody in cytokinesis689,690. At the same time, dynein motors that attach astral microtubules to the plasma membrane pull the spindle poles

towards the cell cortex. AURKA stabilizes astral microtubules, and destruction of

AURKA is required for astral microtubule contraction674. In order for anaphases A

and B to occur, CDK substrates need to be dephosphorylated, and this is

dependent on the destruction of cyclins A and B751-754.

Mitotic spindle disassembly in telophase occurs due to the activity of the APC/C,

AURKB, and kinesin-8755. Following the reformation of the nuclear envelope,

126

nuclear proteins enter the nucleus via nuclear pore complexes, and the nucleus

expands. Mitotic histones are dephosphorylated and subsequently removed from

chromatin along with AURKB. This causes the daughter chromosomes to decondense and allows transcription to resume756-759.

Mitotic kinases also play important roles during cytokinesis. The contractile ring,

assembled during anaphase at the cell equator, is composed of F-actin, the motor

protein myosin II, and associated structural and regulatory proteins760. The formation of the contractile ring is directed by the small GTPase RhoA which localizes to the cell cortex761. During metaphase, cyclin B-CDK1 adds an inhibitory

phosphate to the centralspindlin complex, a protein complex that is responsible for

the activation of RhoA762. When CDK1 levels begin to decline in anaphase, centralspindlin is dephosphorylated and binds to microtubules in the midzone. The

activity of centralspindlin is positively regulated by phosphorylation of its

components by PLK1 and AURKB763-765. Centralspindlin sends signals to RhoA,

activating RhoA which initiates the formation of the contractile ring766,767. This is accomplished through the activation of formins that promote the assembly of actin filaments as well as the kinases Rho-associated kinase (ROCK) and Citron that stimulate myosin II motor activity768-772. Movement of myosin II along actin

filaments causes the ring to contract.

The constriction of the contractile ring forms the midbody, an organelle that is

critical for abscission, the process that separates the daughter cells773. The midbody is composed of the central spindle, which is derived from interpolar microtubules, and a dense matrix. AURKs and PLKs, along with other kinases,

127

play important roles in midbody function and abscission. In cytokinesis, AURKB is involved in central spindle assembly, formation of the midbody, and assembly and constriction of the contractile ring by phosphorylating centralspindlin and the

KIF4A/PRC1 complex695,774,775. AURKB acts at the abscission checkpoint,

delaying abscission if it detects a lagging chromosome and inhibiting cleavage

furrow regression, thereby preventing tetraploidization697. PLK1 also regulates

abscission timing by preventing the localization of abscission factors, including

CEP55, to the midbody until late in cytokinesis665. Both AURKB and PLK1 need to

be inactived for abscission to occur. Abscission also requires the fusion of vesicles

to the membrane which closes the gap in the plasma membrane, forming two

separate daughter cells776. Once cytokinesis has ended and if conditions are

appropriate, the daughter cells will reenter the cell cycle and this process will begin

again.

1.5.5 Targeting mitotic kinases in breast cancer

The uncontrolled growth that defines breast cancer as well as other cancers has

led to the development of anti-mitotic chemotherapies that target highly

proliferative cells. Well-known examples of these drugs are those that target the

microtubules: vinca alkaloids, which prevent microtubule polymerization, and

taxanes, which inhibit microtubule depolymerization. However, these agents are

not selective for cancer cells and are thus associated with side effects such as

neurotoxicity, nausea, and diarrhea, which can be severe and lead to the cessation

of treatment777. Thus, other anti-mitotic agents have been developed targeting

128

proteins that have critical functions in mitosis and are deregulated in cancer, such

as PLK1, AURKA, and AURKB.

1.5.5.1 Polo-like kinase 1

PLK1 expression has been linked to breast cancer pathogenesis. Analysis of 135

breast carcinomas revealed PLK1 was overexpressed in about 40% of the tumors

tested, positively correlated with histological grade and HER2 expression, and

negatively correlated with ER status778. In a separate study by King, et al.,

overexpression of PLK1 in breast tumors was associated with TP53 mutations,

and patients who had tumors with both detectable levels of PLK1 and mutated

TP53 had poor breast cancer-specific and disease-free survival and were more

likely to be triple negative779. Another group found TNBC tumors have higher

expression of PLK1 compared to tumors that fall within the other breast cancer

subtypes, corroborating the findings of King and colleagues780. A smaller study reported high expression of PLK1 alone correlated with shorter survival781. These studies indicate that overexpression of PLK1 is associated with aggressive phenotypes and poor outcomes in breast cancer patients.

In vitro experiments revealed that PLK1 silencing in ER+ MCF7 cells inhibited proliferation and induced apoptosis while having no effect on nontransformed mammary epithelial cells782. Similarly, inhibition of PLK1 using siRNA or the PLK1

inhibitor BI-2536 induced DNA breaks, suppressed growth, reduced clonogenic

potential, and initiated apoptosis in TNBC cells, including the cancer stem cell

(CSC) population780,783. Treatment of two TNBC PDX models with BI-2536

dramatically inhibited tumor growth and did not induce any overt toxicity780. In

129

addition, mice receiving a combination treatment of BI-2536 with doxorubicin +

cyclophosphamide chemotherapy exhibited a shorter time to complete response

and had a lower probability of relapse compared to mice treated with

chemotherapy alone780. Delivery of PLK1 siRNA packaged within nanoparticles

effectively reduced PLK1 mRNA expression in metastatic TNBC cells in mice, and

this suppressed the incidence of lung metastasis and improved overall survival784.

Cancer cells, especially those deficient in p53, were much more sensitive to loss of PLK1 compared to non-transformed cells, and BI-2536 only induced apoptosis

in TNBC cells grown in the extracellular matrix Matrigel and not in non-transformed

mammary epithelial cells780,785. Together, these data suggest that selective PLK1

inhibitors may be an effective, durable strategy to target breast cancer cells,

particularly TNBC, while eliciting little to no toxic side effects. However, while

several PLK1-selective inhibitors have been developed, including volasertib, which

was awarded breakthrough therapy status by the FDA for its observed clinical

benefits in AML patients, they have achieved limited long-lasting success in clinical

trials of various cancers. This is due to dose-limiting toxicities, low intratumoral

drug levels, and the development of resistance786-790. Development of analogs that

can improve PLK1 inhibitor delivery and uptake by tumors as well as combining

PLK1 inhibitors with other agents should improve their utility in cancer, including

breast cancer.

1.5.5.2 Aurora kinase A

As in other tumor types, AURKA amplification and/or overexpression is common

in breast cancer, especially in basal-like tumors, and high AURKA protein

130

expression is associated with chromosomal instability791,792. High AURKA mRNA

expression further correlates with high nuclear grade and Ki67 levels as well as

poor relapse-free survival793. In ER+ breast cancer patients, higher AURKA levels

are linked to shorter breast cancer-specific and metastasis-free survival and poorer

prognosis794,795. Similarly, in TNBC tumors, high expression of AURKA protein is

linked to poor overall and progression-free survival796. In MMTV-AURKA

transgenic mice, AURKA expression in the mammary gland induced chromosomal

abnormalities and activated oncogenic signaling pathways, leading to tumor

initiation797. Elevated AURKA levels in ER+ breast cancer cells activated EMT,

thereby promoting metastasis798. AURKA also maintained the self-renewal and

metastatic properties of breast tumor initiating cells and led to resistance to

endocrine therapy and cytotoxic chemotherapy799-803. These data indicate AURKA

is an oncogene and potential therapeutic target in breast cancer.

Several selective AURKA inhibitors have developed such as MLN8237 (MLN,

Alisertib) and AKI603. In MCF7, MDA-MB-231, and MDA-MB-468 breast cancer

cells, MLN inhibited growth and induced autophagy, apoptosis, and/or

senescence472,804. In addition, combination treatments of MLN and a taxane

(docetaxel or paclitaxel) had an additive or synergistic effect on TNBC xenograft

growth depending on the doses used805. MLN and AKI603 also synergized with tamoxifen and the anthracycline epirubicin and could overcome resistance to both

agents801,803. MLN is the only AURKA selective inhibitor currently being evaluated

in breast cancer clinical trials. In a phase II study, the activity and safety of MLN

was assessed in several solid tumor types, including breast cancer. Of the five

131

types of cancer included in the trial, breast and small-cell lung cancer patients had

the highest rate of response806. 23% of HR+ metastatic breast cancer patients had

either a partial or complete response, and 31% of these patients had stable

disease for at least six months. These data support further evaluation of MLN in

breast cancer (Table 1.7).

1.5.5.3 Aurora kinase B

Similar to AURKA, AURKB expression is linked to worse survival from breast

cancer. Analysis of 312 invasive breast tumors revealed high expression of

AURKB correlated with Ki67 expression, histological grade, poor disease-free

survival, and chemoresistance807. In a separate study using ER+ tumors that had

previously been treated with tamoxifen, patients who had tumors that had a higher

percentage of AURKB-expressing cells had lower disease-free and overall survival, suggesting AURKB expression could be used as a biomarker for aggressive disease808. In vitro, phosphorylation of AURKB, which is thought to

positively regulate its kinase function, was higher in fulvestrant-resistant breast

cancer cells compared to parental cells, and an AURKB selective inhibitor,

AZD1152 (AZD, barasertib), suppressed growth to a greater extent in fulvestrant-

or tamoxifen-resistant cells compared to parental cells808-810. Both AURKA and

AURKB were upregulated in aromatase inhibitor-resistant ER+ breast cancer cells,

and treatment with AZD led to cell cycle arrest and polyploidy802. In addition, AZD

induced multinucleation followed by apoptosis and/or senescence in TNBC

cells472. Similarly, AZD stimulated mitotic catastrophe, polyploidy, and apoptosis in

three TNBC and three HER2+ cell lines in vitro811. In vivo, it inhibited tumor growth

132

in a HER2+ mouse xenograft model and suppressed the colonization of a TNBC

cell line in the lung following tail vein injection811. Together, these data suggest

AURKB could be an effective target in breast cancer. There are no ongoing clinical

trials studying an AURKB selective inhibitor in breast cancer, but several studies

are assessing AURKB inhibitors in other types of cancer. It will be important in the

future to confirm if AURKB can be used as a biomarker to identify aggressive

breast tumors and to assess if inhibition of AURKB poses a viable treatment option

for breast cancer patients in clinical trials.

1.6 Statement of purpose

Several studies have shown BETi suppress growth and induce apoptosis and

senescence in TNBC cells. However, the mechanism by which BETi elicit their

effects in this disease is unclear. It is also unknown how BETi impact the different

subtypes of TNBC. Both are important to understand in order to inform which

patients should receive BETi therapy, predict potential avenues of resistance, and design rational drug combination strategies.

In our first study, we showed BETi suppressed growth followed by two terminal responses, apoptosis and senescence, in TNBC cell lines that represent multiple

TNBC subtypes, thus hinting at the broad clinical relevance of BETi in this disease.

Prior to the activation of apoptosis and senescence pathways, BETi-treated cells

became polyploid and multinucleated. This suggests BETi disrupt mitosis. Indeed,

Aurora kinases, critical mitotic proteins, were found to be downstream targets of

BETi, and the downregulation of these genes was important for the response of

TNBC cells to BETi.

133

The appearance of multinucleated cells followed by apoptosis and senescence

suggested BETi may induce mitotic catastrophe in TNBC cells. Live cell imaging

experiments coupled with gene expression microarray analyses revealed BETi

suppress expression of a large number of mitosis and cytokinesis genes,

increased mitotic timing, and induce cell death either during mitosis or soon after

mitotic exit. These data indicate BETi induce mitotic catastrophe in response to

BETi. The disruption of SEs was not involved in the induction of mitotic

catastrophe, refuting the prevailing theory that BETi activity primarily involves the

dismantling of SEs. It also provides an additional explanation for the sensitivity of

cancer cells to BETi compared to nontransformed cells. In addition, we found that

LIN9 mediated the effects of BETi. Furthermore, LIN9 is amplified and

overexpressed in the majority of TNBC tumors and is associated with poor survival,

indicating its significance in this disease.

Together, these studies identify BETi as an anti-cancer therapy that should be

effective in diverse TNBC tumors, especially those that express high levels of LIN9.

We suggest LIN9 acts as an oncogene in TNBC and can be targeted by BETi. The

data presented here will inform approaches for the effective use of these inhibitors

in the clinic and should improve their utility in TNBC.

134

Table 1.1: Examples of classification systems used to subtype breast tumors

Subtype Method Subtypes References

Receptor- Receptor ER/PR+, HER2+, and triple- based status negative subtypes

Intrinsic Gene Luminal A, luminal B, Perou, et al. (2000); subtypes expression HER2-enriched, basal-like, Prat, et al. (2010); profiling claudin-low, normal-like Hennessy, et al. (2009); Herschkowitz, et al. (2007) Integrative Copy number IntClust1-10 Curtis, et al. (2012) clusters and gene expression profiling Pathway Pathway Subgroups 1-17 Gatza, et al. (2010) subtypes signature expression HER2+ Gene Clusters 1-3 Staaf, et al. (2010) clusters expression profiling TNBCtypes Gene Basal-like 1, basal-like 2, Lehmann, et al. expression immunomodulatory, (2011) profiling mesenchymal, mesenchymal stem-like, luminal androgen receptor TNBCtype-4 Gene Basal-like 1, basal-like 2, Lehmann, et al. expression mesenchymal, luminal (2016) profiling androgen receptor

TNBC DNA and Luminal-AR, mesenchymal, Burstein, et al. subtypes mRNA basal-like immune (2015) profiling suppressed, basal-like immune-activated

135

Figure 1.1. Receptor status-based subtypes are not interchangeable with the intrinsic subtypes of breast cancer.

The central graph represents the percentage of breast tumors within the UNC337 dataset that belong to each of the intrinsic subtypes of breast cancer. The radiating graphs represent the percentage of tumors expressing ER and/or HER2 within each of the intrinsic subtypes. Adapted from Prat, et al.78.

136 Table 1.2: The characteristics of the intrinsic subtypes of breast cancer

Intrinsic Characteristics Common metastatic subtype site

Luminal A ER+ and/or PR+, HER2- Bone, skin Low Ki67 Best prognosis, high survival Luminal B ER+ and/or PR+, HER2+ Bone, skin High Ki67 Poorer prognosis than luminal A, high survival HER2- ER/PR-, HER2+ Bone, liver, lungs, brain enriched Poor prognosis

Basal-like Generally triple-negative Brain, lungs Poor prognosis Express basal cytokeratins High Ki67 Claudin-low Generally triple-negative Unknown Low expression of luminal differentiation markers and claudins Low Ki67 compared to basal-like Enrichment of EMT and stem cell genes Poor prognosis

137 Figure 1.2. The intrinsic subtypes of breast cancer subdivide into the

pathway-derived subgroups of breast cancers.

Analysis of 1143 breast tumors based on pathway activity was used to define 17 subgroups of breast tumors (numbered circles). The circles are color-coded based on which intrinsic subtypes are contained within each pathway-derived subgroup.

138 Table 1.3: FDA-approved therapies for breast cancer

Drug Brand name Drug class Subtype

Abemaciclib Verzenio CDK4/6 inhibitor HR+/HER2-

Ado-trastuzumab Kadcyla Antibody-drug HER2+ emtansine conjugate (monoclonal antibody conjugated to a microtubule inhibitor) Anastrozole Arimidex Aromatase inhibitor HR+

Capecitabine Xeloda 5ʹ- All deoxyribonuceloside Cyclophosphamide Clafen, Cytoxan, Antineoplastic agent All Neosar (alkylating agent) Docetaxel Docefrez, Taxotere, Taxane All Taxotere injection concentrate Doxorubicin Adriamycin, Doxil Anthracycline All hydrochloride Epirubicin Ellence Anthracycline All hydrochloride Eribulin mesylate Halaven Furopyran All

Everolimus Afinitor, Zortress Antineoplastic agent HR+/HER2- (macrolide lactam) Exemestane Aromasin Aromatase inhibitor ER+

Fluorouracil Adrucil Halopyrimidine All

Fulvestrant Faslodex Selective estrogen HR+/HER2- receptor degrader Gemcitabine Gemzar Pyrimidine 2ʹ- HR+ hydrochloride deoxyribonucleoside Ixabepilone Ixempra Antineoplastic agent All (microtubule inhibitor) Lapatinib Tykerb Antineoplastic agent HER2+ ditosylate (tyrosine kinase inhibitor)

139 Letrozole Femara Aromatase inhibitor HR+ Megestrol acetate Megace Progestin All

Methotrexate Abitrexate, Folex, Antimetabolite All Folex PFS, Methotrexate LPF, Mexate, Mexate-AQ, Rheumatrex, Trexall Neratinib maleate Nerlynx Antineoplastic agent HER2+ (tyrosine kinase inhibitor) Paclitaxel Taxol Taxane All

Paclitaxel albumin- Abraxane Taxane All stabilized nanoparticle formulation Palbociclib Ibrance CDK4/6 inhibitor HR+/HER2-

Pertuzumab Perjeta Monoclonal HER2+ antibody Raloxifene Evista, Keoxifene Selective estrogen Preventative receptor modulator Ribociclib Kisqali CDK4/6 inhibitor HR+/HER2-

Tamoxifen citrate Nolvadex Selective estrogen ER+, receptor modulator preventative Thiotepa Thioplex Antineoplastic agent All (alkylating agent) Toremifene Fareston Stilbene ER+

Trastuzumab Herceptin Monoclonal HER2+ antibody Vinblastine sulfate Velban, Velsar Vinca alkaloid All

140 Table 1.4: The core histones and their variants

Core Histone Variant Histone Function

DNA repair, recombination, genome integrity, X H2A H2A.X chromosome inactivation Gene activation, gene silencing, chromosome H2A.Z segregation, DNA repair

macroH2A X chromosome inactivation, gene silencing

H2A.B/H2A.Bbd Active transcription, chromatin unfolding

H2B spH2B Sperm chromatin packaging

H2BFWT Testis-specific

Chromatin-to-nucleoprotamine transition, TSH2B sperm- and testis-specific

H2B.1 Testis-, oocyte-, and zygote-specific

H2B.E Regulation of olfactory neuron function

Spermatogenesis, telomere-associated H2B.W functions

H3 H3.3 Gene activation, chromosome segregation

H3.4 Testis-specific

H3.5 Gene transcription, testis-specific

H3.X Not determined

H3.Y Gene activation

Kinetochore assembly, chromosome CENPA segregation

141 A. Ac

HAT

Ac Ac

Gene Gene

HDAC

Ac Ac

Gene Gene

BET

BET Ac Ac Ac Ac

Gene Gene

Figure 1.3. Histone modification proteins add, remove, and read acetylated histone marks.

(A) Histone acetyltransferase (HAT) proteins add acetyl groups to histone tails, which is typically associated with active transcription. (B) Histone deacetylases

(HDACs) remove acetyl groups from histone tails, reducing transcription of associated genes. (C) Bromodomain and Extraterminal (BET) proteins are epigenetic readers that bind acetylated lysines in histone tails and recruit co-

142 activators and other transcription-related proteins to enhance transcription of the target gene.

143 Table 1.5 Breast cell lines used to study BET proteins and their inhibitors.

Human Cell Lines

Cell Line Origin Comments

MCF10A Fibrocystic disease ER- non-transformed mammary epithelium

Invasive ductal breast carcinoma; MCF7 ER+; luminal A derived from pleural effusion

Invasive ductal breast carcinoma; T47D ER+; luminal A derived from pleural effusion

BT474 Invasive ductal breast carcinoma ER+; luminal B

Invasive ductal breast carcinoma; ZR-75-1 ER+; luminal B derived from ascites

Breast adenocarcinoma; derived MDA-MB-453 HER2+; also subtypes as TNBC from pericardial effusion

HCC70 Ductal breast carcinoma TNBC; basal-like

Primary breast acantholytic HCC1806 TNBC; basal-like squamous cell carcinoma

HCC1937 Primary ductal breast carcinoma TNBC; basal-like; BRCA1 mutation

HCC3153 Ductal breast carcinoma TNBC; basal-like; BRCA1 mutation

Breast adenocarcinoma; derived MDA-MB-468 TNBC; basal-like from pleural effusion

TNBC; basal-like; inflammatory breast SUM149/SUM149PT Invasive ductal breast carcinoma cancer; BRCA1 mutation

BT549 Invasive ductal breast carcinoma TNBC; claudin-low

HCC38 Primary ductal breast carcinoma TNBC; claudin-low

HS578T Invasive ductal breast carcinoma TNBC; claudin-low

Breast adenocarcinoma; derived MDA-MB-231 TNBC; claudin-low; highly metastatic from pleural effusion

Invasive ductal breast carcinoma; MDA-MB-436 TNBC; claudin-low BRCA1 mutation derived from pleural effusion

TNBC; claudin-low; activating PiK3CA SUM159/SUM159PT Anaplastic breast carcinoma mutation; amplified MYC

Invasive ductal breast carcinoma; SUM1315 TNBC; claudin-low; BRCA1 mutation derived from skin

Invasive ductal breast carcinoma; EVSAT ER-/HER2-; unknown subtype derived from ascites

Mouse Cell Lines

Mammary tumor from MMTV- 6DT1 Highly metastatic Myc mouse

144 Mammary tumor from MMTV- H1047R+ Cre MCCL-278 Myc; Pik3ca ; Wap Activating PI3K mutation; amplified Myc mouse

Mammary tumor from MMTV- MCCL-357 flox/flox Cre Deletion of Pten; amplified Myc Myc; Pten ; Wap mouse

Mammary tumor from MMTV- Mvt-1 Highly metastatic Myc-VEGF mouse

145 BRD4

Ac Ac Ac

Gene

+ BETi

BRD4

Ac Ac

Gene

Cancer Inflammatory Hypoxia Metabolism Metastasis Angiogenesis stem cells response

Apoptosis Senescence

Figure 1.4. BET inhibitors suppress several oncogenic pathways in breast cancer.

Binding of BETi to the BRD of BET proteins, such as BRD4, prevent BET proteins from binding to acetylated lysines in histone tails. This blocks BET protein

146 localization to the promoter and enhancer regions of target genes, thus inhibiting their transcription. In breast cancer, outcomes of BETi include suppression of self-

renewal properties in cancer stem cells, inhibition of hypoxia-induced pathways,

altered metabolism, decreased metastatic potential, inhibition of angiogenic

factors, and reduction in the pro-inflammatory response. In response to the

disruption of several of these oncogenic pathways, breast cancer cells either undergo apoptosis or senescence.

147 A. BETi sensitive BETi resistant

Ligand Ligand

RTK RTK Plasma Membrane

PTEN PTEN

PIP2 PIP3 PIP2 PIP3

PI3K PI3K

AKT AKT

mTOR mTOR

BRD4 BRD4 Cell growth, Cell growth, translation, translation, Gene Gene proliferation proliferation

BETi

B. CK2 CK2 P P BRD4 BRD4 BRD4 BRD4

PP2A PP2A

BRD4 Mediator P Mediator BRD4

Ac Ac Ac Ac

Gene Gene

Figure 1.5. Mechanisms of BET inhibitor resistance in breast cancer.

(A) Increased activity of the PI3K/AKT/mTOR pathway. Left: BET inhibition and

normal PI3K/AKT/mTOR signaling. The binding of ligands, such as growth factors,

148 hormones, or cytokines, to receptor tyrosine kinases (RTKs) activates PI3K. PI3K

converts PIP2 to PIP3, and PTEN counteracts PI3K by dephosphorylating PIP3.

PIP3 mediates the phosphorylation, and thus activation, of AKT which leads to the activation of mTOR. mTOR signaling normally drives critical cellular process including cell growth, translation, and proliferation, but BETi suppress expression of the genes responsible for these processes. Right: The

PI3K/AKT/mTOR pathway can remain active if PTEN is mutated, preventing it from

converting PIP3 to PIP2, or if PI3K has an activating mutation. Both mutations

increase phosphorylation of PIP3, maintaining activity of AKT and mTOR and

overcoming BETi-mediated suppression of gene expression. (B) Increased

phosphorylation of BRD4. Left: BRD4 is phosphorylated by CK2 and

dephosphorylated by PP2A. When BETi (blue circle) binds to BRD4, it prevents

localization of BRD4 to the chromatin. BRD4 can thus no longer recruit Mediator

to the DNA, silencing target gene expression. Right: Reduced PP2A activity

increases the abundance of phosphorylated BRD4 (pBRD4). Despite the binding

of BETi to its BRD regions, pBRD4 can still localize to the chromatin and interact

with Mediator, maintaining transcription of target genes.

149

Figure 1.6. Combination treatments with BET inhibitors in breast cancer.

Drugs that have been combined with BETi and resulted in greater efficacy are

listed and color-coded according to the outcome of the combination treatment.

Gray circles: drug synergized with BETi. Green circle: drug synergized with BETi,

and the combination overcame tamoxifen resistance. Orange circle: drug

synergized with BETi, and the combination overcame everolimus resistance.

Purple circle: drug synergized with BETi, the combination overcame BETi

resistance, and the combination prevented reactivation of the PI3K pathway.

Yellow circle: the combination overcame and prevented lapatinib resistance. Red

circle: drug synergized with BETi, and the combination overcame and prevented

MEK inhibitor resistance.

150 Figure 1.7. The response of the hallmarks of cancer to BET inhibitor treatment in breast cancer.

BETi impact most of the hallmarks of cancer described by Hanahan and

Weinberg485. An example of a BETi-induced effect is listed next to the hallmarks

that were altered by BETi in breast cancer cells.

151 A.

P CAK

P Phosphorylation Proliferation CDK CDK Cyclin of substrates CDK Cyclin

Substrates not Cell cycle CDK phosphorylated arrest INK4/ARF CDK Cyclin

Cip/Kip

Substrates not Cell cycle CDK CDK Cyclin phosphorylated arrest CDK Cyclin

Figure 1.8. Activation and inhibition of cyclin-CDK complexes.

(A) Cyclin-dependent kinases (CDKs) are activated once they bind with the appropriate cyclin and are phosphorylated by Cdk-activating kinase (CAK). CAK can phosphorylate a CDK either prior to or after its association with cyclin.

Activated cyclin-CDK complexes can then phosphorylate their substrates, influencing cell cycle progression. (B) CDK inhibitors (CKIs) belonging to the

INK4/ARF family bind CDKs, preventing them from binding to cyclins and thus inhibiting substrate phosphorylation and cell cycle progression. (C) CKIs belonging to the Cip/Kip family bind preformed cyclin-CDK complexes, inactivating CDK and preventing substrate phosphorylation and movement through the cell cycle.

152 Cyclin expression: CDK2 CDK4/6 Cyclin E Cyclin D

G1 S Cyclin-CDK complex formation: CDK2 CDK2 CDK4/6 Cyclin E Cyclin A Cyclin D

Cyclin-CDK complex activity:

P P P P Pre-replication complex assembly P RB CDK4/6 Cyclin D

RB E2F E2F

P CDK2 Cyclin E CDK2 Cyclin E

Cyclin A Cyclin E Centrosome duplication S phase protiens E2F Gene

Figure 1.9. Cyclin-CDK activity in G1 phase.

Top: Expression changes of cyclins D, E, and A in G1 and S phases of the cell cycle. Cyclin D expression increases during G1. Later in G1, cyclin E is upregulated. Following entry into S phase, cyclin E levels decrease while cyclin A

153 levels increase. Middle: The cyclin-CDK complexes that form during G1 and early

S phase. Cyclin D interacts with CDK4/6 and cyclin E forms a complex with CDK2

in G1. Cyclin A-CDK2 complexes form during S phase. Bottom: The activity of cyclin D-CDK4/6 and cyclin E-CDK2 complexes in G1 phase. CyclinD-CDK4/6 phosphorylates Rb, releasing E2Fs which then drive transcription of cyclin A, cyclin

E, and S phase genes. Cyclin E forms a complex with CDK2 which can then further inactivate Rb. Cyclin E-CDK2 also initiates the assembly of the pre-replication complex and centrosome duplication.

154

Cyclin expression: CDK1 CDK2 Cyclin B Cyclin E CDK2 CDK1 CDK2 Cyclin A Cyclin B Cyclin A

S G2 M Cyclin-CDK complex formation: CDK2 CDK1 CDK1 Cyclin A Cyclin A Cyclin B

Cyclin-CDK complex activity:

Pre-replication complex assembly S to G2 transition CDK1 Cyclin A

Transcription of mitotic regulators CDK2 Cyclin A

Regulation of processes CDK1

Cyclin B in early mitosis

DNA replication

Figure 1.10. Cyclin-CDK activity in S, G2, and M phases.

Top: Expression changes of cyclins A, E, and B in S, G2, and M phases. In S phase, cyclin A levels rise while cyclin E levels decrease. Cyclin B expression increases in G2. Both cyclin A and cyclin B are degraded towards the end of mitosis. Middle: The cyclin-CDK complexes that form during S and G2 phases. In

S phase, cyclin A forms a complex with CDK2 (early S phase) and CDK1 (late S

155 phase). Cyclin B interacts with CDK2 during G2 phase. Bottom: The activity of

cyclin A-CDK2, cyclin A-CDK1, and cyclinB-CDK1 complexes in S, G2, and M

phases. Cyclin A-CKD2 initiate DNA replication and prevent the formation of

additional pre-replication complexes. Cyclin A-CDK1 is responsible for the S to G2

transition and transcription of mitotic regulators. Activated cyclin B-CDK1 drives

several processes in early mitosis.

156

Table 1.6. Current and completed clinical trials evaluating AURKA inhibitors in breast cancer.

Combination Study Tumor Type Status Drug

Phase I/II: Advanced solid tumors, Completed, None NCT01045421 including breast results published

Locally recurrent or metastatic Phase I/II: Active, not breast (Phase I only) and Paclitaxel NCT01091428 recruiting ovarian cancer

Phase I: Advanced solid tumors, Completed, no Pazopanib NCT01639911 including breast results posted

Locally recurrent or metastatic Phase II: ER+/HER2 breast cancer and Paclitaxel Recruiting NCT02187991 TNBC

Phase I: Locally advanced or metastatic Active, not Fulvestrant NCT02219789 inoperable HR+ breast cancer recruiting

Advanced solid tumors; Phase I: metastatic breast cancer MLN0128 Recruiting NCT02719691 expansion cohort

Endocrine-resistant locally With or Phase II: advanced or metastatic without Recruiting NCT02860000 ER+/HER2- breast cancer fulvestrant

157

CHAPTER 2: BROMODOMAIN AND EXTRATERMINAL PROTEIN INHIBITION

BLOCKS GROWTH IN TRIPLE-NEGATIVE BREAST CANCERS THROUGH

THE SUPPRESSION OF AURORA KINASES

This research was originally published in the Journal of Biological Chemistry. Sahni

JM, Gayle SS, Weber Bonk KL, Cuellar Vite L, Yori JL, Webb B, Ramos EK,

Seachrist DD, Landis MD, Chang JC, Bradner JE, and Keri RA. Bromodomain and

Extraterminal protein inhibition blocks growth in triple-negative breast cancers through the suppression of Aurora kinases. J Biol Chem. 2016; 291: 23756-68. ©

The American Society for Biochemistry and Molecular Biology

158

2.1 Abstract

Bromodomain and extraterminal (BET) proteins are epigenetic “readers” that recognize acetylated histones and mark areas of the genome for transcription.

BRD4, a BET family member protein, has been implicated in a number of types of cancer, and BET protein inhibitors (BETi) are efficacious in many preclinical cancer models. However, the drivers of response to BETi vary depending on tumor type, and little is known regarding the target genes conveying BETi activity in triple- negative breast cancer (TNBC). Here, we show that BETi repress growth of multiple in vitro and in vivo models of TNBC by inducing two terminal responses: apoptosis and senescence. Unlike in other cancers, response to BETi in TNBC is not dependent upon suppression of MYC. Instead, both end points are preceded by the appearance of polyploid cells caused by the suppression of Aurora kinases

A and B (AURKA/B), which are critical mediators of mitosis. In addition, AURKA/B inhibitors phenocopy the effects of BETi. These results indicate that Aurora kinases play an important role in the growth suppressive activity of BETi in TNBC.

Elucidating the mechanism of response to BETi in TNBC should 1) facilitate the prediction of how distinct TNBC tumors will respond to BETi and 2) inform the rational design of drug combination therapies.

2.2 Introduction

Triple-negative breast cancer (TNBC) is a heterogeneous disease that comprises

∼15–20% of all breast cancers that can be further subdivided into “basal” and

“claudin low” subtypes. As a whole, TNBC is highly aggressive, lacks effective

targeted therapies, and conveys poor clinical outcome58,107. This is in part due to

159 the lack of estrogen, progesterone, and HER2 receptors. It is therefore critical to

develop new targeted treatment strategies that will decrease aggressiveness and

improve patient outcomes.

The bromodomain and extraterminal (BET) family of epigenetic readers is a

potential therapeutic target in many cancers812. This family is comprised of four

proteins (BRD2, BRD3, BRD4, and BRDT) characterized by two bromodomains

that bind acetylated lysines of histone tails and transcription factors385. The most

well characterized BET protein, BRD4, regulates gene transcription and is critical

for cell cycle progression and mitosis813,814. In cancer, it is enriched at super-

enhancers that drive the expression of critical oncogenes366,446. As a result,

inhibition of BRD4 has been shown to selectively suppress key oncogenic

drivers366,446.

BET inhibitors (BETi) compete with acetylated lysines for binding to the

bromodomain pockets of BET proteins, including BRD4431,460. These inhibitors are

effective in mouse models of various diseases, including cancer, inflammation,

HIV, and heart failure812. In cancer, they have been proposed to function by

disrupting super-enhancers that drive expression of key oncogenes such as MYC.

This leads to growth arrest, and, in some cases, apoptosis366,446,815. To date, eight

BETi are being investigated for safety and efficacy in diverse cancers in early

phase clinical trials.

The BET protein gene BRD4 is amplified or overexpressed in ∼17% of basal type

breast cancers, the largest subclass of TNBC21, suggesting that this protein may

be a particularly useful target in TNBC. Recent studies have examined the

160 therapeutic impact of BETi in models of TNBC455,457,536 and also uncovered mechanism(s) of resistance420. However, the utility of BETi in different subtypes of

TNBC and the genes that control BETi activity in these subtypes remains unknown.

Here, we show that BET inhibition halts the growth of multiple TNBC cell lines, independent of their subtype classification as basal and claudin low80 or according to the six subtypes recently described by Lehmann, et al.122. We also show that

BETi inhibit growth of TNBC tumors in multiple mouse models representing different subtypes. Mechanistically, BETi suppress expression of the critical mitosis factors Aurora kinase A and B through direct loss of BRD4 binding to the

Aurora kinase promoters, and this is followed by the generation of multinucleated cells that then either senesce or die via apoptosis. Aurora kinase inhibitors also induce polyploidy and apoptosis/senescence in TNBC cells and thus phenocopy

BETi. Together, these data reveal that suppression of Aurora kinases mediates the activity of BETi in both the basal and claudin low subtypes of TNBC.

2.3 Materials and methods

Cell culture and reagents

The cell lines were obtained from the American Type Culture Collection. The cells were maintained at 37 °C with 5% CO2. MDA-MB-231, HCC38, MDA-MB-468,

HCC1143, BT549, and HCC70 cells were grown in RPMI 1640 supplemented with

10% FBS. For BT549 cells, 0.023 IU/ml insulin was added to the media. MDA-MB-

453 cells were grown in DMEM supplemented with 10% FBS. JQ1, I-BET151

(MedChem Express), I-BET762 (Cayman Chemical), AZD1152 (AdooQ

Bioscience), and MLN8237 (AdooQ Bioscience) were dissolved in DMSO. For

161 growth curves, the cells were treated with the indicated drugs for 72 h and trypsinized. Viable cells were identified by trypan blue exclusion and counted on a hemocytometer.

Flow cytometry

Cell cycle analyses were performed as previously described816 with the following modification: the cells were harvested, fixed in ice-cold 70% ethanol, suspended in propidium iodide/RNase solution, and analyzed using Attune NxT Flow

Cytometer (Thermo Fisher). During analysis, gating was performed to remove doublets from the results.

RNA analysis

RNA was isolated using TRIzol reagent (Ambion; 15596018) and treated with

DNase I (Ambion; AM1906). Reverse transcription was performed using

SuperScript II reverse transcriptase (Invitrogen; 18064-014). Quantitative real time

PCR was performed on an Applied Biosystems Step One Plus real time PCR system using the following TaqMan Gene Expression Assays (Thermo Fisher):

1) AURKA (Hs01582072_m1), 2) AURKB (Hs00945858_g1),

3) MYC (Hs00153408_m1), and 4) GAPDH (Hs02758991_g1).

Western blotting analysis

Western blotting analysis was performed as previously described817 using the following primary and secondary antibodies: c-Myc (Cell Signaling; 9402), p21

(Cell Signaling; 2947), AURKA (Cell Signaling; 4718), AURKB (Cell Signaling;

3094), β-actin (Sigma; A1978), goat anti-rabbit IgG-HRP (Santa Cruz; sc-2054),

162 and goat anti-mouse IgG-HRP (Santa Cruz; sc-2005). The blots were developed using Pierce ECL Western blotting substrate (Thermo Scientific). The bands were quantified using ImageJ software.

Senescence-associated β-galactosidase activity stain

The cells were fixed in PBS-buffered 2% formaldehyde and 0.2% glutaraldehyde for 5 min at room temperature. The cells were then washed three times in PBS and stained (5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 2 mM MgCl2, 150 mM NaCl, 30 mM citric acid/phosphate buffer, pH 6, 1 mg/ml X-gal;

Invitrogen, 15520) for 12–16 h at 37 °C.

Colony forming assay

The cells were treated with vehicle or JQ1 for 14 days and then harvested. For each treatment group, 500 (MDA-MB-231) or 1000 (HCC1143) viable cells were seeded in three 35-mm plates and evenly dispersed. The cells were then grown in drug-free media for 11 days. The colonies were fixed and stained (0.05% crystal violet, 1% formaldehyde, 1% methanol, and 10% 10× PBS) for 20 min at room temperature.

Mouse xenograft studies

All in vivo experiments were performed with approval from the Institutional Animal

Care and Use Committee at Case Western Reserve University, which is certified by the American Association of Accreditation for Laboratory Animal Care. The mice were housed in microisolator units, given standard sterile chow and water ad libitum, and maintained on a 12-h light/dark cycle. Xenografts were implanted into

163 the two inguinal mammary glands of adult (2–4 month old) female NOD/scid/γ

mice. Tumor size was measured twice per week using calipers. Mouse weight was

measured once per week to assess toxicity. Upon completion of treatments, the

tumors were removed and processed for sectioning, RT-qPCR, and Western

blotting analysis.

For MDA-MB-231 xenografts, mice with tumors 120 ± 50 mm3 in size were

randomized into two treatment groups of five mice each, vehicle (1:1 propylene

glycol:water) or JQ1 (50 mg/kg) i.p. daily, and tumor size was measured with

calipers for 28 days. At this time, livers were harvested, and surface

macrometastases were counted. For MDA-MB-468 xenografts, mice with tumors

90 ± 20 mm3 in size were randomized into 2 treatment groups of 10 mice each

vehicle or JQ1 (50 mg/kg) i.p. daily, and tumor size was measured with calipers

for 32 days. For the patient-derived xenograft, BCM-4013, mice with tumors 260 ±

150 mm3 in size were randomized into 2 treatment groups of 10 mice each, vehicle

or JQ1 (50 mg/kg) i.p. daily, and tumors were measured for 30 days.

Analysis of mammary gland morphology, apoptosis, and proliferation

To assess the impact of BETi on normal mammary glands, 10 adult female FVB/N

mice were treated with either vehicle or JQ1 (50 mg/kg i.p. daily) for 1 week. For

each mouse, both inguinal mammary glands were collected and either processed

for whole mounts or sectioned for TUNEL and phospho-histone H3 staining. For

whole mounts, mammary glands were isolated, fixed, and processed as previously

described817. For analysis of apoptosis and proliferation, mammary glands were

collected, fixed in 4% paraformaldehyde for 6 h, paraffin-embedded, and sectioned

164

by the Case Western Reserve University Tissue Procurement and Histology Core

Facility. TUNEL staining was performed according to the manufacturer's

instructions (ApopTag Plus peroxidase in situ apoptosis detection kit; Millipore).

For phospho-histone H3 staining, mammary gland sections were cleared and

hydrated by washing twice each with xylene, 100% ethanol, and 95% ethanol, and

once with PBS. Antigen retrieval was accomplished using a Biocare Medical

decloaking chamber at 125°F for 10 min in 10 mM citrate buffer (pH 6). The

sections were then washed three times with PBS, and peroxidase was blocked

using Dako EnVision Plus kit (Dako; K4011) supplemented with 15 μl/ml goat

serum. Phospho-histone H3 (Ser10) antibody (Cell Signaling; 9701) diluted in

PBST with 5% BSA and 15 μl/ml goat serum was added, and sections were

incubated overnight at 4°C. After two PBS washes, the slides were stained and

developed with Dako Envision + System-HRP (Dako; K4011). Specifically, the

secondary antibody was incubated for 1.5 h at room temperature. The sections

were then washed twice with PBS and counterstained with Hematoxylin (Gill's III;

CS402-1D) and Scott's bluing solution. Sections were dehydrated prior to

mounting by washing once each with 95% ethanol and100% ethanol and twice

with xylene.

Nuclear morphology

The cells were seeded onto sterile glass coverslips and treated with vehicle

(DMSO), JQ1, AZD1152, or MLN8237 for 4 (MDA-MB-468) or 8 (MDA-MB-231

and HCC1143) days. The cells were fixed with 3.7% formaldehyde, 1× PBS,

permeabilized with 0.1% Triton X-100, 1× PBS, and blocked with 1% BSA, 1× PBS.

165

F-actin was labeled with Texas Red-X phalloidin (Invitrogen; T7471). The nuclei

were counterstained with Vectashield hard set mounting medium with DAPI

(Vector Labs; H-1500). To identify cells undergoing apoptosis, the cells were

stained with 10 μM Hoechst 33342 (Thermo Scientific; 62249) for 10 min at 37 °C.

The number of cells with and without pyknotic nuclei were counted, and the

percentage of apoptotic cells was calculated.

Gene-specific chromatin immunoprecipitation

ChIP-PCR was performed as previously described818. MDA-MB-231 cells were

treated for 24 h with vehicle or 500 nM JQ1. Chromatin was immunoprecipitated

with a BRD4-specific antibody (Bethyl Laboratories; A301-985A50) or control

mouse IgG (Sigma, I5281). The following promoter-specific primer sequences

were used: CDKN1A, forward 5′-GCCTCCCTCCATCCCTATG-3′ and reverse 5′-

CAGCCCAAGGACAAAATAGC-3′; AURKA, forward 5′-

AGGACAAGGGCCTTCTTAGG-3′ and reverse 5′-

TAGTGGGTGGGGAGACAGAC-3′; and AURKB, forward 5′-

AGCCGTGAGAAGCAGAGAAA-3′ and reverse 5′-

ATTGGGGCTAGTGTGCTGAC-3′. As a negative control, the following gene-

specific primer sequences designed against regions outside of the promoter region

were used: CDKN1A, forward 5′-GCAGCAGGGGAGGAAAGTAT-3′ and reverse

5′-CCCCATGCTGTTCTCGTAAC-3′; AURKA, forward 5′-

ATCTCTGGCACAGAATTCCAG-3′ and reverse 5′-

TTTGTCTGGTTTCTCCACTGT-3′; and AURKB, forward 5′-

166

TATATCCCAAAGCCCCAGAG-3′ and reverse 5′-ATGTCCCCAGTGAACTCCAA-

3′.

Statistics

Statistical analyses were performed using two-tailed Student's t test (in vitro data)

and Mann-Whitney U test (in vivo data). p values less than 0.05 were considered

statistically significant. All in vitro data are represented as mean values from three

independent experiments performed in triplicate.

2.4 Results

2.4.1 BET inhibition blocks growth of diverse TNBC cells without consistently

down-regulating MYC

Multiple studies have shown that BETi suppress growth of TNBC cells without

specific consideration of the distinct TNBC subtypes420,455,457,536. We validated the

efficacy of BETi in TNBC cells by treating six cell lines representing the two

predominant subtypes of TNBC, basal (HCC1143, MDA-MB-468, and HCC70) and

claudin low (MDA-MB-231, BT549, and HCC38), with the prototypical BETi, JQ1.

We also assessed BETi responsiveness of MDA-MB-453 cells, which are triple-

negative but express androgen receptor and are thus often classified within the

luminal breast cancer subtype. As is typical for TNBC, TP53 (p53-encoding gene)

is mutated to transcriptionally inactive isoforms in all seven cell lines tested819.

Within 72 h, JQ1 inhibited growth of all seven cell lines in a dose-dependent

manner, independent of TNBC subtype (Figure 2.1A–C). Qualitatively similar

effects as JQ1 were observed with two additional BETi, I-BET151, and I-BET762

167

(Figure 2.1 D and E), indicating JQ1-induced growth suppression was due to inhibition of BET protein function and not off target effects.

In numerous cancer types, the effects of BETi are due to their suppression of MYC expression; often MYC transcription is significantly decreased with BET protein inhibition, and overexpression of MYC reverses the effects of these drugs526,527.

To determine whether MYC repression is necessary for the BETi response in

TNBC, we treated the seven TNBC cell lines with JQ1 for 24 h and assessed the response of MYC mRNA and protein expression. Although BETi suppressed growth of all seven cell lines, the impact of JQ1 on MYC mRNA and protein was highly variable (Figure 2.1 F and G), in some cases having no effect, and did not correlate with TNBC subtype or the GI50 for JQ1 (Table 2.1). For example, MDA-

MB-231 and HCC38 cells are among the most growth-suppressed, yet MYC protein expression is unchanged in response to JQ1 treatment. These data indicate that suppression of MYC is not essential for BETi to inhibit the growth of

TNBC cells. This result contrasts with luminal breast cancer models where growth inhibition by BETi is in part mediated by suppression of MYC468.

2.4.2 Sustained BET inhibition induces apoptosis and senescence in TNBC cells

To determine whether the growth suppression of TNBC cells by BETi was due to an induction of apoptosis, we treated four cell lines (MDA-MB-468, HCC1143,

MDA-MB-231, and BT549; two basal and two claudin low, respectively) with vehicle or JQ1 for 72 h and stained them for pyknotic nuclei with Hoechst. JQ1 increased the number of apoptotic cells in all four cell lines (Figure 2.2A). However, two of the lines (MDA-MB-468 and BT549) had a more pronounced apoptotic

168 response than the other two. Notably, the extent of apoptosis was again independent of TNBC subtype; both a claudin low and a basal line were highly apoptotic. Surprisingly, MDA-MB-231 and HCC1143 cells exhibited greater growth suppression than BT549 cells (Figure 2.1A and B), yet they displayed fewer apoptotic cells. This suggested that these two cell lines may undergo a different cellular response in addition to apoptosis. Consistent with this possibility, expression of p21, a protein involved in the induction of senescence, was greatly elevated in MDA-MB-231 and HCC1143 cells within 4 days of JQ1 treatment

(Figure 2.2B). In contrast, the two cell lines displaying a greater degree of JQ1- induced apoptosis (MDA-MB-468 and BT549) only modestly increased p21 levels.

The increase in p21 suggested that MDA-MB-231 and HCC1143 cells may undergo senescence in response to BETi. Supporting this postulate, within 8 days of JQ1 treatment, MDA-MB-231 and HCC1143 cells became flattened, had a greater cytoplasmic to nuclear ratio, and stained positively for senescence- associated β-galactosidase (SA-βgal) activity (Figure 2.2C and D), all of which are hallmarks of senescence. Within 14 days, most JQ1-treated MDA-MB-231 and

HCC1143 cells expressed SA-βgal (Figure 2.2C and D), indicating the majority of cells had undergone senescence. Colony forming assays with MDA-MB-231 and

HCC1143 cells revealed that JQ1-induced senescence is irreversible (Figure

2.2E and F). Thus, sustained BETi exposure induces two terminal responses: apoptosis and senescence. These responses are independent of TNBC subtype, impact on MYC expression, and the GI50 of JQ1 (Table 2.1).

169 2.4.3 BET inhibition abrogates TNBC tumor growth

To determine whether the distinct cell fates (senescence versus apoptosis) would

translate to differential tumor responses in vivo, we compared the impact of BETi

on tumor growth using two xenograft mouse models. First, mice harboring palpable

orthotopic xenografts of one of the cell lines that primarily undergoes senescence

(MDA-MB-231) were treated with vehicle or JQ1 for 28 days. JQ1 suppressed the

growth of these tumors (Figure 2.3A), and at the conclusion of this study, the

average final tumor size was smaller in JQ1-treated mice compared with those in

vehicle-treated animals (p < 0.01; Figure 2.3B). JQ1 also suppressed the

incidence of liver metastasis (p < 0.05; Figure 2.3C) compared with vehicle-treated

mice. A similar study was performed using a cell line that primarily undergoes

apoptosis in response to BETi (MDA-MB-468). These tumors had a more robust

response to JQ1 treatment than those derived from MDA-MB-231 cells. Within 5

days, MDA-MB-468 tumors treated with JQ1 began to regress (Figure 2.3D). Using

RECIST criteria820, JQ1 treatment resulted in one mouse having progressive

disease, five having partial regression, and the remainder having stable disease

after 35 days (Figure 2.3E). Thus, in vivo models mimicked our in

vitro observations, because tumors formed from a predominantly apoptotic cell line

partially regressed in response to BETi, whereas BETi suppressed the growth but

did not elicit regression of tumors generated from the cell line that senesces. In

both tumor models, JQ1 regulation of MYC expression followed a similar pattern

as the in vitro studies (Figure 2.3F and G), providing in vivo evidence that at least

in MDA-MB-231 tumors, MYC repression is unnecessary to mount a growth

170 inhibitory response to BETi. Lastly, to assess the impact of BETi in a more clinically relevant model, we evaluated BETi efficacy in a patient-derived xenograft (PDX) model, PDX BCM-4013821, classified as basal by PAM5082. Similar to MDA-MB-

231 tumors, JQ1 blunted growth of tumors formed from this PDX (Figure 2.3H).

As previously reported for multiple other mouse models, there was no change in body weight for mice with MDA-MB-231 or PDX tumors that were treated with JQ1 for 28–30 days compared with vehicle-treated mice (Figure 2.4A). In contrast, JQ1- treated mice with MDA-MB-468 tumors did weigh less than their vehicle-treated counterparts at 32 days post-treatment (Figure 2.4A). However, this is likely due to extensive tumor growth in the vehicle-treated group that increased the overall weights of these mice. Further demonstrating selectivity of drug response in breast tumors compared with normal tissues, non-tumor-bearing female mice treated with vehicle or JQ1 displayed no differences in mammary gland architecture, proliferation (phospho-histone H3 staining), or apoptosis (TUNEL staining) (Figure

2.4B and C, and data not shown). Thus, BETi lack toxic side effects in the mammary gland. Together, these data reveal that BET protein inhibition selectively suppresses tumor growth in numerous in vivo models of TNBC, as well as metastatic progression with minimal to no toxicity.

2.4.4 Aurora kinases are downstream targets of BETi

To begin to assess the mechanism(s) by which BETi induce apoptosis and senescence in TNBC cells, cellular morphology was examined for several days after exposure to JQ1. This revealed that the cells become multinucleated regardless of whether they undergo apoptosis or senescence, or their particular

171 TNBC subtype, and multinucleation occurred prior to the induction of either cell fate (Figure 2.5A). Within 8 days, nearly half of JQ1-treated MDA-MB-231 cells were tetraploid (40.5 ± 5.9%) (Figure 2.5B). These data suggest that BETi disrupt mitosis and/or cytokinesis, both of which are known to induce apoptosis and/or senescence depending on the cell line being examined.

It is well established that deregulation of Aurora kinases A or B

(AURKA or AURKB) induces polyploidy528-530. Both kinases play critical roles in mitosis531, and a previous study reported that BRD4 stimulates transcription of AURKB414, although it was not shown whether this was a direct effect. This led us to determine whether Aurora kinases may mediate BETi response in TNBC by first examining the expression of Aurora kinases following JQ1 treatment. Both

AURKA and AURKB mRNA and protein were down-regulated within 24 h of JQ1 treatment in four TNBC cell lines, regardless of whether the cell lines ultimately senesce or apoptose in response to the drug (Figure 2.5C and D). JQ1 also disrupted the binding of BRD4 to AURKA and AURKB but did not reduce binding to CDKN1A, the gene encoding p21 (Figure 2.5E). Both Aurora kinase transcripts were also suppressed in JQ1-treated xenografted tumors (Figure 2.5F). The effects of JQ1 on AURKA and AURKB expression and binding of BRD4 to their respective genes were again independent of TNBC subtype. Together, these data indicate that Aurora kinases are direct targets of BET inhibition through disruption of BRD4 binding to their respective genes.

Aurora kinases are precisely regulated throughout the cell cycle. Overexpression or silencing of these proteins elicits mitotic dysfunction, precluding restoring their

172 cyclic expression to assess their involvement in the BETi response822,823. To circumvent this limitation, we used a selective AURKB inhibitor, AZD1152

(AZD)824, to determine whether loss of AURKB can phenocopy the effects of BETi.

Indeed, 100 nM AZD induced multinucleation in MDA-MB-231 and MDA-MB-468 cells (Figure 2.6A) similar to that observed with JQ1. Moreover, 72 h of treatment with 100 nM AZD caused the different cell lines to primarily undergo the same differential cell fates of apoptosis or senescence in a manner that corresponded to their response to JQ1. Specifically, a greater proportion of MDA-MB-468 and

BT549 cells underwent apoptosis compared with MDA-MB-231 and HCC1143 cells (Figure 2.6B). MDA-MB-231 and HCC1143 cells, on the other hand, permanently senesced, as evidenced by increased cytoplasmic/nuclear ratio, enhanced SA-βgal activity, and decreased colony formation (Figure 2.6C and D).

Thus, inhibition of AURKB elicited the same cellular responses as JQ1 (Figure

2.6E). We also tested whether an AURKA inhibitor, MLN8237825, phenocopies

BETi and found it generated similar responses as JQ1 in the MDA-MB-231 and

MDA-MB-468 cell lines (Figure 2.6A, F, and G). Together, these data indicate that suppression of AURKA and/or AURKB activity causes the same effects as BET inhibition, specifically polyploidy, apoptosis, and senescence, and that the particular cellular fate (i.e. senescence or apoptosis) was consistent within each cell line regardless of whether aurora kinases or BET proteins were inhibited.

2.5 Discussion

TNBC is the most aggressive subtype of breast cancer due, in part, to its lack of effective targeted therapies. Although BETi have previously been shown to reduce

173 invasiveness of TNBC cells in vitro455 and inhibit tumor growth in xenograft models420,455,536, the utility in different subtypes of TNBC and the mechanism by which BETi elicit their effects on growth have not been previously established.

Here, we report that BET inhibition results in growth suppression of TNBC cell lines independent of their intrinsic subtype, including claudin low and basal subtypes474,826, as well as five of the six more recently defined TNBC subtypes described by Lehmann, et al.122. Although suppression of MYC expression is essential for the effects of BETi in other cancers468,526,527, this is not the case in a subset of TNBC cell lines, because we and others420 have found that BETi- mediated growth inhibition occurs independently of MYC down-regulation. We also report that BETi reduce tumor growth in three TNBC xenograft models and decrease the incidence of liver metastasis in a highly aggressive model while having no impact on the normal mammary gland. To our knowledge, these data are the first to demonstrate that BETi suppress distal metastasis in vivo for any tumor type. Through the use of diverse in vitro and in vivo models of TNBC, our data together with the recently published study by Polyak and co-workers420 provide clear preclinical evidence of efficacy of BETi in TNBC models, supporting the future assessment of these drugs in clinical trials of patients with this spectrum of disease. Notably, BETi induce minimal to no toxicity in these animal models, suggesting high selectivity for cancers rather than normal tissues. This selectivity may be due to the differential responsiveness of normal and cancer cells to disruption of mitotic and cytokinetic events. Although normal cells maintain an ability to arrest to ensure accurate cell division, cancer cells lack these

174 mechanisms and ultimately abort aberrant cell division by either dying or activating a senescence program535.

Although BETi have been shown to inhibit growth or induce cell death of many tumor models, the mechanism(s) underlying either response to these drugs has not been fully explored. As indicated above, some tumor types are highly reliant on MYC, and BETi have been shown to suppress this gene in these models.

However, in TNBC, the mechanism of action of these drugs appears more complex and likely involves the suppression of AURKA and AURKB and the resulting induction of multinucleation. AURKB has previously been suggested to be a target of BRD4, because its expression is reduced upon BRD4 silencing414,827,828. This serine/threonine kinase is essential for cytokinesis, phosphorylation of histone H3, and appropriate spindle attachment to the kinetochore during metaphase672.

Overexpression or loss of AURKB induces multinucleation followed by either apoptosis or senescence in other cellular models528-530,829. Another family member,

AURKA, also plays a critical role in mitosis by mediating centrosome maturation and duplication, bipolar spindle assembly, alignment of chromosomes, and cytokinesis672. Here, we show BETi rapidly down-regulate both AURKA and AURKB, and this occurs with the reduction of BRD4 binding to their respective promoters. Furthermore, use of selective AURKA and AURKB inhibitors revealed that blocking the activity of either protein phenocopies the polyploidy, apoptosis, and senescence phenotypes induced by BETi in TNBC cells. Notably, the cell-specific fates that occur in response to AURKA/AURKB inhibitors mimicked those of BETi. Together, these data indicate that the

175 mechanism of action of BETi in TNBC involves direct suppression of the AURKA and AURKB genes. What remains unclear is the cell-specific mechanism underlying the choice to undergo apoptosis or senescence in response to BET or Aurora kinase inhibitors. Thus far, we have found that the extent to which cells undergo JQ1-induced apoptosis and/or senescence is independent of TNBC subtype, changes in MYC expression, the extent of AURKA/AURKB suppression, or the JQ1 GI50. It will now be important to identify the pathways driving the choice of cells to undergo BETi-induced senescence versus apoptosis to reveal biomarkers of therapeutic response, as well as to identify approaches that ensure cell death in tumors treated with BETi. It will also be important to determine whether cell lines that are representative of other types of cancer, including luminal breast cancer, also respond to BETi by suppressing AURKA/AURKB and altering ploidy or whether this effect is specific to TNBC.

In summary, inhibition of BET proteins in diverse forms of TNBC reduces the expression of AURKA and AURKB, critical factors for normal cell division. This results in growth arrest and polyploidy. Cells respond to the ploidy defects by either undergoing senescence or cell death. In two mouse models of TNBC, BETi suppresses tumor growth while it induces regression in a third. These responses are consistent with the ability of BETi to induce senescence or apoptosis in a cell line-dependent manner. Lastly, BETi reduce liver metastasis in a highly aggressive

TNBC model. Together, these data reveal that Aurora kinases play a key role in the response of TNBC to BETi and provide preclinical evidence supporting the future use of BETi in diverse subtypes of TNBC to suppress tumor growth and

176 metastasis. Moreover, they suggest that changes in AURKA and/or AURKB expression may serve as valuable biomarkers to predict therapeutic response.

177 A. 1.2 1.2 1.2 HCC1143 MDA-MB-468 HCC70 1 1 1 * * 0.8 0.8 * 0.8 * * * 0.6 * 0.6 0.6 * * * * * 0.4 0.4 * 0.4 0.2 0.2 0.2 Relative Cell Number Relative Cell Number Relative Cell Number 0 0 0 0 100 250 500 750 1000 0 100 250 500 750 1000 0 100 250 500 750 1000 [JQ1] (nM) [JQ1] (nM) [JQ1] (nM)

B. 1.2 1.2 1.2 MDA-MB-231 BT549 HCC38 1 1 * 1 * * * * 0.8 0.8 * 0.8

0.6 0.6 * 0.6 * * 0.4 0.4 0.4 * * * * * 0.2 0.2 0.2 * Relative Cell Number Relative Cell Number Relative Cell Number 0 0 0 0 100 250 500 750 1000 0 100 250 500 750 1000 0 100 250 500 750 1000 [JQ1] (nM) [JQ1] (nM) [JQ1] (nM)

C. 1.2 D. 1.2 E. 1.2 MDA-MB-453 1 1 1 * 0.8 0.8 0.8

0.6 * 0.6 0.6

0.4 * 0.4 0.4 MDA-MB-231 MDA-MB-231 * 0.2 * 0.2 MDA-MB-468 0.2 MDA-MB-468 Relative Cell Number Relative Number Cell Relative HCC70 Number Cell Relative HCC70 0 0 0 0 100 250 500 750 1000 0 100 250 500 750 1000 0 100 250 500 750 1000 [JQ1] (nM) [I-BET151] (nM) [I-BET762] (nM)

F. 1.4 Vehicle JQ1 1.2 1 * 0.8 * 0.6 * * 0.4 * MYC/GAPDH 0.2 0 1143 468 HCC70 231 BT549 HCC38 453 Basal Claudin-low Luminal G.

Figure 2.1. BET inhibition blocks growth of TNBC cells without consistently down-regulating MYC.

178 (A–C) Relative fold change in cell number (compared with initial plating density)

for seven TNBC cell lines representing the basal (A), claudin low (B), and luminal androgen receptor (C) subtypes of TNBC. The cells were treated with increasing concentrations of JQ1, and viable cells were counted after 72 h. (D and E) Growth curves of TNBC cells treated with increasing doses of I-BET151 (D) or I-BET762

(E). Viable cells were counted after 72 h. (F and G) RT-qPCR (F) and Western

blotting analysis (G) of MYC expression levels in TNBC cell lines treated for 24 h

with vehicle or 500 nM JQ1. The values on the Western blot are relative to the

vehicle-treated 1143 sample following normalization to β-actin. For all graphs, the data are means ± S.D. *, p < 0.05 compared with vehicle.

179 Table 2.1: TNBC subtypes and responses to BET inhibitors

GI MYC Gray Pietenpol p53 50 E E D Cell line A A B C Senescence Apoptosis classification classification status (nM) expression

Inactivating HCC1143 Basal A (Basal) Basal-like 1 mutation, 800 0.33 ++ + missense

Inactivating MDA-MB- Basal A (Basal) Basal-like 1 mutation, 470 0.41 - ++ 468 missense

Inactivating HCC70 Basal A (Basal) Basal-like 2 mutation, 930 0.31 - ++ missense

Inactivating MDA-MB- Basal B Mesenchymal mutation, 270 0.81 ++ + 231 (Claudin-low) stem-like missense

Inactivating Basal B HCC38 Basal-like 1 mutation, 240 1.13 - +++ (Claudin-low) missense

Inactivating Basal B BT549 Mesenchymal mutation, 2000 0.97 - ++ (Claudin-low) missense

Luminal Inactivating MDA-MB- Luminal Androgen mutation, 220 0.60 - +++ 453 Receptor deletion

A 7 TNBC cell lines organized according to subtype based on 2 classification

systems: Gray474 and Pietenpol122.

B “Inactivating mutation” is defined as a mutation that lacks the ability to regulate

the transcription of known p53-target genes, such as CDKN1A.

C GI50 of JQ1 after 72 hours of treatment.

D Relative MYC expression after 24 hours of treatment with 500 nM JQ1 compared

to vehicle, measured by qRT-PCR.

180

E Responses to prolonged (up to 8 days) JQ1 treatment. ++, +++ is representative

of the time to, and extent of, response to BET inhibition.

181

A. 40 * B. 30 * * 231 (C) 1143 (B) 468 (B) BT549 (C) * JQ1: - + - + - + - + 20 * * * * * * * p21 10 β-actin % Apoptotic Cells Apoptotic % 0 JQ1: - - - - 231 (C) 1143 (B) 468 (B) BT549 (C)

C. MDA-MB-231 (C) Vehicle 250 nM 500 nM 1000 nM Day 8 Day Day 14 Day

D. HCC1143 (B) Vehicle 250 nM 500 nM 1000 nM Day 8 Day Day 14 Day

E. MDA-MB-231 (C) F. HCC1143 (B) Vehicle 250 nM 500 nM 1000 nM Vehicle 250 nM 500 nM 1000 nM

Figure 2.2. Sustained BET inhibition induces apoptosis and senescence in

TNBC cells.

182 (A) The indicated TNBC cell lines were treated with vehicle or JQ1 (250, 500, or

1000 nM) for 72 h and stained with Hoechst. The percentage of pyknotic nuclei was then determined. The data are means ± S.E. *, p < 0.05 compared with vehicle. (B) Western blots for p21 and β-actin in TNBC cell lines treated for 24 h with vehicle or 500 nM JQ1. (C and D) Top panels, representative phase images

(10x) of MDA-MB-231 (C) and HCC1143 (D) cells treated with vehicle or JQ1 for

8 days. Middle and bottom panels, representative images (4x) of MDA-MB-231 (C) and HCC1143 (D) cells treated with vehicle or JQ1 for 8 days (middle panels) or

14 days (bottom panels) and stained for SA-βgal activity. (E and F)

Representative images of plates of MDA-MB-231 (E) and HCC1143 (F) cells examined for their ability to form colonies after removal of JQ1. Following 14 days of treatment with vehicle or JQ1, 500 (231) or 1000 (1143) live cells were seeded and grown in complete medium in the absence of JQ1 for 11 additional days. The resulting colonies were stained with crystal violet. Triplicate wells are shown for each concentration of JQ1. (B), basal; (C), claudin low.

183 A.B. C. 231 (C) 2000 25 MDA-MB-231 (C) MDA-MB-231 (C) >50 1800 1600 20 30 1400 15 1200 1000 20 Vehicle 10 800 JQ1 600 5 * 10 * * * * 400 RelativeTumor Volume * * 200 Number of Liver Metastases Liverof Number 0 change volumetumor Percent 1 4 7 11 15 18 22 26 28 0 0 Vehicle JQ1 Days of Treatment Vehicle JQ1 D. E. 4 MDA-MB-468 (B) 500 MDA-MB-468 (B) 400 3 300

2 200 PD * 100 1 * * * * * * * * 0 SD Vehicle

RelativeTumorVolume PR

JQ1 Percenttumor volume change 0 -100 1 5 10 13 17 21 24 28 32 35 Days of Treatment Vehicle JQ1 F. G. H. 1.4 2 5 Vehicle JQ1 MDA-MB-231 (C) BCM-4013 (B) 1.8 1.2 1.6 4 1 1.4 * 0.8 1.2 3 1 GAPDH

/ * 0.6 * 0.8 2 * *

MYC 0.6 0.4 MYC/β-actin * * 0.4 1 0.2 Vehicle 0.2 RelativeTumor Volume JQ1 0 0 0 231 (C) 468 (B) Vehicle JQ1 1 5 7 11 14 20 23 27 30 Days of Treatment

Figure 2.3. BET inhibition abrogates tumor growth.

(A) NOD/scid/γ mice with palpable MDA-MB-231 tumors were treated with vehicle or JQ1. The graph shows relative tumor size over 28 days of treatment. The data are means ± S.E. (B) Waterfall plot showing the percentage of change in tumor size for each mouse at the end of the study compared with the first day of treatment. Each bar represents an individual tumor. (C) After 28 days of treatment, surface liver macrometastases were counted. Each dot represents the number of

184 liver metastases within an individual mouse. (D) NOD/scid/γ mice with palpable

MDA-MB-468 tumors were treated with vehicle or JQ1. The graph shows relative tumor size over 35 days of treatment. The dashed line indicates initial tumor size.

The data are means ± S.E. (E) Waterfall plot showing the percentage of change in size for each MDA-MB-468 tumor at the end of the study. RECIST values820 are indicated by horizontal lines and shown to the right of the graph. PD, progressive disease; SD, stable disease; PR, partial regression. (F) RT-qPCR analysis of MYC expression in vehicle and JQ1-treated tumors (MDA-MB-231 tumors n =

5/group, MDA-MB-468 tumors n = 10/group). The data are means ± S.D. *, p <

0.05 compared with vehicle; comparing vehicle- and JQ1-treated MDA-MB-231 tumors, p = 0.21. (G) Quantitation of Western blotting analysis of MYC expression levels in MDA-MB-231 tumors treated with vehicle or JQ1 for 28 days. MYC protein expression was normalized to β-actin. The values are means ± S.D. (H) Relative tumor size of PDX BCM-4013 tumors treated with vehicle or JQ1 (n = 10/group).

The data are means ± S.E. (B), basal; (C), claudin low.

185 A. 34 34 34 MDA-MB-231 (C) MDA-MB-468 (B) BCM-4013 (B) 32 32 32

30 30 30

28 28 * * 28 Weight (g) Weight Weight(g) 26 26 (g) Weight 26 Vehicle Vehicle Vehicle 24 24 24 JQ1 JQ1 JQ1

22 22 22 1 5 9 16 23 28 1 5 10 17 24 32 1 5 7 11 14 20 23 27 30 Days of Treatment Days of Treatment Days of Treatment

B. C. Vehicle JQ1 0.02

0.016

0.012

0.008

% pH3-PositiveCells 0.004

0 Vehicle JQ1

Figure 2.4. BETi lack toxic side effects and do not impact normal adult mouse mammary gland morphology or proliferation.

(A) Mice from all three tumor studies were weighed once per week. The data are means ± S.E. (B) Representative whole mounts of inguinal mammary glands isolated from FVB/N adult female mice treated with vehicle or JQ1 for 1 week. The glands were stained with Carmine alum. (C) Quantitation of phospho-histone H3- positive cells in sections of mammary glands from mice treated with vehicle or JQ1 for 1 week. The data are means ± S.E. *, p < 0.05 compared with vehicle, for all graphs. (B), basal; (C), claudin low.

186 A. Senescent Apoptotic 50 * Vehicle JQ1 MDA-MB-231 (C) HCC1143 (B) MDA-MB-468 (B) 40 * 30

20 *

Vehicle 10 % MultinucleatedCells 0 231 (C) 1143 (B) 468 (B) B. 231 (C)

100

500 nM 500 JQ1 80 60 4N 40 2N

C. % Gated Cells 20 1.2 Vehicle JQ1 0 1 0 500 0.8 [JQ1] (nM) GAPDH

/ 0.6 0.4 * D. * * Senescent Apoptotic

AURKA 0.2 * 231 (C) 1143 (B) 468 (B) BT549 (C) 0 1.2 JQ1: -+ - + - +-+ 1 AURKA 0.8 * * 1.0 0.2 1.0 0.3 1.0 0.4 1.0 0.7

GAPDH * / 0.6 * 0.4 AURKB

AURKB 0.2 1.0 0.1 1.0 0.3 1.0 0.5 1.0 0.5 0 β-actin 231 (C) 1143 (B) 468 (B) BT549 (C) Senescent Apoptotic

E. MDA-MB-231 (C) F. 2.5 1.4 Negative Primer Vehicle JQ1 * 1.2 2 Vehicle JQ1 1 1.5 * 0.8 * * 1 * 0.6 0.4 * 0.5 *

RelativeBRD4 Binding 0.2 Relative mRNA Expression mRNA Relative 0 0 CDKN1A AURKA AURKB 231 (C) 468 (B) 231 (C) 468 (B) AURKA AURKB

Figure 2.5. Aurora kinases are downstream targets of BETi.

(A) Left panel, representative images (20x) of the indicated cell lines treated with vehicle or 500 nM JQ1 for 96 h and stained with DAPI (blue, nuclei) and Texas

Red-X phalloidin (red, actin cytoskeleton). Insets show examples of multinucleated

187 cells. Arrows indicate multinucleated cells. Bars, 50 μm. Right panel, quantitation of multinucleated cells. The data are means ± S.D. (B) MDA-MB-231 cells were treated for 8 days, stained with propidium iodide, and analyzed by flow cytometry.

The data are means ± S.E. (p < 0.05 compared with vehicle for both 2N and 4N populations). (C) RT-qPCR analysis of AURKA and AURKB expression in the indicated TNBC cell lines treated with vehicle or 500 nM JQ1 for 24 h. The data are means ± S.D. (D) Western blotting analysis of AURKA and AURKB expression in four TNBC cell lines treated with vehicle or 500 nM JQ1 for 24 h. The values on the Western blot are relative to untreated samples per cell line following normalization to β-actin. (E) Representative gene-specific ChIP-PCR analysis of

MDA-MB-231 cells assessing binding of BRD4 to AURKA and AURKB.

CDKN1A was used as a control that does not lose BRD4 binding with JQ1 treatment. Negative primers were designed to areas outside the promoter region for each gene. The data are means ± S.D. (F) RT-qPCR analysis of AURKA and AURKB expression in MDA-MB-231 and MDA-MB-468 tumors from mice treated for 28 (MDA-MB-231 cells) or 35 (MDA-MB-468 cells) days with vehicle or JQ1. The data are means ± S.D. *, p < 0.05 compared with vehicle, for all graphs. (B), basal; (C), claudin low.

188 A. C. Vehicle AZD MLN Vehicle JQ1 AZD 231 (C) 231 231 (C) 231 1143 (B) 1143 468 (B) 468 D. MDA-MB-231 (C) B. 60

50 * Vehicle 40 * * 30 * * * AZD 20 * * % ApoptoticCells 10

0 F. 231 (C) 1143 (B) 468 (B) BT549 (C) 60 JQ1 ‐ + ‐ ‐ + ‐ ‐ + ‐ ‐ + ‐ AZD ‐ ‐ + ‐ ‐ + ‐ ‐ + ‐ ‐ + 50 * * Senescent Apoptotic 40 * 30 * * E. 20 * Cell line Subtype JQ1 AZD1152

% ApoptoticCells 10 Senescence/ Senescence/ 0 HCC1143 Basal Apoptosis Apoptosis MDA-MB-231 (C) MDA-MB-468 (B) JQ1 ‐ + ‐ ‐+‐ ‐ ‐ MDA-MB-468 Basal Apoptosis Apoptosis AZD ‐ ‐ + ‐ ‐ ‐ + ‐ MLN ‐‐‐+‐ ‐‐+ HCC70 Basal Apoptosis Apoptosis

Senescence/ Senescence/ MDA-MB-231 Claudin-low Apoptosis Apoptosis G. Vehicle MLN

HCC38 Claudin-low Apoptosis Apoptosis

BT549 Claudin-low Apoptosis Apoptosis 231 (C) 231 MDA-MB-453 Luminal Apoptosis Apoptosis

Figure 2.6. Aurora kinase inhibitors phenocopy BETi.

(A) Representative images (20x) of MDA-MB-231 and MDA-MB-468 cells treated with vehicle, 100 nM AZD1152 (AZD), or 250 nM MLN8237 (MLN) for 96 h. The cells were stained with DAPI (blue, nuclei) and Texas Red-X phalloidin (red, actin cytoskeleton). Arrows indicate multinucleated cells. Bars, 50 μm. (B) The indicated

TNBC cell lines were treated with vehicle, 1000 nM JQ1, or 100 nM AZD for 72 h

189 and stained with Hoechst. The percentages of pyknotic nuclei were then determined. The data are means ± S.E. (C) Representative images (4x) of MDA-

MB-231 (top panels) and HCC1143 (bottom panels) cells treated with vehicle,

1000 nM JQ1, or 100 nM AZD for 8 days and stained for SA-βgal activity. (D)

Representative image of colony forming assays using MDA-MB-231 cells treated for 14 days with vehicle or 100 nM AZD. (E) Summary of the predominant responses of seven TNBC cell lines to JQ1 and AZD treatment. (F) MDA-MB-231 and MDA-MB-468 cells were treated with vehicle, 1000 nM JQ1, or 250 nM MLN8237 for 72 h and stained with Hoechst. The data are means ± S.E. (G)

Representative images (4x) of MDA-MB-231 cells treated with vehicle or 250 nM MLN8237 for 8 days and stained for SA-βgal activity. *, p < 0.05 compared with vehicle, for all graphs. (B), basal; (C), claudin low.

190

CHAPTER 3: MITOTIC VULNERABILITY IN TRIPLE-NEGATIVE BREAST

CANCER ASSOCIATED WITH LIN9 IS TARGETABLE WITH BET INHIBITORS

This research was originally published in Cancer Research. Sahni JM, Gayle SS,

Webb BM, Weber Bonk KL, Seachrist DD, Singh S, Sizemore ST, Restrepo NA,

Bebek G, Scacheri PC, Varadan V, Summers MK, and Keri RA. Mitotic vulnerability in triple-negative breast cancer associated with LIN9 is targetable with

BET inhibitors. Cancer Res. 2017; 77: 5395-408. © The American Association for

Cancer Research

191

3.1 Abstract

Triple-negative breast cancers (TNBC) are highly aggressive, lack FDA-approved

targeted therapies, and frequently recur, making the discovery of novel therapeutic

targets for this disease imperative. Our previous analysis of the molecular

mechanisms of action of bromodomain and extraterminal protein inhibitors (BETi)

in TNBC revealed these drugs cause multinucleation, indicating BET proteins are

essential for efficient mitosis and cytokinesis. Here, using live cell imaging, we

show that BET inhibition prolonged mitotic progression and induced mitotic cell

death, both of which are indicative of mitotic catastrophe. Mechanistically, the

mitosis regulator LIN9 was a direct target of BET proteins that mediated the effects

of BET proteins on mitosis in TNBC. Although BETi have been proposed to

function by dismantling super-enhancers (SE), the LIN9 gene lacks an SE but was

amplified or overexpressed in the majority of TNBCs. In addition, its mRNA

expression predicted poor outcome across breast cancer subtypes. Together,

these results provide a mechanism for cancer selectivity of BETi that extends

beyond modulation of SE-associated genes and suggest that cancers dependent

upon LIN9 overexpression may be particularly vulnerable to BETi.

3.2 Introduction

Proper progression through mitosis is critical for maintaining cell function and

viability. To prevent mitotic defects and subsequent chromosomal instability, the

expression and activity of mitotic proteins are carefully controlled by several

mechanisms, including ubiquitin-mediated protein degradation, phosphorylation,

miRNA regulation, and transcription. FOXM1, E2F family members, the MuvB core

192

complex (composed of LIN9, LIN52, LIN37, LIN54, and RBBP4), B-MYB, and NF-

Y are master transcriptional regulators of mitosis and are responsible for the timely

expression of genes encoding crucial mitosis proteins, including AURKA, AURKB,

PLK1, and CCNB1830. When these transcription factors are dysregulated,

abnormal mitosis occurs that can produce cells with aberrant nuclei (potentially

with damaged DNA) and induce cell death pathways, senescence, and/or

oncogenesis830. One mechanism that avoids genomic instability is mitotic

catastrophe, a process that occurs due to chromosomal abnormalities or abnormal

mitosis, coincides with mitotic arrest, and leads to one of three cell fates:

irreversible senescence, death during mitosis, or death immediately following

mitotic exit831,832. Before the execution of these oncosuppressive mechanisms, a

characteristic early-stage indicator of mitotic catastrophe is the appearance of

multiple nuclei and/or micronuclei832,833. Either early entry into mitosis or failed

mitosis can trigger mitotic catastrophe831,832. In cancer, mitotic catastrophe can be

induced in response to treatment with ionizing radiation and anti-cancer agents,

including microtubule-targeting and DNA-damaging drugs, and the inhibition of

mitotic catastrophe provides a mechanism for tumor initiation and the development

of chemoresistance834-836.

Triple-negative breast cancer (TNBC) is the most aggressive subtype of breast

cancer, and there is a paucity of effective targeted therapies for this disease. These

tumors are treated with traditional chemotherapy such as taxanes and

anthracyclines, and while they initially respond, they frequently recur within 3

years837. It is therefore critical that we develop new treatment strategies for this

193

devastating disease. We and others have recently reported that bromodomain and

extraterminal protein inhibitors (BETi) are efficacious in multiple models of

TNBC420,455,456,472,473. We further discovered that BETi induce the formation of

large, multinucleated cells followed by apoptosis and senescence, suggesting

these drugs cause mitotic catastrophe472. BETi selectively target the BET family of

epigenetic readers by binding to the bromodomain pockets of BET proteins (BRD2,

BRD3, BRD4, and BRDT). This prevents recruitment of these proteins to

chromatin, thus suppressing their transcriptional activity460. BETi are efficacious in

mouse models of diverse cancers812 and are currently being investigated in early

phase clinical trials. The selectivity for cancers and broad therapeutic windows

observed with BETi in mice have been suggested to result from the selective

disruption of super-enhancers (SE), exceptionally large clusters of enhancers that

control expression of cell identity genes and, in cancer, critical oncogenes366,446.

BRD4 disproportionately accumulates at SEs compared with typical enhancers.

Hence, dismantling SEs at oncogenes would have a greater transcriptional effect

and be more impactful in cancer cells that depend on those genes rather than

normal cells. This model provides a mechanism to preferentially silence

oncogenes, which could in turn inhibit tumor formation, growth, and progression,

while sustaining viability of normal tissues. However, it remains unclear whether

the primary mechanism for selectivity of BETi in cancers involves disruption of SEs

at oncogenes, or if cancer cells may be particularly sensitive to the suppression of

viability genes that extend beyond oncogenes and those involved in maintaining

cell identity. Identifying the processes underlying cellular responses to these

194

inhibitors will be essential for improving patient selection for future clinical trials,

predicting therapeutic response and resistance, and rationally discovering optimal

added therapies for evoking synergistic tumor responses.

Here, we show for the first time that suppression of BET protein activity leads to a

significant delay or death in mitosis in TNBC cells. Together with the generation of

multinucleated cells, these findings indicate BETi induce mitotic catastrophe. This

process is initiated by the direct suppression of LIN9 as well as other cell-cycle

regulatory transcription factors, including FOXM1 and MYBL2. None of these

genes contain SEs, disputing the concept that tumor response to BETi solely relies

on the dismantling of such enhancers. Notably, LIN9 is amplified or overexpressed

in the majority of TNBC tumors and its suppression mimics BETi. This indicates

that LIN9 may be an exploitable therapeutic target in TNBC that can be selectively

silenced with BETi.

3.3 Materials and methods

Cell culture and reagents

MDA-MB-231, MDA-MB-468, HCC1143, HCC70, and HCC38 cells from the ATCC

were grown in RPMI-1640 supplemented with 10% FBS and maintained at 37°C

with 5% CO2. MDA-MB-231 cells were authenticated in 2013 by STR profiling

(BDC Molecular Biology Core Facility, University of Colorado). All other cell lines

were purchased from the ATCC between 2008 and 2010. Upon receipt, they were

thawed and expanded for freezing approximately 75, 1 mL vials. From each of

these vials, approximately 75, 1 mL vials were generated. All experiments were

195

then performed with cells that were within 10 passages of these secondary vials.

MDA-MB-231 and HCC1143 cells were tested in 2013 for Mycoplasma

pulmonis and Mycoplasma spp. by IDEXX RADIL (Columbia, MO). JQ1 was

dissolved in dimethyl sulfoxide (DMSO). Transient mRNA silencing was performed

using the following siRNAs (Dharmacon): Nontargeting siRNA #2 (D-001810-02-

20), siFOXM1 (L-009762-00), siE2F2 (L-003260-00), siE2F8 (L-014407-01),

siLIN9 (L-018918-01), and siMYBL2 (L-010444-00).

Caspase 3/7 cleavage stain

Cells were treated with vehicle or JQ1 for 72 hours. Media and cells were then

harvested and stained with CellEvent Caspase-3/7 Green Reagent (Molecular

Probes, R37111) for two hours according to the provided protocol. GFP-positive

(apoptotic) and GFP-negative (live) cells were counted using a Countess II FL

(Thermo Fisher, AMQAF1000) with an EVOS GFP light cube (Thermo Fisher,

AMEP4651).

Senescence-associated β-galactosidase activity stain

Staining was performed as previously described472.

Live cell imaging

HCC38 and MDA-MB-231 cells were treated with vehicle or 1,000 nmol/L JQ1 and

imaged using the IncuCyte Zoom System (Essen BioScience) for 4 days. Images

were collected at ×20 every 20 minutes. Individual cells were tracked from mitotic

entry to mitotic exit and the time course of events determined. Relative cell

proliferation was analyzed using the IncuCyte software Confluence application.

196

RNA analysis

RNA analysis was performed as previously described472 using the following

TaqMan Gene Expression Assays (Thermo

Fisher): CCNB1 (Hs01030099_m1); E2F2 (Hs00231667_m1); E2F8(Hs0022663

5_m1); FOXM1 (Hs01073586_m1); KIF2C (Hs00901710_m1); KIF20A(Hs00993

573_m1); LIN9 (Hs00542748_m1); LIN37 (Hs00375230_m1); MYBL2 (Hs009425

43_m1); PLK1 (Hs00983227_m1); GAPDH (Hs02758991_g1).

In vivo studies

All in vivo experiments were performed with approval from the Institutional Animal

Care and Use Committee at Case Western Reserve University, which is certified

by the American Association of Accreditation for Laboratory Animal Care. Mouse

xenograft studies were previously reported with tissues being used for the current

analysis472. Briefly, MDA-MB-231 or MDA-MB-468 cells were xenografted into the

two inguinal mammary fat pads of adult female NOD/scid/γ (NSG) mice. Once

palpable tumors formed, mice were randomized into two treatment groups: vehicle

(1:1 propylene glycol:water) or JQ1 (50 mg/kg IP daily). After 28 (MDA-MB-231) or

32 (MDA-MB-468) days of treatment, tumors were removed and processed for RT-

qPCR analysis.

Nuclear morphology

FOXM1, E2F2, E2F8, LIN9, or MYBL2 were transiently silenced in MDA-MB-231

cells grown on sterile glass coverslips. After five days, cells were fixed with 3.7%

formaldehyde/1xPBS, permeabilized with 0.1% Triton X-100/1xPBS, and blocked

197

with 1% BSA/1xPBS. Texas Red-X phalloidin (Invitrogen, T7471) was used to label

F-actin. Nuclei were counterstained with Vectashield hard set mounting medium

with DAPI (Vector Labs, H-1500).

Gene expression microarray analysis

MDA-MB-231 and HCC70 cells were treated with vehicle or 500 nmol/L JQ1 for 72

hours and were processed for transcriptional profiling using Human Gene 2.0 ST

expression microarray (Affymetrix). Cells were harvested and RNA was extracted

with TRizol reagent (Ambion, 15596018) and treated with DNase I (Ambion,

AM1906). RNA (50 ng/μL) was delivered to the Gene Expression and Genotyping

Core Facility at Case Western Reserve University. For each sample, 150 ng RNA

was used to synthesize and label cDNA with biotin for hybridization to Human

Gene 2.0 ST expression microarrays using the GeneChip WT Plus labeling kit and

protocol, and the hybridized arrays were automatically stained and scanned with

the Affymetrix standard stain and scan protocol. The microarray data were

processed with RMA (robust multichip average algorithm) as implemented in

Bioconductor package oligo838 where background subtraction, quantile

normalization and summarization (via median-polish) was accomplished. The top

differentially expressed genes were identified using empirical Bayesian procedure

of limma package839. Multiple-testing correction was performed using the

Benjamini-Hochberg method, thus providing FDR estimates per differentially

expressed gene. The final list of differentially expressed genes was defined as

those with P values < 0.05 and FDR < 0.1. All data were submitted to GEO

(GSE79332) using MIAME guidelines. The Reactome database840 was used to

198

identify the top five nonoverlapping biological pathways regulated by BET inhibition

in both cell lines with a P value cutoff of 0.01. Violin and volcano plots were

generated in R version 3.3.3 using RStudio version 1.0.136 (RStudio Inc). Two

sample proportions tests were run in Stata SE 14 (StataCorp) with a significance

cutoff value of P < 0.05.

Gene set enrichment analysis

Gene set enrichment analysis (GSEA)841 was used to determine whether a

priori defined cell-cycle– and breast cancer subtype-associated genes show

statistically significant, concordant differences between vehicle- and JQ1-treated

samples in HCC70 and MDA-MB-231 cell lines. The GSEA portal molecular

signatures database842 was used to define the cell-cycle signatures. The subtype-

associated gene sets were tabulated from Neve and colleagues474 and Charafe-

Jauffret and colleagues475 datasets as previously described843. FDR-

corrected P values were considered to rank gene sets that had significant

enrichment.

Western blot analysis

Western blot analysis was performed as previously described843 using the

following primary and secondary antibodies: LIN9 (Thermo Fisher, PA5-43640), β-

actin (Sigma; A1978), Anti-rabbit IgG HRP-linked (Cell Signaling Technology;

7074), and Anti-mouse IgG HRP-linked (Cell Signaling Technology; 7076).

199

Gene-specific chromatin immuneprecipitation

Chromatin immunoprecipitation (ChIP)-PCR was performed as previously

described818. MDA-MB-231 cells were treated for 24 hours with vehicle or 500

nmol/L JQ1, and chromatin was immunoprecipitated with an anti-BRD4 specific

antibody (Bethyl Laboratories, A301-985A50) or a control mouse IgG (Sigma,

I5281). The following promoter-specific primers were used: FOXM1, forward 5′-

GTAAGATGGAGGCGGTGTTG-3′ and reverse 5′-

GGGTGGCCTACCTTCTTAGG-3′; E2F2, forward 5′-

GACAATAGCAGGCACCCAGTA-3′ and reverse 5′-

AGCACTGGATTGCGAGTCTG-3′; E2F8, forward 5′-

TAGGAAGCACCCACCTGTTC-3′ and reverse 5′-

GGGAGAAATCCAGGCATCTA-3′; LIN9, forward 5′-

GGAACTGCAGGCTGTTTGTT-3′ and reverse 5′-

GGGTTTCGGGAACTGTGAGT-3′; MYBL2, forward 5′-

GTCTTCAAGTCCCAGCCAGT-3′ and reverse 5′-

CCGGAATGTTAAGGAGCAAA-3′.

ChIP-seq

ChIP-seq was performed as previously described818. Briefly, chromatin from MDA-

MB-231 cells was immunoprecipitated with an anti-H3K27Ac specific antibody

(Abcam, ab5079) or control mouse IgG (Sigma, I5281). Input and precipitated DNA

were used to produce libraries and conduct high-throughput sequencing by the

CWRU Genomics Sequencing Core. Sequences were quality-filtered using

FASTX-Toolkit and aligned using Bowtie 2844. MACS (Model-based Analysis of

200

ChIP-Seq)845 was used to identify ChIP-enriched regions with a P value

enrichment threshold of 10−9, and ROSE (Rank Ordering of Super-enhancers)

software366,447 was used to identify enhancer regions. The enhancer element

locations were then annotated using custom annotation scripts, jointly

called Grannotator [unpublished data]. Briefly, Grannotator combines genome-

wide gene/transcript location information obtained from the UCSC Genome

Browser and uses this information to annotate Granges (objects that contain the

enhancer element locations). For each enhancer element, Grannotator provides

the nearest 3′ and 5′ gene for both the forward and reverse strands in addition to

the distance of the gene's transcription start site from the enhancer location. This

information is then used to determine the gene most likely associated with an

enhancer element. All data were submitted to GEO (GSE95222).

Bioinformatics

CBioPortal846,847 was used to identify the percent of breast cancer tumors with

overexpressed (z score (RNA Seq V2 REM) ≥ 2.0) or amplified (GISTIC 2.0 score

= 2) genes as well as genes that are correlated with LIN9 in both The Cancer

Genome Atlas (TCGA)451 and METABRIC27,452 datasets. Relapse free survival

Kaplan–Meier curves in all breast cancer patients were generated with Kaplan–

Meier plotter 2014 edition517 and compared by the log-rank test. Patient groups

corresponding to high/low tumor expression of LIN9 were identified, resulting in an

optimal expression cutoff of 111 to define the low (n = 1127) and high (n = 637)

expression groups. To assess overall survival, gene-expression data for the

METABRIC dataset were retrieved from Oncomine (www.oncomine.com, Thermo

201

Fisher Scientific). Samples (n = 1971) were segregated into high (upper

10th percentile, n = 197) or low (remain 90th percentile, n = 1,774) groups based

upon LIN9 (probe ID ILMN_2137084) expression. The prognostic value of LIN9 on

5-year overall survival in this dataset was analyzed by statistical comparison of

Kaplan–Meier curves by the log-rank test. 95% hazard ratios were calculated using

a Cox regression model.

Statistical methods

Data are represented as mean values from three independent experiments

performed in triplicate. Statistical analyses were performed using the two-tailed

Student t test or χ2 test, and P values less than 0.05 were considered statistically

significant.

3.4 Results

3.4.1 Sustained BET activity is necessary for normal progression through mitosis

We previously reported that the loss of BET activity induces multinucleation,

followed by two distinct terminal responses (apoptosis and senescence)

irrespective of TNBC subtype472. To confirm these outcomes, we treated two

TNBC cell lines (MDA-MB-468 and HCC1143) with vehicle or the prototypical

BETi, JQ1. After 72 hours, activation of caspases 3 and 7 was assessed to quantify

the number of apoptotic cells. While both HCC1143 and MDA-MB-468 cells

underwent apoptosis in response to JQ1, MDA-MB-468 cells had a greater

apoptotic response compared with HCC1143 cells (Figure 3.1A). JQ1 also induced

the expression of senescence-associated β-galactosidase (SA-βgal) activity in

202

HCC1143 cells (Figure 3.1B), confirming that BETi induce senescence and/or

apoptosis in TNBC, depending on the cell line examined.

Cells respond to mitotic catastrophe by activating senescence or apoptosis

pathways831,832. Hence, the ability of BETi to induce these two responses as well

as multinucleation suggested that sustained activity of BET proteins may be

necessary for normal mitotic progression in TNBC cells. To assess the impact of

losing BET function on mitosis, we used live cell imaging to track the fates of

individual cells following treatment with JQ1. Both a primarily apoptotic (HCC38)

and a primarily senescent (MDA-MB-231) cell line were observed over four days

of vehicle or JQ1 exposure. Growth of HCC38 cells was completely arrested within

48 hours whereas MDA-MB-231 cells were initially more tolerant of JQ1 treatment

(Figure 3.2A). In both cases, when cells were tracked through their first mitosis

following an initial 6 hours of drug treatment, JQ1 treatment significantly extended

the time necessary for cells to complete mitosis (Figure 3.2B–D). In addition to

initially increasing the duration of mitosis, JQ1 treatment had a profound effect on

mitotic cell fate for both cell lines. The majority of JQ1-treated HCC38 cells died

immediately following mitotic exit whereas only 6% of control-treated cells died

during mitosis (Figure 3.3A and B). In addition, 23% of JQ1-treated HCC38 cells

died during mitosis compared with only 7% in vehicle-treated cells. Death during

mitosis was also associated with an extended time traversing mitosis compared

with cells that were completed mitosis (Figure 3.3C). Similar to HCC38 cells, the

MDA-MB-231 cell line experienced a large increase (∼8 fold) in the percentage of

cells that died immediately following mitosis with JQ1 treatment (Figure 3.3D).

203

While this cell line did not die during mitosis in response to JQ1, the drug caused

cells to undergo a protracted interphase, meaning they exited mitosis and survived

but never divided again, likely because they were entering senescence. Finally,

only approximately 6% of JQ1-treated MDA-MB-231 cells underwent a second cell

division, and this was associated with a significant increase in post-mitotic timing

(Figure 3.3E). The induction of multinucleation472, followed by apoptosis or

senescence, and the induction of death either in, or immediately after, mitosis

suggests that mitotic catastrophe is the primary mechanism of action of BETi

efficacy in both models of TNBC.

3.4.2 BET activity is necessary for sustained expression of cell-cycle-associated

genes

To discern the molecular mechanism underlying the mitotic defects that occur with

the loss of BET activity, we performed gene expression microarray analyses of cell

lines that primarily undergo apoptosis (HCC70) or senescence (MDA-MB-231) in

response to JQ1 (Figure 3.4A–H). After 72 hours, the expression of 1,271 genes

in both cell lines was significantly altered in the same direction in response to JQ1

compared with vehicle (Figure 3.4A). We found 149 (MDA-MB-231) and 51

(HCC70) Reactome pathways to be statistically significantly altered with BETi (P <

0.01), of which the strongest enrichments were observed in cell cycle/mitosis and

metabolism pathways (Figure 3.4B). GSEA using an established group of

classifying genes for each stage of the cell cycle842 revealed a significant (P < 0.05)

suppression of genes definitive for G2–M and M–G1transitions in both cell lines

following JQ1 treatment (Figure 3.4C). JQ1-induced repression of eight genes that

204

encode proteins involved in cell-cycle regulation, mitosis, and/or cytokinesis was

further confirmed by RT-qPCR in multiple TNBC cell lines (Figure 3.4D–F; Figure

3.5A). These genes were also downregulated in orthotopically xenografted tumors

(MDA-MB-231 and MDA-MB-468) collected from mice treated with either vehicle

or JQ1 (Figure 3.4G; Figure 3.5B and C). In addition, 14 of the 16 kinesins known

to play critical roles in mitosis and cytokinesis in humans848 were suppressed.

Finally, in both MDA-MB-231 and HCC70 cells, there was a skewed JQ1-mediated

downregulation of genes identified by SuperPath849 as critical for mitosis (P <<

0.001; Figure 3.4H). The reduced expression of mitosis-regulating genes in

response to BETi in TNBC cells and tumors further supports a role for BET proteins

in navigating the effective progression through mitosis in this disease.

3.4.3 BET inhibitors fail to induce a luminal differentiation signature

Breast cancer subtype switching has been observed with the manipulation of

multiple transcription factors in breast cancer cell lines843,850. Given the roles of

BET proteins as transcriptional modulators, we assessed whether their inhibition

with JQ1 could induce a shift in breast cancer cell fate from the basal to luminal

subtype. We assessed the potential for subtype switching by GSEA using the Neve

and colleagues474 and Charafe-Jauffret and colleagues475 genesets that classify

breast cancer cell line subtypes. Although the basal signature was diminished

upon JQ1 treatment, TNBC cells failed to gain a luminal signature (Figure 3.6).

Thus, although it has been suggested that a loss of BET activity results in the

differentiation of TNBC cells420, this genome-wide approach indicates BETi fail to

induce full subtype switching, or luminal differentiation, of TNBC cell lines. These

205

data further suggest shifts in breast cancer cell fate are unlikely to underlie the

responsiveness of TNBC cells to BETi.

3.4.4 BET proteins directly modulate the mitotic transcriptional program

Further analysis of the microarray data described above revealed that the

expression of nine genes encoding mitosis-controlling transcription factors was

suppressed by BETi. Of these factors, four (FOXM1, E2F8, LIN9, and MYBL2) are

associated with polyploidy827,851-853, a key response we observed in TNBC cells

following treatment with BETi472. Another, E2F2, plays a critical role in the BETi

response in liver cancer854. These genes are rapidly repressed in MDA-MB-231

cells after just six hours of JQ1 treatment (Figure 3.5A). The suppression of these

genes by JQ1 was further confirmed in tumors from mice orthotopically

xenografted with MDA-MB-231 or MDA-MB-468 cell lines (Figure 3.5B and C).

Notably, suppression of BET protein activity does not reduce all mitosis-regulating

transcription factors. For example, expression of LIN37, a subunit of the MuvB core

complex that regulates transcription of mitosis genes830, was not suppressed with

JQ1 treatment (Figure 3.5A). An additional BETi currently being evaluated in

clinical trials, I-BET762, also regulated the expression of these mitosis regulators

in a similar manner, indicating the suppression of these genes is due to the

inhibition of BET proteins and not off-target effects (Figure 3.7). To determine

whether the five genes are direct targets of BRD4, a BET family member, gene-

specific ChIP assays were used. This approach revealed that BRD4 binds the

promotor regions of all five genes and that BRD4 binding is significantly reduced

with JQ1 exposure (Figure 3.5D). This suggests that the mitotic disruption

206

observed with the loss of BET protein activity may be due to the reduced

expression of one or more of these factors.

To determine whether these five mitosis-regulating transcription factors may be

downstream effectors of BET proteins in TNBC, we used siRNA to simultaneously

silence all five genes in MDA-MB-231 cells. We then examined whether the

combined repression would phenocopy the BETi response by first examining the

same cell-cycle target genes evaluated above. The combined siRNAs effectively

reduced the expression of all five factors (Figure 3.8A), as well as the expression

of cell-cycle genes (Figure 3.8B), in a manner similar to JQ1 treatment (Figure

3.4D–F). Furthermore, expression of CDKN1A, which we previously reported

increases with BETi treatment of TNBC cells472, was also increased with the

silencing of these five factors (Figure 3.8B). Most importantly, suppression of all

five genes generated very large cells with multiple nuclei, a mitotic/cytokinetic

defect that is comparable to that observed when BET proteins are inhibited (Figure

3.8C), indicating that the loss of these five factors can recapitulate BETi treatment.

3.4.5 BETi suppress mitosis transcription factors in the absence of SEs

The increased sensitivity of cancer cells to BETi compared with nontransformed

cells is proposed to result from the disassembly of SEs at oncogenes366,446. Thus,

we assessed whether alterations in SEs may drive the response of the mitosis

transcription factors identified above, and hence, the induction of mitotic

catastrophe in TNBC. We generated an SE map of MDA-MB-231 cells using

H3K27Ac ChIP-seq and compared the list of SE-containing genes with those

regulated by JQ1. In MDA-MB-231 cells, 1,038 genes have putative SEs,

207

accounting for 3% of all genes, whereas 8.6% of genes whose expression is

altered by JQ1 contain SEs (P < 0.01; Figure 3.9A). This confirms prior reports

indicating that BETi-regulated genes are enriched for SEs when interrogating the

full complement of protein coding genes366,446. When examining the subset of 599

genes identified as mitosis-associated genes by SuperPath849, a larger percentage

of these genes contain an SE (∼10%, P < 0.005), yet there was no enrichment of

SEs at mitosis genes regulated by JQ1 compared with those that are not. Thus,

although mitosis-associated genes generally are more likely to contain an SE than

other genes, the presence of an SE does not dictate their response to BETi. Across

the genome, genes that contain an SE undergo greater repression by BETi than

genes that lack an SE (Figure 3.9B). However, when selectively examining mitosis-

associated genes, the extent of repression of these genes is independent of the

presence or absence of an SE (Figure 3.9C). Finally, we found no correlation

between the strength of a gene's enhancer score, defined by H3K27Ac presence

under basal conditions, with its extent of repression following JQ1 treatment

(Figure 3.9D).

To determine whether any of the five mitosis-regulating transcription factors we

identified as early responders to BETi (FOXM1, E2F2, E2F8, LIN9, and MYBL2)

are associated with SEs, we examined the ChIP-seq–binding profiles for H3K27Ac

at these genes. None of these genes have putative SEs (Figure 3.9D and E), even

though an SE at MYC is readily detected. We also examined a dataset of BRD4

binding profiles in a larger panel of TNBC cell lines compiled by Shu and

colleagues and confirmed in this additional dataset that these genes lacked SEs.

208

Together, these data indicate that disruption of SEs is unlikely to be the primary mechanism by which BETi suppress mitosis-regulating transcription factors. Thus,

SE disassembly is not responsible for the induction of mitotic catastrophe in response to BETi in TNBC. Rather, mitotic catastrophe occurs as a result of the direct suppression of master regulators of mitosis through the loss of typical BRD4 binding to the promotor regulatory regions of these genes.

3.4.6 LIN9 is a key downstream effector of BET proteins

Of the five transcription factors evaluated above, three interact to modulate cell- cycle progression and prevent entry into senescence. LIN9 is one of five subunits of the MuvB complex that works with FOXM1 and B-MYB (the protein product of MYBL2) to drive expression of critical mitosis genes and ensure proper progression through the cell cycle653. To assess whether reduced expression of these three genes is responsible for the observed mitotic defect that occurs with the loss of BET activity, or if E2F2 and E2F8 are also necessary, we simultaneously silenced the expression of FOXM1, LIN9, and MYBL2 using siRNA transfection. After 5 days, all three genes remained suppressed. More importantly, the same cell cycle target genes suppressed by the set of five factors were also inhibited with the loss of just FOXM1, LIN9, and MYBL2. Silencing these three factors also caused the cells to become large and multinucleated (Figure 3.10A and B), indicating that suppression of E2F2 and E2F8 were unnecessary for the mitotic response. We then sought to determine which BET protein(s) is necessary for the expression of FOXM1, LIN9, and MYBL2. We used individual and pairwise siRNA transfections to silence BRD2, BRD3, and/or BRD4for 72 hours. Only

209 simultaneous loss of BRD2 and BRD4 suppressed expression of FOXM1, LIN9,

and MYBL2 (Figure 3.10C and D). These data further affirm the selectivity of BETi

in controlling the expression of these three genes. Notably, we also found that

silencing either BRD2 or BRD4 reduces BRD3 expression, and this suppression is

further enhanced when BRD2 and BRD4 are silenced together (Figure 3.10C and

D). To our knowledge, this is the first time BRD2 and BRD4 have been shown to

be necessary for sustained BRD3 expression.

We then determined whether loss of just one of the mitosis-associated

transcription factors is sufficient for mediating the BETi response in TNBC by using

siRNAs individually targeting FOXM1, LIN9, or MYBL2. As controls, we also

individually silenced E2F2 or E2F8. After five days, expression of all targeted

genes was reduced (Figure 3.11A and 3.12A). However, only loss of LIN9 altered

the expression of cell cycle target genes in the same direction as JQ1 (Figure

3.11B and 3.12B). In addition, silencing LIN9 induced significant multinucleation

that is comparable with that observed with JQ1 treatment (Figure 3.11C). This

suggests that LIN9 is a principle target of BET proteins that is necessary for

maintaining mitotic progression.

To assess the clinical significance of LIN9 in human breast cancer, we interrogated

publically available datasets for changes in LIN9 copy number or expression and

found LIN9 is amplified or overexpressed in 24% to 29% of all breast tumors in the

TCGA451 and METABRIC27,452 datasets. In the basal-like subset of breast

cancers, LIN9 is amplified and/or overexpressed in 66% of TCGA samples (19%

amplified, 65% overexpressed). We assessed BETi-induced apoptosis in four

210

TNBC cell lines (MDA-MB-231, HCC1143, MDA-MB-468, and BT549) and found

there is no difference in the extent of apoptosis in cells with amplified versus

nonamplified but overexpressed LIN9 (data not shown). In contrast, other subunits

of the MuvB complex are amplified or overexpressed in less than 5% of all breast

cancers and less than 20% of basal-like cancers. Kaplan–Meier survival analysis

revealed high expression of LIN9 is correlated with lower relapse-free (Figure

3.11D) and overall survival (Figure 3.13) rates in all breast cancer patients.

Together, these data implicate LIN9 as a driver of breast cancer, in general, and

of TNBC, specifically. Given the frequent amplification of LIN9 in TNBC, it was

possible that the absence of an SE at the LIN9 locus was due to DNA-content

normalization used in the analysis of H3K27Ac ChIP-seq data. However, the

Cancer Cell Line Encyclopedia855 revealed that LIN9 is not amplified in MDA-MB-

231 cells. Further indicating that our approach would not eclipse SE assignments

due to amplification of a genomic locus, the MYC gene was identified as having an

SE even though MYCis highly amplified in these cells. Thus, LIN9 amplification

often leads to its overexpression in the absence of a canonical SE and this

expression can be suppressed with BETi.

To infer the global potential of modulating LIN9 expression as a mediator of BETi,

we identified genes whose expression is highly correlated (Pearson coefficient r ≥

0.5) with LIN9 in breast cancer samples from TCGA and determined if these genes

may be particularly sensitive to BETi, reasoning this would implicate LIN9 as being

upstream of these correlated genes. This analysis revealed that most genes whose

expression does not correlate with LIN9 (r < 0.5) in breast cancer samples are also

211

unresponsive to BETi in MDA-MB-231 cells (Figure 3.11E). In stark contrast, the

majority of genes whose expression is highly correlated with LIN9 in breast cancer

samples (r ≥ 0.5) are suppressed by BETi in TNBC cells, with correlated genes

being on average 2.5- to 3.2-fold more repressed by BETi than noncorrelated

genes (Figure 3.11F). Several genes whose expression correlates with LIN9 are

co-amplified with this gene. To determine whether BETi was simply modulating

expression of genes within the LIN9 amplicon, we subdivided the group of LIN9-

correlated genes into those that reside on chromosome 1q and those that do not.

While a subset of genes on 1q are repressed by BETi, most are not. Furthermore,

most of the LIN9-correlated genes that do not reside in this region are generally

much more suppressed by BETi (Figure 3.11E). As a member of the MuvB

complex, LIN9 interacts with 1,379 genes across the genome in HeLa cells653. To

determine if BETi responsive genes can bind to LIN9/MuvB, we interrogated this

dataset and found that 27.4% of genes reported to bind LIN9/MuvB are also

regulated by JQ1 in TNBC cells. In contrast, only 8.9% of non-MuvB bound genes

are modulated by JQ1 (P < 0.01, Figure 3.11G). Together, these data indicate that

LIN9 is downstream of BET proteins and a major mediator of BETi in TNBC cells.

3.5 Discussion

Identifying new therapeutic targets for TNBC is essential for improving outcomes

of patients with this disease. Using genetic and pharmacologic approaches, it is

well-established that the activity of BET proteins is essential for TNBC cell

growth, in vitro and in vivo420,455,456,472,473. However, the mechanism of action of

BETi in this disease has not been fully elucidated. To ensure effective clinical

212

utilization of these drugs, it is critical to develop a detailed understanding of their

mechanisms of action. Such information will be key for predicting which tumors are

likely to respond and for the rational selection of drug combinations. We recently

showed that BETi produce two distinct responses, apoptosis and senescence, and

these are preceded by multinucleation, indicating BETi induce mitotic

dysfunction472. Here, we report that BETi repress TNBC growth by inducing mitotic

catastrophe. In 2012, the International Nomenclature on Cell Death defined mitotic

catastrophe as an oncosuppressive mechanism that occurs in response to

aberrant mitosis, coincides with mitotic arrest, and induces any of three irreversible

cell fates: death during mitosis, apoptosis following mitotic exit, or

senescence832,856. The study described herein provides several lines of evidence

substantiating the role of mitotic catastrophe in the BETi response of TNBC cells.

First, we found BETi suppress a large number of mitosis-regulating genes,

including several (KIF20A, AURKA, AURKB, and FOXM1) whose loss is known to

cause mitotic catastrophe832. Supporting the expression studies herein, Borbely

and colleagues456 also recently reported that BETi repress cell-cycle gene

expression in breast cancer. Second, we found suppression of BET activity

increases the amount of time cells spend traversing mitosis with many becoming

multinucleated, processes that are also associated with mitotic catastrophe832.

Finally, direct evidence of BETi-induced mitotic catastrophe was observed when

individual cells were followed through mitosis. Live cell imaging revealed JQ1

greatly increases the number of cells that die during mitosis as well as the number

that exit mitosis and die soon thereafter. It is well-known that mitotic catastrophe

213

induces senescence or apoptosis, and this depends on the individual cell line832.

Both outcomes were observed in response to BETi but were cell line-specific472.

Together, these data indicate a loss of BET activity conveyed by BETi induces

mitotic catastrophe that leads to apoptosis and senescence in TNBC.

BETi induce apoptosis and senescence in TNBC while having no evident effect on

the normal mammary gland472. This preferential impact on cancer cells compared

with nontransformed cells has been attributed to the disruption of SEs366,410,446.

BRD4 disproportionately binds to SEs, with about 30% to 40% of all bound BRD4

located at these specialized regions of the chromatin366,446. Like other co-

activators, BRD4 exhibits cooperative binding. As a result, BRD4 inhibition causes

greater downregulation of SE-associated genes than typical enhancer-associated

genes366,446. We now report that inhibition of BET proteins also leads to extensive

dysregulation of cell-cycle genes that are particularly active in cancer. Besides

being more sensitive to the disruption of SEs, cancer cells are also more

susceptible to mitotic catastrophe than nontransformed cells535. Thus, we pose an

alternative explanation for the cancer-specificity of BETi: specifically, these drugs

induce mitotic catastrophe, at least in TNBC, and, because cancer cells are

particularly prone to this process, they initiate cell death or senescence pathways.

Notably, none of the genes we found to be functionally involved in the response of

TNBC to BETi are associated with SEs in MDA-MB-231 cells or a larger panel of

TNBC cell lines420. This suggests disruption of SEs may not be essential for the

inhibition of TNBC growth observed when BET protein activity is suppressed.

Indeed, we found that loss of a group of mitosis-regulating transcription factors

214

phenocopied the effects of BET suppression and none of those genes contain a

conventionally-defined SE.

The analysis described herein revealed that suppression of BET activity

downregulates LIN9 as well as four additional mitosis-regulating transcription

factors within 6 hours of treatment. Of the five BETi-regulated transcription factors

we investigated, only individually silencing LIN9 mimics BETi treatment by

repressing mitosis-associated genes and inducing multinucleation. LIN9 is

reported to contribute to a variety of processes, including embryonic development,

852,857,858 progression through G2-phase, mitosis, and cytokinesis , and is a

component of the Mammaprint breast cancer gene signature that predicts

metastasis506. Along with LIN37, LIN52, LIN54, and RBBP4, LIN9 is a subunit of

the MuvB complex830. During S phase, the MuvB complex interacts with B-MYB,

and together they drive the expression of late S-phase genes652,653. They then

652 recruit FOXM1 to the chromatin during G2 phase . B-MYB is phosphorylated and

subsequently degraded by the proteasome653, while FOXM1 is phosphorylated

and activated859. Phosphorylated FOXM1 and the MuvB complex remain bound to

the DNA and regulate expression of genes important for the G2–M transition and

successful completion of mitosis, including AURKA, AURKB, PLK1,

and KIF20A 652,827,857, all of which are also suppressed with BETi treatment. In

addition, six mitotic kinesins (KIF4A, KIF23, KIF2C, KIF14, KIFC1, and KIF20A)

and two microtubule-associated non-motor proteins (PRC1 and CEP55) that have

known functions in mitosis and cytokinesis are direct targets of MuvB, B-MYB, and

FOXM1 in breast cancer860. Expression of genes encoding these proteins as well

215

as eight additional mitotic kinesins848 were downregulated following BETi

treatment. Prior studies have shown that loss of LIN9 expression produces

polyploid cells with aberrant nuclei and induces senescence857,861. Thus, data

presented herein not only confirm the impact of LIN9 on mitosis, but also implicate

LIN9 as an intermediate between BET proteins and effective mitosis in TNBC.

Using the TCGA and METABRIC datasets, we found that LIN9 is amplified or

overexpressed in 24% to 29% of all breast tumors and in 66% of basal-like tumors.

High expression of LIN9 is also associated with poor survival of breast cancer

patients. Supporting a role for LIN9 in mediating the effects of BET proteins, we

found that genes highly correlated with LIN9 are more susceptible to BET

suppression than those that are not. Furthermore, genes harboring a MuvB binding

site, indicative of LIN9 association, were also more likely to be suppressed by JQ1.

Together, these data suggest that loss of LIN9 expression in response to BETi

initiates a cascade of events wherein many cell-cycle–regulated genes are

suppressed, mitosis is disrupted, and cells either become multinucleated, die, or

enter a prolonged interphase (G0). These results also argue that breast cancers

with amplified or overexpressed LIN9 may be more susceptible to BET inhibition

and this possibility could be evaluated in clinical trials currently underway

examining the efficacy of BETi in TNBC patients.

In summary, direct BET protein binding is necessary to sustain expression

of LIN9 as well as several additional mitosis-regulating transcription factors in

TNBC. Loss of BET protein activity through the use of BETi suppresses key cell-

cycle and mitosis genes causing mitotic catastrophe. Reduced expression

216

of LIN9 alone can mimic the effects of BET protein suppression, suggesting that it

is a primary mediator of BETi. Notably, LIN9, as well as genes encoding the other

mitosis-associated transcription factors evaluated in this study, lack SEs,

indicating that disruption of such elements is unnecessary for the mitotic

dysfunction observed with these inhibitors. Given the high rate of overexpression

of LIN9 in breast cancers, these data suggest that LIN9 may be a key vulnerability

in breast cancers that can be targeted with BETi. They further suggest that BETi

may be particularly effective when combined with additional agents that increase

cancer cell sensitivity to mitotic dysfunction.

217

A. 45 B. * Vehicle 250 nM 40 35 30 25 * 20 * 15 * *

* HCC1143 % ApoptoticCells 10 5 0 JQ1: - - 468 1143

Figure 3.1. BET inhibition induces apoptosis and senescence in TNBC cells.

(A) MDA-MB-468 and HCC1143 cells were treated with vehicle or JQ1 (250 nM,

500 nM, or 1000 nM) for 72 hours, stained with CellEvent Caspase 3/7 Reagent,

and counted. The percent of caspase 3/7-positive (apoptotic) cells were then calculated. Data are means ± SD (*=p<0.05 compared to vehicle). (B)

Representative images (4x) of HCC1143 cells treated with vehicle or 250 nM JQ1

for eight days and stained for SA-βgal activity.

218 A. B. 250 Vehicle JQ1 200 * * 150

100

50 Duration of Mitosis (min) 0 HCC38 231

C. HCC38 Min: 0 10 20 30 50 70 Vehicle

0 10 60 80 90 150

0 20 40 120 150 160 1000 nM1000 JQ1

D. MDA-MB-231 Min: 0 40 60 80 100 120 Vehicle

0 60 80 100 180 200

0 100 120 140 160 180 1000 nM1000 JQ1

Figure 3.2. Sustained BET activity is necessary for timely progression through mitosis.

219 HCC38 and MDA-MB-231 cells were treated with vehicle or 1000 nM JQ1 and

observed via live-cell imaging for 96 hours. (A) Average percentage confluency of

HCC38 (left) and MDA-MB-231 (right) cells over time. Data are means ±

SEM. (B) Quantitation of the length of time required by individual TNBC cells to complete mitosis 6 hours after the addition of vehicle or JQ1. Data are means ±

SEM (*, p < 0.05 compared with vehicle). (C and D) Representative live-cell

images (20x) of HCC38 (C) and MDA-MB-231 (D) cells treated with vehicle or 1000 nM JQ1. Numbers indicate minutes following the initiation of mitosis.

220

A. HCC38 D. MDA-MB-231 Vehicle JQ1 Vehicle JQ1 (n=70) (n=74) (n=59) (n=66) 7% 8% 3% 6% 6% 23% 24%

70% 87% 69% 97%

B.C. E.

Die in Exit and Exit and Mitosis Die Divide

Figure 3.3. BET inhibitors promote mitosis-associated death or prolonged interphase.

(A) Pie chart showing the percentage of vehicle- and JQ1-treated HCC38 cells that underwent different mitosis-associated cell fates: exit and divide (blue), exit and die (black), or die in mitosis (white). Vehicle versus JQ1 for all three outcomes, p << 0.001. (B) Quantitation of the length of time for HCC38 cells to divide again (blue dots) or die (black dots) following mitotic exit. p < 0.05 compared with vehicle. (C) Comparison of the duration of mitosis of individual JQ1-treated

HCC38 cells that die in mitosis, exit mitosis and die, or exit mitosis and divide. Die in mitosis versus exit and die, p < 0.05. (D) Pie chart showing the percentage of vehicle- and JQ1-treated MDA-MB-231 cells that underwent different post-mitotic cell fates: exit and divide (blue), exit and die (black), or prolonged interphase

221 (gray). Vehicle versus JQ1 for all three outcomes, p << 0.001. (E) Quantitation of

the length of time for MDA-MB-231 cells to divide again (blue dots) or die (black

dots) following mitotic exit. p < 0.05 compared with vehicle. For all graphs, each

dot represents an individual cell, and red lines are mean ± SEM.

222

A. B. MDA-MB-231 HCC70 Pathway p-value q-value Pathway p-value q-value Mitotic 1.7E-23 6.2E-21 Metabolism 2.9E-06 3.0E-3 Prometaphase Cell Cycle, Mitotic 3.4E-05 0.01 S Phase 2.7E-19 4.9E-17 Assembly of the Primary 5.1E-05 0.01 Signaling by Rho 1.9E-10 1.1E-8 Cilium GTPases Apoptotic Execution 0.00014 0.03 Mitotic G2-G2/M 3.8E-10 2.0E-8 Phase Phases Non-Integrin Membrane- 0.00022 0.03 Kinesins 3.1E-06 9.8E-5 ECM Interactions C. CELL-CYCLE_G1_S CELL-CYCLE_G2 CELL-CYCLE_G2_M CELL-CYCLE_M_G1

NES = -1.20 NES = -1.17 NES = -1.21 NES = -1.36

231 p = 0.10 p << 0.001 p << 0.001 p << 0.001 Enrichment Score Enrichment

NES = -1.19 NES = -1.32 NES = -1.37 NES = -1.38 p = 0.19 p = 0.08 p << 0.001 p << 0.001 HCC70 Enrichment Score Enrichment D. E. F. 1.5 1.5 1.5 Vehicle JQ1 Vehicle JQ1 Vehicle JQ1

1 1 1 GAPDH GAPDH / / GAPDH * * * * / * * * 0.5 * * 0.5 * 0.5 * PLK1 * KIF20A CCNB1

0 0 0 231 1143 468 BT549 231 1143 468 BT549 231 1143 468 BT549 G. H. 1.6 Vehicle JQ1 1.2 * 0.8 * * * * * * Expression 0.4 * RelativemRNA 0 Tumor: 231 468 231 468 231 468 231 468 CCNB1 KIF20A PLK1 KIF2C

Figure 3.4. BET activity is necessary for sustained expression of cell cycle- associated genes.

(A–C) MDA-MB-231 and HCC70 cells were treated for 72 hours with vehicle or

500 nM JQ1 and transcriptomes were analyzed using Affymetrix Human Gene 2.0

ST expression microarrays. (A) Venn diagram showing the number of genes

223 whose expression significantly changed in each cell line as well as the number of

genes commonly altered in both. (B) Top 5 nonoverlapping Reactome terms for

MDA-MB-231 (left) and HCC70 (right) cells. (C) GSEA of cell-cycle–classifying

genes whose expression was altered by JQ1 in MDA-MB-231 and HCC70

cells. (D–F) RT-qPCR analysis of three cell cycle/mitosis genes (CCNB1, KIF20A, and PLK1) in four TNBC cell lines treated with vehicle or 500 nmol/L JQ1 for 24 hours. (G) RT-qPCR analysis of cell cycle and mitosis genes (CCNB1, KIF20A,

PLK1, and KIF2C) in MDA-MB-231 (n = 5) and MDA-MB-468 (n = 10) tumors from orthotopically xenografted mice treated with vehicle or JQ1. (H) Volcano plots depicting mRNA log2 fold changes versus the corresponding log10 p values for

genes whose expression significantly changes in response to JQ1 in MDA-MB-

231 (right) and HCC70 (left) cells after 72 hours. Red dots, genes that are critical

for mitosis. For all bar graphs, data are presented as means ± SD (*, p < 0.05

compared with vehicle).

224 A. B. 1.4 MDA-MB-231 Vehicle JQ1 1.8 MDA-MB-231 Vehicle JQ1 1.2 1.6 (tumor) 1.4 1 1.2 0.8 * * 1 * 0.6 * * 0.8 0.6 Expression Expression 0.4 * * * Relative mRNA Relative RelativemRNA * 0.4 * 0.2 0.2 0 0 LIN37 FOXM1 E2F2 E2F8 LIN9 MYBL2 FOXM1 E2F2 E2F8 LIN9 MYBL2 C. D. 1.8 1.2 MDA-MB-468 MDA-MB-231 Vehicle JQ1 1.6 Vehicle JQ1 (tumor) 1 1.4 1.2 0.8 * 1 * 0.6 0.8 * * * 0.6 * * 0.4 * * Expression *

RelativemRNA 0.4 0.2 0.2 RelativeBRD4 Binding 0 0 FOXM1 E2F2 E2F8 LIN9 MYBL2 FOXM1 E2F2 E2F8 LIN9 MYBL2

Figure 3.5. BET proteins directly modulate the mitotic transcriptional program.

(A) RT-qPCR quantitation of the expression of selected mitosis-controlling transcription factors in MDA-MB-231 cells treated with vehicle or 500 nM JQ1 for

6 hours. (B and C) RT-qPCR analysis of BETi-repressed genes in tumors from vehicle- or JQ1-treated mice harboring MDA-MB-231 (n = 5; B) or MDA-MB-468

(n = 10; C) tumors. (D) Representative gene-specific ChIP-PCR analysis of MDA-

MB-231 cells assessing binding of BRD4 to selected genes encoding mitosis- controlling transcription factors following treatment with vehicle or 500 nM JQ1 for

24 hours. For all graphs, data are presented as means ± SD (*, p < 0.05 compared with vehicle or siNS).

225 A. LUMINAL GENES NON-BASAL GENES BASAL A GENES BASAL B GENES

NES = -1.05 NES = -1.05 NES = -1.31 NES = -1.36

231 p = 0.26 p = 0.26 p << 0.001 p << 0.001 EnrichmentScore

NES = -1.31 NES = -1.26 NES = -1.15 NES = -1.25 p << 0.001 p << 0.001 p << 0.001 p << 0.001 Gray Gray BC Subtype HCC70 Enrichment Score Enrichment

B. LUMINAL (B) GENES LUMINAL (M) GENES BASAL GENES MESENCHYMAL GENES

NES = -1.36 NES = -1.25 NES = -1.31 NES = -1.28

231 p << 0.001 p << 0.001 p << 0.001 p << 0.001 Enrichment Score Enrichment

NES = -1.38 NES = -1.38 NES = -1.29 NES = -1.28 p << 0.001 p << 0.001 p << 0.001 p << 0.001 HCC70 Bertucci BC Subtype Enrichment Score Enrichment

Figure 3.6. BET inhibitors fail to induce a luminal differentiation signature.

GSEA of breast cancer subtype classifying gene sets in MDA-MB-231 and HCC70 cells using Gray474 (A) and Bertucci475 (B) classifiers.

226 1.8 Vehicle I-BET762 1.6 1.4 1.2 1 * 0.8 * * 0.6 * 0.4 * RelativeExpressionmRNA 0.2 0 LIN37 FOXM1 E2F2 E2F8 LIN9 MYBL2

3.7. I-BET762 suppresses expression of master regulators of mitosis.

RT-qPCR quantitation of the expression of selected mitosis-controlling transcription factors in MDA-MB-231 cells treated with vehicle or 1000 nM I-

BET762 for 6 hours.

227 A. B. 1.2 2 siNS siF/E2/E8/L/M siNS siF/E2/E8/L/M 1.8 * 1 1.6 0.8 1.4 * 1.2 0.6 1 * * * 0.8 * 0.4 * * 0.6 * 0.2 0.4 0.2 RelativemRNA Expression RelativemRNAExpression 0 0 FOXM1 E2F2 E2F8 LIN9 MYBL2 CCNB1 KIF20A PLK1 CDKN1A C. siNS siF/E2/E8/L/M JQ1 60 * 50 * 40

10 % Multinucleated Cells Multinucleated % 0 siNS siF/E2/E8/L/M JQ1

Figure 3.8. Simultaneous knockdown of FOXM1, E2F2, E2F8, LIN9, and

MYBL2 phenocopies BET inhibition.

FOXM1, E2F2, E2F8, LIN9, and MYBL2 were simultaneously knocked down for five days in MDA-MB-231 cells using gene-specific siRNA. (A) Confirmation of

siRNA-mediated knockdown. (B) RT-qPCR analysis of BETi-regulated genes. (C)

Left: Representative images (20x) of MDA-MB-231 cells stained with DAPI (blue,

nuclei) and Texas Red-X phalloidin (red, actin cytoskeleton). MDA-MB-231 cells

treated with 500 nM JQ1 are included as a comparison. Arrows indicate

multinucleated cells. Right: Quantitation of multinucleated cells. siF/E2/E8/L/M =

simultaneous knockdown of FOXM1, E2F2, E2F8, LIN9, and MYBL2. For all

graphs, data are means ± SD (*=p<0.05 compared to siNS).

228 A. D. E. 5 kb 135

E2F2

135

B. E2F8

135

FOXM1 RHNO1

135

C. LIN9

135

SEs

MYBL2

206

-3.8 0 3.4

Log 2 Fold Change

MYC

Figure 3.9. BETi suppress mitosis transcription factors in the absence of

SEs.

Binding of H3K27Ac in MDA-MB-231 cells was analyzed using ChIP-seq. These data were compared with data generated from JQ1-treated MDA-MB-231 cells analyzed by gene expression microarray. (A) Number of genes in MDA-MB-231 cells that are associated with super-enhancers (SE genes) or lack super- enhancers (Non-SE genes). Total genes, all genes expressed in MDA-MB-231 cells based on microarray data. Mitosis genes, 599 genes identified by SuperPath

229 as critical for mitosis. Percentages indicate the percent of genes that contain

SEs. (B) Violin plots depicting fold change in gene expression of SE-associated genes and non–SE-associated genes for all genes expressed in MDA-MB-231 cells (p = 4.2 × 10−15). (C) Violin plots depicting fold change in gene expression of

SE-associated genes and non-SE-associated genes for mitosis genes. (D) Heatmap showing log2 fold change in expression of all JQ1-regulated genes in MDA-MB-231 cells ranked according to enhancer strength. The black line indicates the cutoff value between SEs and typical enhancers. Four of the five

BETi-regulated mitosis-regulating transcription factors are listed. MYBL2 is absent because it lacks a specific enrichment of H3K27Ac binding within 10 kbp of its promoter. (E) ChIP-seq–binding profiles for H3K27Ac at the promoter regions for five mitosis-regulating transcription factors. MYC is shown as a representative SE- associated gene.

230 A. 2 MDA-MB-231 siNS siF/L/M * 1.8 1.6 1.4 1.2 1 0.8 * Expression 0.6 * RelativemRNA 0.4 * * * 0.2 * 0 FOXM1 LIN9 MYBL2 CCNB1 KIF20A PLK1 CDKN1A

B. siNS siF/L/M JQ1 70 60 * 50 * 40 30 20 10 % Multinucleated Cells Multinucleated % 0 siNS siF/L/M JQ1 C. D. 2.5 1.4 siNS siBRD2 siBRD3 siBRD4 1.2 siNS siBRD2/3 siBRD3/4 siBRD2/4 2 1 1.5 * 0.8 * 1 0.6 * * * * * * * 0.4 0.5 * ** 0.2 ** * * RelativemRNAExpression 0 RelativemRNAExpression 0 BRD2 BRD3 BRD4 FOXM1 LIN9 MYBL2 BRD2 BRD3 BRD4 FOXM1 LIN9 MYBL2

Figure 3.10. Simultaneous knockdown of FOXM1, LIN9, and MYBL2 phenocopies BET inhibition.

(A and B) FOXM1, LIN9, and MYBL2 were simultaneously knocked down for five days in MDA-MB-231 cells using gene-specific siRNA. (A) RT-qPCR analysis of siRNA-targeted genes and BETi-target genes. (B) Left: Representative images

(20x) of MDA-MB-231 cells stained with DAPI (blue, nuclei) and Texas Red-X phalloidin (red, actin cytoskeleton). MDA-MB-231 cells treated with 500 nM JQ1 are included as a comparison. Arrows indicate multinucleated cells. Right:

Quantitation of multinucleated cells. siF/L/M = simultaneous knockdown of

231 FOXM1, LIN9, and MYBL2. (C and D) Individual or pairwise knockdown of BRD2,

BRD3, and BRD4 was performed in MDA-MB-231 cells using gene-specific siRNA

for 72 hours. Graphs show RT-qPCR analysis of siRNA-targeted and BETi-

regulated genes following individual (C) or pairwise (D) knockdown of BET- encoding genes. For all graphs, data are presented as means ± SD (*=p<0.05 compared to siNS).

232 1.2 siNS siLIN9 A. B. 2 D. HR = 1.39 (1.19 – 1.63) 1 siNS siLIN9 * logrank P = 3.7e-05 LIN9 0.8 1.5 β-Actin 0.6 1 *

0.4 Expression *

* Probability LIN9/GAPDH Relative mRNA Relative 0.5 * Expression 0.2 Low High

0 0 0.0 0.2 0.4 0.6 0.8 1.0 siNS siLIN9 CCNB1 KIF20A PLK1 CDKN1A 0 50 100 150 200 250 C. Time (months) siNS siLIN9 JQ1 60 * 50 * 40

10 % Multinucleated Cells 0 siNS siLIN9 JQ1 E. F. G.

4 LIN9 Correlation 30 3.5 r < 0.5 r ≥ 0.5 25 3 20 2.5 2 15 1.5 10 1 5 % Genes Δ with JQ1

Fold Change with JQ1 0.5 0 0 Bound by Not Bound TCGA METABRIC LIN9 by LIN9

Figure 3.11. LIN9 mediates the effects of BET inhibition.

(A) Confirmation of siRNA-mediated knockdown of LIN9 in MDA-MB-231 cells after 5 days. Inset is a representative Western blot showing suppression of LIN9 substantially reduces LIN9 protein. (B) RT-qPCR analysis of BETi-target genes in MDA-MB-231 cells following LIN9 silencing. (C) Representative images

(20x) of MDA-MB-231 cells following LIN9 silencing that were stained with DAPI

(blue, nuclei) and Texas Red-X phalloidin (red, actin cytoskeleton). MDA-MB-231 cells treated with 500 nmol/L JQ1 are included as a comparison. Arrows, multinucleated cells. (D) Kaplan–Meier curve of relapse-free survival for breast cancer patients with high and low expression of LIN9. (E) Violin plots depicting

233 BETi-induced expression (log2 fold change) of genes that are not correlated (r <

0.5) or are correlated (r ≥ 0.5) with LIN9 expression. The right side of the panel depicts subdivision of genes that were correlated with LIN9 (r ≥ 0.5) according to residence on chromosome 1q versus their responsiveness to JQ1 (p value for r <

0.5 vs. r ≥ 0.5 = 1.8 × 10−47; p value for r ≥ 0.5 on 1q vs. not on 1q = 1.2 × 10−5). r =

Pearson coefficient. (F) Absolute fold change in expression following JQ1

treatment of genes that are highly correlated with LIN9 expression (r ≥ 0.5) and those that are moderately or not correlated (r < 0.5) in the TCGA and METABRIC datasets. (G) The percentage of genes changed with JQ1 that are bound by LIN9 or not bound by LIN9 (p << 0.001). For A–C, data are presented as means ± SD

(*, p < 0.05 compared with vehicle or siNS).

234

A. 1.2 1.2 1.2 1.2 1 1 1 1 0.8 0.8 0.8 0.8 0.6 0.6 * 0.6 * 0.6 0.4 * 0.4 0.4 0.4 0.2 0.2 0.2 0.2 * E2F2/GAPDH E2F8/GAPDH MYBL2/GAPDH FOXM1/GAPDH 0 0 0 0 siNS siFOXM1 siNS siE2F2 siNS siE2F8 siNS siMYBL2

B. 1.6 1.6 siNS siFOXM1 1.4 * 1.4 * * siNS siE2F2 1.2 1.2 1 * 1 * 0.8 * 0.8 0.6 0.6 Expression Expression 0.4 0.4 RelativemRNA RelativemRNA 0.2 0.2 0 0 CCNB1 KIF20A PLK1 CDKN1A CCNB1 KIF20A PLK1 CDKN1A

1.8 1.8 * siNS siE2F8 siNS siMYBL2 * 1.6 * 1.6 1.4 1.4 1.2 1.2 1 1 * * 0.8 0.8 0.6 0.6 Expression Expression

RelativemRNA 0.4 RelativemRNA 0.4 0.2 0.2 0 0 CCNB1 KIF20A PLK1 CDKN1A CCNB1 KIF20A PLK1 CDKN1A

Figure 3.12. Individual knockdown of FOXM1, E2F2, E2F8, and MYBL2 does not phenocopy BET inhibition.

FOXM1, E2F2, E2F8, and MYBL2 were individually knocked down for five days in

MDA-MB-231 cells using siRNA. (A) Confirmation of knockdowns using RT-qPCR.

(B) RT-qPCR analysis of BETi-regulated genes. For all graphs, data are means ±

SD (*=p<0.05 compared to siNS).

235

Probability Expression Low HR = 1.52 (1.21 – 2.09) High logrank P = 6.8e-04 0 0.2 0.4 0.6 0.8 1 0 500 1000 1500 2000 Time (days)

Figure 3.13. High expression of LIN9 is correlated with lower overall survival rates.

Kaplan-Meier curve of overall survival for breast cancer patients with high and low expression of LIN9.

236

CHAPTER 4: DISCUSSION AND FUTURE DIRECTIONS

237

4.1 Summary

Breast cancer is a diverse collection of tumors that can be subdivided using

multiple subtyping systems, although the only method currently used in the clinic

to inform treatment strategy is based on ER, PR, and HER2 expression. The most

aggressive subtype of breast cancer is referred to as triple-negative breast cancer

(TNBC) because tumors within this subtype lack expression of ER and PR and do

not have amplified HER2. As a result, therapies that have been designed against

these three receptors are ineffective against this group of tumors. Instead, TNBC

patients are treated with cytotoxic chemotherapy, and, while their tumors initially

respond, they often recur. It is therefore of vital importance to create new

therapeutic strategies that can target this diverse group of tumors in order to

improve patient outcome.

BET inhibitors have been investigated as a potential therapy in TNBC by several

groups, revealing that these agents suppress growth and colony formation and

stimulate apoptosis and senescence in vitro and in vivo420,421,424,425,456,473.

However, these studies did not explore the mechanism(s) by which BETi elicit their

effects or the efficacy of BETi in the different subtypes of TNBC. Our first study

described in Chapter 2 found that in seven diverse TNBC cell lines representing

both the claudin-low and basal-like subtypes as well as five of the six

TNBCtypes122, BETi inhibited growth and induced two distinct terminal responses,

apoptosis and senescence, and these cell fates were reflected in tumor responses

in vivo472. BETi also led to the appearance of multinucleated and polyploid cells in

several cell lines, regardless of their terminal response to BETi. This suggests

238

BETi disrupt mitosis and/or cytokinesis across different TNBC subtypes. We

discovered that Aurora kinases A and B, which control several critical steps during

mitosis and cytokinesis, are directly bound by the BET protein BRD4, and this

interaction is disrupted by BETi, leading to the downregulation of both kinases.

Selective inhibitors of AURKA and AURKB phenocopied BETi. These data indicate

that Aurora kinases are downstream targets of BETi and that the mechanism of

action of BETi in TNBC involves the suppression of these mitotic kinases. They

also suggest that Aurora kinases may serve as biomarkers of response to BETi.

The observation of multinucleated cells following BETi treatment led us to

hypothesize that BETi induce mitotic catastrophe, an oncosuppressive mechanism

characterized by multinucleation followed by senescence or death in mitosis or

soon after mitotic exit831,832. Indeed, as described in Chapter 3, BETi disrupted

normal progression through mitosis, which resulted in an increase in the duration

of mitosis, the induction of apoptosis during or after mitosis, and the arrest of cells

in interphase411. Mitotic dysfunction was preceded by the suppression of a large

number of genes that regulate mitosis and cytokinesis. These data indicated that

BETi-treated TNBC cells undergo mitotic catastrophe prior to apoptosis and

senescence. Mechanistically, BETi rapidly and selectively suppressed expression

of five critical mitosis regulators (FOXM1, E2F2, E2F8, LIN9, and MYBL2) by

disrupting the localization of BRD4 to their promoter regions. siRNA-mediated

silencing of these five genes together recapitulated the effects of BETi, highlighting

their importance in the BETi response in TNBC. None of these genes had a

putative SE. While others have attributed the selectivity of BETi for cancer cells

239

compared to normal cells to the disassembly of SEs at oncogenes366,410,446, our

data indicate that an alternative mechanism of selectivity exists in TNBC. Cancer

cells have been shown to be more sensitive to mitotic catastrophe than

nontransformed cells535. We therefore suggest that TNBC cells are preferentially

targeted by BETi due to the induction of mitotic catastrophe following the direct

suppression of master mitosis regulators.

Our studies also identified LIN9 as a critical mediator of the activity of BETi in

TNBC (Figure 4.1). Individual silencing of the five master mitosis transcription

factors revealed that only loss of LIN9 expression induced similar responses to

BETi, indicating LIN9 is crucial for mitotic progression. In addition, genes that were

highly correlated with LIN9 and/or possessed a LIN9 binding site were more likely

to be suppressed by BETi. We also discovered LIN9 is amplified or overexpressed

in 66% of basal-like tumors in the TCGA dataset compared to 24% of all breast

tumors, and high expression of LIN9 is linked to lower overall and relapse-free

survival. In contrast, the other MuvB complex subunits are amplified and/or

overexpressed in <5% of all breast cancers and <20% of basal-like breast cancers,

indicating the observed oncogenic effects are due specifically to LIN9 and not the

MuvB complex. Together, these data reveal that LIN9 is an oncogenic driver in

TNBC and that suppression of LIN9 induces mitotic catastrophe in this disease.

This suggests that tumors with high LIN9 expression may be particularly sensitive

to BET protein inhibition, a hypothesis that can be rapidly examined in ongoing

clinical trials assessing the response of TNBC to BETi.

240

In summary, the data presented here define the mechanism of action of BETi in

TNBC as mitotic catastrophe, indicating combining BETi with agents that further

sensitize cancer cells to defects in mitosis may provide additional benefit in this

diverse disease. They also identify LIN9 as a novel potential oncogene that is

targetable by BETi. However, several questions still remain. These additional

areas of research as well as potential experiments and their expected outcomes

are described in the following sections.

4.2 The BET inhibitor-induced mitotic defect

BETi suppress RNA and protein expression of critical regulators of mitosis,

including AURKA, AURKB, PLK1, and cyclin B1411,472. In addition, live cell imaging

experiments revealed that BETi increase the duration of mitosis, suggesting these

cells encounter a disruption during mitosis411. However, it is unknown when during

mitosis this defect occurs. The expression of many mitotic genes is altered by

BETi. Thus, it is distinctly possible that BETi disrupt several processes during

mitosis instead of a single step. If this is the case, it is likely that the mitotic event

that is altered by BETi in individual cells will depend on where the cells are in the

cell cycle when BETi is added, as has already been shown to occur with inhibition

of AURKA691.

We performed a preliminary experiment to determine if BETi impact early mitotic

processes. Staining for mitotic spindles (α-tubulin) and centrosomes (γ-tubulin) in

vehicle- and JQ1-treated HCC1143 cells revealed there was no consistent

increase in monoploar or multipolar spindles with JQ1 treatment (unpublished

data). However, the experiment was only performed once using a single cell line.

241

To draw any conclusions about the impact of BETi on events during early mitosis,

this experiment will need to be repeated, and a second TNBC cell line should be

used to ensure the observations are not cell line-dependent.

Processes that occur later in mitosis may also be deregulated by BETi. For

example, AURKB is essential for the spindle assembly checkpoint (SAC) as it

removes improperly attached mitotic spindles from the chromosome kinetochores,

providing the cell with an opportunity to form bioriented chromosomes673. PLK1

also plays a role in the SAC by localizing AURKB to kinetochores. We have shown

BETi suppress both AURKB and PLK1 transcription411,472. Thus, it is possible that

BETi override the SAC through the loss of AURKB and PLK1 expression. One of

the ways to directly assess if the SAC is active is to observe the alignment of

chromosomes at the metaphase plate and chromosome segregation during

anaphase700. To accomplish this, we would express GFP-tagged histone H2B862

in MDA-MB-231 and MDA-MB-468 cells for higher resolution of chromosomal

dynamics and use live cell imaging to visualize the movement of chromosomes in

these cell lines beginning six hours after the addition of JQ1. This time point is

consistent with the time point we used in our live cell imaging studies described in

Chapter 3 and, as we have previously shown, is enough time for JQ1 to induce a

mitotic defect. If BETi override the SAC, not all of the chromosomes will align along

the metaphase plate or move into two separate daughter nuclei.

If BETi disrupt the SAC, abscission (the final step of cytokinesis) could be delayed.

This would occur if chromatin became trapped within the chromatin bridge

connecting two daughter cells. Abscission may also be prevented due to the

242

suppression of genes that are critical for this process, including CEP55, ANLN,

and CNTRL863-865, by BETi. We observed downregulation of these genes in JQ1-

treated MDA-MB-231 and HCC70 cells411. Thus, together with our observation that

BETi-treated cells become multinucleated, this suggests BETi could delay or inhibit

abscission. This possibility can be directly assessed by observing the length of

time it takes for cells to break chromatin bridges. To accomplish this, we would use

GFP-tagged LAP2β, a marker of the inner nuclear envelope that localizes to

chromatin in anaphase and permits visualization of chromatin bridges697. GFP-

LAP2β would be expressed in MDA-MB-231 and MDA-MD-468 cells, and the time

it takes for these bridges to disappear would be measured using live cell imaging.

If BETi prevent the destruction of chromatin bridges, the length of time the bridges

are present would be higher in BETi-treated cells compared to vehicle-treated

cells.

Another line of evidence that would support the hypothesis that BETi delay or halt

abscission is the regression of the cleavage furrow, which occurs following AURKB

inhibition700. Cleavage furrow ingression could be prevented and reversed if the

cleavage furrow encounters remaining chromatin bridges. The movement of the

cleavage furrow during live cell imaging is particularly noticeable in MDA-MB-468

cells and does not require labeling to be visualized. While performing live cell

imaging using vehicle- or JQ1-treated MDA-MB-468 cells, I noticed occasionally

that, although two daughter cells began to form after mitosis (cleavage furrow was

visible), the cleavage furrow eventually regressed, ultimately forming a single

daughter cell. This occurred in both treatment groups (Figure 4.2). However, the

243

number of these events was too low to determine if JQ1 significantly increased the

incidence of aborted cleavage furrow movement. In the future, if live cell imaging

experiments are performed using GFP-H2B- or GFP-LAP2β-expressing MDA-MB-

468 cells, we would track the mitotic outcome of many more cells to permit

thorough statistical analysis. If cleavage furrow ingression is aborted more often in

JQ1-treated cells, it could suggest chromatin bridges are maintained following JQ1

treatment due to the inhibition of abscission.

4.3 The role of Bcl-xL in the response of TNBC cells to BETi-induced mitotic

catastrophe

The choice of TNBC cells to undergo BETi-mediated apoptosis or senescence

following mitotic catastrophe is not associated with TNBC subtype, alterations in

the expression of MYC, the extent of suppression of AURKA/B, or the GI50 of

BETi472. Instead, in collaboration with Dr. Sylvia Gayle, data that is soon to be

submitted for publication reveal that basal levels of expression of the anti-apoptotic

protein, Bcl-xL, dictate response to BETi in TNBC. BETi-treated TNBC cells that

express high levels of Bcl-xL senesce while cells that express lower levels undergo

apoptosis. In addition, manipulation of Bcl-xL can alter the BETi-induced terminal

response of TNBC cells, although they still undergo mitotic catastrophe. For

example, MDA-MB-231 cells, which have high expression of Bcl-xL, normally

senesce following BETi treatment. However, BETi induce apoptosis in these cells

following pharmacological inhibition or siRNA-mediated knockdown of Bcl-xL. This

indicates that TNBC tumors with lower expression of Bcl-xL may have a more

244

robust response to BETi, and combining BETi and inhibitors of Bcl-xL could

increase the sensitivity of tumors expressing high levels of Bcl-xL to BETi.

4.4 The mechanism of action of BETi in other subtypes of breast cancer

The BET protein-regulated transcriptome and thus the gene targets of BETi vary

among different cancer types and subtypes. This indicates that the unique cancer

transcriptome will dictate how a tumor responds to BETi. It also suggests BETi

may have different mechanisms of action that are tumor type-dependent. We were

the first to identify mitotic catastrophe as the mechanism of action of BETi in any

cancer411. It remains to be determined if other cancers initiate a similar response

to BETi.

Basal-like breast tumors are more molecularly similar to squamous cell lung cancer

and a subtype of bladder cancer compared to other breast cancer subytpes113,114,

suggesting BETi may have a different mechanism of action in ER+/luminal and

HER2+ tumors than TNBCs. To begin to assess how BETi elicit their anti-tumor

activity in luminal breast cancers, we treated MCF7 and T47D cells with vehicle or

JQ1 and observed them for eight days, assessing any changes in morphology.

There was no overt evidence of multinucleation following JQ1 treatment. However,

staining with DAPI (nuclei) and phalloidin (actin cytoskeleton) must be performed

before any conclusion is reached regarding the impact of BETi on mitosis and

cytokinesis in luminal breast cancer cells. On the other hand, JQ1 did seem to

impact cellular growth patterns. While T47D cells and, to a lesser extent, MCF7

cells normally grew in colonies, JQ1-treated cells did not (Figure 4.3), indicating

JQ1 likely reduces cell-cell junctions (unpublished data). In addition, MCF7 cells

245

seemed to adopt a more mesenchymal phenotype. These data suggest BETi may

induce EMT in luminal breast cancer cells. Based on these observations, we

performed RT-qPCR analysis of EMT and luminal differentiation markers using

RNA harvested from MCF7 and T47D cells that were treated with vehicle or JQ1

for four days. We also assessed the impact of JQ1 on stem cell markers, as EMT

has been linked to the generation of cancer stem cells866. While some

mesenchymal markers, such as CDH2, SNAI1, and SNAI2, as well as the stem

cell marker SOX9 were upregulated by JQ1 in both cell lines, JQ1 also upregulated

the differentiation markers FOXA1 and GATA3. Analysis of individual markers

therefore did not reveal a consistent upregulation of mesenchymal and stem cell

genes or loss of differentiation markers. However, BETi may initiate a partial

transition to a mesenchymal state, and a second hit could facilitate the transition

to a full mesenchymal phenotype.

Additional analyses should be performed to determine if the BETi-mediated

alterations in EMT genes are sufficient to increase motility and/or invasion in ER+

breast cancer cells. To accomplish this, we will treat MCF7 and T47D cells with

vehicle or JQ1. After four days, we will replate the cells in modified Boyden

chambers either with or without a Matrigel coating to assess the impact of JQ1 on

invasion or migration, respectively. Serum would be used as a chemoattractant. If

BETi do indeed push luminal breast cancer cells into a less differentiated state and

allow them to become more motile and invasive, this would suggest BETi have the

potential to make luminal tumors more aggressive and increase their ability to

246

metastasize. Thus, these results would caution against using BETi to treat luminal

breast cancers.

The MCF7 and T47D cell lines represent luminal A breast cancer. While our

preliminary data indicate BETi may induce a partial EMT phenotype in luminal A

breast cancer cells, we did not assess whether luminal B breast cancer cells

respond in a similar manner. Others have shown that BETi or siRNA-mediated

silencing of BRD2 or BRD4 inhibited growth of luminal B breast cancer

cells420,457,468, but they have not commented on the impact of BETi on EMT or the

migratory/invasive capacity in these cells. It is, however, notable that cells that

acquire EMT phenotypes are often less proliferative867,868. Hence, the decreased

growth of the cells in these studies could reflect an underlying mesenchymal

differentiation process. To assess whether luminal B cells may also undergo EMT-

like changes in response to BETi, we would repeat the RT-qPCR, migration, and

invasion experiments described above using the luminal B cell lines ZR-75-1 and

BT474. It is also possible that luminal B breast cancer cells respond differently to

BETi compared to luminal A breast cancer cells, because luminal A and B breast

cancers have different gene expression patterns and behaviors77.

To gain a more complete understanding of the impact of BETi on the luminal breast

cancer transcriptome, we would perform gene expression microarray analysis of

luminal A (MCF7 and T47D) and luminal B (ZR-75-1 and BT474) cells treated with

vehicle or JQ1 for 24 hours. This should reveal early responders to BETi. The

Reactome database840 would be used to identify the biological pathways and

genes within those pathways that are altered by BETi treatment. This information

247

will inform further functional analyses. For example, if BETi suppress expression

of genes within the angiogenesis pathway, we would confirm downregulation of

key angiogenic genes following BETi treatment in several luminal breast cancer

cell lines as well as mouse xenografts. The use of xenografts would also provide

tissues for directly assessing the impact of BETi on microvessel density using

immunostaining for the blood vessel marker CD31.

Elucidating how BETi elicit their effects in the various subtypes of breast cancer

will provide important information about these drugs that will translate into

understanding and improving their clinical utility in these diseases. If we know how

each tumor subgroup will respond to BETi, we can better predict which patients

are likely to receive the most clinical benefit from these drugs. We can also design

appropriate drug combination therapies to improve their anti-tumor impact.

4.5 The role of LIN9 in BET inhibitor-induced apoptosis and senescence

We have found that silencing LIN9 in the TNBC cell line MDA-MB-231 mimics BETi

treatment in that the loss of LIN9 induces multinucleation, suppresses several

mitosis genes, and increases the expression of CDKN1A (p21, a marker of

senescence)411. This indicates a role for LIN9 in BETi-induced apoptosis and

senescence. However, while we showed suppression of LIN9 mimics some of the

outcomes of BETi treatment, it remains to be determined if loss of LIN9 expression

can lead to apoptosis and senescence in TNBC cells. To test if alteration of LIN9

expression can induce apoptosis in TNBC, we would use gene-specific siRNA to

silence LIN9 in two cell lines that undergo apoptosis in response to BETi: MDA-

MB-468 and HCC70. After five days, cells will be harvested and counted using

248

trypan blue exclusion assay on a Countess II FL to determine if loss of LIN9 results

in fewer cells overall (indicating growth suppression) and fewer viable cells

(indicating cell death). In addition, protein would be isolated and western blots

performed for cleaved PARP and cleaved caspase 3 to assess if LIN9 silencing

activates the intrinsic apoptotic pathway. If LIN9 mediates the BETi response in

TNBC, we would expect LIN9 silencing to induce apoptosis in these two cell lines,

as evidenced by fewer viable cells and increased cleaved PARP and caspase 3.

We would also determine if suppression of LIN9 expression alone can induce

senescence. We have already seen in one cell line (MDA-MB-231) that LIN9

silencing upregulated CDKN1A, a marker of senescence411. This experiment will

be repeated in a second cell line (HCC1143) to confirm the increase in CDKN1A

expression was not cell line-dependent. However, direct assessment of

senescence will be more difficult because of the length of time it takes for many

cells to commit to irreversible senescence. It is possible that a single transfection

with siRNA targeting LIN9 will not suppress LIN9 expression long enough to induce

irreversible senescence. In that case, we would perform a second transfection six

days after the initial transfection. A non-targeting control siRNA would be used to

confirm this additional transfection does not cause any unanticipated effects, such

as decreased viability. Within eight days after the initial transfection of LIN9-

targeting siRNA, cells that have been transfected should have a visible increase in

their cytoplasmic to nuclear ratio, and, if these cells have initiated senescence,

cells will stain positive (stain blue) for SA-β gal.

249

Clarifying the role of LIN9 in BETi-induced apoptosis and senescence is important

for understanding the function of LIN9 in TNBC. If loss of LIN9 is directly

responsible for the initiation of apoptotic and senescence pathways, this would

support the theory that LIN9 acts as an oncogene in this disease. Future

experiments that directly assess if LIN9 is an oncogene in TNBC are discussed

later in this chapter.

4.6 The impact of LIN9 in BET inhibitor-induced mitotic catastrophe

We showed LIN9 mediates BETi-induced mitotic catastrophe in MDA-MB-231

cells. However, we did not assess the involvement of LIN9 in the BETi response

in additional cell lines. To ensure the observed effects of BETi are shared with

other TNBC cell lines, we can repeat siRNA-mediated knockdown of LIN9 in an

additional cell line (MDA-MB-468) and perform RT-qPCR analysis to determine if

loss of LIN9 suppresses expression of mitosis genes. We can also stain these cells

with DAPI (nuclei) and phalloidin (actin cytoskeleton) to assess whether

suppression of LIN9 expression induces multinculeation.

While our data suggest that LIN9 suppression causes mitotic catastrophe, we did

not collect direct evidence of this response. Live cell imaging following LIN9

silencing would allow us to track individual cells through mitosis and identify their

outcome. This would be performed using two cell lines: MDA-MB-231 cells, which

senesce in response to BETi, and MDA-MB-468 cells, which undergo apoptosis

following BETi treatment. After transfecting cells with siRNA targeting LIN9 or a

nontargeting control, individual MDA-MB-231 and MDA-MB-468 cells would be

tracked through mitosis using live cell imaging. The length of time it takes for cells

250

to traverse mitosis as well as the mitotic outcome of both parental and daughter

cells would be recorded. If cells with LIN9 knockdown die in or immediately after

mitosis or enter a prolonged interphase, this will indicate loss of LIN9 induces

mitotic catastrophe.

It is possible that, in addition to TNBC, other cancers that are dependent on LIN9

expression may initiate mitotic catastrophe in response to BETi and may be

particularly responsive to BETi treatment. To classify cancers that fall within this

group, we interrogated publically available datasets using cBioPortal and found

that LIN9 is most often amplified/overexpressed in TNBC compared to other types

of cancer in the TCGA dataset (Figure 4.4). This indicates that TNBC tumors, as

a class, may be more vulnerable to BETi compared to other types of cancer.

However, some cancers also have a significant number of tumors with high LIN9

expression, suggesting BETi may induce mitotic defects and could be an effective

treatment strategy in these tumors. For example, while only 12% of lung squamous

cell carcinomas have amplified/overexpressed LIN9, 30% of tumors within the

primitive expression subtype overexpress LIN9. To examine if BETi-treated tumors

with high LIN9 expression induce undergo mitotic catastrophe, we would identify

cell lines that represent the primitive subtype of lung squamous cell carcinoma that

overexpress LIN9 and those that do not. We would then treat these two groups of

cell lines with vehicle or BETi and 1) perform RT-qPCR analysis to determine if

BETi suppress expression of mitosis genes and 2) stain cells with DAPI and

phalloidin to reveal if BETi induce multinucleation in these cells. If BETi

downregulate mitosis genes and lead to the formation of multinucleated cells

251

specifically in the cell lines with high LIN9 expression, this would support our

hypothesis that LIN9 can be used to predict tumors that will respond to BETi.

By interrogating the TCGA dataset, I found LIN9 is amplified and/or overexpressed

in a significant number of non-basal-like breast cancers (Figure 4.4). If high LIN9

expression in TNBC confers sensitivity to BETi, it is possible that the same is true

for LIN9-overexpressing non-TNBC tumors. To assess if LIN9 is predictive of the

response to BETi, we would use qRT-PCR and western blotting analysis to

determine LIN9 mRNA and protein expression across cell lines representing

luminal A and luminal B breast cancers. We would then treat these cell lines with

BETi and compare the robustness of their response to BETi with the level of LIN9

expression. For example, if we found BETi induce EMT in ER+ breast cancer cells,

we would compare the ability of BETi to alter expression of canonical EMT genes

and induce EMT-associated morphology changes in high LIN9- and low LIN9-

expressing cells. We could then modulate LIN9 expression in these cell lines by

silencing LIN9 in high LIN9-expressing cells and overexpressing LIN9 in cells with

low expression of LIN9 to determine if we could alter the sensitivity of these cells

to BETi. If high LIN9 expression confers enhanced sensitivity to BETi, this would

suggest LIN9 could be used as a biomarker to predict responsiveness to these

agents across breast cancer subtypes.

4.7 The LIN9-controlled transcriptome

While genome-wide LIN9 binding sites have been identified in HeLa cells653, it is

unknown what genes are directly regulated by LIN9 in TNBC. It is expected that

genes involved in the G2/M transition and mitosis will be direct targets of LIN9 in

252

TNBC cells, as this has been shown in zebrafish, embryonic stem cells, and

several mouse and human cell lines653,852,858,869,870. In addition, in HeLa cells, LIN9

is enriched at the promoters of the mitosis genes CCNB1 and PLK1. This will likely

be the same in TNBC cells, especially considering LIN9 suppression in MDA-MB-

231 cells downregulated both of these genes411. However, uncovering the entire

LIN9-controlled transcriptome in this disease will provide insight into the specific

role LIN9 plays in TNBC development and progression.

To identify the direct targets of LIN9 in TNBC, we would begin by performing gene

expression microarray analyses. We would silence LIN9 in MDA-MB-231 and

MDA-MB-468 cells for 24 and 48 hours, harvest RNA, and perform RT-qPCR

analysis to determine at which time point LIN9 and its target genes, such as

CCNB1 and PLK1, are suppressed. Once we have selected the optimal time point,

we would conduct the microarray analyses. We would also perform ChIP-seq

studies to identify LIN9 binding sites in genomes of TNBC cells. We would use

chromatin from MDA-MB-231 and MDA-MB-468 cells immunoprecipitated with an

anti-LIN9 specific antibody or control mouse IgG. Combining the list of genes

altered with LIN9 silencing with the list of genes harboring a LIN9 binding site will

identify direct targets of LIN9. Using this list of LIN9 target genes, Reactome

pathway analysis would be conducted to reveal the cellular pathways that are

regulated by LIN9 in both cell lines, which could provide insight into how LIN9 acts

as an oncogene in TNBC.

4.8 Alternative LIN9-containing complexes

Very few studies have examined the function of LIN9 in normal or cancer cells. A

253

few analyses have shown that LIN9 prevents genomic instability and acts as a

tumor suppressor in human and mouse fibroblasts and colon carcinoma cell

lines861,871,872. In addition, Reichert, et al. found that Lin9 heterozygous mice were

more likely to develop lung tumors in response to expression of oncogenic c-Raf,

supporting the role of LIN9 as a tumor suppressor857. On the other hand, our

studies revealed LIN9 appears to have oncogenic activity in TNBC. Notably, LIN9

is also a member of the Mammaprint gene signature that predicts breast cancer

metastasis506.

Interestingly, Wiseman, et al. found that in esophageal adenocarcinoma cells, LIN9

and FOXM1 were co-overexpressed, while an additional component of the MuvB

complex, LIN54, was downregulated873. These data are difficult to reconcile,

considering both LIN9 and LIN54 are two of the core components of the MuvB

complex and are normally thought to co-occur. It is conceivable that FOXM1 and

LIN9 form a complex without LIN54. Supporting this possibility, FOXM1 and LIN9

cooperated to regulate a subset of genes in esophageal adenocarcinoma cells. In

addition, FOXM1 could directly interact with LIN9 to form a complex, and this

interaction was required for the recruitment of FOXM1 to the DNA653,873. LIN54 is

thought to be responsible for the localization of the MuvB complex to its target

genes874. The exclusion of LIN54 from a FOXM1-LIN9 complex could allow

FOXM1-LIN9 to regulate a distinct set of genes from MuvB, possibly those that

enhance a cell’s oncogenic capacity.

The potential of LIN9 to form alternative transcriptional regulatory complexes to

the MuvB complex raises the possibility that LIN9 behaves as an oncogene in

254

TNBC by either acting alone or by associating with currently unknown binding

partners to control transcription of genes critical for TNBC behavior. To test this

hypothesis, we could use an additional antibody against LIN54 when performing

the LIN9 ChIP-seq experiments described above. We would then compare the list

of genes bound by LIN54 to the list of LIN9-bound genes. If LIN9 only localizes to

specific genes as a subunit of the MuvB complex, the two gene lists will be

identical. However, if there is a significant discrepancy between LIN54- and LIN9-

bound genes, this could suggest LIN9 impacts transcription either alone or as part

of a complex separate from MuvB. Pathway analysis could be used to identify the

pathways regulated by this unknown LIN9-containing complex, which would

provide insight into the function of this complex. To identify the other subunits of

this LIN9 complex, we would use tandem affinity purification coupled to mass

spectrometry. Briefly, we would express FLAG-tagged LIN9 in MDA-MB-231 cells

and analyze whole cell extracts using electrospray ionization (ESI) coupled to LC-

MS/MS following separation of proteins by 1D SDS-PAGE875. This would identify

all LIN9-interacting proteins. To confirm our findings from this experiment, we

would perform co-immunoprecipitation assays with a LIN9-specific antibody using

lysates from MDA-MB-231 and MDA-MB-468 cells followed by western blot

analysis.

It is also possible oncogenic LIN9 acts via the MuvB complex and that the observed

effects of LIN9 silencing is simply a marker of the disruption of the MuvB complex.

However, amplification/overexpression of LIN9 does not coincide with higher

expression of the other components of the complex, as these other subunits, unlike

255

LIN9, are rarely amplified/overexpressed in TNBC tumors. However, the

expression of the MuvB components at the protein level remains completely

unexplored. To determine if LIN9 protein overexpression is associated with

increased expression of other MuvB components, we would evaluate the protein

expression of the individual subunits of the MuvB complex by western blot in

several TNBC cell lines. This would reveal if 1) increased LIN9 protein expression

coincides with elevated LIN9 mRNA expression and 2) protein expression of the

other subunits of the MuvB complex increases commensurate with LIN9

expression. This would indicate an increase in MuvB complex formation. If true,

gene-specific ChIP-PCR analyses of several TNBC cell lines using antibodies

specific for the MuvB complex subunits as well as B-MYB and FOXM1 would

reveal if the production of more MuvB complexes leads to a buildup of MuvB and

its binding partners at the promoters of target genes. Specific ER+ cell lines that

lack LIN9 overexpression would be used as controls.

4.9 LIN9 and tumorigenesis

While LIN9 is overexpressed and/or amplified in two-thirds of basal breast cancers

and high expression of LIN9 is linked to poor survival in breast cancer, it is

unknown if overexpression of LIN9 alone can induce tumorigenesis. To test this,

we would use a GFP-LIN9 expression vector to overexpress LIN9 in the

nontransformed mammary epithelial cell lines MCF10A and MCF12A. We would

then perform a soft agar colony formation assay to determine if cells

overexpressing LIN9 have adopted the ability to grow in an anchorage-

independent manner. Nontransformed cells cannot form colonies in this setting

256

and die by anoikis. On the other hand, if outgrowths occur with LIN9

overexpression, this would signal that overexpressing LIN9 has transformed

MCF10A and MCF12A cells, providing direct evidence of LIN9 acting as an

oncogene. The oncogenic potential of LIN9 would also be assessed in vivo. To do

this we would generate a mouse model with targeted, doxycycline-inducible

overexpression of LIN9 in the mammary gland. We would use an inducible system

in order to prevent developmental defects, as LIN9 is important during

embryogenesis857. Following induction of the transgene, mice would be observed

for the development of mammary tumors. If tumors arise in mice that overexpress

LIN9, a subset would be harvested and analyzed to confirm LIN9 expression and

to evaluate expression of LIN9-targeted genes. We would continue to monitor

additional tumors to determine if LIN9-driven tumors tend to metastasize. These

experiments would allow us to characterize LIN9-driven tumors and identify the

pathways activated during LIN9-mediated transformation. They would also directly

reveal whether LIN9 is a bona fide oncogene.

4.10 Additional inhibitors of LIN9

We have shown that BETi quickly downregulate LIN9, suggesting BETi may be

particularly effective against LIN9-overexpressing tumors, such as TNBCs.

However, de novo or acquired resistance to BETi would eliminate the utility of BETi

in LIN9-driven cancers. Thus, it will be important to identify additional agents that

suppress LIN9 expression. It is likely that other drugs that target BET proteins,

such as BET degraders and dual kinase/BET inhibitors, will also reduce expression

of LIN9. In addition, CDK7 inhibitors may function in a similar matter to BETi and

257

may therefore also result in the downregulation of LIN9. We would examine all

three of these drug classes to determine if any result in the loss of LIN9 expression.

BI-2536 (PLK inhibitor), TG-101348 (JAK inhibitor), and GSK2636771 (PI3K-

mTOR inhibitor) act as dual kinase/BET inhibitors in other cancers574. We would

treat TNBC cells with these inhibitors as well as the BET degrader BETd260 and

the CDK7 inhibitors THZ1 and BS-181 and assess if any of these drugs suppress

LIN9 expression. Any agents that downregulate LIN9 will be further studied to

determine if they induce mitotic catastrophe followed by apoptosis or senescence,

similar to BETi, or if they have a different impact on TNBC cells. To confirm BI-

2536, TG-101348, and GSK2636771 target BET proteins in breast cancer, we

would treat TNBC cells with each of these three inhibitors and perform LIN9 gene-

specific ChIP using a BRD4-specific antibody. If BRD4 no longer localizes to the

LIN9 gene following treatment with one of these inhibitors, this would suggest the

drugs inhibit BRD4 function. Together, these experiments would provide

information regarding additional potential treatment options for patients who have

LIN9-overexpressing tumors.

4.11 Conclusions

TNBC represents a diverse collection of highly aggressive tumors associated with

poor outcome. Efforts are currently underway to develop new targeted therapies

for this disease that will improve survival and prevent recurrence. Our studies

contribute to a growing body of evidence indicating that BETi may be an effective

treatment strategy for TNBC patients. We found BETi inhibited growth and induced

apoptosis and/or senescence in multiple cell lines that represent different subtypes

258

of TNBC, suggesting BETi should provide clinical benefit to a broad spectrum of

TNBC patients. In addition, we discovered BETi suppressed expression of the

mitotic regulators Aurora kinases A and B. We propose these kinases could be

used as markers of clinical response to BETi, which can be rapidly assessed in

ongoing clinical trials. BETi also downregulated LIN9, and we identified LIN9 as a

mitotic vulnerability in TNBC that mediates BETi activity. We suggest BETi may be

particularly effective against tumors that express high levels of LIN9, and LIN9 can

potentially serve as a predictor of response to BETi. This possibility should be

evaluated in clinical trials studying the activity of BETi in breast cancer as well as

other tumor types.

As with any drug, there is a risk of acquired resistance if BETi are used as single

agents. Therefore, it will be critical to develop effective combination strategies. The

identification of mitotic catastrophe as the mechanism of action of BETi in TNBC

can be used to predict optimal combination treatments, such as combining BETi

with drugs that further sensitize cancer cells to mitotic defects. For example,

unpublished data from our lab revealed that the addition of a Bcl-xL inhibitor to

BETi treatment increases the apoptotic response of TNBC cells to BETi. It is also

likely that adding BETi following pretreatment with an agent that prevents mitotic

entry, such as CDK4/6 inhibitors which arrest cells in G1, will prevent BETi from

inducing mitotic catastrophe. This will either render BETi ineffective or change the

way tumor cells respond to BETi. Indeed, ongoing work in the Keri laboratory has

revealed that BETi are highly synergistic with CDK4/6 inhibitors in TNBC in vitro

and in vivo.

259

Our discovery that LIN9 is a potential novel oncogene that is particularly important

in TNBC opens up an entirely new area of research. Our future studies will focus

on elucidating the LIN9-controlled transcriptome, identifying LIN9 binding sites in

TNBC cells, and determining if overexpression of LIN9 alone can initiate

tumorigenesis in the mammary gland. We will also examine if other drug classes

can be used to suppress expression of LIN9, suggesting they could serve as

alternatives to BETi therapy. Data generated from these studies will be applicable

not just for the treatment of TNBC tumors but for other tumor types that have

amplified/overexpressed LIN9.

In conclusion, the studies described here have several important clinical

implications for the treatment of TNBC and other LIN9-driven tumors. Our findings

provide insight into the response of TNBC cells to BETi and identify which tumors

are most likely to respond to BETi treatment. This should assist in the selection of

appropriate patients for this line of therapy, improving the utility of BETi as an anti-

cancer agent.

260

Cell Cycle Progression BRD4 Ac Ac G0/G1 M

LIN9 S G2

+BETi

BRD4 Apoptosis Mitotic Catastrophe Ac Ac

LIN9

Senescence

Figure 4.1. BETi induce mitotic catastrophe in TNBC by suppressing

expression of LIN9.

LIN9 expression is positively regulated by BRD4 which enhances progression

through the cell cycle. BETi bind the bromodomain regions of BRD4, preventing it

from localizing to the DNA. This reduces expression of LIN9, inducing mitotic

catastrophe which induces either apoptosis or senescence. Adapted from Gayle,

et al.876.

261

Numberof Cells 1

0 Vehicle JQ1

Figure 4.2. JQ1 may lead to aborted cleavage furrow ingression.

MDA-MB-468 cells were treated with vehicle or 1000 nM JQ1 and observed via

live-cell imaging. 87 vehicle-treated and 83 JQ1-treated cells were tracked mitosis,

and the number of cells that exited mitosis and formed a visible cleavage furrow

that eventually regressed, producing a single daughter cell, were counted.

262

Vehicle 1000 nM JQ1 MCF7 T47D

Figure 4.3. BETi alter growth patterns and cellular morphology of luminal

breast cancer cells.

Representative phase images (10x) of MCF7 and T47D cells that were treated with

vehicle or 1000 nM JQ1 for six days.

263

AMP/OE 60 LIN9

20 % Tumors with with % Tumors

Breast cancer

Figure 4.4. LIN9 is amplified and overexpressed in cancer and particularly in

TNBC.

LIN9 expression was analyzed in various cancer types using cBioPortal, with

breast cancer stratified into five groups: all breast cancer, luminal A, luminal B,

HER2-enriched, and basal-like. LIN9 is amplified and/or overexpressed in a

greater percentage of TNBC tumors compared to all other cancer types. AMP/OE,

amiplification/overexpression.

264

REFERENCES

1 Kohler, B. A. et al. Annual report to the nation on the status of cancer, 1975-

2011, featuring incidence of breast cancer subtypes by race/ethnicity,

poverty, and state. J Natl Cancer Inst 107, djv048 (2015).

2 Jemal, A. et al. Cancer statistics, 2003. CA Cancer J Clin 53, 5-26 (2003).

3 DeSantis, C., Ma, J., Bryan, L. & Jemal, A. Breast cancer statistics, 2013.

CA Cancer J Clin 64, 52-62 (2014).

4 Han, W. et al. Young age: an independent risk factor for disease-free

survival in women with operable breast cancer. BMC Cancer 4, 82 (2004).

5 Arvold, N. D. et al. Age, breast cancer subtype approximation, and local

recurrence after breast-conserving therapy. J Clin Oncol 29, 3885-3891

(2011).

6 DeSantis, C. E. et al. Breast cancer statistics, 2015: Convergence of

incidence rates between black and white women. CA Cancer J Clin 66, 31-

42 (2016).

7 Fedewa, S. A., Sauer, A. G., Siegel, R. L. & Jemal, A. Prevalence of major

risk factors and use of screening tests for cancer in the United States.

Cancer Epidemiol Biomarkers Prev 24, 637-652 (2015).

8 Costa, M. & Saldanha, P. Risk reduction strategies in breast cancer

prevention. Eur J Breast Health 13, 103-112 (2017).

9 Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer

and breastfeeding: collaborative reanalysis of individual data from 47

265

epidemiological studies in 30 countries, including 50302 women with breast

cancer and 96973 women without the disease. Lancet 360, 187-195 (2002).

10 Lambertini, M. et al. Reproductive behaviors and risk of developing breast

cancer according to tumor subtype: A systematic review and meta-analysis

of epidemiological studies. Cancer Treat Rev 49, 65-76 (2016).

11 Claus, E. B., Schildkraut, J. M., Thompson, W. D. & Risch, N. J. The genetic

attributable risk of breast and ovarian cancer. Cancer 77, 2318-2324 (1996).

12 Collaborative Group on Hormonal Factors in Breast Cancer. Familial breast

cancer: collaborative reanalysis of individual data from 52 epidemiological

studies including 58,209 women with breast cancer and 101,986 women

without the disease. Lancet 358, 1389-1399 (2001).

13 Couch, F. J., Nathanson, K. L. & Offit, K. Two decades after BRCA: setting

paradigms in personalized cancer care and prevention. Science 343, 1466-

1470 (2014).

14 Ozsoy, A. et al. The relationship between breast cancer and risk factors: a

single-center study. Eur J Breast Health 13, 145-149 (2017).

15 Venkitaraman, A. R. Linking the cellular functions of BRCA genes to cancer

pathogenesis and treatment. Annu Rev Pathol 4, 461-487 (2009).

16 Yun, M. H. & Hiom, K. CtIP-BRCA1 modulates the choice of DNA double-

strand break repair pathway throughout the cell cycle. Nature 459, 460-463

(2009).

17 Antoniou, A. et al. Average risks of breast and ovarian cancer associated

with BRCA1 or BRCA2 mutations detected in case Series unselected for

266

family history: a combined analysis of 22 studies. Am J Hum Genet 72,

1117-1130 (2003).

18 Balmaña, J., Díez, O., Castiglione, M. & ESMO Guidelines Working Group.

BRCA in breast cancer: ESMO clinical recommendations. Ann Oncol 20,

19-20 (2009).

19 Rubio-Perez, C. et al. In silico prescription of anticancer drugs to cohorts of

28 tumor types reveals targeting opportunities. Cancer Cell 27, 382-396

(2015).

20 Gonzalez-Perez, A. et al. IntOGen-mutations identifies cancer drivers

across tumor types. Nat Methods 10, 1081-1082 (2013).

21 Cancer Genome Atlas Network. Comprehensive molecular portraits of

human breast tumours. Nature 490, 61-70, doi:10.1038/nature11412

(2012).

22 Banerji, S. et al. Sequence analysis of mutations and translocations across

breast cancer subtypes. Nature 486, 405-409 (2012).

23 Stephens, P. J. et al. The landscape of cancer genes and mutational

processes in breast cancer. Nature 486, 400-404 (2012).

24 Ellis, M. J. et al. Whole-genome analysis informs breast cancer response to

aromatase inhibition. Nature 486, 353-360 (2012).

25 Shah, S. P. et al. The clonal and mutational evolution spectrum of primary

triple-negative breast cancers. Nature 486, 395-399 (2012).

26 Wood, L. D. et al. The genomic landscapes of human breast and colorectal

cancers. Science 318, 1108-1113 (2007).

267

27 Curtis, C. et al. The genomic and transcriptomic architecture of 2,000 breast

tumours reveals novel subgroups. Nature 486, 346-352 (2012).

28 Paterni, I., Granchi, C., Katzenellenbogen, J. A. & Minutolo, F. Estrogen

receptors alpha (ERα) and beta (ERβ): subtype-selective ligands and

clinical potential. Steroids 90, 13-29 (2014).

29 Stein, R. A. & McDonnell, D. P. Estrogen-related receptor alpha as a

therapeutic target in cancer. Endocr Relat Cancer 13, S25-S32 (2006).

30 Frasor, J. et al. Profiling of estrogen up- and down-regulated gene

expression in human breast cancer cells: insights into gene networks and

pathways underlying estrogenic control of proliferation and cell phenotype.

Endocrinology 144, 4562-4574 (2003).

31 Tora, L. et al. The human estrogen receptor has two independent nonacidic

transcriptional activation functions. Cell 59, 477-487 (1989).

32 Lees, J. A., Fawell, S. E. & Parker, M. G. Identification of two transactivation

domains in the mouse oestrogen receptor. Nucleic Acids Res 17, 5477-

5488 (1989).

33 Saville, B. et al. Ligand-, cell-, and estrogen receptor subtype (alpha/beta)-

dependent activation at GC-rich (Sp1) promoter elements. J Biol Chem 275,

5379-5387 (2000).

34 Kushner, P. J. et al. Estrogen receptor pathways to AP-1. J Steroid Biochem

Mol Biol 74, 311-317 (2000).

268

35 McKay, L. I. & Cidlowski, J. A. Molecular control of immune/inflammatory

responses: interactions between nuclear factor-kappa B and steroid

receptor-signaling pathways. Endocr Rev 20, 435-459 (1999).

36 Le Goff, P., Montano, M. M., Schodin, D. J. & Katzenellenbogen, B. S.

Phosphorylation of the human estrogen receptor. Identification of hormone-

regulated sites and examination of their influence on transcriptional activity.

J Biol Chem 269, 4458-4466 (1994).

37 Bardou, V.-J., Arpino, G., Elledge, R. M., Osborne, C. K. & Clark, G. M.

Progesterone receptor status significantly improves outcome prediction

over estrogen receptor status alone for adjuvant endocrine therapy in two

large breast cancer databases. J Clin Oncol 21, 1973-1979 (2003).

38 McGuire, W. L. Hormone receptors: their role in predicting prognosis and

response to endocrine therapy. Semin Oncol 5, 428-433 (1978).

39 Lv, Q. et al. Molecular mechanisms and translational therapies for Human

Epidermal Receptor 2 positive breast cancer. Int J Mol Sci 17, 2095 (2016).

40 Graus-Porta, D., Beerli, R. R., Daly, J. M. & Hynes, N. E. ErbB-2, the

preferred heterodimerization partner of all ErbB receptors, is a mediator of

lateral signaling. EMBO J 16, 1647-1655 (1997).

41 Yarden, Y. & Sliwkowski, M. X. Untangling the ErbB signalling network. Nat

Rev Mol Cell Biol 2, 127-137 (2001).

42 Slamon, D. J. et al. Human breast cancer: correlation of relapse and survival

with amplification of the HER-2/neu oncogene. Science 235, 177-182

(1987).

269

43 Carey, L. A. et al. The triple negative paradox: primary tumor

chemosensitivity of breast cancer subtypes. Clin Cancer Res 13, 2329-

2334 (2007).

44 Holbro, T. et al. The ErbB2/ErbB3 heterodimer functions as an oncogenic

unit: ErbB2 requires ErbB3 to drive breast tumor cell proliferation. Proc Natl

Acad Sci USA 100, 8933-8938 (2003).

45 Hellyer, N. J., Cheng, K. & Koland, J. G. ErbB3 (HER3) interaction with the

p85 regulatory subunit of phosphoinositide 3-kinase. Biochem J 333, 757-

763 (1998).

46 Pinkas-Kramarski, R. et al. Diversification of Neu differentiation factor and

epidermal growth factor signaling by combinatorial receptor interactions.

EMBO J 15, 2452–2467 (1996).

47 Xia, W. et al. Regulation of survivin by ErbB2 signaling: therapeutic

implications for ErbB2-overexpressing breast cancers. Cancer Res 66,

1640-1647 (2006).

48 Makino, K., Day, C.-P., Wang, S.-C., Li, Y. M. & Hung, M.-C. Upregulation

of IKKalpha/IKKbeta by integrin-linked kinase is required for HER2/neu-

induced NF-kappaB antiapoptotic pathway. Oncogene 23, 3883-3887

(2004).

49 Laughner, E., Taghavi, P., Chiles, K., Mahon, P. C. & Semenza, G. L. HER2

(neu) signaling increases the rate of hypoxia-inducible factor 1alpha (HIF-

1alpha) synthesis: novel mechanism for HIF-1-mediated vascular

endothelial growth factor expression. Mol Cell Biol 21, 3995-4004 (2001).

270

50 Bagheri-Yarmand, R., Vadlamudi, R. K., Wang, R.-A., Mendelsohn, J. &

Kumar, R. Vascular endothelial growth factor up-regulation via p21-

activated kinase-1 signaling regulates heregulin-beta1-mediated

angiogenesis. J Biol Chem 275, 39451-39457 (2000).

51 Yen, L. et al. Heregulin selectively upregulates vascular endothelial growth

factor secretion in cancer cells and stimulates angiogenesis. Oncogene 19,

3460-3469 (2000).

52 Staaf, J. et al. High-resolution genomic and expression analyses of copy

number alterations in HER2-amplified breast cancer. Breast Cancer Res

12, R25 (2010).

53 Holm, K. et al. Molecular subtypes of breast cancer are associated with

characteristic DNA methylation patterns. Breast Cancer Res 12, R36

(2010).

54 Staaf, J. et al. Identification of subtypes in human epidermal growth factor

receptor 2--positive breast cancer reveals a gene signature prognostic of

outcome. J Clin Oncol 28, 1813-1820 (2010).

55 Gatza, M. L. et al. Analysis of tumor environmental response and oncogenic

pathway activation identifies distinct basal and luminal features in HER2-

related breast tumor subtypes. Breast Cancer Res 13, R62 (2011).

56 Gatza, M. L. et al. A pathway-based classification of human breast cancer.

Proc Natl Acad Sci USA 107, 6994-6999 (2010).

57 Morris, G. J. et al. Differences in breast carcinoma characteristics in newly

diagnosed African-American and Caucasian patients: a single-institution

271

compilation compared with the National Cancer Institute's Surveillance,

Epidemiology, and End Results database. Cancer 110, 876-884 (2007).

58 Bauer, K. R., Brown, M., Cress, R. D., Parise, C. A. & Caggiano, V.

Descriptive analysis of estrogen receptor (ER)-negative, progesterone

receptor (PR)-negative, and HER2-negative invasive breast cancer, the so-

called triple-negative phenotype: a population-based study from the

California cancer Registry. Cancer 109, 1721-1728 (2007).

59 Lund, M. J. et al. Race and triple negative threats to breast cancer survival:

a population-based study in Atlanta, GA. Breast Cancer Res Treat 113, 357-

370 (2009).

60 Stead, L. A. et al. Triple-negative breast cancers are increased in black

women regardless of age or body mass index. Breast Cancer Res 11, R18

(2009).

61 Liedtke, C. et al. Response to neoadjuvant therapy and long-term survival

in patients with triple-negative breast cancer. J Clin Oncol 26, 1275-1281

(2008).

62 von Minckwitz, G. et al. Definition and impact of pathologic complete

response on prognosis after neoadjuvant chemotherapy in various intrinsic

breast cancer subtypes. J Clin Oncol 30, 1796-1804 (2012).

63 Haffty, B. G. et al. Locoregional relapse and distant metastasis in

conservatively managed triple negative early-stage breast cancer. J Clin

Oncol 24, 5652-5657 (2006).

272

64 Dent, R. et al. Triple-negative breast cancer: clinical features and patterns

of recurrence. Clin Cancer Res 13, 4429-4434, doi:10.1158/1078-

0432.CCR-06-3045 (2007).

65 Malorni, L. et al. Clinical and biologic features of triple-negative breast

cancers in a large cohort of patients with long-term follow-up. Breast Cancer

Res Treat 136, 795-804 (2012).

66 Bonotto, M. et al. Measures of outcome in metastatic breast cancer: insights

from a real-world scenario. Oncologist 19, 608-615 (2014).

67 Rimawi, M. F. et al. Epidermal growth factor receptor expression in breast

cancer association with biologic phenotype and clinical outcomes. Cancer

116, 1234-1242 (2010).

68 Viale, G. et al. Invasive ductal carcinoma of the breast with the "triple-

negative" phenotype: prognostic implications of EGFR immunoreactivity.

Breast Cancer Res Treat 116, 317-328 (2009).

69 Richard, J. et al. Epidermal-growth-factor receptor status as predictor of

early recurrence of and death from breast cancer. Lancet 1, 1398-1402

(1987).

70 Atchley, D. P. et al. Clinical and pathologic characteristics of patients with

BRCA-positive and BRCA-negative breast cancer. J Clin Oncol 26, 4282-

4288 (2008).

71 Mittendorf, E. A. et al. PD-L1 expression in triple-negative breast cancer.

Cancer Immunol Res 2, 361-370 (2014).

273

72 Wimberly, H. et al. PD-L1 expression correlates with tumor-infiltrating

lymphocytes and response to neoadjuvant chemotherapy in breast cancer.

Cancer Immunol Res 3, 326–332 (2015).

73 Craig, D. W. et al. Genome and transcriptome sequencing in prospective

metastatic triple-negative breast cancer uncovers therapeutic

vulnerabilities. Mol Cancer Ther 12, 104-116 (2013).

74 Kim, J.-Y. et al. Association between mutation and expression of TP53 as

a potential prognostic marker of triple-negative breast cancer. Cancer Res

Treat 48, 1338-1350 (2016).

75 Horiuchi, D. et al. MYC pathway activation in triple-negative breast cancer

is synthetic lethal with CDK inhibition. J Exp Med 209, 679-696 (2012).

76 Gucalp, A. & Traina, T. A. Triple-negative breast cancer: role of the

androgen receptor. Cancer J 16, 62-65 (2010).

77 Perou, C. M. et al. Molecular portraits of human breast tumours. Nature 406,

747-752, doi:10.1038/35021093 (2000).

78 Prat, A. & Perou, C. M. Deconstructing the molecular portraits of breast

cancer. Mol Oncol 5, 5-23 (2011).

79 Prat, A. et al. Phenotypic and molecular characterization of the claudin-low

intrinsic subtype of breast cancer. Breast Cancer Res 12, R68 (2010).

80 Herschkowitz, J. I. et al. Identification of conserved gene expression

features between murine mammary carcinoma models and human breast

tumors. Genome Biol 8, R76, doi:10.1186/gb-2007-8-5-r76 (2007).

274

81 Hennessy, B. T. et al. Characterization of a naturally occurring breast

cancer subset enriched in epithelial-to-mesenchymal transition and stem

cell characteristics. Cancer Res 69, 4116-4124 (2009).

82 Parker, J. S. et al. Supervised risk predictor of breast cancer based on

intrinsic subtypes. J Clin Oncol 27, 1160-1167,

doi:10.1200/JCO.2008.18.1370 (2009).

83 Prat, A. et al. Clinical implications of the intrinsic molecular subtypes of

breast cancer. Breast 24, S26-35 (2015).

84 Hess, K. R. et al. Pharmacogenomic predictor of sensitivity to preoperative

chemotherapy with paclitaxel and fluorouracil, doxorubicin, and

cyclophosphamide in breast cancer. J Clin Oncol 24, 4236-4244 (2006).

85 Prat, A. et al. Prognostic significance of progesterone receptor-positive

tumor cells within immunohistochemically defined luminal A breast cancer.

J Clin Oncol 31, 203-209 (2013).

86 Prat, A., Parker, J. S., Fan, C. & Perou, C. M. PAM50 assay and the three-

gene model for identifying the major and clinically relevant molecular

subtypes of breast cancer. Breast Cancer Res Treat 135, 301-306 (2012).

87 Usary, J. et al. Predicting drug responsiveness in human cancers using

genetically engineered mice. Clin Cancer Res 19, 4889–4899 (2013).

88 von Minckwitz, G. et al. Response-guided neoadjuvant chemotherapy for

breast cancer. J Clin Oncol 31, 3623-3630 (2013).

89 Rouzier, R. et al. Breast cancer molecular subtypes respond differently to

preoperative chemotherapy. Clin Cancer Res 11, 5678-5685 (2005).

275

90 Martín, M. et al. PAM50 proliferation score as a predictor of weekly

paclitaxel benefit in breast cancer. Breast Cancer Res Treat 138, 457–466

(2013).

91 Dowsett, M. et al. Comparison of PAM50 risk of recurrence score with

oncotype DX and IHC4 for predicting risk of distant recurrence after

endocrine therapy. J Clin Oncol 31, 2783-2790 (2013).

92 Gnant, M. et al. Predicting distant recurrence in receptor-positive breast

cancer patients with limited clinicopathological risk: using the PAM50 Risk

of Recurrence score in 1478 postmenopausal patients of the ABCSG-8 trial

treated with adjuvant endocrine therapy alone. Ann Oncol 25, 339-345

(2014).

93 Cheang, M. C. U. et al. Responsiveness of intrinsic subtypes to adjuvant

anthracycline substitution in the NCIC.CTG MA.5 randomized trial. Clin

Cancer Res 18, 2402-2412 (2012).

94 Chia, S. K. et al. A 50-gene intrinsic subtype classifier for prognosis and

prediction of benefit from adjuvant tamoxifen. Clin Cancer Res 18, 4465-

4472 (2012).

95 Filipits, M. et al. The PAM50 risk-of-recurrence score predicts risk for late

distant recurrence after endocrine therapy in postmenopausal women with

endocrine-responsive early breast cancer. Clin Cancer Res 20, 1298-1305

(2014).

96 Cheang, M. C. U. et al. Ki67 index, HER2 status, and prognosis of patients

with luminal B breast cancer. J Natl Cancer Inst 101, 736-750 (2009).

276

97 Nielsen, T. O. et al. A comparison of PAM50 intrinsic subtyping with

immunohistochemistry and clinical prognostic factors in tamoxifen-treated

estrogen receptor-positive breast cancer. Clin Cancer Res 16, 5222-5232

(2010).

98 Dowsett, M. et al. Prediction of risk of distant recurrence using the 21-gene

recurrence score in node-negative and node-positive postmenopausal

patients with breast cancer treated with anastrozole or tamoxifen: a

TransATAC study. J Clin Oncol 28, 1829-1834 (2010).

99 Harrell, J. C. et al. Genomic analysis identifies unique signatures predictive

of brain, lung, and liver relapse. Breast Cancer Res Treat 132, 523–535

(2012).

100 Knutson, T. P. & Lange, C. A. Tracking progesterone receptor-mediated

actions in breast cancer. Pharmacol Ther 142, 114-125 (2014).

101 Sørlie, T. et al. Gene expression patterns of breast carcinomas distinguish

tumor subclasses with clinical implications. Proc Natl Acad Sci USA 98,

10869-10874, doi:10.1073/pnas.191367098 (2001).

102 Roberts, S. A. et al. An APOBEC cytidine deaminase mutagenesis pattern

is widespread in human cancers. Nat Genet 45, 970-976 (2013).

103 Rakha, E. A., Reis-Filho, J. S. & Ellis, I. O. Basal-like breast cancer: a critical

review. J Clin Oncol 26, 2568-2581 (2008).

104 Ali, H. R. et al. PD-L1 protein expression in breast cancer is rare, enriched

in basal-like tumours and associated with infiltrating lymphocytes. Ann

Oncol 26, 1488-1493 (2015).

277

105 Sørlie, T. et al. Repeated observation of breast tumor subtypes in

independent gene expression data sets. Proc Natl Acad Sci USA 100, 8418-

8423 (2003).

106 Fulford, L. G. et al. Specific morphological features predictive for the basal

phenotype in grade 3 invasive ductal carcinoma of breast. Histopathology

49, 22-34 (2006).

107 Carey, L. A. et al. Race, breast cancer subtypes, and survival in the Carolina

Breast Cancer Study. JAMA 295, 2492-2502 (2006).

108 Prat, A. et al. Molecular features of the basal-like breast cancer subtype

based on BRCA1 mutation status. Breast Cancer Res Treat 147, 185-191

(2014).

109 Foulkes, W. D. et al. Germline BRCA1 mutations and a basal epithelial

phenotype in breast cancer. J Natl Cancer Inst 95, 1482-1485 (2003).

110 Herschkowitz, J. I., He, X., Fan, C. & Perou, C. M. The functional loss of the

retinoblastoma tumour suppressor is a common event in basal-like and

luminal B breast carcinomas. Breast Cancer Res 10, R75 (2008).

111 Chandriani, S. et al. A core MYC gene expression signature is prominent in

basal-like breast cancer but only partially overlaps the core serum

response. PLoS One 4, e6693 (2009).

112 Collins, L. C. et al. Androgen receptor expression in breast cancer in relation

to molecular phenotype: results from the Nurses’ Health Study. Mod Pathol

24, 924–931 (2011).

278

113 Prat, A. et al. Genomic analyses across six cancer types identify basal-like

breast cancer as a unique molecular entity. Sci Rep 3 (2013).

114 Hoadley, K. A. et al. Multiplatform analysis of 12 cancer types reveals

molecular classification within and across tissues of origin. Cell 158, 929-

944 (2014).

115 Nielsen, T. O. et al. Immunohistochemical and clinical characterization of

the basal-like subtype of invasive breast carcinoma. Clin Cancer Res 10,

5367-5374 (2004).

116 Chollet-Hinton, L. et al. Breast cancer biologic and etiologic heterogeneity

by young age and menopausal status in the Carolina Breast Cancer Study:

a case-control study. Breast Cancer Res 18, 79 (2016).

117 Calza, S. et al. Intrinsic molecular signature of breast cancer in a population-

based cohort of 412 patients. Breast Cancer Res 8, R34 (2006).

118 Cheang, M. C. U. et al. Basal-like breast cancer defined by five biomarkers

has superior prognostic value than triple-negative phenotype. Clin Cancer

Res 14, 1368-1376 (2008).

119 Lim, E. et al. Aberrant luminal progenitors as the candidate target population

for basal tumor development in BRCA1 mutation carriers. Nat Med 15, 907-

913 (2009).

120 Taube, J. H. et al. Core epithelial-to-mesenchymal transition interactome

gene-expression signature is associated with claudin-low and metaplastic

breast cancer subtypes. Proc Natl Acad Sci USA 107, 15449-15454 (2010).

279

121 Creighton, C. J. et al. Residual breast cancers after conventional therapy

display mesenchymal as well as tumor-initiating features. Proc Natl Acad

Sci USA 106, 13820-13825 (2009).

122 Lehmann, B. D. et al. Identification of human triple-negative breast cancer

subtypes and preclinical models for selection of targeted therapies. J Clin

Invest 121, 2750-2767, doi:10.1172/JCI45014 (2011).

123 Masuda, H. et al. Differential response to neoadjuvant chemotherapy

among 7 triple-negative breast cancer molecular subtypes. Clin Cancer Res

19, 5533-5540 (2013).

124 Lehmann, B. D. & Pietenpol, J. A. Identification and use of biomarkers in

treatment strategies for triple negative breast cancer subtypes. J Pathol

232, 142-150 (2014).

125 Lehmann, B. D. et al. Refinement of triple-negative breast cancer molecular

subtypes: Implications for neoadjuvant chemotherapy selection. PLoS One

11, e0157368, doi:10.1371/journal.pone.0157368 (2016).

126 Lehmann, B. D. et al. PIK3CA mutations in androgen receptor-positive triple

negative breast cancer confer sensitivity to the combination of PI3K and

androgen receptor inhibitors. Breast Cancer Res 16, 406 (2014).

127 Burstein, M. D. et al. Comprehensive genomic analysis identifies novel

subtypes and targets of triple-negative breast cancer. Clin Cancer Res 21,

1688-1698 (2015).

280

128 Stemke-Hale, K. et al. An integrative genomic and proteomic analysis of

PIK3CA, PTEN, and AKT mutations in breast cancer. Cancer Res 68, 6084-

6091 (2008).

129 Katso, R. et al. Cellular function of phosphoinositide 3-kinases: implications

for development, homeostasis, and cancer. Annu Rev Cell Dev Biol 17, 615-

675 (2001).

130 Finn, R. S., Aleshin, A. & Slamon, D. J. Targeting the cyclin-dependent

kinases (CDK) 4/6 in estrogen receptor-positive breast cancers. Breast

Cancer Res 18, 17 (2016).

131 Peddi, P. F. & Hurvitz, S. A. Trastuzumab emtansine: the first targeted

chemotherapy for treatment of breast cancer. Future Oncol 9, 319-326

(2013).

132 Erickson, H. K. et al. Antibody-maytansinoid conjugates are activated in

targeted cancer cells by lysosomal degradation and linker-dependent

intracellular processing. Cancer Res 66, 4426-4433 (2006).

133 Remillard, S., Rebhun, L. I., Howie, G. A. & Kupchan, S. M. Antimitotic

activity of the potent tumor inhibitor maytansine. Science 189, 1002-1005

(1975).

134 Pang, B. et al. Drug-induced histone eviction from open chromatin

contributes to the chemotherapeutic effects of doxorubicin. Nat Commun 4,

1908 (2013).

281

135 Gewirtz, D. A. A critical evaluation of the mechanisms of action proposed

for the antitumor effects of the anthracycline antibiotics adriamycin and

daunorubicin. Biochem Pharmacol 57, 727-741 (1999).

136 Citron, M. L. et al. Randomized trial of dose-dense versus conventionally

scheduled and sequential versus concurrent combination chemotherapy as

postoperative adjuvant treatment of node-positive primary breast cancer:

first report of Intergroup Trial C9741/Cancer and Leukemia Group B Trial

9741. J Clin Oncol 21, 1431-1439 (2003).

137 Berry, D. A. et al. Estrogen-receptor status and outcomes of modern

chemotherapy for patients with node-positive breast cancer. JAMA 295,

1658-1667 (2006).

138 Gluz, O. et al. Triple-negative high-risk breast cancer derives particular

benefit from dose intensification of adjuvant chemotherapy: results of WSG

AM-01 trial. Ann Oncol 19, 861-870 (2008).

139 Balko, J. M. et al. Molecular profiling of the residual disease of triple-

negative breast cancers after neoadjuvant chemotherapy identifies

actionable therapeutic targets. Cancer Discov 4, 232-245 (2014).

140 Bianchini, G., Balko, J. M., Mayer, I. A., Sanders, M. E. & Gianni, L. Triple-

negative breast cancer: challenges and opportunities of a heterogeneous

disease. Nat Rev Clin Oncol 13, 674-690 (2016).

141 Dantzer, F. et al. Base excision repair is impaired in mammalian cells

lacking Poly(ADP-ribose) polymerase-1. Biochemistry 39, 7559-7569

(2000).

282

142 Saleh-Gohari, N. et al. Spontaneous homologous recombination is induced

by collapsed replication forks that are caused by endogenous DNA single-

strand breaks. Mol Cell Biol 25, 7158-7169 (2005).

143 Farmer, H. et al. Targeting the DNA repair defect in BRCA mutant cells as

a therapeutic strategy. Nature 434, 917-921 (2005).

144 Patel, A. G., Sarkaria, J. N. & Kaufmann, S. H. Nonhomologous end joining

drives poly(ADP-ribose) polymerase (PARP) inhibitor lethality in

homologous recombination-deficient cells. Proc Natl Acad Sci USA 108,

3406–3411 (2011).

145 Rottenberg, S. et al. High sensitivity of BRCA1-deficient mammary tumors

to the PARP inhibitor AZD2281 alone and in combination with platinum

drugs. Proc Natl Acad Sci USA 105, 17079-17084 (2008).

146 Robson, M. et al. Olaparib for metastatic breast cancer in patients with a

germline BRCA mutation. N Engl J Med 377, 523-533 (2017).

147 de Bono, J. et al. Phase I, dose-escalation, two-part trial of the PARP

inhibitor talazoparib in patients with advanced germline BRCA1/2 mutations

and selected sporadic cancers. Cancer Discov 7, 620-629 (2017).

148 Tutt, A. et al. Oral poly(ADP-ribose) polymerase inhibitor olaparib in patients

with BRCA1 or BRCA2 mutations and advanced breast cancer: a proof-of-

concept trial. Lancet 376, 235-244 (2010).

149 Gucalp, A. et al. Phase II trial of bicalutamide in patients with androgen

receptor-positive, estrogen receptor-negative metastatic breast cancer. Clin

Cancer Res 19, 5505-5512 (2013).

283

150 Gonzalez-Angulo, A. M. et al. Androgen receptor levels and association with

PIK3CA mutations and prognosis in breast cancer. Clin Cancer Res 15,

2472-2478 (2009).

151 Asghar, U. S. et al. Single-cell dynamics determines response to CDK4/6

inhibition in triple-negative breast cancer. Clin Cancer Res 23, 5561-5572

(2017).

152 Finn, R. S. et al. PD 0332991, a selective cyclin D kinase 4/6 inhibitor,

preferentially inhibits proliferation of luminal estrogen receptor-positive

human breast cancer cell lines in vitro. Breast Cancer Res 11, R77 (2009).

153 Burness, M. L., Grushko, T. A. & Olopade, O. I. Epidermal growth factor

receptor in triple-negative and basal-like breast cancer: promising clinical

target or only a marker? Cancer J 16, 23-32 (2010).

154 Rakha, E. A. et al. Prognostic markers in triple-negative breast cancer.

Cancer 109, 25-32 (2007).

155 Guérin, M. et al. Structure and expression of c-erbB-2 and EGF receptor

genes in inflammatory and non-inflammatory breast cancer: prognostic

significance. Int J Cancer 43, 201-208 (1989).

156 Nakai, K., Hung, M.-C. & Yamaguchi, H. A perspective on anti-EGFR

therapies targeting triple-negative breast cancer. Am J Cancer Res 6, 1609-

1623 (2016).

157 Silver, D. P. et al. Efficacy of neoadjuvant Cisplatin in triple-negative breast

cancer. J Clin Oncol 28, 1145-1153 (2010).

284

158 Sikov, W. M. et al. Impact of the addition of carboplatin and/or bevacizumab

to neoadjuvant once-per-week paclitaxel followed by dose-dense

doxorubicin and cyclophosphamide on pathologic complete response rates

in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance). J

Clin Oncol 33, 13-21 (2015).

159 von Minckwitz, G. et al. Neoadjuvant carboplatin in patients with triple-

negative and HER2-positive early breast cancer (GeparSixto; GBG 66): a

randomised phase 2 trial. Lancet Oncol 15, 747-756 (2014).

160 Bhattacharyya, A., Ear, U. S., Koller, B. H., Weichselbaum, R. R. & Bishop,

D. K. The breast cancer susceptibility gene BRCA1 is required for

subnuclear assembly of Rad51 and survival following treatment with the

DNA cross-linking agent cisplatin. J Biol Chem 275, 23899-23903 (2000).

161 Kennedy, R. D., Quinn, J. E., Mullan, P. B., Johnston, P. G. & Harkin, D. P.

The role of BRCA1 in the cellular response to chemotherapy. J Natl Cancer

Inst 96, 1659-1668 (2004).

162 Isakoff, S. J. et al. TBCRC009: a multicenter phase II clinical trial of platinum

monotherapy with biomarker assessment in metastatic triple-negative

breast cancer. J Clin Oncol 33, 1902-1909 (2015).

163 Byrski, T. et al. Pathologic complete response to neoadjuvant cisplatin in

BRCA1-positive breast cancer patients. Breast Cancer Res Treat 147, 401-

405 (2014).

164 Telli, M. L. et al. Phase II Study of gemcitabine, carboplatin, and iniparib as

neoadjuvant therapy for triple-negative and BRCA1/2 mutation-associated

285

breast cancer with assessment of a tumor-based measure of genomic

instability: PrECOG 0105. J Clin Oncol 33, 1895-1901 (2015).

165 Colleoni, M. et al. Annual hazard rates of recurrence for breast cancer

during 24 years of follow-up: results from the International Breast Cancer

Study Group Trials I to V. J Clin Oncol 34, 927-935 (2016).

166 Sanfilippo, O., Ronchi, E., De Marco, C., Di Fronzo, G. & Silvestrini, R.

Expression of P-glycoprotein in breast cancer tissue and in vitro resistance

to doxorubicin and vincristine. Eur J Cancer 27, 155-158 (1991).

167 Murray, S., Briasoulis, E., Linardou, H., Bafaloukos, D. & Papadimitriou, C.

Taxane resistance in breast cancer: mechanisms, predictive biomarkers

and circumvention strategies. Cancer Treat Rev 38, 890-903 (2012).

168 Aas, T. et al. Specific p53 mutations are associated with de novo resistance

to doxorubicin in breast cancer patients. Nat Med 2, 811-814 (1996).

169 Brown, I. et al. Reduced expression of p27 is a novel mechanism of

docetaxel resistance in breast cancer cells. Breast Cancer Res 6, R601-

607 (2004).

170 Sharifi, S., Barar, J., Hejazi, M. S. & Samadi, N. Roles of the Bcl-2/Bax ratio,

caspase-8 and 9 in resistance of breast cancer cells to paclitaxel. Asian Pac

J Cancer Prev 15, 8617-8622 (2014).

171 Ibrado, A. M., Liu, L. & Bhalla, K. Bcl-xL overexpression inhibits progression

of molecular events leading to paclitaxel-induced apoptosis of human acute

myeloid leukemia HL-60 cells. Cancer Res 57, 1109-1115 (1997).

286

172 Strobel, T., Swanson, L., Korsmeyer, S. & Cannistra, S. A. BAX enhances

paclitaxel-induced apoptosis through a p53-independent pathway. Proc

Natl Acad Sci USA 93, 14094-14099 (1996).

173 Strobel, T., Tai, Y.-T., Korsmeyer, S. & Cannistra, S. A. BAD partly reverses

paclitaxel resistance in human ovarian cancer cells. Oncogene 17, 2419-

2427 (1998).

174 Rouzier, R. et al. Microtubule-associated protein tau: a marker of paclitaxel

sensitivity in breast cancer. Proc Natl Acad Sci USA 102, 8315-8320 (2005).

175 Han, E. K. et al. Modulation of drug resistance by alpha-tubulin in paclitaxel-

resistant human lung cancer cell lines. Eur J Cancer 36, 1565-1571 (2000).

176 Cabral, F., Abraham, I. & Gottesman, M. M. Isolation of a taxol-resistant

Chinese hamster ovary cell mutant that has an alteration in alpha-tubulin.

Proc Natl Acad Sci USA 78, 4388-4391 (1981).

177 Weaver, B. A. A. & Cleveland, D. W. Decoding the links between mitosis,

cancer, and chemotherapy: the mitotic checkpoint, adaptation, and cell

death. Cancer Cell 8, 7-12 (2005).

178 Sudo, T., Nitta, M., Saya, H. & Ueno, N. T. Dependence of paclitaxel

sensitivity on a functional spindle assembly checkpoint. Cancer Res 64,

2502-2508 (2004).

179 Chabalier, C. et al. BRCA1 downregulation leads to premature inactivation

of spindle checkpoint and confers paclitaxel resistance. Cell Cycle 5, 1001-

1007 (2006).

287

180 Bièche, I. et al. Overexpression of the stathmin gene in a subset of human

breast cancer. Br J Cancer 78, 701-709 (1998).

181 Alli, E., Bash-Babula, J., Yang, J.-M. & Hait, W. N. Effect of stathmin on the

sensitivity to antimicrotubule drugs in human breast cancer. Cancer Res 62,

6864-6869 (2002).

182 Takahashi, T. et al. Cyclin A-associated kinase activity is needed for

paclitaxel sensitivity. Mol Cancer Ther 4, 1039-1046 (2005).

183 Anand, S., Penrhyn-Lowe, S. & Venkitaraman, A. R. AURORA-A

amplification overrides the mitotic spindle assembly checkpoint, inducing

resistance to Taxol. Cancer Cell 3, 51-62 (2003).

184 Nakayama, S. et al. Prediction of paclitaxel sensitivity by CDK1 and CDK2

activity in human breast cancer cells. Breast Cancer Res 11, R12 (2009).

185 Bavetsias, V. & Linardopoulos, S. Aurora kinase inhibitors: current status

and outlook. Front Oncol 5, 278 (2015).

186 Scharer, C. D. et al. Aurora kinase inhibitors synergize with paclitaxel to

induce apoptosis in ovarian cancer cells. J Transl Med 6, 79 (2008).

187 Scharer, C. D. et al. Paclitaxel and CYC3, an aurora kinase A inhibitor,

synergise in pancreatic cancer cells but not bone marrow precursor cells.

Br J Cancer 107, 1692-1701 (2012).

188 Bush, T. L. et al. AMG 900, a small-molecule inhibitor of aurora kinases,

potentiates the activity of microtubule-targeting agents in human metastatic

breast cancer models. Mol Cancer Ther 12, 2356-2366, doi:10.1158/1535-

7163.MCT-12-1178 (2013).

288

189 Nielsen, D., Maare, C. & Skovsgaard, T. Cellular resistance to

anthracyclines. Gen Pharmacol 27, 251-255 (1996).

190 Lage, H. & Dietel, M. Involvement of the DNA mismatch repair system in

antineoplastic drug resistance. J Cancer Res Clin Oncol 125, 156-165

(1999).

191 Fedier, A. et al. Resistance to topoisomerase poisons due to loss of DNA

mismatch repair. Int J Cancer 93, 571-576 (2001).

192 Doroshow, J. H. Prevention of doxorubicin-induced killing of MCF-7 human

breast cancer cells by oxygen radical scavengers and iron chelating agents.

Biochem Biophys Res Commun 135, 330-335 (1986).

193 Harbottle, A., Daly, A. K., Atherton, K. & Campbell, F. C. Role of glutathione

S-transferase P1, P-glycoprotein and multidrug resistance-associated

protein 1 in acquired doxorubicin resistance. Int J Cancer 92, 777-783

(2001).

194 Maston, G. A., Evans, S. K. & Green, M. R. Transcriptional regulatory

elements in the human genome. Annu Rev Genomics Hum Genet 7, 29-59

(2006).

195 Yang, C., Bolotin, E., Jiang, T., Sladek, F. M. & Martinez, E. Prevalence of

the initiator over the TATA box in human and yeast genes and identification

of DNA motifs enriched in human TATA-less core promoters. Gene 389, 52-

65 (2007).

289

196 Butler, J. E. F. & Kadonaga, J. T. The RNA polymerase II core promoter: a

key component in the regulation of gene expression. Genes Dev 16, 2583-

2592 (2002).

197 Venter, J. C. et al. The sequence of the human genome. Science 291, 1304-

1351 (2001).

198 Bird, A. P. CpG islands as gene markers in the vertebrate nucleus. Trends

Genet 3, 342–347 (1987).

199 Kron, K. J., Bailey, S. D. & Lupien, M. Enhancer alterations in cancer: a

source for a cell identity crisis. Genome Med 6, 77 (2014).

200 Vilar, J. M. G. & Saiz, L. DNA looping in gene regulation: from the assembly

of macromolecular complexes to the control of transcriptional noise. Curr

Opin Genet Dev 15, 136-144 (2005).

201 Kagey, M. H. et al. Mediator and cohesin connect gene expression and

chromatin architecture. Nature 467, 430-435 (2010).

202 Heinz, S., Romanoski, C. E., Benner, C. & Glass, C. K. The selection and

function of cell type-specific enhancers. Nat Rev Mol Cell Biol 16, 144-154

(2015).

203 Wang, A. et al. Epigenetic priming of enhancers predicts developmental

competence of hESC-derived endodermal lineage intermediates. Cell Stem

Cell 16, 386–399 (2015).

204 Buecker, C. & Wysocka, J. Enhancers as information integration hubs in

development: lessons from genomics. Trends Genet 28, 276-284 (2012).

290

205 Heintzman, N. D. et al. Distinct and predictive chromatin signatures of

transcriptional promoters and enhancers in the human genome. Nat Genet

39, 311-318 (2007).

206 Creyghton, M. P. et al. Histone H3K27ac separates active from poised

enhancers and predicts developmental state. Proc Natl Acad Sci USA 107,

21931-21936 (2010).

207 Zentner, G. E., Tesar, P. J. & Scacheri, P. C. Epigenetic signatures

distinguish multiple classes of enhancers with distinct cellular functions.

Genome Res 21, 1273-1283 (2011).

208 Rada-Iglesias, A. et al. A unique chromatin signature uncovers early

developmental enhancers in humans. Nature 470, 279-283 (2011).

209 Ali, T., Renkawitz, R. & Bartkuhn, M. Insulators and domains of gene

expression. Curr Opin Genet Dev 37, 17-26 (2016).

210 Bell, A. C., West, A. G. & Felsenfeld, G. The protein CTCF is required for

the enhancer blocking activity of vertebrate insulators. Cell 98, 387-396

(1999).

211 Li, Q., Peterson, K. R., Fang, X. & Stamatoyannopoulos, G. Locus control

regions. Blood 100, 3077-3086 (2002).

212 Khatter, H., Vorländer, M. K. & Müller, C. W. RNA polymerase I and III:

similar yet unique. Curr Opin Struct Biol 47, 88-94 (2017).

213 Chapman, R. D., Heidemann, M., Hintermair, C. & Eick, D. Molecular

evolution of the RNA polymerase II CTD. Trends Genet 24, 289-296 (2008).

291

214 Zhang, J. & Corden, J. L. Identification of phosphorylation sites in the

repetitive carboxyl-terminal domain of the mouse RNA polymerase II largest

subunit. J Biol Chem 266, 2290-2296 (1991).

215 Cadena, D. L. & Dahmus, M. E. Messenger RNA synthesis in mammalian

cells is catalyzed by the phosphorylated form of RNA polymerase II. J Biol

Chem 262, 12468-12474 (1987).

216 Kang, M. E. & Dahmus, M. E. RNA polymerases IIA and IIO have distinct

roles during transcription from the TATA-less murine dihydrofolate

reductase promoter. J Biol Chem 268, 25033-25040 (1993).

217 Harlen, K. M. & Churchman, L. S. The code and beyond: transcription

regulation by the RNA polymerase II carboxy-terminal domain. Nat Rev Mol

Cell Biol 18, 263-273 (2017).

218 Schroeder, S. C., Schwer, B., Shuman, S. & Bentley, D. Dynamic

association of capping enzymes with transcribing RNA polymerase II.

Genes Dev 14, 2435-2440 (2000).

219 Komarnitsky, P., Cho, E.-J. & Buratowski, S. Different phosphorylated forms

of RNA polymerase II and associated mRNA processing factors during

transcription. Genes Dev 14, 2452-2460 (2000).

220 Phatnani, H. P. & Greenleaf, A. L. Phosphorylation and functions of the RNA

polymerase II CTD. Genes Dev 20, 2922-2936 (2006).

221 Cho, E.-J., Takagi, T., Moore, C. R. & Buratowski, S. mRNA capping

enzyme is recruited to the transcription complex by phosphorylation of the

292

RNA polymerase II carboxy-terminal domain. Genes Dev 11, 3319-3326

(1997).

222 McCracken, S. et al. 5'-Capping enzymes are targeted to pre-mRNA by

binding to the phosphorylated carboxy-terminal domain of RNA polymerase

II. Genes Dev 11, 3306-3318 (1997).

223 McCracken, S. et al. The C-terminal domain of RNA polymerase II couples

mRNA processing to transcription. Nature 385, 357-361 (1997).

224 Birse, C. E., Minvielle-Sebastia, L., Lee, B. A., Keller, W. & Proudfoot, N. J.

Coupling termination of transcription to messenger RNA maturation in

yeast. Science 280, 298-301 (1998).

225 Mortillaro, M. J. et al. A hyperphosphorylated form of the large subunit of

RNA polymerase II is associated with splicing complexes and the nuclear

matrix. Proc Natl Acad Sci USA 93, 8253-8257 (1996).

226 Yuryev, A. et al. The C-terminal domain of the largest subunit of RNA

polymerase II interacts with a novel set of serine/arginine-rich proteins. Proc

Natl Acad Sci USA 93, 6975-6980 (1996).

227 Li, J., Moazed, D. & Gygi, S. P. Association of the histone methyltransferase

Set2 with RNA polymerase II plays a role in transcription elongation. J Biol

Chem 277, 49383-49388 (2002).

228 Krogan, N. J. et al. Methylation of histone H3 by Set2 in Saccharomyces

cerevisiae is linked to transcriptional elongation by RNA polymerase II. Mol

Cell Biol 23, 4207-4218 (2003).

293

229 Zawel, L. & Reinberg, D. Common themes in assembly and function of

eukaryotic transcription complexes. Annu Rev Biochem 64, 533-561 (1995).

230 Weake, V. M. & Workman, J. L. Inducible gene expression: diverse

regulatory mechanisms. Nat Rev Genet 11, 426-437 (2010).

231 Ansari, S. A. et al. Mediator, TATA-binding protein, and RNA polymerase II

contribute to low histone occupancy at active gene promoters in yeast. J

Biol Chem 289, 14981-14995 (2014).

232 Adkins, M. W., Williams, S. K., Linger, J. & Tyler, J. K. Chromatin

disassembly from the PHO5 promoter is essential for the recruitment of the

general transcription machinery and coactivators. Mol Cell Biol 27, 6372-

6382 (2007).

233 Vicent, G. P. et al. Four enzymes cooperate to displace histone H1 during

the first minute of hormonal gene activation. Genes Dev 25, 845-862 (2011).

234 Stewart, M. D. & Wong, J. Nuclear receptor repression: regulatory

mechanisms and physiological implications. Prog Mol Biol Transl Sci 87,

235-259 (2009).

235 Chen, L. & Widom, J. Mechanism of transcriptional silencing in yeast. Cell

120, 37-48 (2005).

236 Harris, M. B., Mostecki, J. & Rothman, P. B. Repression of an interleukin-4-

responsive promoter requires cooperative BCL-6 function. J Biol Chem 280,

13114-13121 (2005).

237 Srinivasan, L. & Atchison, M. L. YY1 DNA binding and PcG recruitment

requires CtBP. Genes Dev 18, 2596-2601 (2004).

294

238 Stargell, L. A. & Struhl, K. Mechanisms of transcriptional activation in vivo:

two steps forward. Trends Genet 12, 311-315 (1996).

239 Buratowski, S., Hahn, S., Guarente, L. & Sharp, P. A. Five intermediate

complexes in transcription initiation by RNA polymerase II. Cell 56, 549-561

(1989).

240 Flores, O. et al. The small subunit of transcription factor IIF recruits RNA

polymerase II into the preinitiation complex. Proc Natl Acad Sci USA 88,

9999-10003 (1991).

241 Goodrich, J. A. & Tjian, R. Transcription factors IIE and IIH and ATP

hydrolysis direct promoter clearance by RNA polymerase II. Cell 77, 145-

156 (1994).

242 Holstege, F. C. P., van der Vliet, P. C. & Timmers, H. T. M. Opening of an

RNA polymerase II promoter occurs in two distinct steps and requires the

basal transcription factors IIE and IIH. EMBO J 15, 1666-1677 (1996).

243 Matsui, T., Segall, J., Weil, P. A. & Roeder, R. G. Multiple factors required

for accurate initiation of transcription by purified RNA polymerase II J Biol

Chem 255, 11992-11996 (1980).

244 Jeronimo, C. & Robert, F. The Mediator complex: at the nexus of RNA

polymerase II transcription. Trends Cell Biol 27, 765-783 (2017).

245 Petrenko, N., Jin, Y., Wong, K. H. & Struhl, K. Mediator undergoes a

compositional change during transcriptional activation. Mol Cell 64, 443-454

(2016).

295

246 Esnault, C. et al. Mediator-dependent recruitment of TFIIH modules in

preinitiation complex. Mol Cell 31, 337-346 (2008).

247 Kim, Y.-J., Björklund, S., Li, Y., Sayre, M. H. & Kornberg, R. D. A

multiprotein mediator of transcriptional activation and its interaction with the

C-terminal repeat domain of RNA polymerase II. Cell 77, 599-608 (1994).

248 Saunders, A., Core, L. J. & Lis, J. T. Breaking barriers to transcription

elongation. Nat Rev Mol Cell Biol 7, 557-567 (2006).

249 Marshall, N. F., Peng, J., Xie, Z. & Price, D. H. Control of RNA polymerase

II elongation potential by a novel carboxyl-terminal domain kinase. J Biol

Chem 271, 27176-27183 (1996).

250 Marshall, N. F. & Price, D. H. Purification of P-TEFb, a transcription factor

required for the transition into productive elongation. J Biol Chem 270,

12335-12338 (1995).

251 Peng, J., Marshall, N. F. & Price, D. H. Identification of a cyclin subunit

required for the function of Drosophila P-TEFb. J Biol Chem 273, 13855-

13860 (1998).

252 Eberhardy, S. R. & Farnham, P. J. Myc recruits P-TEFb to mediate the final

step in the transcriptional activation of the cad promoter. J Biol Chem 277,

40156-40162 (2002).

253 Wittmann, B. M., Fujinaga, K., Deng, H., Ogba, N. & Montano, M. M. The

breast cell growth inhibitor, estrogen down regulated gene 1, modulates a

novel functional interaction between estrogen receptor alpha and

transcriptional elongation factor cyclin T1. Oncogene 24, 5576-5588 (2005).

296

254 Nowak, D. E. et al. RelA Ser276 phosphorylation is required for activation

of a subset of NF-kappaB-dependent genes by recruiting cyclin-dependent

kinase 9/cyclin T1 complexes. Mol Cell Biol 28, 3623-3638 (2008).

255 Jang, M. K. et al. The bromodomain protein Brd4 is a positive regulatory

component of P-TEFb and stimulates RNA polymerase II-dependent

transcription. Mol Cell 19, 523-534, doi:10.1016/j.molcel.2005.06.027

(2005).

256 Luo, Z., Lin, C. & Shilatifard, A. The super elongation complex (SEC) family

in transcriptional control. Nat Rev Mol Cell Biol 13, 543-547 (2012).

257 Jonkers, I. & Lis, J. T. Getting up to speed with transcription elongation by

RNA polymerase II. Nat Rev Mol Cell Biol 16, 167-177 (2015).

258 Flanagan, J. F. et al. Double chromodomains cooperate to recognize the

methylated histone H3 tail. Nature 438, 1181-1185 (2005).

259 Li, H. et al. Molecular basis for site-specific read-out of histone H3K4me3

by the BPTF PHD finger of NURF. Nature 442, 91-95 (2006).

260 Zentner, G. E. & Henikoff, S. Regulation of nucleosome dynamics by

histone modifications. Nat Struct Mol Biol 20, 259-266 (2013).

261 Saha, A., Wittmeyer, J. & Cairns, B. R. Chromatin remodelling: the industrial

revolution of DNA around histones. Nat Rev Mol Cell Biol 7, 437-447 (2006).

262 Erdel, F., Krug, J., Längst, G. & Rippe, K. Targeting chromatin remodelers:

signals and search mechanisms. Biochim Biophys Acta 1809, 497-508

(2011).

297

263 Chereji, R. V. & Morozov, A. V. Functional roles of nucleosome stability and

dynamics. Brief Funct Genomics 14, 50-60 (2015).

264 Clapier, C. R. & Cairns, B. R. The biology of chromatin remodeling

complexes. Annu Rev Biochem 78, 273-304 (2009).

265 Krasnov, A. N., Mazina, M. Y., Nikolenko, J. V. & Vorobyeva, N. E. On the

way of revealing coactivator complexes cross-talk during transcriptional

activation. Cell Biosci 6, 15 (2016).

266 Roeder, R. G. The role of general initiation factors in transcription by RNA

polymerase II. Trends Biochem Sci 21, 327-335 (1996).

267 Lu, H., Flores, O., Weinmann, R. & Reinberg, D. The nonphosphorylated

form of RNA polymerase II preferentially associates with the preinitiation

complex. Proc Natl Acad Sci USA 88, 10004-10008 (1991).

268 Laybourn, P. J. & Dahmus, M. E. Phosphorylation of RNA polymerase IIA

occurs subsequent to interaction with the promoter and before the initiation

of transcription. J Biol Chem 265, 13165-13173 (1990).

269 Vermeulen, M. et al. Selective anchoring of TFIID to nucleosomes by

trimethylation of histone H3 lysine 4. Cell 131, 58-69 (2007).

270 Jacobson, R. H., Ladurner, A. G., King, D. S. & Tjian, R. Structure and

function of a human TAFII250 double bromodomain module. Science 288,

1422-1425 (2000).

271 Orphanides, G., Lagrange, T. & Reinberg, D. The general transcription

factors of RNA polymerase II. Genes Dev 10, 2657-2683 (1996).

298

272 Søgaard, T. M. M. & Svejstrup, J. Q. Hyperphosphorylation of the C-

terminal repeat domain of RNA polymerase II facilitates dissociation of its

complex with mediator. J Biol Chem 282, 14113-14120 (2007).

273 Hengartner, C. J. et al. Temporal regulation of RNA polymerase II by Srb10

and Kin28 cyclin-dependent kinases. Mol Cell 2, 43-53 (1998).

274 Jeronimo, C. & Robert, F. Kin28 regulates the transient association of

Mediator with core promoters. Nat Struct Mol Biol 21, 449-455 (2014).

275 Wong, K. H., Jin, Y. & Struhl, K. TFIIH phosphorylation of the Pol II CTD

stimulates mediator dissociation from the preinitiation complex and

promoter escape. Mol Cell 54, 601-612 (2014).

276 Wada, T. et al. DSIF, a novel transcription elongation factor that regulates

RNA polymerase II processivity, is composed of human Spt4 and Spt5

homologs. Genes Dev 12, 343-356 (1998).

277 Yamaguchi, Y. et al. NELF, a multisubunit complex containing RD,

cooperates with DSIF to repress RNA polymerase II elongation. Cell 97, 41-

51 (1999).

278 Wu, C. H. et al. NELF and DSIF cause promoter proximal pausing on the

hsp70 promoter in Drosophila. Genes Dev 17, 1402-1414 (2003).

279 Ni, Z. et al. P-TEFb is critical for the maturation of RNA polymerase II into

productive elongation in vivo. Mol Cell Biol 28, 1161-1170 (2008).

280 Yamada, T. et al. P-TEFb-mediated phosphorylation of hSpt5 C-terminal

repeats is critical for processive transcription elongation. Mol Cell 21, 227-

237 (2006).

299

281 Fujinaga, K. et al. Dynamics of human immunodeficiency virus transcription:

P-TEFb phosphorylates RD and dissociates negative effectors from the

transactivation response element. Mol Cell Biol 24, 787-795 (2004).

282 Perales, R. & Bentley, D. "Cotranscriptionality": the transcription elongation

complex as a nexus for nuclear transactions. Mol Cell 36, 178-191 (2009).

283 Colgan, D. F. & Manley, J. L. Mechanism and regulation of mRNA

polyadenylation. Genes Dev 11, 2755-2766 (1997).

284 West, S., Gromak, N. & Proudfoot, N. J. Human 5' --> 3' exonuclease Xrn2

promotes transcription termination at co-transcriptional cleavage sites.

Nature 432, 522-525 (2004).

285 Kim, M. et al. The yeast Rat1 exonuclease promotes transcription

termination by RNA polymerase II. Nature 432, 517-522 (2004).

286 Rosonina, E., Kaneko, S. & Manley, J. L. Terminating the transcript:

breaking up is hard to do. Genes Dev 20, 1050-1056 (2006).

287 Kornberg, R. D. Structure of chromatin. Annu Rev Biochem 46, 931-954

(1977).

288 McGhee, J. D. & Felsenfeld, G. Nucleosome structure. Annu Rev Biochem

49, 1115-1156 (1980).

289 Radman-Livaja, M. & Rando, O. J. Nucleosome positioning: how is it

established, and why does it matter? Dev Biol 339, 258-266 (2010).

290 Kunkel, G. R. & Martinson, H. G. Nucleosomes will not form on double-

stranded RNa or over poly(dA).poly(dT) tracts in recombinant DNA. Nucleic

Acids Res 9, 6869-6888 (1981).

300

291 Nelson, H. C. M., Finch, J. T., F., L. B. & Klug, A. The structure of an

oligo(dA).oligo(dT) tract and its biological implications. Nature 330, 221-226

(1987).

292 Simpson, R. T. & Künzler, P. Cromatin and core particles formed from the

inner histones and synthetic polydeoxyribonucleotides of defined sequence.

Nucleic Acids Res 6, 1387-1415 (1979).

293 Takai, D. & Jones, P. A. Comprehensive analysis of CpG islands in human

chromosomes 21 and 22. Proc Natl Acad Sci USA 99, 3740-3745 (2002).

294 Sharma, S., Kelly, T. K. & Jones, P. A. Epigenetics in cancer.

Carcinogenesis 31, 27-36, doi:10.1093/carcin/bgp220 (2010).

295 Bird, A. DNA methylation patterns and epigenetic memory. Genes Dev 16,

6-21 (2002).

296 Prendergast, G. C. & Ziff, E. B. Methylation-sensitive sequence-specific

DNA binding by the c-Myc basic region. Science 251, 186-189 (1991).

297 Watt, F. & Molloy, P. L. Cytosine methylation prevents binding to DNA of a

HeLa cell transcription factor required for optimal expression of the

adenovirus major late promoter. Genes Dev 2, 1136-1143 (1988).

298 Nan, X. et al. Transcriptional repression by the methyl-CpG-binding protein

MeCP2 involves a histone deacetylase complex. Nature 393, 386-389

(1998).

299 Jones, P. L. et al. Methylated DNA and MeCP2 recruit histone deacetylase

to repress transcription. Nat Genet 19, 187-191 (1998).

301

300 Rountree, M. R., Bachman, K. E. & Baylin, S. B. DNMT1 binds HDAC2 and

a new co-repressor, DMAP1, to form a complex at replication foci. Nat

Genet 25, 269-277 (2000).

301 Wade, P. A. et al. Mi-2 complex couples DNA methylation to chromatin

remodelling and histone deacetylation. Nat Genet 23, 62-66 (1999).

302 Pradhan, S., Bacolla, A., Wells, R. D. & Roberts, R. J. Recombinant human

DNA (cytosine-5) methyltransferase. I. Expression, purification, and

comparison of de novo and maintenance methylation. J Biol Chem 274,

33002-33010 (1999).

303 Okano, M., Bell, D. W., Haber, D. A. & Li, E. DNA methyltransferases

Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian

development. Cell 99, 247-257 (1999).

304 Kriaucionis, S. & Heintz, N. The nuclear DNA base 5-

hydroxymethylcytosine is present in Purkinje neurons and the brain.

Science 324, 929-930 (2009).

305 Tahiliani, M. et al. Conversion of 5-methylcytosine to 5-

hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science

324, 930-935 (2009).

306 Ito, S. et al. Role of Tet proteins in 5mC to 5hmC conversion, ES-cell self-

renewal and inner cell mass specification. Nature 466, 1129-1133 (2010).

307 He, Y.-F. et al. Tet-mediated formation of 5-carboxylcytosine and its

excision by TDG in mammalian DNA. Science 333, 1303-1307 (2011).

302

308 Ito, S. et al. Tet proteins can convert 5-methylcytosine to 5-formylcytosine

and 5-carboxylcytosine. Science 333, 1300-1303 (2011).

309 Pfaffeneder, T. et al. The discovery of 5-formylcytosine in embryonic stem

cell DNA. Angew Chem Int Ed Engl 50, 7008-7012 (2011).

310 Nabel, C. S. et al. AID/APOBEC deaminases disfavor modified cytosines

implicated in DNA demethylation. Nat Chem Biol 8, 751-758 (2012).

311 Wu, H. & Zhang, Y. Reversing DNA methylation: mechanisms, genomics,

and biological functions. Cell 156, 45-68 (2014).

312 Baylin, S. B. & Jones, P. A. A decade of exploring the cancer epigenome -

biological and translational implications. Nat Rev Cancer 11, 726-734

(2011).

313 Esteller, M. Cancer epigenomics: DNA methylomes and histone-

modification maps. Nat Rev Genet 8, 286-298 (2007).

314 Feinberg, A. P. & Vogelstein, B. Hypomethylation distinguishes genes of

some human cancers from their normal counterparts. Nature 301, 89-92

(1983).

315 Greger, V., Passarge, E., Höpping, W., Messmer, E. & Horsthemke, B.

Epigenetic changes may contribute to the formation and spontaneous

regression of retinoblastoma. Hum Genet 83, 155-158 (1989).

316 Sakai, T. et al. Allele-specific hypermethylation of the retinoblastoma tumor-

suppressor gene. Am J Hum Genet 48, 880-888 (1991).

303

317 Ohtani-Fujita, N. et al. CpG methylation inactivates the promoter activity of

the human retinoblastoma tumor-suppressor gene. Oncogene 8, 1063-

1067 (1993).

318 Qu, G.-z., Grundy, P. E., Narayan, A. & Ehrlich, M. Frequent

hypomethylation in Wilms tumors of pericentromeric DNA in chromosomes

1 and 16. Cancer Genet Cytogenet 109, 34-39 (1999).

319 Eden, A., Gaudet, F., Waghmare, A. & Jaenisch, R. Chromosomal instability

and tumors promoted by DNA hypomethylation. Science 300, 455 (2003).

320 Ko, M. et al. Modulation of TET2 expression and 5-methylcytosine oxidation

by the CXXC domain protein IDAX. Nature 497, 122-126 (2013).

321 Song, S. J. et al. MicroRNA-antagonism regulates breast cancer stemness

and metastasis via TET-family-dependent chromatin remodeling. Cell 154,

311-324 (2013).

322 Song, S. J. et al. The oncogenic microRNA miR-22 targets the TET2 tumor

suppressor to promote hematopoietic stem cell self-renewal and

transformation. Cell Stem Cell 13, 87-101 (2013).

323 Ley, T. J. et al. DNMT3A mutations in acute myeloid leukemia. N Engl J

Med 363, 2424-2433 (2010).

324 Patel, J. P. et al. Prognostic relevance of integrated genetic profiling in acute

myeloid leukemia. N Engl J Med 366, 1079-1089 (2012).

325 Delhommeau, F. et al. Mutation in TET2 in myeloid cancers. N Engl J Med

360, 2289-2301 (2009).

304

326 Takiguchi, M., Achanzar, W. E., Qu, W., Li, G. & Waalkes, M. P. Effects of

cadmium on DNA-(Cytosine-5) methyltransferase activity and DNA

methylation status during cadmium-induced cellular transformation. Exp

Cell Res 286, 355-365 (2003).

327 Okoji, R. S., Yu, R. C., Maronpot, R. R. & Froines, J. R. Sodium arsenite

administration via drinking water increases genome-wide and Ha-ras DNA

hypomethylation in methyl-deficient C57BL/6J mice. Carcinogenesis 23,

777-785 (2002).

328 Rhee, I. et al. DNMT1 and DNMT3b cooperate to silence genes in human

cancer cells. Nature 416, 552-556 (2002).

329 Tsai, H. C. et al. Transient low doses of DNA-demethylating agents exert

durable antitumor effects on hematological and epithelial tumor cells.

Cancer Cell 21, 430-446 (2012).

330 Oki, Y., Jelinek, J., Shen, L., Kantarjian, H. M. & Issa, J.-P. J. Induction of

hypomethylation and molecular response after decitabine therapy in

patients with chronic myelomonocytic leukemia. Blood 111, 2382-2384

(2008).

331 Jones, P. A., Issa, J.-P. J. & Baylin, S. B. Targeting the cancer epigenome

for therapy. Nat Rev Genet 17, 630-641 (2016).

332 Stewart, D. J. et al. Decitabine effect on tumor global DNA methylation and

other parameters in a phase I trial in refractory solid tumors and lymphomas.

Clin Cancer Res 15, 3881-3888 (2009).

305

333 Prébet, T. et al. Outcome of high-risk myelodysplastic syndrome after

azacitidine treatment failure. J Clin Oncol 29, 3322-3327 (2011).

334 Issa, J. J. et al. Safety and tolerability of guadecitabine (SGI-110) in patients

with myelodysplastic syndrome and acute myeloid leukaemia: a

multicentre, randomised, dose-escalation phase 1 study. Lancet Oncol 16,

1099-1110 (2015).

335 Jiang, C. & Pugh, B. F. Nucleosome positioning and gene regulation:

advances through genomics. Nat Rev Genet 10, 161-172 (2009).

336 Yuan, G.-C. et al. Genome-scale identification of nucleosome positions in

S. cerevisiae. Science 309, 626-630 (2005).

337 Shivaswamy, S. et al. Dynamic remodeling of individual nucleosomes

across a eukaryotic genome in response to transcriptional perturbation.

PLoS Biol 6, e65 (2008).

338 Lin, J. C. et al. Role of nucleosomal occupancy in the epigenetic silencing

of the MLH1 CpG island. Cancer Cell 12, 432-444 (2007).

339 Schones, D. E. et al. Dynamic regulation of nucleosome positioning in the

human genome. Cell 132, 887-898 (2008).

340 Ahmad, K. & Henikoff, S. The histone variant H3.3 marks active chromatin

by replication-independent nucleosome assembly. Mol Cell 9, 1191-1200

(2002).

341 Wu, R. S., Tsai, S. & Bonner, W. M. Patterns of histone variant synthesis

can distinguish G0 from G1 cells. Cell 31, 367-374 (1982).

306

342 Sarma, K. & Reinberg, D. Histone variants meet their match. Nat Rev Mol

Cell Biol 6, 139-149 (2005).

343 Palmer, D. K., O'Day, K., Wener, M. H., Andrews, B. S. & Margolis, R. L. A

17-kD centromere protein (CENP-A) copurifies with nucleosome core

particles and with histones. J Cell Biol 104, 805-815 (1987).

344 Régnier, V. et al. CENP-A is required for accurate chromosome segregation

and sustained kinetochore association of BubR1. Mol Cell Biol 25, 3967-

3981 (2005).

345 Santisteban, M. S., Kalashnikova, T. & Smith, M. M. Histone H2A.Z

regulates transcription and is partially redundant with nucleosome

remodeling complexes. Cell 103, 411-422 (2000).

346 Sarcinella, E., Zuzarte, P. C., Lau, P. N. I., Draker, R. & Cheung, P.

Monoubiquitylation of H2A.Z distinguishes its association with euchromatin

or facultative heterochromatin. Mol Cell Biol 27, 6457-6468 (2007).

347 Venkatesh, S. & Workman, J. L. Histone exchange, chromatin structure and

the regulation of transcription. Nat Rev Mol Cell Biol 16, 178-189 (2015).

348 Svotelis, A., Gévry, N. & Gaudreau, L. Regulation of gene expression and

cellular proliferation by histone H2A.Z. Biochem Cell Biol 87, 179-188

(2009).

349 Witcher, M. & Emerson, B. M. Epigenetic silencing of the p16(INK4a) tumor

suppressor is associated with loss of CTCF binding and a chromatin

boundary. Mol Cell 34, 271-284 (2009).

307

350 Schwartzentruber, J. et al. Driver mutations in histone H3.3 and chromatin

remodelling genes in paediatric glioblastoma. Nature 482, 226-231 (2012).

351 Wu, G. et al. Somatic histone H3 alterations in pediatric diffuse intrinsic

pontine gliomas and non-brainstem glioblastomas. Nat Genet 44, 251-253

(2012).

352 Jiao, Y. et al. DAXX/ATRX, MEN1, and mTOR pathway genes are

frequently altered in pancreatic neuroendocrine tumors. Science 331, 1199-

1203 (2011).

353 Reisman, D., Glaros, S. & Thompson, E. A. The SWI/SNF complex and

cancer. Oncogene 28, 1653-1668 (2009).

354 Glaros, S. et al. The reversible epigenetic silencing of BRM: implications for

clinical targeted therapy. Oncogene 26, 7058-7066 (2007).

355 Kouzarides, T. Chromatin modifications and their function. Cell 128, 693-

705 (2007).

356 Hebbes, T. R., Thorne, A. W. & Crane-Robinson, C. A direct link between

core histone acetylation and transcriptionally active chromatin. EMBO J 7,

1395-1402 (1988).

357 Jenuwein, T. & Allis, C. D. Translating the histone code. Science 293, 1074-

1080 (2001).

358 Simó-Riudalbas, L. & Esteller, M. Targeting the histone orthography of

cancer: drugs for writers, erasers and readers. Br J Pharmacol 172, 2716-

2732 (2015).

308

359 Dawson, M. A. & Kouzarides, T. Cancer epigenetics: from mechanism to

therapy. Cell 150, 12-27 (2012).

360 Kurdistani, S. K. Histone modifications in cancer biology and prognosis.

Prog Drug Res 67, 91-106 (2011).

361 Seligson, D. B. et al. Global histone modification patterns predict risk of

prostate cancer recurrence. Nature 435, 1262-1266 (2005).

362 Manuyakorn, A. et al. Cellular histone modification patterns predict

prognosis and treatment response in resectable pancreatic

adenocarcinoma: results from RTOG 9704. J Clin Oncol 28, 1358-1365

(2010).

363 Wei, Y. et al. Loss of trimethylation at lysine 27 of histone H3 is a predictor

of poor outcome in breast, ovarian, and pancreatic cancers. Mol Carcinog

47, 701-706 (2008).

364 Fraga, M. F. et al. Loss of acetylation at Lys16 and trimethylation at Lys20

of histone H4 is a common hallmark of human cancer. Nat Genet 37, 391-

400 (2005).

365 Nguyen, C. T. et al. Histone H3-lysine 9 methylation is associated with

aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-2'-

deoxycytidine. Cancer Res 62, 6456-6461 (2002).

366 Lovén, J. et al. Selective inhibition of tumor oncogenes by disruption of

super-enhancers. Cell 153, 320-334, doi:10.1016/j.cell.2013.03.036 (2013).

367 Weinstein, J. N. et al. The Cancer Genome Atlas Pan-Cancer analysis

project. Nat Genet 45, 1113-1120 (2013).

309

368 Kretzmer, H. et al. DNA methylome analysis in Burkitt and follicular

lymphomas identifies differentially methylated regions linked to somatic

mutation and transcriptional control. Nat Genet 47, 1316-1325 (2015).

369 You, J. S. & Jones, P. A. Cancer genetics and epigenetics: two sides of the

same coin? Cancer Cell 22, 9-20 (2012).

370 Halkidou, K. et al. Upregulation and nuclear recruitment of HDAC1 in

hormone refractory prostate cancer. Prostate 59, 177-189 (2004).

371 Song, J. et al. Increased expression of histone deacetylase 2 is found in

human gastric cancer. APMIS 113, 264-268 (2005).

372 Kondo, Y. et al. Alterations of DNA methylation and histone modifications

contribute to gene silencing in hepatocellular carcinomas. Hepatol Res 37,

974-983 (2007).

373 Marks, P. A. et al. Histone deacetylases and cancer: causes and therapies.

Nat Rev Cancer 1, 194-202 (2001).

374 Halsall, J. A. & Turner, B. M. Histone deacetylase inhibitors for cancer

therapy: An evolutionarily ancient resistance response may explain their

limited success. Bioessays 38, 1102-1110 (2016).

375 Cameron, E. E., Bachman, K. E., Myöhänen, S., Herman, J. G. & Baylin, S.

B. Synergy of demethylation and histone deacetylase inhibition in the re-

expression of genes silenced in cancer. Nat Genet 21, 103-107 (1999).

376 Yamashita, K. et al. Pharmacologic unmasking of epigenetically silenced

tumor suppressor genes in esophageal squamous cell carcinoma. Cancer

Cell 2, 485-495 (2002).

310

377 Fiskus, W. et al. Highly active combination of BRD4 antagonist and histone

deacetylase inhibitor against human acute myelogenous leukemia cells.

Mol Cancer Ther 13, 1142-1154, doi:10.1158/1535-7163.MCT-13-0770

(2014).

378 Heinemann, A. et al. Combining BET and HDAC inhibitors synergistically

induces apoptosis of melanoma and suppresses AKT and YAP signaling.

Oncotarget 6, 21507-21521 (2015).

379 Bhadury, J. et al. BET and HDAC inhibitors induce similar genes and

biological effects and synergize to kill in Myc-induced murine lymphoma.

Proc Natl Acad Sci USA 111, E2721-2730 (2014).

380 Loosveld, M. et al. Therapeutic targeting of c-Myc in T-cell acute

lymphoblastic leukemia, T-ALL. Oncotarget 5, 3168-3172 (2014).

381 Mazur, P. K. et al. Combined inhibition of BET family proteins and histone

deacetylases as a potential epigenetics-based therapy for pancreatic ductal

adenocarcinoma. Nat Med 21, 1163-1171, doi:10.1038/nm.3952 (2015).

382 Stronach, E. A. et al. HDAC4-regulated STAT1 activation mediates

platinum resistance in ovarian cancer. Cancer Res 71, 4412-4422 (2011).

383 Stuhlmiller, T. J. et al. Inhibition of lapatinib-induced kinome reprogramming

in ERBB2-positive breast cancer by targeting BET family bromodomains.

Cell Rep 11, 390-404, doi:10.1016/j.celrep.2015.03.037 (2015).

384 Mujtaba, S., Zeng, L. & Zhou, M. M. Structure and acetyl-lysine recognition

of the bromodomain. Oncogene 26, 5521-5527 (2007).

311

385 Dhalluin, C. et al. Structure and ligand of a histone acetyltransferase

bromodomain. Nature 399, 491-496 (1999).

386 Fujisawa, T. & Filippakopoulos, P. Functions of bromodomain-containing

proteins and their roles in homeostasis and cancer. Nat Rev Mol Cell Biol

18, 246-262 (2017).

387 Fukazawa, H. & Masumi, A. The conserved 12-amino acid stretch in the

inter-bromodomain region of BET family proteins functions as a nuclear

localization signal. Biol Pharm Bull 35, 2064-2068 (2012).

388 Platt, G. M., Simpson, G. R., Mittnacht, S. & Schulz, T. F. Latent nuclear

antigen of Kaposi’s sarcoma-associated herpesvirus interacts with RING3,

a homolog of the Drosophila female sterile homeotic (fsh) gene. J Virol 73,

9789-9795 (1999).

389 Denis, G. V., Vaziri, C., Guo, N. & Faller, D. V. RING3 kinase transactivates

promoters of cell cycle regulatory genes through E2F. Cell Growth Differ 11,

417-424 (2000).

390 Bisgrove, D. A., Mahmoudi, T., Henklein, P. & Verdin, E. Conserved P-

TEFb-interacting domain of BRD4 inhibits HIV transcription. Proc Natl Acad

Sci USA 104, 13690–13695 (2007).

391 Floyd, S. R. et al. The bromodomain protein Brd4 insulates chromatin from

DNA damage signalling. Nature 498, 246-250 (2013).

392 Alsarraj, J. et al. Deletion of the proline-rich region of the murine metastasis

susceptibility gene Brd4 promotes epithelial-to-mesenchymal transition-

and stem cell-like conversion. Cancer Res 71, 3121-3131 (2011).

312

393 Alsarraj, J. et al. BRD4 short isoform interacts with RRP1B, SIPA1 and

components of the LINC complex at the inner face of the nuclear

membrane. PLoS One 8, e80746 (2013).

394 Crawford, N. P. S. et al. Bromodomain 4 activation predicts breast cancer

survival. Proc Natl Acad Sci USA 105, 6380-6385 (2008).

395 Crawford, N. P. S. et al. The Diasporin Pathway: a tumor progression-

related transcriptional network that predicts breast cancer survival. Clin Exp

Metastasis 25, 357-369 (2008).

396 Fernandez, P. et al. Transformation resistance in a premature aging

disorder identifies a tumor-protective function of BRD4. Cell Rep 9, 248-260

(2014).

397 Dey, A., Chitsaz, F., Abbasi, A., Misteli, T. & Ozato, K. The double

bromodomain protein Brd4 binds to acetylated chromatin during interphase

and mitosis. Proc Natl Acad Sci USA 100, 8758-8763 (2003).

398 Nagarajan, S. et al. Bromodomain protein BRD4 is required for estrogen

receptor-dependent enhancer activation and gene transcription. Cell Rep 8,

460-469 (2014).

399 Rahman, S. et al. The Brd4 extraterminal domain confers transcription

activation independent of pTEFb by recruiting multiple proteins, including

NSD3. Mol Cell Biol 31, 2641-2652 (2011).

400 Liu, W. et al. Brd4 and JMJD6-associated anti-pause enhancers in

regulation of transcriptional pause release. Cell 155, 1581-1595 (2013).

313

401 Shen, C. et al. NSD3-short Is an adaptor protein that couples BRD4 to the

CHD8 chromatin remodeler. Mol Cell 60, 847-859 (2015).

402 Fierz, B. et al. Histone H2B ubiquitylation disrupts local and higher-order

chromatin compaction. Nat Chem Biol 7, 113-119 (2011).

403 Pirngruber, J. et al. CDK9 directs H2B monoubiquitination and controls

replication-dependent histone mRNA 3'-end processing. EMBO Rep 10,

894-900 (2009).

404 Pavri, R. et al. Histone H2B monoubiquitination functions cooperatively with

FACT to regulate elongation by RNA polymerase II. Cell 125, 703-717

(2006).

405 Yang, Z. et al. Recruitment of P-TEFb for stimulation of transcriptional

elongation by the bromodomain protein Brd4. Mol Cell 19, 535-545 (2005).

406 Nagarajan, S., Benito, E., Fischer, A. & Johnsen, S. A. H4K12ac is

regulated by estrogen receptor-alpha and is associated with BRD4 function

and inducible transcription. Oncotarget 6, 7305–7317 (2015).

407 Devaiah, B. N. et al. BRD4 is an atypical kinase that phosphorylates serine2

of the RNA polymerase II carboxy-terminal domain. Proc Natl Acad Sci USA

109, 6927-6932 (2012).

408 Patel, M. C. et al. BRD4 coordinates recruitment of pause release factor P-

TEFb and the pausing complex NELF/DSIF to regulate transcription

elongation of interferon-stimulated genes. Mol Cell Biol 33, 2497-2507

(2013).

314

409 LeRoy, G., Rickards, B. & Flint, S. J. The double bromodomain proteins

Brd2 and Brd3 couple histone acetylation to transcription. Mol Cell 30, 51-

60, doi:10.1016/j.molcel.2008.01.018 (2008).

410 Dawson, M. A. et al. Inhibition of BET recruitment to chromatin as an

effective treatment for MLL-fusion leukaemia. Nature 478, 529-533,

doi:10.1038/nature10509 (2011).

411 Sahni, J. M. et al. Mitotic vulnerability in triple-negative breast cancer

associated with LIN9 is targetable with BET inhibitors. Cancer Res 77,

5395–5408 (2017).

412 Crowley, T. E., Kaine, E. M., Yoshida, M., Nandi, A. & Wolgemuth, D. J.

Reproductive cycle regulation of nuclear import, euchromatic localization,

and association with components of Pol II mediator of a mammalian double-

bromodomain protein. Mol Endocrinol 16, 1727-1737 (2002).

413 Denis, G. V. et al. Identification of transcription complexes that contain the

double bromodomain protein Brd2 and chromatin remodeling machines. J

Proteome Res 5, 502-511 (2006).

414 You, J. et al. Regulation of Aurora B expression by the bromodomain protein

Brd4. Mol Cell Biol 29, 5094-5103, doi:10.1128/MCB.00299-09 (2009).

415 Kanno, T. et al. Selective recognition of acetylated histones by

bromodomain proteins visualized in living cells. Mol Cell 13, 33-43 (2004).

416 Dey, A., Nishiyama, A., Karpova, T., McNally, J. & Ozato, K. Brd4 marks

select genes on mitotic chromatin and directs postmitotic transcription. Mol

Biol Cell 20, 4899-4909 (2009).

315

417 Yang, Z., He, N. & Zhou, Q. Brd4 recruits P-TEFb to chromosomes at late

mitosis to promote G1 gene expression and cell cycle progression. Mol Cell

Biol 28, 967-976 (2008).

418 Mochizuki, K. et al. The bromodomain protein Brd4 stimulates G1 gene

transcription and promotes progression to S phase. J Biol Chem 283, 9040–

9048 (2008).

419 Farina, A. et al. Bromodomain protein Brd4 binds to GTPase-activating

SPA-1, modulating its activity and subcellular localization. Mol Cell Biol 24,

9059-9069 (2004).

420 Shu, S. et al. Response and resistance to BET bromodomain inhibitors in

triple-negative breast cancer. Nature 529, 413-417,

doi:10.1038/nature16508 (2016).

421 Vázquez, R. et al. The bromodomain inhibitor OTX015 (MK-8628) exerts

anti-tumor activity in triple-negative breast cancer models as single agent

and in combination with everolimus. Oncotarget 8, 7598-7613 (2017).

422 Andrieu, G., Tran, A. H., Strissel, K. J. & Denis, G. V. BRD4 regulates breast

cancer dissemination through Jagged1/Notch1 signaling. Cancer Res 76,

6555-6567 (2016).

423 Sengupta, S., Biarnes, M. C., Clarke, R. & Jordan, V. C. Inhibition of BET

proteins impairs estrogen-mediated growth and transcription in breast

cancers by pausing RNA polymerase advancement. Breast Cancer Res

Treat 150, 265-278 (2015).

316

424 Pérez-Peña, J. et al. In silico analysis guides selection of BET inhibitors for

triple-negative breast cancer treatment. Mol Cancer Ther 15, 1823-1833

(2016).

425 Choi, S. K. et al. JQ1, an inhibitor of the epigenetic reader BRD4,

suppresses the bidirectional MYC-AP4 axis via multiple mechanisms. Oncol

Rep 35, 1186-1194 (2016).

426 Nieto-Jiménez, C. et al. Targeting basal-like breast tumors with

bromodomain and extraterminal domain (BET) and polo-like kinase

inhibitors. Oncotarget 8, 19478-19490 (2017).

427 Pinz, S. et al. Deacetylase inhibitors repress STAT5-mediated transcription

by interfering with bromodomain and extra-terminal (BET) protein function.

Nucleic Acids Res 43, 3524-3545 (2015).

428 Wang, F. et al. Brd2 disruption in mice causes severe obesity without Type

2 diabetes. Biochem J 425, 71-83 (2009).

429 Belkina, A. C., Nikolajczyk, B. S. & Denis, G. V. BET protein function is

required for inflammation: Brd2 genetic disruption and BET inhibitor JQ1

impair mouse macrophage inflammatory responses. J Immunol 190, 3670-

3678 (2013).

430 Huang, B., Yang, X.-D., Zhou, M.-M., Ozato, K. & Chen, L.-F. Brd4

coactivates transcriptional activation of NF-κB via specific binding to

acetylated RelA. Mol Cell Biol 29, 1375-1387 (2009).

431 Nicodeme, E. et al. Suppression of inflammation by a synthetic histone

mimic. Nature 468, 1119-1123, doi:10.1038/nature09589 (2010).

317

432 Ferri, E., Petosa, C. & McKenna, C. E. Bromodomains: structure, function

and pharmacology of inhibition. Biochem Pharmacol 106, 1-18 (2016).

433 Tsume, M. et al. Brd2 is required for cell cycle exit and neuronal

differentiation through the E2F1 pathway in mouse neuroepithelial cells.

Biochem Biophys Res Commun 425, 762-768 (2012).

434 Shang, E., Want, X., Wen, D., Greenberg, D. A. & Wolgemuth, D. J. Double

bromodomain-containing gene Brd2 is essential for embryonic development

in mouse. Dev Dyn 238, 908-917 (2009).

435 Di Micco, R. et al. Control of embryonic stem cell identity by BRD4-

dependent transcriptional elongation of super-enhancer-associated

pluripotency genes. Cell Rep 9, 234-247, doi:10.1016/j.celrep.2014.08.055

(2014).

436 Gonzales-Cope, M., Sidoli, S., Bhanu, N. V., Won, K.-J. & Garcia, B. A.

Histone H4 acetylation and the epigenetic reader Brd4 are critical regulators

of pluripotency in embryonic stem cells. BMC Genomics 17, 95 (2016).

437 Liu, W. et al. BRD4 regulates Nanog expression in mouse embryonic stem

cells and preimplantation embryos. Cell Death Differ 21, 1950-1960 (2014).

438 Horne, G. A. et al. Nanog requires BRD4 to maintain murine embryonic

stem cell pluripotency and is suppressed by bromodomain inhibitor JQ1

together with Lefty1. Stem Cells Dev 24, 879-891 (2015).

439 Wu, T., Pinto, H. B., Kamikawa, Y. F. & Donohoe, M. E. The BET family

member BRD4 interacts with OCT4 and regulates pluripotency gene

expression. Stem Cell Reports 4, 390-403 (2015).

318

440 Houzelstein, D. et al. Growth and early postimplantation defects in mice

deficient for the bromodomain-containing protein Brd4. Mol Cell Biol 22,

3794-3802 (2002).

441 Jones, M. H., Numata, M. & Shimane, M. Identification and characterization

of BRDT: A testis-specific gene related to the bromodomain genes RING3

and Drosophila fsh. Genomics 45, 529-534 (1997).

442 Shang, E., Nickerson, H. D., Wen, D., Wang, X. & Wolgemuth, D. J. The

first bromodomain of Brdt, a testis-specific member of the BET sub-family

of double-bromodomain-containing proteins, is essential for male germ cell

differentiation. Development 134, 3507-3515 (2007).

443 Matzuk, M. M. et al. Small-molecule inhibition of BRDT for male

contraception. Cell 150, 673-684 (2012).

444 Anand, P. et al. BET bromodomains mediate transcriptional pause release

in heart failure. Cell 154, 569-582, doi:10.1016/j.cell.2013.07.013 (2013).

445 Wang, C.-Y. & Filippakopoulos, P. Beating the odds: BETs in disease.

Trends Biochem Sci 40, 468-479 (2015).

446 Chapuy, B. et al. Discovery and characterization of super-enhancer-

associated dependencies in diffuse large B cell lymphoma. Cancer Cell 24,

777-790, doi:10.1016/j.ccr.2013.11.003 (2013).

447 Whyte, W. A. et al. Master transcription factors and mediator establish

super-enhancers at key cell identity genes. Cell 153, 307-319,

doi:10.1016/j.cell.2013.03.035 (2013).

319

448 Hnisz, D. et al. Super-enhancers in the control of cell identity and disease.

Cell 155, 934-947, doi:10.1016/j.cell.2013.09.053 (2013).

449 Pott, S. & Lieb, J. D. What are super-enhancers? Nat Genet 47, 8-12,

doi:10.1038/ng.3167 (2015).

450 Hnisz, D., Shrinivas, K., Young, R. A., Chakraborty, A. K. & Sharp, P. A. A

phase separation model for transcriptional control. Cell 169, 13-23 (2017).

451 Ciriello, G. et al. Comprehensive molecular portraits of invasive lobular

breast cancer. Cell 163, 506-519 (2015).

452 Pereira, B. et al. The somatic mutation profiles of 2,433 breast cancers

refines their genomic and transcriptomic landscapes. Nat Commun 7,

11479, doi:10.1038/ncomms11479 (2016).

453 Yang, L. et al. Repression of BET activity sensitizes homologous

recombination-proficient cancers to PARP inhibition. Sci Transl Med 9,

eaal1645 (2017).

454 Kadota, M. et al. Identification of novel gene amplifications in breast cancer

and coexistence of gene amplification with an activating mutation of

PIK3CA. Cancer Res 69, 7357–7365 (2009).

455 Shi, J. et al. Disrupting the interaction of BRD4 with diacetylated Twist

suppresses tumorigenesis in basal-like breast cancer. Cancer Cell 25, 210-

225, doi:10.1016/j.ccr.2014.01.028 (2014).

456 Borbely, G., Haldosen, L.-A., Dahlman-Wright, K. & Zhao, C. Induction of

USP17 by combining BET and HDAC inhibitors in breast cancer cells.

Oncotarget 6, 33623-33635, doi:10.18632/oncotarget.5601 (2015).

320

457 Marcotte, R. et al. Functional genomic landscape of human breast cancer

drivers, vulnerabilities, and resistance. Cell 164, 293-309 (2016).

458 Carroll, J. S. et al. Genome-wide analysis of estrogen receptor binding sites.

Nat Genet 38, 1289-1297 (2006).

459 Bhar, A. et al. Coexpression and coregulation analysis of time-series gene

expression data in estrogen-induced breast cancer cell. Algorithms Mol Biol

8, 9 (2013).

460 Filippakopoulos, P. et al. Selective inhibition of BET bromodomains. Nature

468, 1067-1073 (2010).

461 Chung, C.-w. et al. Discovery and characterization of small molecule

inhibitors of the BET family bromodomains. J Med Chem 54, 3827-3838

(2011).

462 Zhang, G. et al. Down-regulation of NF-κB transcriptional activity in HIV-

associated kidney disease by BRD4 inhibition. J Biol Chem 287, 28840-

28851 (2012).

463 Boi, M. et al. The BET bromodomain inhibitor OTX015 affects pathogenetic

pathways in preclinical B-cell tumor models and synergizes with targeted

drugs. Clin Cancer Res 21, 1628-1638 (2015).

464 Liu, Z. et al. Drug discovery targeting bromodomain-containing protein 4. J

Med Chem 60, 4533–4558 (2017).

465 Belkina, A. C. & Denis, G. V. BET domain co-regulators in obesity,

inflammation and cancer. Nat Rev Cancer 12, 465-477 (2012).

321

466 Bihani, T. et al. Resistance to everolimus driven by epigenetic regulation of

MYC in ER+ breast cancers. Oncotarget 6, 2407-2420 (2015).

467 Pérez-Salvia, M. et al. Bromodomain inhibition shows antitumoral activity in

mice and human luminal breast cancer. Oncotarget 8, 51621-51629 (2017).

468 Feng, Q. et al. An epigenomic approach to therapy for tamoxifen-resistant

breast cancer. Cell Res 24, 809-819, doi:10.1038/cr.2014.71 (2014).

469 Eeckhoute, J. et al. Positive cross-regulatory loop ties GATA-3 to estrogen

receptor α expression in breast cancer. Cancer Res 67, 6477-6483 (2007).

470 Guy, C. T., Cardiff, R. D. & Muller, W. J. Induction of mammary tumors by

expression of polyomavirus middle T oncogene: a transgenic mouse model

for metastatic disease. Mol Cell Biol 12, 954–961 (1992).

471 Lin, E. Y. et al. Progression to malignancy in the polyoma middle T

oncoprotein mouse breast cancer model provides a reliable model for

human diseases. Am J Pathol 163, 2113-2126 (2003).

472 Sahni, J. M. et al. Bromodomain and extraterminal protein inhibition blocks

growth of triple-negative breast cancers through the suppression of Aurora

kinases. J Biol Chem 291, 23756-23768, doi:10.1074/jbc.M116.738666

(2016).

473 da Motta, L. L. et al. The BET inhibitor JQ1 selectively impairs tumour

response to hypoxia and downregulates CA9 and angiogenesis in triple

negative breast cancer. Oncogene 36, 122-132, doi:10.1038/onc.2016.184

(2016).

322

474 Neve, R. M. et al. A collection of breast cancer cell lines for the study of

functionally distinct cancer subtypes. Cancer Cell 10, 515-527,

doi:10.1016/j.ccr.2006.10.008 (2006).

475 Charafe-Jauffret, E. et al. Gene expression profiling of breast cell lines

identifies potential new basal markers. Oncogene 25, 2273-2284,

doi:10.1038/sj.onc.1209254 (2006).

476 Vaupel, P., Höckel, M. & Mayer, A. Detection and characterization of tumor

hypoxia using pO2 histography. Antioxid Redox Signal 9, 1221-1235

(2007).

477 McIntyre, A. & Harris, A. L. Metabolic and hypoxic adaptation to anti-

angiogenic therapy: a target for induced essentiality. EMBO Mol Med 7,

368–379 (2015).

478 Gilkes, D. M. Implications of hypoxia in breast cancer metastasis to bone.

Int J Mol Sci 17, E1669 (2016).

479 Tan, E. Y. et al. The key hypoxia regulated gene CAIX is upregulated in

basal-like breast tumours and is associated with resistance to

chemotherapy. Br J Cancer 100, 405-411 (2009).

480 Swietach, P., Patiar, S., Supuran, C. T., Harris, A. L. & Vaughan-Jones, R.

D. The role of carbonic anhydrase 9 in regulating extracellular and

intracellular ph in three-dimensional tumor cell growths. J Biol Chem 284,

20299-20310 (2009).

323

481 Chiche, J. et al. Hypoxia-inducible carbonic anhydrase IX and XII promote

tumor cell growth by counteracting acidosis through the regulation of the

intracellular pH. Cancer Res 69, 358-368 (2009).

482 Lou, Y. et al. Targeting tumor hypoxia: suppression of breast tumor growth

and metastasis by novel carbonic anhydrase IX inhibitors. Cancer Res 71,

3364-3376 (2011).

483 Dubois, L. et al. Specific inhibition of carbonic anhydrase IX activity

enhances the in vivo therapeutic effect of tumor irradiation. Radiother Oncol

99, 424-431 (2011).

484 Gieling, R. G. et al. Inhibition of carbonic anhydrase activity modifies the

toxicity of doxorubicin and melphalan in tumour cells in vitro. J Enzyme Inhib

Med Chem 28, 360-369 (2013).

485 Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation.

Cell 144, 646-674 (2011).

486 Maxwell, P. H. et al. Hypoxia-inducible factor-1 modulates gene expression

in solid tumors and influences both angiogenesis and tumor growth. Proc

Natl Acad Sci USA 94, 8104-8109 (1997).

487 Aalders, K. C., Tryfonidis, K., Senkus, E. & Cardoso, F. Anti-angiogenic

treatment in breast cancer: Facts, successes, failures and future

perspectives. Cancer Treat Rev 53, 98-110 (2017).

488 Bid, H. K. et al. The bromodomain BET inhibitor JQ1 suppresses tumor

angiogenesis in models of childhood sarcoma. Mol Cancer Ther 15, 1018–

1028 (2016).

324

489 Jostes, S. et al. The bromodomain inhibitor JQ1 triggers growth arrest and

apoptosis in testicular germ cell tumours in vitro and in vivo. J Cell Mol Med

21, 1300-1314 (2017).

490 McIntyre, A. et al. Carbonic anhydrase IX promotes tumour growth and

necrosis in vivo and inhibition enhances anti-VEGF therapy. Clin Cancer

Res 18, 3100-3111 (2012).

491 Nilendu, P., Kumar, A., Kumar, A., Pal, J. K. & Sharma, N. K. Breast cancer

stem cells as last soldiers eluding therapeutic burn: A hard nut to crack. Int

J Cancer 142, 7-17 (2018).

492 Herrmann, H. et al. Small-molecule inhibition of BRD4 as a new potent

approach to eliminate leukemic stem- and progenitor cells in acute myeloid

leukemia (AML). Oncotarget 3, 1588–1599 (2012).

493 Cheng, Z. et al. Inhibition of BET bromodomain targets genetically diverse

glioblastoma. Clin Cancer Res 19, 1748-1759 (2013).

494 Venkataraman, S. et al. Inhibition of BRD4 attenuates tumor cell self-

renewal and suppresses stem cell signaling in MYC driven

medulloblastoma. Oncotarget 5, 2355-2371 (2014).

495 Steeg, P. S. Metastasis suppressors alter the signal transduction of cancer

cells. Nat Rev Cancer 3, 55-63 (2003).

496 Singh, M., Yelle, N., Venugopal, C. & Singh, S. K. EMT: mechanisms and

therapeutic implications. Pharmacol Ther,

doi:10.1016/j.pharmthera.2017.08.009 (2017).

325

497 Hu, Y. et al. BRD4 inhibitor inhibits colorectal cancer growth and metastasis.

Int J Mol Sci 16, 1928-1948 (2015).

498 Chang, H. et al. Synergistic action of master transcription factors controls

epithelial-to-mesenchymal transition. Nucleic Acids Res 44, 2514-2527

(2016).

499 Wang, L. et al. BRD4 inhibition suppresses cell growth, migration and

invasion of salivary adenoid cystic carcinoma. Biol Res 50, 19 (2017).

500 Zhang, P. et al. BRD4 promotes tumor growth and epithelial-mesenchymal

transition in hepatocellular carcinoma. Int J Immunopathol Pharmacol 28,

36-44 (2015).

501 Wu, X. et al. Inhibition of BRD4 suppresses cell proliferation and induces

apoptosis in renal cell carcinoma. Cell Physiol Biochem 41, 1947-1956

(2017).

502 Chen, Q. K., Lee, K., Radisky, D. C. & Nelson, C. M. Extracellular matrix

proteins regulate epithelial-mesenchymal transition in mammary epithelial

cells. Differentiation 86, 126-132 (2013).

503 Yang, H. et al. Metastasis predictive signature profiles pre-exist in normal

tissues. Clin Exp Metastasis 22, 593–603 (2005).

504 Ramaswamy, S., Ross, K. N., Lander, E. S. & Golub, T. R. A molecular

signature of metastasis in primary solid tumors. Nat Genet 33, 49-54 (2003).

505 van de Vijver, M. J. et al. A gene-expression signature as a predictor of

survival in breast cancer. N Engl J Med 347, 1999-2009 (2002).

326

506 van 't Veer, L. J. et al. Gene expression profiling predicts clinical outcome

of breast cancer. Nature 415, 530-536 (2002).

507 Yang, H. et al. Caffeine suppresses metastasis in a transgenic mouse

model: a prototype molecule for prophylaxis of metastasis. Clin Exp

Metastasis 21, 719–735 (2004).

508 Gupta, R. A. et al. Long non-coding RNA HOTAIR reprograms chromatin

state to promote cancer metastasis. Nature 464, 1071-1076 (2010).

509 Deng, J. et al. Long non-coding RNA HOTAIR regulates the proliferation,

self-renewal capacity, tumor formation and migration of the cancer stem-

like cell (CSC) subpopulation enriched from breast cancer cells. PLoS One

12, e0170860 (2017).

510 Sørensen, K. P. et al. Long non-coding RNA HOTAIR is an independent

prognostic marker of metastasis in estrogen receptor-positive primary

breast cancer. Breast Cancer Res Treat 142, 529-536 (2013).

511 Zhang, L. et al. Circulating DNA of HOTAIR in serum is a novel biomarker

for breast cancer. Breast Cancer Res Treat 152, 199-208 (2015).

512 Li, M. et al. Induction of a novel isoform of the lncRNA HOTAIR in Claudin-

low breast cancer cells attached to extracellular matrix. Mol Oncol 11, 1698-

1710, doi:10.1002/1878-0261.12133 (2017).

513 Jackstadt, R. et al. AP4 is a mediator of epithelial–mesenchymal transition

and metastasis in colorectal cancer. J Exp Med 210, 1331-1350 (2013).

514 Gong, H., Han, S., Yao, H., Zhao, H. & Wang, Y. AP-4 predicts poor

prognosis in non-small cell lung cancer. Mol Med Rep 10, 336-340 (2014).

327

515 Chen, S. & Chiu, S.-K. AP4 activates cell migration and EMT mediated by

p53 in MDA-MB-231 breast carcinoma cells. Mol Cell Biochem 407, 57-68

(2015).

516 Li, D., Masiero, M., Banham, A. H. & Harris, A. L. The notch ligand Jagged1

as a target for anti-tumor therapy. Front Oncol 4, 254 (2014).

517 Györffy, B. et al. An online survival analysis tool to rapidly assess the effect

of 22,277 genes on breast cancer prognosis using microarray data of 1,809

patients. Breast Cancer Res Treat 123, 725-731 (2010).

518 Liu, Q. et al. IL-6 promotion of glioblastoma cell invasion and angiogenesis

in U251 and T98G cell lines. J Neurooncol 100, 165-176 (2010).

519 Sullivan, N. J. et al. Interleukin-6 induces an epithelial-mesenchymal

transition phenotype in human breast cancer cells. Oncogene 28, 2940-

2947 (2009).

520 Ravishankaran, P. & Karunanithi, R. Clinical significance of preoperative

serum interleukin-6 and C-reactive protein level in breast cancer patients.

World J Surg Oncol 9, 18 (2011).

521 Yang, J. et al. Twist, a master regulator of morphogenesis, plays an

essential role in tumor metastasis. Cell 117, 927-939 (2004).

522 Asem, M. S., Buechler, S., Wates, R. B., Miller, D. L. & Stack, M. S. Wnt5a

signaling in cancer. Cancers (Basel) 8, 79 (2016).

523 Yang, Y. et al. Immunocompetent mouse allograft models for development

of therapies to target breast cancer metastasis. Oncotarget 8, 30621-30643

(2017).

328

524 Pan, D. et al. Metabolic response to XD14 treatment in human breast

cancer cell line MCF-7. Int J Mol Sci 17, 1772 (2016).

525 Bai, L. et al. Targeted degradation of BET proteins in triple-negative breast

cancer. Cancer Res 77, 2476-2487 (2017).

526 Delmore, J. E. et al. BET bromodomain inhibition as a therapeutic strategy

to target c-Myc. Cell 146, 904-917, doi:10.1016/j.cell.2011.08.017 (2011).

527 Tolani, B., Gopalakrishnan, R., Punj, V., Matta, H. & Chaudhary, P.

Targeting Myc in KSHV-associated primary effusion lymphoma with BET

bromodomain inhibitors. Oncogene 33, 2928-2937,

doi:10.1038/onc.2013.242 (2014).

528 Ditchfield, C. et al. Aurora B couples chromosome alignment with anaphase

by targeting BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 161,

267-280 (2003).

529 Dreier, M. R., Grabovich, A. Z., Katusin, J. D. & Taylor, W. R. Short and

long-term tumor cell responses to Aurora kinase inhibitors. Exp Cell Res

315, 1085-1099, doi:10.1016/j.yexcr.2009.02.008 (2009).

530 Gizatullin, F. et al. The Aurora kinase inhibitor VX-680 induces

endoreduplication and apoptosis preferentially in cells with compromised

p53-dependent postmitotic checkpoint function. Cancer Res 66, 7668-7677

(2006).

531 Carmena, M. & Earnshaw, W. C. The cellular geography of aurora kinases.

Nat Rev Mol Cell Biol 4, 842–854 (2003).

329

532 Zawistowski, J. S. et al. Enhancer remodeling during adaptive bypass to

MEK inhibition Is attenuated by pharmacologic targeting of the P-TEFb

complex. Cancer Discov 7, 302-321 (2017).

533 Shen, H. et al. Suppression of enhancer overactivation by a RACK7-histone

demethylase complex. Cell 165, 331-342 (2017).

534 Betancur, P. A. et al. A CD47-associated super-enhancer links pro-

inflammatory signalling to CD47 upregulation in breast cancer. Nat

Commun 8, 14802 (2017).

535 Janssen, A., Kops, G. J. P. L. & Medema, R. H. Elevating the frequency of

chromosome mis-segregation as a strategy to kill tumor cells. Proc Natl

Acad Sci USA 106, 19108-19113, doi:10.1073/pnas.0904343106 (2009).

536 Stratikopoulos, E. E. et al. Kinase and BET inhibitors together clamp

inhibition of PI3K signaling and overcome resistance to therapy. Cancer Cell

27, 837-851, doi:10.1016/j.ccell.2015.05.006 (2015).

537 Zeng, H. et al. Feedback activation of leukemia inhibitory factor receptor

limits response to histone deacetylase inhibitors in breast cancer. Cancer

Cell 30, 459-473 (2016).

538 Korb, E., Herre, M., Zucker-Scharff, I., Darnell, R. B. & Allis, C. D. BET

protein Brd4 activates transcription in neurons and BET inhibitor Jq1 blocks

memory in mice. Nat Neurosci 18, 1464–1473 (2016).

539 Sternlicht, M. D., Kouros-Mehr, H., Lu, P. & Werb, Z. Hormonal and local

control of mammary branching morphogenesis. Differentiation 74, 365-381

(2006).

330

540 Bolden, J. E. et al. Inducible in vivo silencing of Brd4 identifies potential

toxicities of sustained BET protein inhibition. Cell Rep 8, 1919-1929,

doi:10.1016/j.celrep.2014.08.025 (2014).

541 Stathis, A. et al. Clinical response of carcinomas harboring the BRD4-NUT

oncoprotein to the targeted bromodomain inhibitor OTX015/MK-8628.

Cancer Discov 6, 492-500 (2016).

542 Berthon, C. et al. Bromodomain inhibitor OTX015 in patients with acute

leukaemia: a dose-escalation, phase 1 study. Lancet Haematol 3, e186-195

(2016).

543 Amorim, S. et al. Bromodomain inhibitor OTX015 in patients with lymphoma

or multiple myeloma: a dose-escalation, open-label, pharmacokinetic,

phase 1 study. Lancet Haematol 3, e196-204 (2016).

544 French, C. A. et al. BRD4-NUT fusion oncogene: a novel mechanism in

aggressive carcinoma. Cancer Res 63, 304-307 (2003).

545 French, C. A. et al. BRD-NUT oncoproteins: a family of closely related

nuclear proteins that block epithelial differentiation and maintain the growth

of carcinoma cells. Oncogene 27, 2237-2242 (2008).

546 Yokoyama, Y. et al. BET inhibitors suppress ALDH activity by targeting

ALDH1A1 super-enhancer in ovarian cancer. Cancer Res 76, 6320-6330

(2016).

547 Bryant, H. E. et al. Specific killing of BRCA2-deficient tumours with inhibitors

of poly(ADP-ribose) polymerase. Nature 434, 913-917 (2005).

331

548 Gonzalez-Angulo, A. M. et al. Incidence and outcome of BRCA mutations

in unselected patients with triple receptor-negative breast cancer. Clin

Cancer Res 17, 1082–1089 (2011).

549 Eckschlager, T., Plch, J., Stiborova, M. & Hrabeta, J. Histone deacetylase

inhibitors as anticancer drugs. Int J Mol Sci 18, 1414 (2017).

550 Lamb, J. et al. The Connectivity Map: using gene-expression signatures to

connect small molecules, genes, and disease. Science 313, 1929-1935

(2006).

551 Ramakrishna, S. et al. Hyaluronan binding motifs of USP17 and SDS3

exhibit anti-tumor activity. PLoS One 7, e37772 (2012).

552 Burrows, J. F. et al. USP17 regulates Ras activation and cell proliferation

by blocking RCE1 activity. J Biol Chem 284, 9587-9595 (2009).

553 Amemiya, S., Yamaguchi, T., Hashimoto, Y. & Noguchi-Yachide, T.

Synthesis and evaluation of novel dual BRD4/HDAC inhibitors. Bioorg Med

Chem 25, 3677-3684 (2017).

554 Shao, M. et al. Structure-based design, synthesis and in vitro

antiproliferative effects studies of novel dual BRD4/HDAC inhibitors. Bioorg

Med Chem Lett S0960-894X, 30761-30768 (2017).

555 Zhang, Z. et al. Targeting epigenetic reader and eraser: Rational design,

synthesis and in vitro evaluation of dimethylisoxazoles derivatives as

BRD4/HDAC dual inhibitors. Bioorg Med Chem Lett 26, 2931-2935 (2016).

556 Ring, A. & Dowsett, M. Mechanisms of tamoxifen resistance. Endocr Relat

Cancer 11, 643-658 (2004).

332

557 Tricker, E. M. et al. Combined EGFR/MEK inhibition prevents the

emergence of resistance in EGFR-mutant lung cancer. Cancer Discov 5,

960-971 (2015).

558 Mahadevan, D. et al. A novel tyrosine kinase switch is a mechanism of

imatinib resistance in gastrointestinal stromal tumors. Oncogene 26, 3909-

3919 (2007).

559 Richardson, A. L. et al. X chromosomal abnormalities in basal-like human

breast cancer. Cancer Cell 9, 121–132 (2006).

560 Alimonti, A. et al. Subtle variations in Pten dose determine cancer

susceptibility. Nat Genet 42, 454-458 (2010).

561 Yu, K. et al. mTOR, a novel target in breast cancer: the effect of CCI-779,

an mTOR inhibitor, in preclinical models of breast cancer. Endocr Relat

Cancer 8, 249-258 (2001).

562 Sparano, J. A. et al. Weekly paclitaxel in the adjuvant treatment of breast

cancer. N Engl J Med 358, 1663-1671, doi:10.1056/NEJMoa0707056

(2008).

563 Royce, M. E. & Osman, D. Everolimus in the treatment of metastatic breast

cancer. Breast Cancer (Auckl) 9, 73-79 (2015).

564 Miller, T. W. et al. Hyperactivation of phosphatidylinositol-3 kinase promotes

escape from hormone dependence in estrogen receptor–positive human

breast cancer. J Clin Invest 120, 2406–2413 (2010).

333

565 Singla, H. et al. Recent advances in HER2 positive breast cancer

epigenetics: Susceptibility and therapeutic strategies. Eur J Med Chem

(2017).

566 Duncan, J. S. et al. Dynamic reprogramming of the kinome in response to

targeted MEK inhibition in triple-negative breast cancer. Cell 149, 307–321

(2012).

567 Neklesa, T. K., Winkler, J. D. & Crews, C. M. Targeted protein degradation

by PROTACs. Pharmacol Ther 174, 138-144 (2017).

568 Winter, G. E. et al. Phthalimide conjugation as a strategy for in vivo target

protein degradation. Science 348, 1376-1381 (2015).

569 Lu, J. et al. Hijacking the E3 ubiquitin ligase cereblon to efficiently target

BRD4. Chem Biol 22, 755-763 (2015).

570 Raina, K. et al. PROTAC-induced BET protein degradation as a therapy for

castration-resistant prostate cancer. Proc Natl Acad Sci USA 113, 7124-

7129 (2016).

571 Sun, B. et al. BET protein proteolysis targeting chimera (PROTAC) exerts

potent lethal activity against mantle cell lymphoma cells. Leukemia (2017).

572 Trusolino, L. & Bertotti, A. Compensatory pathways in oncogenic kinase

signaling and resistance to targeted therapies: six degrees of separation.

Cancer Discov 2, 876-880 (2012).

573 Haber, D. A., Gray, N. S. & Baselga, J. The evolving war on cancer. Cell

145, 19-24 (2011).

334

574 Ciceri, P. et al. Dual kinase-bromodomain inhibitors for rationally designed

polypharmacology. Nat Chem Biol 10, 305-312 (2014).

575 Allen, B. K. et al. Large-scale computational screening identifies first in class

multitarget inhibitor of EGFR kinase and BRD4. Sci Rep 5, 16924 (2015).

576 Zhang, Z. et al. BET bromodomain inhibition as a therapeutic strategy in

ovarian cancer by downregulating FoxM1. Theranostics 6, 219-230,

doi:10.7150/thno.13178 (2016).

577 Chen, L. et al. BRD4 structure-activity relationships of dual PLK1

kinase/BRD4 bromodomain inhibitor BI-2536. ACS Med Chem Lett 6, 764-

769 (2015).

578 Andrews, F. H. et al. Dual-activity PI3K–BRD4 inhibitor for the orthogonal

inhibition of MYC to block tumor growth and metastasis. Proc Natl Acad Sci

USA 114, E1072-E1080 (2017).

579 Ember, S. W. et al. Potent dual BET bromodomain-kinase inhibitors as

value-added multitargeted chemical probes and cancer therapeutics. Mol

Cancer Ther 16, 1054-1067 (2017).

580 Ember, S. W. J. et al. Acetyl-lysine binding site of bromodomain-containing

protein 4 (BRD4) interacts with diverse kinase inhibitors. ACS Chem Biol 9,

1160-1171 (2014).

581 Dittmann, A. et al. The commonly used PI3-kinase probe LY294002 is an

inhibitor of BET bromodomains. ACS Chem Biol 9, 495-502 (2014).

335

582 Christensen, C. L. et al. Targeting transcriptional addictions in small cell

lung cancer with a covalent CDK7 inhibitor. Cancer Cell 26, 909-922,

doi:10.1016/j.ccell.2014.10.019 (2014).

583 Jiang, Y.-Y. et al. Targeting super-enhancer-associated oncogenes in

oesophageal squamous cell carcinoma. Gut 66, 1358-1368 (2017).

584 Kwiatkowski, N. et al. Targeting transcription regulation in cancer with a

covalent CDK7 inhibitor. Nature 511, 616–620 (2014).

585 Chipumuro, E. et al. CDK7 inhibition suppresses super-enhancer-linked

oncogenic transcription in MYCN-driven cancer. Cell 159, 1126-1139

(2014).

586 Whittaker, S. R., Mallinger, A., Workman, P. & Clarke, P. A. Inhibitors of

cyclin-dependent kinases as cancer therapeutics. Pharmacol Ther 173, 83-

105 (2017).

587 Patel, H. et al. Expression of CDK7, Cyclin H, and MAT1 is elevated in

breast cancer and is prognostic in estrogen receptor-positive breast cancer.

Clin Cancer Res 22, 5929-5938 (2016).

588 Li, B. et al. Therapeutic rationale to target highly expressed CDK7

conferring poor outcomes in triple-negative breast cancer. Cancer Res 77,

3834-3845 (2017).

589 Ali, S. et al. The development of a selective cyclin-dependent kinase

inhibitor that shows antitumor activity. Cancer Res 69, 6208-6215 (2009).

590 Węsierska-Gądek, J., Gritsch, D., Zulehner, N., Komina, O. & Maurer, M.

Roscovitine, a selective CDK inhibitor, reduces the basal and estrogen-

336

induced phosphorylation of ER-α in human ER-positive breast cancer cells.

J Cell Biochem 112, 761-772 (2011).

591 Xie, G. e. et al. The cyclin-dependent kinase inhibitor SNS-032 induces

apoptosis in breast cancer cells via depletion of Mcl-1 and X-linked inhibitor

of apoptosis protein and displays antitumor activity in vivo. Int J Oncol 45,

804-812 (2014).

592 Kaliszczak, M. et al. Development of a cyclin-dependent kinase inhibitor

devoid of ABC transporter-dependent drug resistance. Br J Cancer 109,

2356-2367 (2013).

593 Wojciechowski, J., Horky, M., Gueorguieva, M. & Węsierska-Gądek, J.

Rapid onset of nucleolar disintegration preceding cell cycle arrest in

roscovitine-induced apoptosis of human MCF-7 breast cancer cells. Int J

Cancer 106, 486-495 (2003).

594 Nagaraja, S. et al. Transcriptional dependencies in diffuse intrinsic pontine

glioma. Cancer Cell 31, 635-652 (2017).

595 Dominguez-Brauer, C. et al. Targeting mitosis in cancer: emerging

strategies. Mol Cell 60, 524-536 (2015).

596 Massagué, J. G1 cell-cycle control and cancer. Nature 432, 298-306 (2004).

597 Peters, J.-M. & Nishiyama, T. Sister chromatid cohesion. Cold Spring Harb

Perspect Biol 4, pii: a011130 (2012).

598 Azimzadeh, J. & Bornens, M. Structure and duplication of the centrosome.

J Cell Sci 120, 2139-2142 (2007).

337

599 Willis, N. & Rhind, N. Regulation of DNA replication by the S-phase DNA

damage checkpoint. Cell Div 4, 13 (2009).

600 Couch, F. B. et al. ATR phosphorylates SMARCAL1 to prevent replication

fork collapse. Genes Dev 27, 1610-1623 (2013).

601 Ragland, R. L. et al. RNF4 and PLK1 are required for replication fork

collapse in ATR-deficient cells. Genes Dev 27, 2259-2273 (2013).

602 Lopes, M. et al. The DNA replication checkpoint response stabilizes stalled

replication forks. Nature 412, 557-561 (2001).

603 Matsuoka, S. et al. ATM and ATR substrate analysis reveals extensive

protein networks responsive to DNA damage. Science 316, 1160-1166

(2007).

604 Zou, L. & Elledge, S. J. Sensing DNA damage through ATRIP recognition

of RPA-ssDNA complexes. Science 300, 1542-1548 (2003).

605 Manchado, E., Guillamot, M. & Malumbres, M. Killing cells by targeting

mitosis. Cell Death Differ 19, 369-377 (2012).

606 Morgan, D. O. Cyclin-dependent kinases: engines, clocks, and

microprocessors. Annu Rev Cell Dev Biol 13, 261-291 (1997).

607 Fisher, R. P. & Morgan, D. O. A novel cyclin associates with MO15/CDK7

to form the CDK-activating kinase. Cell 78, 713-724 (1994).

608 Espinoza, F. H., Farrell, A., Erdjument-Bromage, H., Tempst, P. & Morgan,

D. O. A cyclin-dependent kinase-activating kinase (CAK) in budding yeast

unrelated to vertebrate CAK. Science 273, 1714-1717 (1996).

338

609 Kaldis, P., Sutton, A. & Solomon, M. J. The Cdk-activating kinase (CAK)

from budding yeast. Cell 86, 553-564 (1996).

610 Sherr, C. J. & Roberts, J. M. CDK inhibitors: positive and negative

regulators of G1-phase progression. Genes Dev 13, 1501-1512 (1999).

611 Alcorta, D. A. et al. Involvement of the cyclin-dependent kinase inhibitor p16

(INK4a) in replicative senescence of normal human fibroblasts. Proc Natl

Acad Sci USA 93, 13742-13747 (1996).

612 El-Deiry, W. S. et al. WAF1, a potential mediator of p53 tumor suppression.

Cell 75, 817-825 (1993).

613 Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J. M. & Pagano,

M. Human cyclin E, a nuclear protein essential for the G1-to-S phase

transition. Mol Cell Biol 15, 2612-2624 (1995).

614 Lipinski, M. M. & Jacks, T. The retinoblastoma gene family in differentiation

and development. Oncogene 18, 7873-7882 (1999).

615 Duronio, R. J., Brook, A., Dyson, N. & O'Farrell, P. H. E2F-induced S phase

requires cyclin E. Genes Dev 10, 2505-2513 (1996).

616 Kitagawa, M. et al. The consensus motif for phosphorylation by cyclin D1-

Cdk4 is different from that for phosphorylation by cyclin A/E-Cdk2. EMBO J

15, 7060-7069 (1996).

617 Connell-Crowley, L., Harper, J. W. & Goodrich, D. W. Cyclin D1/Cdk4

regulates retinoblastoma protein-mediated cell cycle arrest by site-specific

phosphorylation. Mol Biol Cell 8, 287-301 (1997).

339

618 Harbour, J. W., Luo, R. X., Dei Santi, A., Postigo, A. A. & Dean, D. C. Cdk

phosphorylation triggers sequential intramolecular interactions that

progressively block Rb functions as cells move through G1. Cell 98, 859-

869 (1999).

619 Lundberg, A. S. & Weinberg, R. A. Functional inactivation of the

retinoblastoma protein requires sequential modification by at least two

distinct cyclin-cdk complexes. Mol Cell Biol 18, 753-761 (1998).

620 Matsumoto, Y., Hayashi, K. & Nishida, E. Cyclin-dependent kinase 2 (Cdk2)

is required for centrosome duplication in mammalian cells. Curr Biol 9, 429-

432 (1999).

621 Coverley, D., Laman, H. & Laskey, R. A. Distinct roles for cyclins E and A

during DNA replication complex assembly and activation. Nat Cell Biol 4,

523-528 (2002).

622 Pagano, M., Pepperkok, R., Verde, F., Ansorge, W. & Draetta, G. Cyclin A

is required at two points in the human cell cycle. EMBO J 11, 961-971

(1992).

623 Wierstra, I. & Alves, J. FOXM1c is activated by cyclin E/Cdk2, cyclin

A/Cdk2, and cyclin A/Cdk1, but repressed by GSK-3α. Biochem Biophys

Res Commun 348, 99-108 (2006).

624 Toyoshima-Morimoto, F., Taniguchi, E., Shinya, N., Iwamatsu, A. & Nishida,

E. Polo-like kinase 1 phosphorylates cyclin B1 and targets it to the nucleus

during prophase. Nature 410, 215-220 (2001).

340

625 Yuan, J. et al. Cooperative phosphorylation including the activity of polo-like

kinase 1 regulates the subcellular localization of cyclin B1. Oncogene 21,

8282-8292 (2002).

626 Jackman, M., Lindon, C., Nigg, E. A. & Pines, J. Active cyclin B1-Cdk1 first

appears on centrosomes in prophase. Nat Cell Biol 5, 143-148 (2003).

627 Larochelle, S. et al. Requirements for Cdk7 in the assembly of Cdk1/cyclin

B and activation of Cdk2 revealed by chemical genetics in human cells. Mol

Cell 25, 839-850 (2007).

628 Larochelle, S., Pandur, J., Fisher, R. P., Salz, H. K. & Suter, B. Cdk7 is

essential for mitosis and for in vivo Cdk-activating kinase activity. Genes

Dev 12, 370-381 (1998).

629 Liu, F., Stanton, J. J., Wu, Z. & Piwnica-Worms, H. The human Myt1 kinase

preferentially phosphorylates Cdc2 on threonine 14 and localizes to the

endoplasmic reticulum and Golgi complex. Mol Cell Biol 17, 571-583 (1997).

630 Parker, L. L., Atherton-Fessler, S. & Piwnica-Worms, H. p107wee1 is a

dual-specificity kinase that phosphorylates p34cdc2 on tyrosine 15. Proc

Natl Acad Sci USA 89, 2917-2921 (1992).

631 Atherton-Fessler, S. et al. Cell cycle regulation of the p34cdc2 inhibitory

kinases. Mol Biol Cell 5, 989-1001 (1994).

632 Qian, Y. W., Erikson, E., Li, C. & Maller, J. L. Activated polo-like kinase Plx1

is required at multiple points during mitosis in Xenopus laevis. Mol Cell Biol

18, 4262-4271 (1998).

341

633 Toyoshima-Morimoto, F., Taniguchi, E. & Nishida, E. Plk1 promotes nuclear

translocation of human Cdc25C during prophase. EMBO Rep 3, 341-348

(2002).

634 Strausfeld, U. et al. Dephosphorylation and activation of a p34cdc2/cyclin B

complex in vitro by human CDC25 protein. Nature 351, 242-245 (1991).

635 Enserink, J. M. & Kolodner, R. D. An overview of Cdk1-controlled targets

and processes. Cell Div 5, 11 (2010).

636 Castro, A., Bernis, C., Vigneron, S., Labbé, J.-C. & Lorca, T. The anaphase-

promoting complex: a key factor in the regulation of cell cycle. Oncogene

24, 314-325 (2005).

637 Serrano, M., Hannon, G. J. & Beach, D. A new regulatory motif in cell-cycle

control causing specific inhibition of cyclin D/CDK4. Nature 366, 704-707

(1993).

638 Kamb, A. et al. A cell cycle regulator potentially involved in genesis of many

tumor types. Science 264, 436-440 (1994).

639 Nobori, T. et al. Deletions of the cyclin-dependent kinase-4 inhibitor gene in

multiple human cancers. Nature 368, 753-756 (1994).

640 Porter, P. L. et al. Expression of cell-cycle regulators p27Kip1 and cyclin E,

alone and in combination, correlate with survival in young breast cancer

patients. Nat Med 3, 222-225 (1997).

641 van Diest, P. J. et al. Cyclin D1 expression in invasive breast cancer.

Correlations and prognostic value. Am J Pathol 150, 705-711 (1997).

342

642 Lundgren, K. et al. Effects of cyclin D1 gene amplification and protein

expression on time to recurrence in postmenopausal breast cancer patients

treated with anastrozole or tamoxifen: a TransATAC study. Breast Cancer

Res 14, R57 (2012).

643 Kawamoto, H., Koizumi, H. & Uchikoshi, T. Expression of the G2-M

checkpoint regulators cyclin B1 and cdc2 in nonmalignant and malignant

human breast lesions: immunocytochemical and quantitative image

analyses. Am J Pathol 150, 15-23 (1997).

644 Rudolph, P. et al. Differential prognostic impact of the cyclins E and B in

premenopausal and postmenopausal women with lymph node-negative

breast cancer. Int J Cancer 105, 674-680 (2003).

645 Aaltonen, K. et al. High cyclin B1 expression is associated with poor survival

in breast cancer. Br J Cancer 100, 1055-1060 (2009).

646 Suzuki, T. et al. Nuclear cyclin B1 in human breast carcinoma as a potent

prognostic factor. Cancer Sci 98, 644-651 (2007).

647 Sung, W.-W. et al. High nuclear/cytoplasmic ratio of Cdk1 expression

predicts poor prognosis in colorectal cancer patients. BMC Cancer 14, 951

(2014).

648 Zhao, P., Hu, Y.-C. & Talbot, I. C. Expressing patterns of p16 and CDK4

correlated to prognosis in colorectal carcinoma. World J Gastroenterol 9,

2202-2206 (2003).

343

649 Law, M. E., Corsino, P. E., Narayan, S. & Law, B. K. Cyclin-dependent

kinase inhibitors as anticancer therapeutics. Mol Pharmacol 88, 846-852

(2015).

650 O'Leary, B., Finn, R. S. & Turner, N. C. Treating cancer with selective

CDK4/6 inhibitors. Nat Rev Clin Oncol 13, 417-430 (2016).

651 Wierstra, I. The transcription factor FOXM1 (Forkhead box M1):

proliferation-specific expression, transcription factor function, target genes,

mouse models, and normal biological roles. Adv Cancer Res 118, 97-398,

doi:10.1016/B978-0-12-407173-5.00004-2 (2013).

652 Litovchick, L. et al. Evolutionarily conserved multisubunit RBL2/p130 and

E2F4 protein complex represses human cell cycle-dependent genes in

quiescence. Mol Cell 26, 539-551 (2007).

653 Sadasivam, S., Duan, S. & DeCaprio, J. A. The MuvB complex sequentially

recruits B-Myb and FoxM1 to promote mitotic gene expression. Genes Dev

26, 474-489, doi:10.1101/gad.181933.111 (2012).

654 Saville, M. K. & Watson, R. J. The cell-cycle regulated transcription factor

B-Myb is phosphorylated by cyclin A/Cdk2 at sites that enhance its

transactivation properties. Oncogene 17, 2679-2689 (1998).

655 Sadasivam, S. & DeCaprio, J. A. The DREAM complex: master coordinator

of cell cycle-dependent gene expression. Nat Rev Cancer 13, 585-595

(2013).

344

656 Lane, H. A. & Nigg, E. A. Antibody microinjection reveals an essential role

for human polo-like kinase 1 (Plk1) in the functional maturation of mitotic

centrosomes. J Cell Biol 135, 1701-1713 (1996).

657 Yarm, F. R. Plk phosphorylation regulates the microtubule-stabilizing

protein TCTP. Mol Cell Biol 22, 6209-6221 (2002).

658 Casenghi, M. et al. Polo-like kinase 1 regulates Nlp, a centrosome protein

involved in microtubule nucleation. Dev Cell 5, 113-125 (2003).

659 Qian, Y.-W., Erikson, E. & Maller, J. L. Mitotic effects of a constitutively

active mutant of the Xenopus polo-like kinase Plx1. Mol Cell Biol 19, 8625-

8632 (1999).

660 Kumagai, A. & Dunphy, W. G. Purification and molecular cloning of Plx1, a

Cdc25-regulatory kinase from Xenopus egg extracts. Science 273, 1377-

1380 (1996).

661 Watanabe, N. et al. M-phase kinases induce phospho-dependent

ubiquitination of somatic Wee1 by SCFβ-TrCP. Proc Natl Acad Sci USA

101, 4419-4424 (2004).

662 Manic, G., Corradi, F., Sistigu, A., Siteni, S. & Vitale, I. Molecular regulation

of the spindle assembly checkpoint by kinases and phosphatases. Int Rev

Cell Mol Biol 328, 105-161 (2017).

663 Strebhardt, K. & Ullrich, A. Targeting polo-like kinase 1 for cancer therapy.

Nat Rev Cancer 6, 321-330 (2006).

345

664 Zitouni, S., Nabais, C., Jana, S. C., Guerrero, A. & Bettencourt-Dias, M.

Polo-like kinases: structural variations lead to multiple functions. Nat Rev

Mol Cell Biol 15, 433-452 (2014).

665 Nunes Bastos, R. & Barr, F. A. Plk1 negatively regulates Cep55 recruitment

to the midbody to ensure orderly abscission. J Cell Biol 191, 751-760

(2010).

666 Steegmaier, M. et al. BI 2536, a potent and selective inhibitor of polo-like

kinase 1, inhibits tumor growth in vivo. Curr Biol 17, 316-322 (2007).

667 Warnke, S. et al. Polo-like kinase-2 is required for centriole duplication in

mammalian cells. Curr Biol 14, 1200-1207 (2004).

668 Habedanck, R., Stierhof, Y. D., Wilkinson, C. J. & Nigg, E. A. The Polo

kinase Plk4 functions in centriole duplication. Nat Cell Biol 7, 1140-1146

(2005).

669 Iida, M., Matsuda, M. & Komatani, H. Plk3 phosphorylates topoisomerase

IIalpha at Thr(1342), a site that is not recognized by Plk1. Biochem J 411,

27-32 (2008).

670 López-Sánchez, I., Sanz-García, M. & Lazo, P. A. Plk3 interacts with and

specifically phosphorylates VRK1 in Ser342, a downstream target in a

pathway that induces Golgi fragmentation. Mol Cell Biol 29, 1189-1201

(2009).

671 de Cárcer, G. et al. Plk5, a polo box domain-only protein with specific roles

in neuron differentiation and glioblastoma suppression. Mol Cell Biol 31,

1225-1239 (2011).

346

672 Bolanos-Garcia, V. M. Aurora kinases. Int J Biochem Cell Biol 37, 1572-

1577, doi:10.1016/j.biocel.2005.02.021 (2005).

673 Tang, A. et al. Aurora kinases: novel therapy targets in cancers. Oncotarget

8, 23937–23954 (2017).

674 Floyd, S., Pines, J. & Lindon, C. APC/C Cdh1 targets aurora kinase to

control reorganization of the mitotic spindle at anaphase. Curr Biol 18, 1649-

1658 (2008).

675 Lindon, C. & Pines, J. Ordered proteolysis in anaphase inactivates Plk1 to

contribute to proper mitotic exit in human cells. J Cell Biol 164, 233-241

(2004).

676 Stewart, S. & Fang, G. Destruction box-dependent degradation of aurora B

is mediated by the anaphase-promoting complex/cyclosome and Cdh1.

Cancer Res 65, 8730-8735 (2005).

677 Nguyen, H. G., Chinnappan, D., Urano, T. & Ravid, K. Mechanism of

Aurora-B degradation and its dependency on intact KEN and A-boxes:

identification of an aneuploidy-promoting property. Mol Cell Biol 25, 4977-

4992 (2005).

678 Reboutier, D., Benaud, C. & Prigent, C. Aurora A's functions during mitotic

exit: the guess who game. Front Oncol 5, 290 (2015).

679 Sugimoto, K. et al. Molecular dynamics of Aurora-A kinase in living mitotic

cells simultaneously visualized with histone H3 and nuclear membrane

protein importinalpha. Cell Struct Funct 27, 457-467 (2002).

347

680 Seki, A., Coppinger, J. A., Jang, C.-Y., Yates, J. R., III & Fang, G. Bora and

the kinase Aurora a cooperatively activate the kinase Plk1 and control

mitotic entry. Science 320, 1655-1658 (2008).

681 Dutertre, S. et al. Phosphorylation of CDC25B by Aurora-A at the

centrosome contributes to the G2-M transition. J Cell Sci 117, 2523-2531

(2004).

682 Macůrek, L. et al. Polo-like kinase-1 is activated by aurora A to promote

checkpoint recovery. Nature 455, 119-123 (2008).

683 Cazales, M. et al. CDC25B phosphorylation by Aurora-A occurs at the G2/M

transition and is inhibited by DNA damage. Cell Cycle 4, 1233-1238 (2005).

684 Marumoto, T. et al. Roles of aurora-A kinase in mitotic entry and G2

checkpoint in mammalian cells. Genes Cells 7, 1173-1182 (2002).

685 Berdnik, D. & Knoblich, J. A. Drosophila Aurora-A is required for

centrosome maturation and actin-dependent asymmetric protein

localization during mitosis. Curr Biol 12, 640-647 (2002).

686 Hannak, E., Kirkham, M., Hyman, A. A. & Oegema, K. Aurora-A kinase is

required for centrosome maturation in Caenorhabditis elegans. J Cell Biol

155, 1109-1116 (2001).

687 Cowley, D. O. et al. Aurora-A kinase is essential for bipolar spindle

formation and early development. Mol Cell Biol 29, 1059-1071 (2009).

688 Pinyol, R., Scrofani, J. & Vernos, I. The role of NEDD1 phosphorylation by

Aurora A in chromosomal microtubule nucleation and spindle function. Curr

Biol 23, 143-149 (2013).

348

689 Lioutas, A. & Vernos, I. Aurora A kinase and its substrate TACC3 are

required for central spindle assembly. EMBO Rep 14, 829-836 (2013).

690 Reboutier, D. et al. Aurora A is involved in central spindle assembly through

phosphorylation of Ser 19 in P150Glued. J Cell Biol 201, 65-79 (2013).

691 Marumoto, T. et al. Aurora-A kinase maintains the fidelity of early and late

mitotic events in HeLa cells. J Biol Chem 278, 51786-51795 (2003).

692 Hoar, K. et al. MLN8054, a small-molecule inhibitor of Aurora A, causes

spindle pole and chromosome congression defects leading to aneuploidy.

Mol Cell Biol 27, 4513-4525 (2007).

693 Manfredi, M. G. et al. Antitumor activity of MLN8054, an orally active small-

molecule inhibitor of Aurora A kinase. Proc Natl Acad Sci USA 104, 4106-

4111 (2007).

694 Carmena, M., Wheelock, M., Funabiki, H. & Earnshaw, W. C. The

chromosomal passenger complex (CPC): from easy rider to the godfather

of mitosis. Nat Rev Mol Cell Biol 13, 789-803 (2012).

695 Nunes Bastos, R. et al. Aurora B suppresses microtubule dynamics and

limits central spindle size by locally activating KIF4A. J Cell Biol 202, 605-

621 (2013).

696 Kitagawa, M., Fung, S. Y., Onishi, N., Saya, H. & Lee, S. H. Targeting

Aurora B to the equatorial cortex by MKlp2 is required for cytokinesis. PLoS

One 8, e64826 (2013).

697 Steigemann, P. et al. Aurora B-mediated abscission checkpoint protects

against tetraploidization. Cell 136, 473-484 (2009).

349

698 Terada, Y. et al. AIM-1: a mammalian midbody-associated protein required

for cytokinesis. EMBO J 17, 667-676 (1998).

699 Kaestner, P., Stolz, A. & Bastians, H. Determinants for the efficiency of

anticancer drugs targeting either Aurora-A or Aurora-B kinases in human

colon carcinoma cells. Mol Cancer Ther 8, 2046-2056 (2009).

700 Hauf, S. et al. The small molecule Hesperadin reveals a role for Aurora B in

correcting kinetochore-microtubule attachment and in maintaining the

spindle assembly checkpoint. J Cell Biol 161, 281-294 (2003).

701 Goto, H., Yasui, Y., Nigg, E. A. & Inagaki, M. Aurora-B phosphorylates

Histone H3 at serine28 with regard to the mitotic chromosome

condensation. Genes Cells 7, 11-17 (2002).

702 Giet, R. & Glover, D. M. Drosophila Aurora B kinase is required for histone

H3 phosphorylation and condensin recruitment during chromosome

condensation and to organize the central spindle during cytokinesis. J Cell

Biol 152, 669-682 (2001).

703 Gadea, B. B. & Ruderman, J. V. Aurora kinase inhibitor ZM447439 blocks

chromosome-induced spindle assembly, the completion of chromosome

condensation, and the establishment of the spindle integrity checkpoint in

Xenopus egg extracts. Mol Biol Cell 16, 1305-1318 (2005).

704 Kalitsis, P., Zhang, T., Marshall, K. M., Nielsen, C. F. & Hudson, D. F.

Condensin, master organizer of the genome. Chromosome Res 25, 61-76

(2017).

350

705 Abe, S. et al. The initial phase of chromosome condensation requires Cdk1-

mediated phosphorylation of the CAP-D3 subunit of condensin II. Genes

Dev 25, 863-874 (2011).

706 Bertran, M. T. et al. Nek9 is a Plk1-activated kinase that controls early

centrosome separation through Nek6/7 and Eg5. EMBO J 30, 2634-2647

(2011).

707 Slangy, A. et al. Phosphorylation by p34cdc2 regulates spindle association

of human Eg5, a kinesin-related motor essential for bipolar spindle

formation in vivo. Cell 83, 1159-1169 (1995).

708 Nam, H.-J. & van Deursen, J. M. Cyclin B2 and p53 control proper timing of

centrosome separation. Nat Cell Biol 16, 538-549 (2014).

709 Linder, M. I. et al. Mitotic disassembly of nuclear pore complexes involves

CDK1- and PLK1-mediated phosphorylation of key interconnecting

nucleoporins. Dev Cell 43, 141-156.e147 (2017).

710 Laurell, E. et al. Phosphorylation of Nup98 by multiple kinases is crucial for

NPC disassembly during mitotic entry. Cell 144, 539-550 (2011).

711 Güttinger, S., Laurell, E. & Kutay, U. Orchestrating nuclear envelope

disassembly and reassembly during mitosis. Nat Rev Mol Cell Biol 10, 178-

191 (2009).

712 Jang, C.-Y., Coppinger, J. A., Seki, A., Yates, J. R., III & Fang, G. Plk1 and

Aurora A regulate the depolymerase activity and the cellular localization of

Kif2a. J Cell Sci 122, 1334-1341 (2009).

351

713 Jang, C.-Y. & Fang, G. DDA3 associates with MCAK and controls

chromosome congression. Biochem Biophys Res Commun 407, 610-614

(2011).

714 Zhang, X., Ems-McClung, S. C. & Walczak, C. E. Aurora A phosphorylates

MCAK to control ran-dependent spindle bipolarity. Mol Biol Cell 19, 2752-

2765 (2008).

715 Asteriti, I. A., Giubettini, M., Lavia, P. & Guarguaglini, G. Aurora-A

inactivation causes mitotic spindle pole fragmentation by unbalancing

microtubule-generated forces. Mol Cancer 10, 131 (2011).

716 De Luca, M. et al. Aurora-A and ch-TOG act in a common pathway in control

of spindle pole integrity. Oncogene 27, 6539-6549 (2008).

717 Prosser, S. L. & Pelletier, L. Mitotic spindle assembly in animal cells: a fine

balancing act. Nat Rev Mol Cell Biol 18, 187-201 (2017).

718 Grill, S. W. & Hyman, A. A. Spindle positioning by cortical pulling forces.

Dev Cell 8, 461-465 (2005).

719 Musacchio, A. & Salmon, E. D. The spindle-assembly checkpoint in space

and time. Nat Rev Mol Cell Biol 8, 379-393 (2007).

720 Zich, J. & Hardwick, K. G. Getting down to the phosphorylated ‘nuts and

bolts’ of spindle checkpoint signalling. Trends Biochem Sci 35, 18-27

(2010).

721 O'Connor, A. et al. Requirement for PLK1 kinase activity in the maintenance

of a robust spindle assembly checkpoint. Biol Open 5, 11-19 (2015).

352

722 Chu, Y. et al. Aurora B kinase activation requires survivin priming

phosphorylation by PLK1. J Mol Cell Biol 3, 260-267 (2011).

723 Ahonen, L. J. et al. Polo-like kinase 1 creates the tension-sensing 3F3/2

phosphoepitope and modulates the association of spindle-checkpoint

proteins at kinetochores. Curr Biol 15, 1078-1089 (2005).

724 Chi, Y. H. et al. Requirements for protein phosphorylation and the kinase

activity of polo-like kinase 1 (Plk1) for the kinetochore function of mitotic

arrest deficiency protein 1 (Mad1). J Biol Chem 283, 35834-35844 (2008).

725 Elowe, S., Hümmer, S., Uldschmid, A., Li, X. & Nigg, E. A. Tension-sensitive

Plk1 phosphorylation on BubR1 regulates the stability of kinetochore

microtubule interactions. Genes Dev 21, 2205-2219 (2007).

726 Qi, W., Tang, Z. & Yu, H. Phosphorylation- and polo-box-dependent binding

of Plk1 to Bub1 is required for the kinetochore localization of Plk1. Mol Biol

Cell 17, 3705-3716 (2006).

727 Sudakin, V., Chan, G. K. T. & Yen, T. J. Checkpoint inhibition of the APC/C

in HeLa cells is mediated by a complex of BUBR1, BUB3, CDC20, and

MAD2. J Cell Biol 154, 925-936 (2001).

728 Krenn, V. & Musacchio, A. The Aurora B kinase in chromosome bi-

orientation and spindle checkpoint signaling. Front Oncol 5, 225 (2015).

729 Knowlton, A. L., Lan, W. & Stukenberg, P. T. Aurora B is enriched at

merotelic attachment sites, where it regulates MCAK. Curr Biol 16, 1705-

1710 (2006).

353

730 DeLuca, K. F., Lens, S. M. & DeLuca, J. G. Temporal changes in Hec1

phosphorylation control kinetochore-microtubule attachment stability during

mitosis. J Cell Sci 124, 622-634 (2011).

731 Zhu, T. et al. Phosphorylation of microtubule-binding protein Hec1 by mitotic

kinase Aurora B specifies spindle checkpoint kinase Mps1 signaling at the

kinetochore. J Biol Chem 288, 36149-36159 (2013).

732 Chan, Y. W., Jeyaprakash, A. A., Nigg, E. A. & Santamaria, A. Aurora B

controls kinetochore-microtubule attachments by inhibiting Ska complex-

KMN network interaction. J Cell Biol 196, 563-571 (2012).

733 Schmidt, J. C. et al. Aurora B kinase controls the targeting of the Astrin-

SKAP complex to bioriented kinetochores. J Cell Biol 191, 269-280 (2010).

734 Acquaviva, C. & Pines, J. The anaphase-promoting complex/cyclosome:

APC/C. J Cell Sci 119, 2401-2404 (2006).

735 Kraft, C. et al. Mitotic regulation of the human anaphase‐promoting complex

by phosphorylation. EMBO J 22, 6598-6609 (2003).

736 Kramer, E. R., Scheuringer, N., Podtelejnikov, A. V., Mann, M. & Peters, J.-

M. Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol

Biol Cell 11, 1555-1569 (2000).

737 Fang, G., Yu, H. & Kirschner, M. W. Direct binding of CDC20 protein family

members activates the anaphase-promoting complex in mitosis and G1.

Mol Cell 2, 163-171 (1998).

738 Moshe, Y., Boulaire, J., Pagano, M. & Hershko, A. Role of Polo-like kinase

in the degradation of early mitotic inhibitor 1, a regulator of the anaphase

354

promoting complex/cyclosome. Proc Natl Acad Sci USA 101, 7937-7942

(2004).

739 Hansen, D. V., Loktev, A. V., Ban, K. H. & Jackson, P. K. Plk1 regulates

activation of the anaphase promoting complex by phosphorylating and

triggering SCFbetaTrCP-dependent destruction of the APC Inhibitor Emi1.

Mol Biol Cell 15, 5623-5634 (2004).

740 Cohen-Fix, O., Peters, J. M., Kirschner, M. W. & Koshland, D. Anaphase

initiation in Saccharomyces cerevisiae is controlled by the APC-dependent

degradation of the anaphase inhibitor Pds1p. Genes Dev 10, 3081-3093

(1996).

741 Funabiki, H. et al. Cut2 proteolysis required for sister-chromatid seperation

in fission yeast. Nature 381, 438-441 (1996).

742 Peters, J.-M. The anaphase-promoting complex: proteolysis in mitosis and

beyond. Mol Cell 9, 931-943 (2002).

743 Zou, H., McGarry, T. J., Bernal, T. & Kirschner, M. W. Identification of a

vertebrate sister-chromatid separation inhibitor involved in transformation

and tumorigenesis. Science 285, 418-422 (1999).

744 Hauf, S., Waizenegger, I. C. & Peters, J. M. Cohesin cleavage by separase

required for anaphase and cytokinesis in human cells. Science 293, 1320-

1323 (2001).

745 Uhlmann, F., Wernic, D., Poupart, M. A., Koonin, E. V. & Nasmyth, K.

Cleavage of cohesin by the CD clan protease separin triggers anaphase in

yeast. Cell 103, 375-386 (2000).

355

746 Sudakin, V. et al. The cyclosome, a large complex containing cyclin-

selective ubiquitin ligase activity, targets cyclins for destruction at the end

of mitosis. Mol Biol Cell 6, 185-197 (1995).

747 Clute, P. & Pines, J. Temporal and spatial control of cyclin B1 destruction

in metaphase. Nat Cell Biol 1, 82-87 (1999).

748 King, R. W. et al. A 20S complex containing CDC27 and CDC16 catalyzes

the mitosis-specific conjugation of ubiquitin to cyclin B. Cell 81, 279-288

(1995).

749 Asbury, C. L. Anaphase A: disassembling microtubules move

chromosomes toward spindle poles. Biology (Basel) 6, pii: E15 (2017).

750 Scholey, J. M., Civelekoglu-Scholey, G. & Brust-Mascher, I. Anaphase B.

Biology (Basel) 5, pii: E51 (2016).

751 Verde, F., Dogterom, M., Stelzer, E., Karsenti, E. & Leibler, S. Control of

microtubule dynamics and length by cyclin A- and cyclin B-dependent

kinases in Xenopus egg extracts. J Cell Biol 118, 1097-1108 (1992).

752 Higuchi, T. & Uhlmann, F. Stabilization of microtubule dynamics at

anaphase onset promotes chromosome segregation. Nature 433, 171-176

(2005).

753 Su, K. C. et al. A regulatory switch alters chromosome motions at the

metaphase-to-anaphase transition. Cell Rep 17, 1728-1738 (2016).

754 Brust-Mascher, I. & Scholey, J. M. Mitotic motors and chromosome

segregation: the mechanism of anaphase B. Biochem Soc Trans 39, 1149-

1153 (2011).

356

755 Woodruff, J. B., Drubin, D. G. & Barnes, G. Mitotic spindle disassembly

occurs via distinct subprocesses driven by the anaphase-promoting

complex, Aurora B kinase, and kinesin-8. J Cell Biol 191, 795-808 (2010).

756 Ramadan, K. et al. Cdc48/p97 promotes reformation of the nucleus by

extracting the kinase Aurora B from chromatin. Nature 450, 1258-1262

(2007).

757 Qian, J., Lesage, B., Beullens, M., Van Eynde, A. & Bollen, M. PP1/Repo-

man dephosphorylates mitotic histone H3 at T3 and regulates chromosomal

aurora B targeting. Curr Biol 21, 766-773 (2011).

758 Landsverk, H. B., Kirkhus, M., Bollen, M., Küntziger, T. & Collas, P. PNUTS

enhances in vitro chromosome decondensation in a PP1-dependent

manner. Biochem J 390, 709-717 (2005).

759 Antonin, W. & Neumann, H. Chromosome condensation and

decondensation during mitosis. Curr Opin Cell Biol 40, 15-22 (2016).

760 Miller, A. L. The contractile ring. Curr Biol 21, R976-978 (2011).

761 Wadsworth, P. Cytokinesis: Rho marks the spot. Curr Biol 15, R871-874

(2005).

762 Mishima, M., Pavicic, V., Grüneberg, U., Nigg, E. A. & Glotzer, M. Cell cycle

regulation of central spindle assembly. Nature 430, 908-913 (2004).

763 Minoshima, Y. et al. Phosphorylation by aurora B converts MgcRacGAP to

a RhoGAP during cytokinesis. Dev Cell 4, 549-560 (2003).

357

764 Kim, H., Guo, F., Brahma, S., Xing, Y. & Burkard, M. E. Centralspindlin

assembly and 2 phosphorylations on MgcRacGAP by Polo-like kinase 1

initiate Ect2 binding in early cytokinesis. Cell Cycle 13, 2952-2961 (2014).

765 Guse, A., Mishima, M. & Glotzer, M. Phosphorylation of ZEN-4/MKLP1 by

aurora B regulates completion of cytokinesis. Curr Biol 15, 778-786 (2005).

766 Zhao, W. M. & Fang, G. MgcRacGAP controls the assembly of the

contractile ring and the initiation of cytokinesis. Proc Natl Acad Sci USA

102, 13158-13163 (2005).

767 Yüce, O., Piekny, A. & Glotzer, M. An ECT2-centralspindlin complex

regulates the localization and function of RhoA. J Cell Biol 170, 571-582

(2005).

768 Watanabe, S. et al. Loss of a Rho-regulated actin nucleator, mDia2, impairs

cytokinesis during mouse fetal erythropoiesis. Cell Rep 5, 926-932 (2013).

769 Watanabe, N. et al. p140mDia, a mammalian homolog of Drosophila

diaphanous, is a target protein for Rho small GTPase and is a ligand for

profilin. EMBO J 16, 3044-3056 (1997).

770 Kühn, S. & Geyer, M. Formins as effector proteins of Rho GTPases. Small

GTPases 5, e29513 (2014).

771 Kimura, K. et al. Regulation of myosin phosphatase by Rho and Rho-

associated kinase (Rho-kinase). Science 273, 245-248 (1996).

772 Yamashiro, S. et al. Citron kinase, a Rho-dependent kinase, induces di-

phosphorylation of regulatory light chain of myosin II. Mol Biol Cell 14, 1745-

1756 (2003).

358

773 Mierzwa, B. & Gerlich, D. W. Cytokinetic abscission: molecular mechanisms

and temporal control. Dev Cell 31, 525-538 (2014).

774 Severson, A. F., Hamill, D. R., Carter, J. C., Schumacher, J. & Bowerman,

B. The Aurora-related kinase AIR-2 recruits ZEN-4/CeMKLP1 to the mitotic

spindle at metaphase and is required for cytokinesis. Curr Biol 10, 1162-

1171 (2000).

775 Kaitna, S., Mendoza, M., Jantsch-Plunger, V. & Glotzer, M. Incenp and an

Aurora-like kinase form a complex essential for chromosome segregation

and efficient completion of cytokinesis. Curr Biol 10, 1172-1181 (2000).

776 Finger, F. P. & White, J. G. Fusion and fission: membrane trafficking in

animal cytokinesis. Cell 108, 727-730 (2002).

777 Jordan, M. A. & Wilson, L. Microtubules as a target for anticancer drugs.

Nat Rev Cancer 4, 253-265 (2004).

778 Weichert, W. et al. Polo-like kinase isoforms in breast cancer: expression

patterns and prognostic implications. Virchows Arch 446, 442-450 (2005).

779 King, S. I. et al. Immunohistochemical detection of Polo-like kinase-1

(PLK1) in primary breast cancer is associated with TP53 mutation and poor

clinical outcome. Breast Cancer Res 14, R40 (2012).

780 Maire, V. et al. Polo-like kinase 1: a potential therapeutic option in

combination with conventional chemotherapy for the management of

patients with triple-negative breast cancer. Cancer Res 73, 813-823 (2013).

359

781 Donizy, P. et al. Augmented expression of Polo-like kinase 1 is a strong

predictor of shorter cancer-specific overall survival in early stage breast

cancer at 15-year follow-up. Oncol Lett 12, 1667-1674 (2016).

782 Spänkuch-Schmitt, B., Bereiter-Hahn, J. r., Kaufmann, M. & Strebhardt, K.

Effect of RNA silencing of polo-like kinase-1 (PLK1) on apoptosis and

spindle formation in human cancer cells. J Natl Cancer Inst 94, 1863-1877

(2002).

783 Hu, K., Law, J. H., Fotovati, A. & Dunn, S. E. Small interfering RNA library

screen identified polo-like kinase-1 (PLK1) as a potential therapeutic target

for breast cancer that uniquely eliminates tumor-initiating cells. Breast

Cancer Res 14, R22 (2012).

784 Morry, J. et al. Targeted treatment of metastatic breast cancer by PLK1

siRNA delivered by an antioxidant nanoparticle platform. Mol Cancer Ther

16, 763-772 (2017).

785 Liu, X., Lei, M. & Erikson, R. L. Normal cells, but not cancer cells, survive

severe Plk1 depletion. Mol Cell Biol 26, 2093-2108 (2006).

786 Haupenthal, J. et al. Reduced efficacy of the Plk1 inhibitor BI 2536 on the

progression of hepatocellular carcinoma due to low intratumoral drug levels.

Neoplasia 14, 410-419 (2012).

787 Gutteridge, R. E., Ndiaye, M. A., Liu, X. & Ahmad, N. Plk1 inhibitors in

cancer therapy: from laboratory to clinics. Mol Cancer Ther 15, 1427-1435

(2016).

360

788 Degenhardt, Y. & Lampkin, T. Targeting Polo-like kinase in cancer therapy.

Clin Cancer Res 16, 384-389 (2010).

789 Van den Bossche, J. et al. Spotlight on volasertib: preclinical and clinical

evaluation of a promising Plk1 inhibitor. Med Res Rev 36, 749-786 (2016).

790 Liu, Z., Sun, Q. & Wang, X. PLK1, a potential target for cancer therapy.

Transl Oncol 10, 22-32 (2017).

791 Staff, S., Isola, J., Jumppanen, M. & Tanner, M. Aurora-A gene is frequently

amplified in basal-like breast cancer. Oncol Rep 23, 307-312 (2010).

792 Miyoshi, Y., Iwao, K., Egawa, C. & Noguchi, S. Association of centrosomal

kinase STK15/BTAK mRNA expression with chromosomal instability in

human breast cancers. Int J Cancer 92, 370-373 (2001).

793 Yamamoto, S., Yamamoto-Ibusuki, M., Yamamoto, Y., Fujiwara, S. &

Iwase, H. A comprehensive analysis of Aurora A; transcript levels are the

most reliable in association with proliferation and prognosis in breast

cancer. BMC Cancer 13, 217 (2013).

794 Ali, H. R. et al. Aurora kinase A outperforms Ki67 as a prognostic marker in

ER-positive breast cancer. Br J Cancer 106, 1798-1806 (2012).

795 Siggelkow, W. et al. Expression of aurora kinase A is associated with

metastasis-free survival in node-negative breast cancer patients. BMC

Cancer 12, 562 (2012).

796 Xu, J. et al. Aurora-A identifies early recurrence and poor prognosis and

promises a potential therapeutic target in triple negative breast cancer.

PLoS One 8, e56919 (2013).

361

797 Wang, X. et al. Overexpression of aurora kinase A in mouse mammary

epithelium induces genetic instability preceding mammary tumor formation.

Oncogene 25, 7148-7158 (2006).

798 D'Assoro, A. B. et al. The mitotic kinase Aurora-A promotes distant

metastases by inducing epithelial-to-mesenchymal transition in ERα+

breast cancer cells. Oncogene 33, 599-610 (2014).

799 Eterno, V. et al. AurkA controls self-renewal of breast cancer-initiating cells

promoting wnt3a stabilization through suppression of miR-128. Sci Rep 6,

28436 (2016).

800 Opyrchal, M. et al. Aurora-A mitotic kinase induces endocrine resistance

through down-regulation of ERα expression in initially ERα+ breast cancer

cells. PLoS One 9, e96995 (2014).

801 Zheng, X. Q. et al. Aurora-A is a determinant of tamoxifen sensitivity through

phosphorylation of ERa in breast cancer. Oncogene 33, 4985-4996 (2014).

802 Hole, S., Pedersen, A. M., Lykkesfeldt, A. E. & Yde, C. W. Aurora kinase A

and B as new treatment targets in aromatase inhibitor-resistant breast

cancer cells. Breast Cancer Res Treat 149, 715-726 (2015).

803 Zheng, F. M. et al. A novel small molecule aurora kinase inhibitor attenuates

breast tumor-initiating cells and overcomes drug resistance. Mol Cancer

Ther 13, 1991-2003 (2014).

804 Li, J. P. et al. The investigational Aurora kinase A inhibitor alisertib

(MLN8237) induces cell cycle G2/M arrest, apoptosis, and autophagy via

362

p38 MAPK and Akt/mTOR signaling pathways in human breast cancer cells.

Drug Des Devel Ther 9, 1627-1652 (2015).

805 Huck, J. J. et al. Translational exposure-efficacy modeling to optimize the

dose and schedule of taxanes combined with the investigational Aurora A

kinase inhibitor MLN8237 (alisertib). Mol Cancer Ther 13, 2170-2183

(2014).

806 Melichar, B. et al. Safety and activity of alisertib, an investigational aurora

kinase A inhibitor, in patients with breast cancer, small-cell lung cancer,

non-small-cell lung cancer, head and neck squamous-cell carcinoma, and

gastro-oesophageal adenocarcinoma: a five-arm phase 2 study. Lancet

Oncol 16, 395-405 (2015).

807 Zhang, Y. et al. Elevated Aurora B expression contributes to

chemoresistance and poor prognosis in breast cancer. Int J Clin Exp Pathol

8, 751-757 (2015).

808 Larsen, S. L. et al. Aurora kinase B is important for antiestrogen resistant

cell growth and a potential biomarker for tamoxifen resistant breast cancer.

BMC Cancer 15, 239, doi:10.1186/s12885-015-1210-4 (2015).

809 Sugiyama, K. et al. Aurora-B associated protein phosphatases as negative

regulators of kinase activation. Oncogene 21, 3103-3111 (2002).

810 Yasui, Y. et al. Autophosphorylation of a newly identified site of Aurora-B is

indispensable for cytokinesis. J Biol Chem 279, 12997-13003 (2004).

811 Gully, C. P. et al. Antineoplastic effects of an Aurora B kinase inhibitor in

breast cancer. Mol Cancer 9, 42 (2010).

363

812 Shi, J. & Vakoc, C. R. The mechanisms behind the therapeutic activity of

BET bromodomain inhibition. Mol Cell 54, 728-736,

doi:10.1016/j.molcel.2014.05.016 (2014).

813 Dey, A. et al. A bromodomain protein, MCAP, associates with mitotic

chromosomes and affects G(2)-to-M transition. Mol Cell Biol 20, 6537-6549

(2000).

814 Wu, S.-Y. & Chiang, C.-M. The double bromodomain-containing chromatin

adaptor Brd4 and transcriptional regulation. J Biol Chem 282, 13141-13145

(2007).

815 Wang, H. et al. NOTCH1–RBPJ complexes drive target gene expression

through dynamic interactions with superenhancers. Proc Natl Acad Sci USA

111, 705-710, doi:10.1073/pnas.1315023111 (2014).

816 Montañez-Wiscovich, M. et al. LMO4 is an essential mediator of

ErbB2/HER2/Neu-induced breast cancer cell cycle progression. Oncogene

28, 3608-3618, doi:10.1038/onc.2009.221 (2009).

817 Bernardo, G. M. et al. FOXA1 is an essential determinant of ERα expression

and mammary ductal morphogenesis. Development 137, 2045-2054,

doi:10.1242/dev.043299 (2010).

818 Yori, J. L., Johnson, E., Zhou, G., Jain, M. K. & Keri, R. A. Kruppel-like factor

4 inhibits epithelial-to-mesenchymal transition through regulation of E-

cadherin gene expression. J Biol Chem 285, 16854-16863,

doi:10.1074/jbc.M110.114546 (2010).

364

819 Petitjean, A. et al. Impact of mutant p53 functional properties on TP53

mutation patterns and tumor phenotype: Lessons from recent

developments in the IARC TP53 database. Hum Mutat 28, 622-629 (2007).

820 Eisenhauer, E. A. et al. New response evaluation criteria in solid tumours:

revised RECIST guideline (version 1.1). Eur J Cancer 45, 228-247,

doi:10.1016/j.ejca.2008.10.026 (2009).

821 Zhang, X. et al. A renewable tissue resource of phenotypically stable,

biologically and ethnically diverse, patient-derived human breast cancer

xenograft models. Cancer Res 73, 4885-4897, doi:10.1158/0008-

5472.CAN-12-4081 (2013).

822 Tatsuka, M. et al. Multinuclearity and increased ploidy caused by

overexpression of the Aurora- and Ipl1-like midbody-associated protein

mitotic kinase in human cancer cells. Cancer Res 58, 4811-4816 (1998).

823 Meraldi, P., Honda, R. & Nigg, E. A. Aurora-A overexpression reveals

tetraploidization as a major route to centrosome amplification in p53–/–

cells. EMBO J 21, 483-492 (2002).

824 Wilkinson, R. W. et al. AZD1152, a selective inhibitor of Aurora B kinase,

inhibits human tumor xenograft growth by inducing apoptosis. Clin Cancer

Res 13, 3682-3688 (2007).

825 Görgün, G. et al. A novel Aurora-A kinase inhibitor MLN8237 induces

cytotoxicity and cell-cycle arrest in multiple myeloma. Blood 115, 5202-

5213, doi:10.1182/blood-2009-12-259523 (2010).

365

826 Herschkowitz, J. I. et al. Comparative oncogenomics identifies breast

tumors enriched in functional tumor-initiating cells. Proc Natl Acad Sci USA

109, 2778-2783, doi:10.1073/pnas.1018862108 (2012).

827 Wang, I.-C. et al. Forkhead box M1 regulates the transcriptional network of

genes essential for mitotic progression and genes encoding the SCF (Skp2-

Cks1) ubiquitin ligase. Mol Cell Biol 25, 10875-10894 (2005).

828 Kimura, M., Uchida, C., Takano, Y., Kitagawa, M. & Okano, Y. Cell cycle-

dependent regulation of the human aurora B promoter. Biochem Biophys

Res Commun 316, 930-936 (2004).

829 Sadaie, M. et al. Cell-based screen for altered nuclear phenotypes reveals

senescence progression in polyploid cells after Aurora kinase B inhibition.

Mol Biol Cell 26, 2971-2985, doi:10.1091/mbc.E15-01-0003 (2015).

830 Nath, S., Ghatak, D., Das, P. & Roychoudhury, S. Transcriptional control of

mitosis: deregulation and cancer. Front Endocrinol 6, 60,

doi:10.3389/fendo.2015.00060 (2015).

831 Denisenkoa, T. V., Sorokina, I. V., Gogvadze, V. & Zhivotovsky, B. Mitotic

catastrophe and cancer drug resistance: A link that must to be broken. Drug

Resist Updat 24, 1-12, doi:10.1016/j.drup.2015.11.002 (2016).

832 Vitale, I., Galluzzi, L., Castedo, M. & Kroemer, G. Mitotic catastrophe: a

mechanism for avoiding genomic instability. Nat Rev Mol Cell Biol 12, 385-

392 (2011).

366

833 Roninson, I. B., Broude, E. V. & Chang, B.-D. If not apoptosis, then what?

Treatment-induced senescence and mitotic catastrophe in tumor cells. Drug

Resist Updat 4, 303-313 (2001).

834 Jiang, L. et al. iASPP and chemoresistance in ovarian cancers: effects on

paclitaxel-mediated mitotic catastrophe. Clin Cancer Res 17, 6924-6933,

doi:10.1158/1078-0432.CCR-11-0588 (2011).

835 Nitta, M. et al. Spindle checkpoint function is required for mitotic catastrophe

induced by DNA-damaging agents. Oncogene 23, 6548-6558,

doi:10.1038/sj.onc.1207873 (2004).

836 Vakifahmetoglu, H., Olsson, M. & Zhivotovsky, B. Death through a tragedy:

Mitotic catastrophe. Cell Death Differ 15, 1153-1162,

doi:10.1038/cdd.2008.47 (2008).

837 Jitariu, A.-A., Cîmpean, A. M., Ribatti, D. & Raica, M. Triple negative breast

cancer: the kiss of death. Oncotarget, doi:10.18632/oncotarget.16938

(2017).

838 Carvalho, B. S. & Irizarry, R. A. A framework for oligonucleotide microarray

preprocessing. Bioinformatics 26, 2363-2367,

doi:10.1093/bioinformatics/btq431 (2010).

839 Ritchie, M. E. et al. limma powers differential expression analyses for RNA-

sequencing and microarray studies. Nucleic Acids Res 43, e47,

doi:10.1093/nar/gkv007 (2015).

840 Joshi-Tope, G. et al. Reactome: a knowledgebase of biological pathways.

Nucleic Acids Res 33, D428-432, doi:10.1093/nar/gki072 (2005).

367

841 Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based

approach for interpreting genome-wide expression profiles. Proc Natl Acad

Sci USA 102, 15545-15550, doi:10.1073/pnas.0506580102 (2005).

842 Liberzon, A. et al. Molecular signatures database (MSigDB) 3.0.

Bioinformatics 27, 1739-1740 (2011).

843 Bernardo, G. M. et al. FOXA1 represses the molecular phenotype of basal

breast cancer cells. Oncogene 32, 554-563, doi:10.1038/onc.2012.62

(2013).

844 Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2.

Nat Methods 9, 357-359 (2012).

845 Zhang, Y. et al. Model-based analysis of ChIP-Seq (MACS). Genome Biol

9, R137, doi:10.1186/gb-2008-9-9-r137 (2008).

846 Gao, J. et al. Integrative analysis of complex cancer genomics and clinical

profiles using the cBioPortal. Sci Signal 6, pl1 (2013).

847 Cerami, E. et al. The cBio cancer genomics portal: an open platform for

exploring multidimensional cancer genomics data. Cancer Discov 2, 401-

404 (2012).

848 Rath, O. & Kozielski, F. Kinesins and cancer. Nat Rev Cancer 12, 527-539

(2012).

849 Belinky, F. et al. PathCards: Multi-source consolidation of human biological

pathways. Database 2015, doi:10.1093/database/bav006 (2015).

368

850 Chu, I. M. et al. GATA3 inhibits lysyl oxidase-mediated metastases of

human basal triple-negative breast cancer cells. Oncogene 31, 2017-2027,

doi:10.1038/onc.2011.382 (2012).

851 Pandit, S. K. et al. E2F8 is essential for polyploidization in mammalian cells.

Nat Cell Biol 14, 1181-1191 (2012).

852 Esterlechner, J. et al. LIN9, a subunit of the DREAM complex, regulates

mitotic gene expression and proliferation of embryonic stem cells. PLoS

One 8, e62882, doi:10.1371/journal.pone.0062882 (2013).

853 Tarasov, K. V. et al. B-MYB is essential for normal cell cycle progression

and chromosomal stability of embryonic stem cells. PLoS One 3, e2478

(2008).

854 Hong, S. H. et al. Epigenetic reader BRD4 inhibition as a therapeutic

strategy to suppress E2F2-cell cycle regulation circuit in liver cancer.

Oncotarget 7, 32628-32640 (2016).

855 Barretina, J. et al. The Cancer Cell Line Encyclopedia enables predictive

modelling of anticancer drug sensitivity. Nature 483, 603-607 (2012).

856 Galluzzi, L. et al. Molecular definitions of cell death subroutines:

recommendations of the Nomenclature Committee on Cell Death 2012. Cell

Death Differ 19, 107-120, doi:10.1038/cdd.2011.96 (2012).

857 Reichert, N. et al. Lin9, a subunit of the mammalian DREAM complex, is

essential for embryonic development, for survival of adult mice, and for

tumor suppression. Mol Cell Biol 30, 2896-2908, doi:10.1128/MCB.00028-

10 (2010).

369

858 Osterloh, L. et al. The human synMuv-like protein LIN-9 is required for

transcription of G2/M genes and for entry into mitosis. EMBO J 26, 144-157

(2007).

859 Laoukili, J. et al. Activation of FoxM1 during G2 requires cyclin A/Cdk-

dependent relief of autorepression by the FoxM1 N-terminal domain. Mol

Cell Biol 28, 3076-3087, doi:10.1128/MCB.01710-07 (2008).

860 Wolter, P., Hanselmann, S., Pattschull, G., Schruf, E. & Gaubatz, S. Central

spindle proteins and mitotic kinesins are direct transcriptional targets of

MuvB, B-MYB and FOXM1 in breast cancer cell lines and are potential

targets for therapy. Oncotarget 8, 11160-11172 (2017).

861 Hauser, S. et al. Loss of LIN9, a member of the DREAM complex,

cooperates with SV40 large T antigen to induce genomic instability and

anchorage-independent growth. Oncogene 31, 1859-1868 (2012).

862 Kanda, T., Sullivan, K. F. & Wahl, G. M. Histone-GFP fusion protein enables

sensitive analysis of chromosome dynamics in living mammalian cells. Curr

Biol 8, 377-385 (1998).

863 Gromley, A. et al. A novel human protein of the maternal centriole is

required for the final stages of cytokinesis and entry into S phase. J Cell

Biol 161, 535-545 (2003).

864 Fabbro, M. et al. Cdk1/Erk2- and Plk1-dependent phosphorylation of a

centrosome protein, Cep55, is required for its recruitment to midbody and

cytokinesis. Dev Cell 9, 477-488 (2005).

370

865 Straight, A. F., Field, C. M. & Mitchison, T. J. Anillin binds nonmuscle myosin

II and regulates the contractile ring. Mol Biol Cell 16, 193-201 (2005).

866 Scheel, C. & Weinberg, R. A. Cancer stem cells and epithelial-

mesenchymal transition: concepts and molecular links. Semin Cancer Biol

22, 396-403 (2012).

867 Vega, S. et al. Snail blocks the cell cycle and confers resistance to cell

death. Genes Dev 18, 1131-1143 (2004).

868 Mejlvang, J. et al. Direct repression of cyclin D1 by SIP1 attenuates cell

cycle progression in cells undergoing an epithelial mesenchymal transition.

Mol Biol Cell 18, 4615-4624 (2007).

869 Kleinschmidt, M. A. et al. lin9 is required for mitosis and cell survival during

early zebrafish development. J Biol Chem 284, 13119-13127 (2009).

870 Pilkinton, M., Sandoval, R., Song, J., Ness, S. A. & Colamonici, O. R.

Mip/LIN-9 regulates the expression of B-Myb and the induction of cyclin A,

cyclin B, and CDK1. J Biol Chem 282, 168-175 (2007).

871 Gagrica, S. et al. Inhibition of oncogenic transformation by mammalian Lin-

9, a pRB-associated protein. EMBO J 23, 4627-4638 (2004).

872 Begley, U. et al. A human tRNA methyltransferase 9-like protein prevents

tumour growth by regulating LIN9 and HIF1-α. EMBO Mol Med 5, 366-383

(2013).

873 Wiseman, E. F. et al. Deregulation of the FOXM1 target gene network and

its coregulatory partners in oesophageal adenocarcinoma. Mol Cancer 14,

69 (2015).

371

874 Schmit, F., Cremer, S. & Gaubatz, S. LIN54 is an essential core subunit of

the DREAM/LINC complex that binds to the cdc2 promoter in a sequence-

specific manner. FEBS J 276, 5703-5716 (2009).

875 Ewing, R. M. et al. Large-scale mapping of human protein-protein

interactions by mass spectrometry. Mol Syst Biol 3, 89 (2007).

876 Gayle, S. S., Sahni, J. M. & Keri, R. A. BETi induction of mitotic catastrophe:

towing the LIN9. Oncoscience 4, 128-130 (2017).

372