Sp1) Regulates the Centrochromatin Landscape and Centromeric Transcription During Mitosis by Aislinn Rebecca Sowash Molinari

Transcription Factor Specificity Protein 1 (Sp1) Regulates the Centrochromatin Landscape and Centromeric Transcription During Mitosis By Aislinn Rebecca Sowash Molinari

July 2016

A Dissertation Presented to the Faculty of Drexel University College of Medicine in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in Molecular and Cellular Biology and Genetics

Chairperson Dr. Jane Clifford, Ph.D. Dr. Eishi Noguchi, Ph.D. Professor and Chair; Associate Associate Professor; Director, Dean for Medical Student Research Graduate Program in Molecular & Department of Biochemistry & Cellular Biology & Genetics Molecular Biology Department of Biochemistry & Molecular Biology

Dr. Michael Bouchard, Ph.D. Dr. Elias Spiliotis, Ph.D. Director, Division of Biomedical Associate Professor; Director of the Science Programs; Associate Cell Imaging Center Professor Department of Biology Department of Biochemistry & Drexel University Molecular Biology

Dr. Timothy Yen, Ph.D. Professor; Facility Director, Biological Imaging Facility Fox Chase Cancer Center

Transcription Factor Specificity Protein 1 (Sp1) Regulates the Centrochromatin

Landscape and Centromeric Transcription During Mitosis

Aislinn Rebecca Sowash Molinari

July 2016

A Dissertation Presented to the Faculty of

Drexel University College of Medicine

in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy in Molecular and Cellular Biology and Genetics

Aislinn Rebecca Sowash Molinari

2016

DEDICATION

To my brother Graham, who is a rock star and inspiration, and to my son Harrison, who

helped me finish.

ACKNOWLEDGEMENTS

I would like to begin by thanking my thesis advisor, Dr. Jane Azizkhan-Clifford, for the opportunity to work in her lab and to grow as a scientist under her mentorship. Dr.

Clifford taught me how to be a critical, independent thinker, providing continued support and encouragement. She provided me with invaluable scientific advice, but only after I had exhausted all of my own resources. She also supported me while I navigated some complicated personal matters in a professional, yet comfortingly maternal manner. Dr.

Clifford had continued faith in me, even when I didn’t have faith in myself, and for that I am deeply grateful. I would also like to thank the members of my thesis committee, including Dr. Eishi Noguchi, Dr. Michael Bouchard, Dr. Timothy Yen, and Dr. Elias

Spiliotis, for your time, suggestions, critiques, and encouragements over the years.

Finally, I would like to thank the office staff, including Jenny Sherwood, Kate Maum, and

Lucia Boyer, for your resources, support, and friendship.

To both current and past Clifford lab members – Thank you for your friendship and your support over the years, and for making the Clifford Lab an enjoyable place to work. I would especially like to thank Bill Donegan, who I have had the pleasure of working with for the majority of my time in graduate school. I have held your scientific suggestions and advice in the highest regard, and I have appreciated your willingness to listen to my grievances, both scientific and otherwise. I would also like to thank three of my mentees, Carol Stojinski, Sam Flashner, and Jacob Havens, for providing me with

ii the opportunity to grow as a mentor and teacher. Finally, I would like to thank Yi Guo for your friendship through our greatest experiment yet.

Lastly, I would like to thank my family for their love and support. Most importantly, thank you to my husband and best friend, Jonathan Molinari. You have loved me, supported me, and believed in me, and I could not have done any of this without you.

iii

TABLE OF CONTENTS

Abstract…………………………………………………………………………………………xiii

Chapter 1: Whole Chromosomal Instability……………………………………………….1

Introduction………………………………………………………………………………2

What is Numerical/Whole CIN (W-CIN)?…………..…………………………………3

Causes for Generation of W-CIN……………………………………………………...5

Centrosome Amplification……………………………………………………..5

Disengagement………………………………………………………...7

Procentriolar formation……………………………………………...... 8

Elongation………………………………………………………………8

Disjunction/Maturation and Movement……………………………....9

Causes for Centrosome Amplification……………………………….9

Centrosome Amplification and Chromosomal Instability…………10

Sister Chromatid Cohesion Defects………………………………………...10

Sister-Chromatid-Cohesin Complexes……………………………..11

Loading and Establishment of Sister Chromatid Cohesion……...12

Removal of Sister-Chromatid-Cohesin Complexes………………13

Sister Chromatid Cohesion Defects and Chromosomal

Instability……………………………………………………………....13

Improper Kinetochore-Microtubule Attachment……………………………14

Aurora B Kinase and the Chromosomal Passenger Complex for

Attachment Correction……………………………………………….16

Improper Kinetochore-Microtubule Attachment and Chromosomal

Instability………………………………………………………………17

Weakening of the Spindle Assembly Checkpoint………………………….18

The Mitotic Checkpoint Complex…………………………………...19

Silencing the SAC……………………………………………….……20

Weakening of the SAC and Chromosomal Instability…………….21

Clinical Significance of W-CIN………………………………………………………..22

Functional Consequences of W-CIN………………………………………..23

Clinical Consequences……………………………………………………….25

Clinical Diagnosis of W-CIN………………………………………………….26

Exploitation of W-CIN for Anti-Cancer Therapies………………………….29

Conclusions…………………………………………………………………………….30

Chapter 2: Centromere Biology…………………………………………………………….37

Introduction……………………………………………………………………………..38

Human Centromeric DNA…………………………………………………………….40

Defining Centromeres: Centromeric Protein A (CENP-A)……………………… ..42

The human CENP-A Gene and Protein…………………………………….42

The CENP-A Nucleosome…………………………………………………...43

CENP-A Deposition at Centromeres………………………………………..46

Holliday junction recognition protein (HJURP)……………………47

The Mis18 Complex………………………………………………….48

CENP-C and CENP-I………………………………………….……..50

Other DNA Binding Proteins at Centromeres………………………………………51

Centromeric Protein B (CENP-B)……………………………………………51

Centromeric Protein C (CENP-C)………………………………………...…52

The CENP-T-W-S-X Heterotetramer………………………………………..54

The Centrochromatin: Centromere-Specific Histone Modifications…………..….55

Histone Modifications at the Core Centromere…………………………….56

Histone Modifiers………………………………………………………….…..57

Transcription Through Core Centromeres…………………………………………..58

Evidence for Transcription at Core Centromeres in Human Cells……….58

Transcription Factors at Centromeres in Human Cells…………………...59

Timing of Core Centromere Transcription………………………………….59

Human Artificial Chromosomes (HACs)…………………………………….60

A Functional Requirement for Transcription at Core Centromeres……...62

Dysregulation of Centromeric Transcription………………………………..64

Conclusions…………………………………………………………………………….65

Chapter 3: Specificity Protein 1……………………………………………………….……71

Introduction………………………………………………………………………..……72

The Specificity Protein/Krüppel-like Factor (SP/KLF) Transcription Factor

Family…………………………………………………………………………………...73

The Sp1 Gene and Protein(s)………………………………………………………..74

Sp1 and Gene Regulation…………………………………………………………….75

Sp1 as a General Transcription Factor……………………………………..75

Sp1 as a Specific Transcription Factor……………………………………..76

Post-Translational Modifications (PTMs) for Sp1…………………………………..78

Sp1 and Mitosis………………………………………………………………………..80

Sp1 and Whole Chromosomal Instability (W-CIN)…………………………………81

Sp1 and Centrosome Regulation……………………………………………82

Sp1 and Survivin………………………………………………………………82

Sp1 and Cancer………………………………………………………………………..84

Conclusions…………………………………………………………………………….85

Chapter 4: Transcription Factor Sp1 Regulates the Centrochromatin Landscape and Centromeric Transcription During Mitosis………………………………………….90

Introduction………………………………………………………………………….....91

Sp1 Localizes to Centromeres in Mitotic Cells……………………………………..92

Sp1 Binds to Centromeres and Pericentromeres in MCF 10A Cells…………….93

Sp1 Binding to Centromeres is Not Dependent on the Sp1 Zinc-Finger,

DNA-Binding Domain………………………………………………………………....93

ATM Activity Is Required for Sp1 Localization to Centromeres…………………..94

Sp1 Knockdown Results in a Decrease in CENP-A at Centromeres…………….95

Decreased CENP-A at Centromeres is Not Due to Reduced Expression of

CENP-A…………………………………………………………………………………96

Sp1 Knockdown Results in a Decrease in CENP-C at Centromeres……………97

Sp1 Contributes to Regulation of α-Satellite-Derived Long Non-Coding RNAs...98

Over-Expression of Sp1 Results in a Decrease in CENP-A at Centromeres….100

Sp1 Depletion Disrupts the Centrochromatin Landscape………………………..101

Conclusions…………………………………………………………………………...102

Chapter 5: Sp11-182 May Be Sufficient for Functional Centromeres…………………121

Introduction……………………………………………………………………………122

Sp11-182 Is Sufficient for DNA Double-Strand-Break Repair……………………...124

Sp11-182 Is Sufficient for Centrosome Regulation………………………………….124

Sp11-182 Rescues Centromere Distance Phenotype………………………………125

Sp11-182 Rescues CENP-A Intensity Phenotype by Immunofluorescence

Imaging………………………………………………………………………………..126

Sp11-182 Partially Rescues CENP-A Binding by Chromatin

Immunoprecipitation………………………………………………………………….126

Conclusion…………………………………………………………………………….127

vii

Chapter 6: Discussion and Future Directions………………………………………….132

Introduction……………………………………………………………………………133

Discussion of Presented Results…………………………………………………...133

Transcription Factor Sp1 Regulates the Centrochromatin Landscape

and Centromeric Transcription During Mitosis…………………………...134

Sp11-182 May Be Sufficient for Functional Centromeres………………….140

Current Model………………………………………………………………..142

Future Directions……………………………………………………………………..142

Identifying Sp1 Interacting Partners at Centromeres……………………142

Linking Histone Modifications to Transcription…………………………...144

Demonstrating a Decrease in CENP-A Over Time………………………145

Completing the Sp11-182 Story………………………………………………145

Linking the Function of Sp1 at Centromeres to Chromosomal

Instability……………………………………………………………………..146

Identifying Post-Translational Modifications Required for

Localization…………………………………………………………………..147

Effects of Sp1 Knockdown on the Chromosomal Passenger

Complex………………………………………………………………………148

Understanding the Centromeric Cohesion Defect in Sp1-depleted

Cells…………………………………………………………………………..149

Chapter 7: Experimental Procedures……………………………………………………155

Cell Lines and Culture Conditions………………………………………………….156

Plasmids and Viral Infections……………………………………………………….156

Transduction of MCF 10A Cells…………………………………………………….157

DNA Spreads and Immunofluorescence…………………………………………..157

Western Blotting……………………………………………………………………...158

viii

Chromatin Immunoprecipitation…………………………………………………….159

Co-Immunoprecipitation…………………………………………………………...... 160

RNA Isolation, cDNA Synthesis and Quantitative PCR………………………… 161

Sp1183-785 Stabilization Experiment………………………………………………….161

List of References…………………………………………………………………………...162

LIST OF TABLES

Table 1.1: Proteins Associated with W-CIN………………………………………...34

Table 1.2: Clinical relevance of W-CIN……………………………………………...36

Table 3.1: Sp1 Post-Translational Modifications…………………………………...88

Table 3.2: Cancer-related genes regulated by Sp1………………………………..89

Table 4.1: Primers Used in Described Experiments……………………………...120

Table 6.1: Sp1 SQ/TQ Sites………………………………………………………...154

LIST OF FIGURES

Figure 1.1: Acquired Capabilities of Cancer………………………………………..32

Figure 1.2: Genomic Instability……………………………………………………….33

Figure 2.1: Overview of Mitotic Chromatin………………………………………….67

Figure 2.2: The Constitutive-Centromere-Associated Network (CCAN) and the

Kinetochore………………………………………………………………………….....68

Figure 2.3: Human A-satellite Consensus Sequence……………………………...69

Figure 2.4: Histone Modifications at the Centrochromatin………………………...70

Figure 3.1: Domain Features of Sp1, Sp2, Sp3, and Sp4…………………………87

Figure 4.1: Sp1 Localizes to Centromeres in Mitotic Cells………………………104

Figure 4.2: Temporal Localization of Sp1 at Centromeres………………………106

Figure 4.3: Transcription Factor Sp1 Localizes and Binds to Centromeres and

Pericentromeres in Mitotic Cells……………………………………………………107

Figure 4.4: Sp1 Binding to Centromeres is Not Dependent on the Sp1 Zinc-

Finger, DNA-Binding Domain……………………………………………………….108

Figure 4.5: ATM Activity Is Required for Sp1 Localization to Centromeres……109

Figure 4.6: Sp1 Knockdown Results in a Decrease in CENP-A at

Centromeres………………………………………………………………………….110

Figure 4.7: Sp1 Binds to CENP-A Promoter, but Does Not Influence CENP-A mRNA or Protein Levels Upon Knockdown ………………………………………112

Figure 4.8: Sp1 Knockdown Results in a Decrease in CENP-C at

Centromeres………………………………………………………………………….113

Figure 4.9: Sp1 Contributes to Regulation of α-Satellite-Derived Long Non-

Coding RNAs…………………………………………………………………………114

Figure 4.10: Sp1 Depletion Results in an Increase in RNA Polymerase IIpS5

Binding at Centromeres……………………………………………………………..115

Figure 4.11: Over-Expression of Sp1 Results in a Decrease in CENP-A at

Centromeres……………………………………………………………………….....116

Figure 4.12: Sp1 Depletion Disrupts the Centrochromatin Landscape………...117

Figure 4.13: Control Western Blots………………………………………………...119

Figure 5.1: Cleavage of Sp1 at Aspartic Acid 183 Stabilizes Sp1183-785………..128

Figure 5.2: Sp11-182 Rescues Centromere Distance Phenotype…………………129

Figure 5.3: Sp11-182 Rescues CENP-A Intensity Phenotype by

Immunofluorescence Imaging……………………………………………………...130

Figure 5.4: Sp11-182 Partially Rescues CENP-A Binding by Chromatin

Immunoprecipitation …………………………………………………………………131

Figure 6.1: Model…………………………………………………………………….152

Figure 6.2: Understanding the Centromeric Cohesion Defect in Sp1-depleted

Cells…………………………………………………………………………….……..153

xii

ABSTRACT

Transcription Factor Specificity Protein 1 (Sp1) Regulates the Centrochromatin Landscape and Centromeric Transcription During Mitosis

Aislinn Rebecca Sowash Molinari

Jane Azizkhan-Clifford, Ph.D.

Chromosomal instability (CIN) is a dynamic and continual gain or loss of whole chromosomes, or parts of chromosomes, during cell division. It is associated with poor patient outcome in multiple cancer types, as well as tumor heterogeneity and resistance to multiple chemotherapeutics, underscoring its clinical importance. Despite its prevalence and clinical significance, the exact mechanisms that lead to CIN remain to be determined.

The transcription factor Specificity Protein 1 (Sp1) regulates the transcription of genes involved with many cellular processes, including differentiation, cell cycle progression, DNA repair, apoptosis, and senescence. Sp1 binds to specific GC-rich elements through its highly conserved carboxy-terminal zinc-finger, DNA-binding domain, and recruits various factors to chromatin to influence transcription. Our previous work shows that Sp1 is important for maintaining chromosomal stability during mitosis.

We have shown that loss of Sp1 results in abnormal chromosome alignment along the metaphase plate, creation of micronuclei, and aneuploidy, as well as lagging chromosomes and anaphase bridges, all of which are phenotypes consistent with CIN.

xiii

We now show that Sp1 localizes and binds to centromeres during mitosis. Rapid localization is dependent on ATM (ataxia telangiectasia mutated) activity, and does not require the Sp1 DNA-binding domain. Loss of Sp1 results in disrupted centrochromatin, including changes in histone modifications and transcription of α-satellite arrays. Further, loss of Sp1 results in defects in centromeric cohesion, as well as a decrease in

Centromeric Protein C (CENP-C) and Centromeric Protein A (CENP-A) binding at centromeres. These data suggest that Sp1 is an important factor for maintaining the structure, function, and identity of centromeres, thereby maintaining chromosomal stability.

xiv

Chapter 1: Whole Chromosomal Instability

Introduction

A fundamental process important for the health and survival of an organism is the ability of the cells to accurately replicate their genome, and to pass a complete and intact copy of their genome on to daughter cells. There are several surveillance systems within the cell to ensure that this occurs with little genomic variation. DNA replication is monitored to ensure minimal errors during DNA synthesis, and the DNA damage response marks and repairs areas of damaged DNA. Additionally, the cell employs cell cycle checkpoints to respond to errors in DNA replication and/or DNA damage, and stalls the cell cycle to allow for repair. Finally, a checkpoint during mitosis ensures that chromosomes are divided evenly between the two daughter cells. When these governing mechanisms break down, a cell can acquire a variety of genomic abnormalities, collectively classified as genomic instability. The consequences of genomic instability can be devastating to the health of the organism, and can result in pathological disease, including cancer.

Genomic instability is described as an increased tendency for the genome to acquire mutations [1], and is a characteristic of almost all human cancers [2, 3]. When all cellular surveillance systems are functioning, the frequency of normal genetic variation is extremely low, at 10-8 bases per generation [4]. However, an increased mutation frequency results when one or more of these mechanisms becomes dysfunctional [1].

Genomic instability provides a cell with the opportunity to acquire the “Hallmarks of

Cancer” as described by Hanahan and Weinberg [5], a collection of characteristics that allows a normal cell to transform into a cancerous cell (Figure 1.1). The defective surveillance mechanism will dictate the type of mutation that results, and can include (a) changes on the nucleotide level, such as point mutations and microsatellite instability,

(b) changes in chromosome structure, including gross chromosomal rearrangements and copy number variations, and (c) changes in the number of whole chromosomes in

2 the nucleus [6]. The latter two types of mutation are classified as chromosomal instability

(CIN), including structural/segmental CIN (S-CIN) and numerical/whole CIN (W-CIN), respectively (Figure 1.2).

We have previously shown that Specificity Protein 1 (Sp1, discussed in detail in

Chapter 3) is required for maintaining chromosomal stability. In this study, depletion of

Sp1 resulted in several phenotypes consistent with W-CIN, including the formation of micronucleated cells, mis-alignment of chromosomes along the metaphase plate, and aneuploidy [7]. Although this study identified supernumerary centrosomes and a multipolar mitotic spindle as a likely contributing factor to the generation of W-CIN in Sp1- depleted cells, my work described in Chapters 4 and 5 suggests that Sp1 is also involved with maintaining centromere identity and function, which is the foundation upon which the mechanisms for preventing W-CIN are built. To bring these recent findings into context and to appreciate the importance of this function for Sp1, this chapter will focus on the significance of W-CIN, including a description of this phenomenon, its clinical relevance, current understanding of the mechanisms that may generate W-CIN in cancers, and its potential as an anti-cancer therapeutic target.

What is Numerical/Whole CIN (W-CIN)?

W-CIN is the dynamic and continual gain or loss of whole chromosomes at an elevated rate during cell division [8]. This continuous chromosome mis-segregation is a major cause for aneuploidy, defined as a deviation from an exact multiple of the normal haploid number of chromosomes in a cell [9]. Over 100 years ago, German zoologist

Theodor Boveri discovered while studying sea urchin embryos that aneuploidy can have a detrimental effect on cell physiology, and proposed that abnormal chromosome content may promote cancer development [10]. Aneuploidy is now known to be a common characteristic of human cancers [11]. While W-CIN always results in

3 aneuploidy, it is important to note that W-CIN and aneuploidy are not synonymous. By definition, W-CIN is dynamic and continual. Therefore, a population of cells that has stably acquired or lost whole chromosomes can be aneuploid, but not necessarily chromosomally unstable. Examples of this include hyperdiploid acute lymphoblastic leukemia and near-triploid neuroblastoma, both of which have stable aneuploid karyotypes [12, 13].

There are several different mechanisms that govern proper chromosome segregation during mitosis (discussed in detail below), and it is clear that when these mechanisms are experimentally disrupted, chromosomes can mis-segregate, resulting in

W-CIN. Whether these mechanisms become defective and are a driving force for cancer development, or simply a consequence or side effect of cell transformation, remains a hotly debated topic. Some contest that W-CIN is an initiating event [14], with mathematical modeling in support of this theory [15]. The resulting karyotype from the initial event sets the stage for further W-CIN, generating an increasingly abnormal and eventually cancer-causing combination of chromosomes. Others believe that the expression of oncogenes and/or the loss of tumor suppressor genes are responsible for cell transformation, with W-CIN developing as a consequence [16]. It has been challenging to determine scientifically if W-CIN is an initiating event, in part because W-

CIN creates highly heterogeneous tumors, making identifying an original mis-segregated chromosome difficult. Further, tumors with W-CIN often have other chromosomal abnormalities, including S-CIN. In these cases, it is less likely assumed that W-CIN is responsible, as many specific S-CIN rearrangements have been directly implicated in malignancy. Large scale sequencing of cancer genomes has started to provide additional information as to what mutations are commonly associated with W-CIN. These studies have identified specific “CIN genes”. However, the question remains whether mutations or changes in expression levels of these CIN genes actually drive

4 tumorigenesis, or if these changes are simply a consequence. The gene products identified as being involved with CIN are summarized in Table 1.1 [36]. Thus, although

W-CIN is a characteristic of almost all cancers, a consistent basis for how W-CIN develops in tumors is still not clear. Further, as discussed below, W-CIN is associated with poor patient prognosis. Therefore, there is much to be gained from elucidating the cause(s) for W-CIN, with great potential for therapeutic applications.

Causes for Generation of W-CIN

Within the W-CIN field, most researchers agree that there are four main mechanisms that govern proper chromosome segregation, and when these mechanisms are disrupted, chromosome mis-segregate, leading to W-CIN. These disruptions include centrosome amplification, sister chromatid cohesion defects, improper kinetochore- microtubule attachment, and weakening of the mitotic checkpoint. The research presented in Chapters 4 and 5 identify Sp1 as an important factor for maintaining the architecture of centromeres (centromeres are discussed in detail in Chapter 2). Each of the mechanisms governing chromosome segregation relies heavily on active centromeres for function, suggesting that Sp1 may be a vital upstream factor for preventing chromosome mis-segregation and W-CIN. To bring our findings into context, this section summarizes each of the mechanisms that ensure faithful chromosome segregation, as well as how disruption of these mechanisms contributes to W-CIN.

Centrosome Amplification:

Centrosomes are the microtubule organizing centers (MTOC) of the cell. These organelles are composed of a pair of centrioles, surrounded by a matrix of hundreds of proteins called the pericentriolar material (PCM). The PCM consists of proteins that are important for many cellular processes, and include cell cycle regulators and signaling molecules, as well as proteins that are important for the nucleation, growth, and

5 organization of microtubules [17]. As the MTOC, centrosomes are central in orchestrating several cellular processes through the nucleation and anchoring of microtubules. These processes include cell motility, cell signaling, adhesion, protein trafficking, cell polarity, and the division of chromosomes during mitosis [18].

When a cell enters mitosis, it should contain two centrosomes that migrate to opposite ends of the cell with the assistance of microtubule-anchored motor proteins, to create a bipolar mitotic spindle. At each pole, centrosomes nucleate a radial array of dynamic astral microtubules, which stochastically probe the cytoplasm with their plus ends in search of a kinetochore. Once the plus end of a microtubule comes into contact with a kinetochore, the microtubule is “captured” and stabilized [19]. Each chromosome must attach to microtubules emanating from both centrosomes, as these attachments create the tension necessary to align the chromosomes along the metaphase plate.

Once all chromosomes are correctly attached and aligned, the spindle assembly checkpoint (SAC) is satisfied (reviewed in detail below), and sister chromatids are pulled in opposite directions (anaphase), towards the centrosomes at the poles of the cell. This is followed by cytokinesis, generating two daughter cells with symmetrically segregated chromosomes.

The bipolarity of the mitotic spindle is necessary for ensuring that each daughter cell receives one of each pair of sister chromatids. Cancer cells frequently have more than two centrosomes, creating the potential for a multipolar spindle [20]. For example, nearly 80% of invasive breast tumor cells have an amplified number of centrosomes

[20]. Additionally, centrosome amplification was shown to be a characteristic of pancreatic, prostate, colorectal, lung, and gall bladder cancers, among others [21-25]. A multipolar spindle created from more than two centrosomes during mitosis can result in the segregation of chromosomes into more than two daughter cells, generating severe aneuploidy in those cells. Studies have shown that such a severe aneuploidy produces

6 daughter cells that are inviable, and therefore do not pose a threat to the health of the organism [26]. In order to survive, cancer cells cluster their extra centrosomes into a pseudo-bipolar configuration, allowing for segregation into two viable daughter cells.

However, this clustering process often creates improper kinetochore-microtubule attachments (reviewed in detail below), which promotes chromosome mis-segregation and W-CIN. As such, numerical centrosome amplification has emerged as an important mechanism for the generation of W-CIN.

Upon the completion of cytokinesis, each daughter cell must receive one centrosome, and that centrosome must only duplicate once in the subsequent cell cycle.

The centrosome duplication cycle consists of four stages, including disengagement in late mitosis and early G1, procentriolar formation in later G1 and S phases, elongation in later S and G2 phases, and finally disjunction or maturation as the cell enters mitosis

[27].

Disengagement:

In early G1, the centrosome contains two centrioles, tightly orthogonally oriented to one another and attached by electron dense fibers [28]. Disengagement, the first stage of the duplication cycle, is the so-called licensing step, and describes the cleavage of the fibrous material connecting the two centrioles so that the distance between the centrioles increases. Some studies have suggested that the connecting material consists of sister-chromatid-cohesin (SCC) complexes, and that separase, the same protease that cleaves centromeric SCC molecules during anaphase, acts at the centrosome.

Shockel, et al. showed that disengagement was suppressed by replacing endogenous

Scc1 (cohesin subunit) with non-cleavable Scc1. Additionally, when a cleavage-inducible form of Scc1 was used, disengagement could be artificially triggered [29]. However, other experiments, including those in Drosophila, have challenged this theory, where centriolar disengagement could not be triggered by cleavage-inducible Rad21, the

Drosophila Scc1 ortholog [30]. More recently, additional separase targets involved with centriolar disengagement have been identified, including kendrin [31]. Thus, a complete understanding of the centriolar disengagement process is still lacking. Once centrioles have disengaged, they begin the process of duplication.

Procentriolar formation:

The recently disengaged centrioles are known as the “mother” centrioles. One new centriole (a procentriole or daughter centriole) begins to grow in an orthogonal angle at the proximal end of each licensed mother centriole, maintaining a tight ‘side-to- base’ connection between the mother and daughter centriole [32]. The location of the daughter centriole in relation to the mother centriole is known as the origin, and once the origin is established, the cartwheel, a ninefold symmetric template structure inside of the proximal end of the daughter centriole, begins to form. Although how the origin is established is not fully understood, three proteins appear to play a vital role, and include

PLK4, SAS-6, and STIL. These proteins localize to the origin in later G1 and S phases during the time of procentriole initiation. Although there is currently no direct link between PLK4 kinase activity and centriole initiation, it’s believed that PLK4 phosphorylation of downstream proteins in centriole assembly initiates the assembly process [33].

Elongation:

Daughter centriole elongation begins during S phase, and centrioles reach ∼80% the length of the maternal centriole in late G2 [34]. Centrioles reach a maximum length of about 500 nanometers, and a width of about 200 nanometers through the addition of heterodimers consisting of α-tubulin and β-tubulin [35]. The mechanisms that control centriole elongation and length determination are still relatively unclear. In recent years, two proteins have been shown to be important regulators of elongation, including the

8 centriolar protein CPAP and the distal end-capping protein CP110. Overexpression of

CPAP results in elongated centriolar structures, as does depletion of CP110 [36].

Disjunction/Maturation and Movement:

During late G2 and early mitosis, the two sister centrosomes separate from each other in a process called disjunction or maturation, then migrate to opposite poles of the cell. During the process of centrosome duplication, two linker proteins, C-Nap1 and rootletin, join the centrosomes together. At the beginning of mitosis, these two proteins are phosphorylated by Nek2A kinase, displacing C-Nap1 and rootletin from centrosomes, a process necessary for their subsequent migration. Nek2A is regulated upstream by PLK1 [37]. Centrosomes are then pushed apart by the force of microtubule- dependent motor proteins. This occurs at the very beginning of mitosis, simultaneously with nuclear envelope breakdown. One motor protein that appears to be essential for this process is the plus end directed motor Eg5 [38].

Causes for Centrosome Amplification:

Centrosome amplification can result from a variety of abnormalities, including cytokinesis failure, mitotic slippage, cell–cell fusion, overduplication of centrioles, and de novo centriole assembly. However, the major cause for centrosome amplification is disruption in the regulation of the duplication cycle. The duplication cycle is regulated at each of the different stages, with PLK4 as the “master regulator” of centrosome duplication. PLK4 plays a major role in procentriolar formation, and it is clear that the cellular levels of PLK4 must be tightly controlled to ensure proper centrosome number.

When PLK4 is experimentally decreased, centrosome number decreases as well [39].

Likewise, when PLK4 is increased, centrosome numbers increase [40]. PLK4 regulation is achieved through several different mechanisms. First, the SCF/Slimb ubiquitin ligase binds to and facilitates the degradation of PLK4 to prevent unintended duplication. When this ubiquitin ligase is absent, PLK4 protein accumulates, and the resulting phenotype

9 includes an increase in daughter centrioles [41]. Further, PLK4 mRNA levels are negatively regulated by p53 through recruitment of HDAC (histone deacetylases) repressors to the PLK4 promoter [42]. In mouse cells, loss of p53 has been associated with an increase in centrosome number [43]. Increased levels of CP110, the distal end- capping protein important for regulating centriole elongation, has also been shown to result in centrosome amplification [44]. Finally, over-expression of protein components of the PCM, the matrix of hundreds of proteins surrounding the centrioles, can also result in centrosome amplification. For example, over-expression of pericentrin results in centrosome amplification [45], as well as increased levels of γ-tubulin within the PCM

[46].

Centrosome Amplification and Chromosomal Instability:

Centrosome amplification is frequently detected in W-CIN positive cancers, including breast, prostate, colon, ovarian, and pancreatic cancers [21, 47-49]. In addition to these solid tumors, centrosome amplification has been described in hematological malignancies, including multiple myeloma, non-Hodgkin's and Hodgkin's lymphomas, and acute and chronic myeloid leukemia [50, 51]. Studies suggest that centrosome amplification may be a major player in the progression from early to advanced stages of carcinogenesis, as this phenotype has been characterized in both pre-neoplastic lesions and tissue that presents as histopathologically normal. One mechanism by which centrosome amplification may contribute to W-CIN is by increasing the frequency of merotelic attachments (discussed below) [52]. A recent review summarizes in great detail the existing clinical data for centrosome abnormalities and cancer (refer to reference [53]).

Sister Chromatid Cohesion Defects:

During prometaphase, spindle microtubules probe the cytoplasm of the cell in search of kinetochores, which function to connect each set of sister chromatids to the

10 bipolar mitotic spindle. The sisters must connect to microtubules emanating from opposite poles of the cell, which serves two purposes. First, the tension created from the bipolar connections allows the sister chromatids to align along the metaphase plate.

Second, when the mitotic checkpoint is satisfied and the cell proceeds into anaphase, the sister chromatids are pulled in opposite directions, ensuring faithful segregation. This mechanism of chromosome segregation critically depends on sister chromatids remaining physically connected to each other from the time of DNA replication, until the onset of anaphase. This is achieved by a complex of proteins called the sister- chromatid-cohesin (SCC) complex. If sister chromatid cohesion fails, sister chromatids can separate prematurely, resulting in W-CIN.

Sister-Chromatid-Cohesin Complexes:

The somatic mammalian SCC complex consists of three main proteins that form a ring-like structure, including Smc1, Smc3, and Scc1. Smc1 and Smc3 are part of the structural maintenance of chromosome family (Smc), a family of proteins characterized by a distinct domain organization. Proteins in the Smc family contain a hinge domain, allowing for the proteins to fold back on themselves at this region, resulting in an antiparallel coiled-coil structure between the hinge domain, and the N- and C-termini of the proteins [54]. The N- and C-termini of the proteins come together to form an ATPase

“head”. The hinge domains from Smc1 and Smc3 bind to each other, and the ATPase heads from each of these proteins bind the third component of the ring structure, Scc1

[55]. When assembled, the outer diameter of the cohesin ring structure is approximately

50 nanometers [56]. A fourth subunit of the SCC complex is the SA1 (stromalin antigen

1) or SA2 protein, which associates with Scc1. Interestingly, complexes either contain

SA1 or SA2, but never both [57]. It is predicted that in vertebrate somatic cells, there are three fold more complexes containing SA2 than complexes containing SA1, but the functional difference between these two associated proteins is not yet clear [58].

Meiotic cells contain cohesin complexes that are distinct from those found in somatic cells. This is likely because during meiosis I, kinetochores on sister chromatids must attach to the same spindle pole for separation of homologous chromosomes, thus giving meiotic SCC complexes a slightly different function. These complexes consist of

Smc1β in place of Smc1, Rec8 in place of Scc1, and STAG3 in place of SA1 or SA2 [59,

60].

Loading and Establishment of Sister Chromatid Cohesion:

In most eukaryotes, SCC complexes are loaded onto DNA prior to DNA replication. Although it has been demonstrated that no specific DNA sequence is required for cohesin complex loading, the density of SCC complexes is highest around centromeres, as well as in regions called cohesin-associated regions (CARs). CARs span about 1 kilobase of DNA, and tend to be adenine and thymine rich [61, 62]. SCC complexes appear to be loaded onto DNA by the Scc2/Scc4 complex, as this complex is required for cohesin association with chromatin, and also requires the activity of the acetyltransferase Ctf7/Eco1, but much about how the complexes are loaded remains unclear [63]. This process likely involves transient opening of the Smc1 and Smc3 hinge, as tethering of these regions together abolishes loading [64].

There are four main models for how SCC complexes associate with DNA and maintain cohesion between sisters, including the one-ring model, the two-ring model, the bracelet model, and the handcuff model. The one-ring model is most simple, and predicts that Smc1, Smc3, and Scc1 form a ring around DNA prior to DNA replication, and sister chromatid cohesion is established when the replication fork passes through the ring. This model is largely supported by the fact that it has been difficult to identify protein-protein interactions between cohesin complexes. However, the size of the ring is thought to be too small to accommodate two chromatids [65]. The two ring model suggests that each sister chromatid is bound by the Smc1-Smc3 heterodimer, and

12 cohesion is formed by a single Scc1 molecule during DNA replication [66]. The bracelet model is similar to the two ring model, with the added feature that Scc1 connects several

SCC complexes together to form extended filaments [67]. Finally, in the handcuff model, each chromatid is embraced by a single SCC complex, and the Scc1 and SA1/2 proteins on the respective complexes bind in order to pair the sisters together [68].

Removal of Sister-Chromatid-Cohesin Complexes:

SCC complexes are removed from chromatin through two separate mechanisms.

First, arm complexes are removed at the beginning of mitosis, during prophase. Here, the SA1/2 subunits are phosphorylated by kinases Plk1 (polo-like kinase 1) and Aurora

B. This results in the destabilization of the complexes, and their subsequent removal. In cells containing SA1/2 that could not be phosphorylated, cohesin complexes failed to dissociate from chromosome arms [69]. Cohesin is maintained at centromeres until the mitotic checkpoint is satisfied. Once the mitotic checkpoint is satisfied, the protease separase is activated, and subsequently cleaves Scc1 to release centromere cohesion and allow sister chromatids to be pulled to opposite poles of the cell. Centromeric cohesion is protected from Plk1 and Aurora B-mediated dissociation through the activities of Sgo1 (Shugoshin 1) and PP2A (protein phosphatase 2A). Sgo1 is thought to recruit PP2A to centromeres, and together, these proteins function to continuously dephosphorylate SA1/2 at this specific region, protecting centromeric SCC complexes from dissociating prematurely [70].

Sister Chromatid Cohesion Defects and Chromosomal Instability:

Defects in sister chromatid cohesion resulting in precocious separation of sister chromatids has been described in a variety of diseases. For example, Smc1, Smc3 and

SA1/2, as well as the cohesin loading complex component Scc2, have all been found to be mutated in colon cancers. In a 2008 study, DNA from 132 W-CIN positive colorectal tumors were sequenced and compared to matched normal DNA from the same patients

13 to eliminate the possibility of germ line polymorphisms. Results showed four tumors with mutations in Smc1, one tumor with mutated Smc3, one tumor with mutated SA2, and four tumors with mutated Scc2. To confirm their hypothesis that W-CIN likely resulted from reduced activity of the gene products that were found to be mutated, the authors used RNAi to decrease the expression of each protein in the W-CIN negative colorectal carcinoma cell line HTC 116, and subsequently analyzed cells for W-CIN by flow cytometry and evaluation of metaphase spreads. Flow cytometry results suggested a heterogeneous population of cells with extra chromosomes, and metaphase spread analysis confirmed these results [71]. In addition to colorectal cancers, an increased distance between the primary constriction site in Giemsa stained chromosome spreads, as well as aneuploidy, was observed in Wilms tumors in several infant patients, implicating sister chromatid cohesion defects [72-74]. In one of these studies, one patient also presented with rhabdomyosarcoma, a rare malignant tumor of striated muscle tissue, while two other patients also presented with acute leukemia [74]. Another study examined the relevance of cohesion defects and subsequent aneuploidy in lymphocyte cultures from individuals at high risk for familial breast cancer. This study found that a significant proportion of breast cancer patients, patients with benign breast lesions, and unaffected members from breast cancer families presented with cohesion defects and aneuploidy [75]. Premature separation of sister chromatids has also been described in Roberts syndrome [76], Fanconi Anemia, Ataxia Teleangiectasia [77],

Alzheimer disease [78, 79], and Tuberous Sclerosis [80].

Improper Kinetochore-Microtubule Attachment:

Live cell imaging has demonstrated that one of the most common causes of W-

CIN is the persistence of errors in the attachment of spindle microtubules to chromosomes [81]. In order for chromosomes to become bi-oriented on the metaphase plate, sister kinetochores must attach to microtubules emanating from opposite poles of

14 the cell. Further, the kinetochore on one sister chromatid must attach to microtubules originating from a single spindle pole. This type of correct attachment is called amphitelic

[82]. Because the spindle microtubules probe the cytoplasm at random in search for kinetochore attachment sites, errors in the orientation of microtubule attachment can occur, particularly in the early stages of mitosis. One prominent error is when the kinetochore on a single chromosome attaches to microtubules from both spindle poles.

This type of attachment error is called merotelic [82, 83], and has the potential to be the most devastating for maintaining genomic integrity for two reasons. First, merotelic attachments can fail to activate the spindle assembly checkpoint (SAC, discussed in more detail below), a signaling cascade employed by the cell to identify and remove improperly attached microtubules. Rather, studies show that cells possessing merotelic attachments proceed into anaphase without significant delay [84, 85]. As such, this type of mal-attachment is the most likely defect to escape detection and persist into anaphase. Further, this type of attachment, if not corrected, will pull chromosomes towards both poles of the cell during anaphase, resulting in lagging chromosomes and anaphase bridges. Then, the cleavage furrow can push that lagging chromosome into either daughter cell, resulting in chromosome mis-segregation in 50% of cases, thus contributing to W-CIN [86]. Other types of incorrect microtubule attachments include monotelic and syntelic attachments. Monotelic describes when one sister kinetochore is attached to one spindle pole and the other sister kinetochore remains unattached, while syntelic describes when both sister kinetochores are attached to microtubules from the same spindle pole [82]. A monoletic attachment will activate the SAC, which will delay mitotic progression so that the error can be corrected. Syntelic attachments are likely to be destabilized by the Chromosomal Passenger Complex (CPC, discussed below), which in turn also activates the SAC.

Aurora B Kinase and the Chromosomal Passenger Complex for Attachment

Correction:

The CPC is a complex of proteins consisting of Aurora B kinase, INCENP (inner centromere protein), Borealin, and Survivin, and is thought to have multiple roles during mitosis. During early stages of mitosis, the CPC localizes to chromosome arms and centromeres, as Aurora B kinase together with Plk1 phosphorylates arm SCC complexes to allow for their dissociation from chromosomes. As the cell progresses into metaphase, localization is restricted to centromeres, where the CPC is thought to assist in recruiting other centromere-associated proteins (such as SAC proteins [87] the Sgo1

[88]), and to assist in destabilizing incorrectly attached microtubules.

Studies have shown that when CPC function is compromised, a cell fails to detach incorrectly attached microtubules, resulting in an increased persistence of merotelic and syntelic attachments. For example, in Saccharomyces cerevisiae, both

Ipl1 and Sli15 mutant strains (the Aurora kinase and INCENP orthologs, respectively) showed an increase in syntelic attachments as compared to control strains [89]. Further, in Ptk1 cells (female rat kangaroo kidney epithelial cells), partial Aurora kinase inhibition using the specific inhibitor ZM447439 increased the frequency of merotelic kinetochores in late metaphase. In this study, immunofluorescence analysis showed that treatment with the Aurora kinase inhibitor suppressed kinetochore-microtubule turnover in prometaphase [90]. By slowly removing reversible small-molecule Aurora B kinase inhibitors to carefully control Aurora B kinase activation in Ptk2 cells (male rat kangaroo kidney epithelial cells), another group confirmed that upon Aurora B activation, mal- oriented microtubules were selectively disassembled [91]. As such, the CPC is most abundant at kinetochores in early stages of mitosis when incorrect attachments are likely to occur, and becomes enriched at kinetochores showing merotelic and syntelic attachments [92].

How a cell detects improper microtubule attachment, and how Aurora B kinase destabilizes those incorrect attachments is not completely understood. The most widely accepted model proposes that the physical distance between Aurora B kinase and its substrate determines whether the microtubule connections are maintained. The Ndc80 complex, a component of the outer kinetochore KMN Network and a major attachment module for microtubules, is a target substrate for Aurora B kinase for this process.

Ndc80 phosphorylation by Aurora B kinase decreases its ability to bind microtubules.

When a kinetochore makes an attachment to a microtubule emanating from the correct spindle pole, tension is created across the kinetochore complex, much like a spring that is being stretched. Because Aurora B kinase is located at centromeres, the tension created by a correct attachment pulls the Ndc80 complex away from centromeres, and out of the kinase’s reach. Thus, Aurora B kinase is not able to phosphorylate the Ndc80 complex, and correct microtubule attachments are stabilized. However, when a kinetochore makes an attachment to a microtubule from the opposite spindle pole, the kinetochore “spring” is not stretched, and this lack of tension places the Ndc80 complex within the reach of Aurora B kinase, resulting in Aurora B kinase-dependent phosphorylation and microtubule destabilization [82]. This theory is supported by studies showing that in mammalian cells, a difference of about 1 to 3 micrometers is observed between a kinetochore that is under tension as compared to a kinetochore in a relaxed state [93]. The CPC is about 45 nanometers in length, allowing for the possibility that tension can physically separate the CPC and Ndc80 [94].

Improper Kinetochore-Microtubule Attachment and Chromosomal Instability:

Centrosome amplification, discussed above, can cause increased incidences of merotelic attachments and a resulting increase in lagging chromosomes and subsequently W-CIN [52]. However, in some cancers, lagging chromosomes occur in a larger percentage of cells than does centrosome amplification, suggesting that

17 centrosome amplification isn’t the only cause for merotelic attachments. For example, in oral cancer cells, 5.5–23% exhibit multipolar spindles from centrosome amplification, but lagging chromosomes occur in as many as 20-40% of cells [95, 96]. Further, 8% of the human breast cancer cell line MX-1 contained anaphase bridges with no evidence of centrosome amplification [97]. In a panel of W-CIN positive cancer cell lines, one study examined the stability of kinetochore-microtubule attachments, and found that the attachments are significantly more stable in the W-CIN positive cancer cell lines as compared to non-transformed control cells. Because merotelic attachments do not significantly delay anaphase onset, an increase in kinetochore-microtubule attachment stability results in reduced correction efficiency. As such, this increased stability correlated with elevated frequencies of lagging chromosomes in anaphase [98]. This same group also showed that over-expressing proteins required for kinetochore- microtubule detachment could decrease the incidence of merotelic attachments in W-

CIN positive cancer cells, regardless of the upstream cause for the mal-attachments. In this study, they over-expressed the microtubule-depolymerizing kinesins Kif2b and

MCAK, and then tested the impact of over-expression on chromosome segregation and

W-CIN. They found that over-expression of either kinesin significantly suppressed the incidence of lagging chromosomes in two W-CIN positive cancer cells lines, U2OS and

MCF-7 cell. Using fluorescence in situ hybridization (FISH) with chromosome-specific

AS DNA probes, they also quantified chromosome mis-segregation events in cells over- expressing Kif2b and MCAK, and found that mis-segregation events were significantly reduced in over-expressing cells as compared to controls [99].

Weakening of the Spindle Assembly Checkpoint:

The spindle assembly checkpoint (SAC), also known as the mitotic or metaphase checkpoint, functions to prevent mitotic cells from transitioning from metaphase to anaphase with kinetochores that lack microtubule attachment, or have incorrectly

18 attached microtubules. If a kinetochore is unattached or improperly attached, the SAC halts progression and thus safeguards the cell against chromosome mis-segregation and W-CIN by inhibiting the APC/C (anaphase promoting complex/cyclosome). The

APC/C is an E3 ubiquitin ligase that drives centromeric SCC complex cleavage and thus mitotic exit though two mechanisms. First, it polyubiquitinates and targets the protein securin for proteasomal degradation by the 26S proteasome. Securin binds to and inhibits separase, the protease that cleaves SCC complexes at centromeres.

Additionally, the APC/C also polyubiquitinates and targets cyclin B for degradation, and the resulting drop in cyclin B levels inactivates CDK1, driving mitotic exit. The SAC inhibits the APC/C by binding to the APC/C coactivator protein Cdc20, blocking the

Cdc20 substrate binding sites, thereby repositioning Cdc20 away from the APC/C thus blocking its activation [100]. As long as the SAC is active, the ACP/C remains inactive,

SCC complexes remain intact, and cyclin B levels remain elevated, collectively preventing mitotic exit. The complex of proteins that functions to inhibit the APC/C is known as the mitotic checkpoint complex (MCC).

The Mitotic Checkpoint Complex:

The MCC is a complex of proteins consisting of Mad2, the kinase BubR1, and

Bub3, as well as Cdc20, all of which are present at unattached kinetochores during mitosis. Whether these proteins exist as a complex at kinetochores, as they do in the cytosol, or are associated independently of each other remains unclear. The MCC proteins are recruited to unattached kinetochores through Aurora B and Msp1 kinase signaling, and bind to Cdc20 to prevent ACP/C activation. Models for how the MCC is recruited to unattached kinetochores are complex, and the mechanism by which the

MCC activates the SAC is still slightly controversial. As mentioned previously, Aurora B kinase localizes to centromeres as a component of the CPC to facilitate the release of mal-attached microtubules by phosphorylating the KMN Network component Ndc80.

Aurora B kinase activity is also required for recruitment of Mps1 kinase to kinetochores.

Once at kinetochores, Mps1 kinase phosphorylates the KMN Network component Knl1 at several Met-Glu-Leu-Thr (MELT) motifs, which creates a docking site for recruitment of additional SAC proteins [101, 102]. Bub1 kinase and its binding partner Bub3

(Bub1:Bub3 complex) recognize and bind to phosphorylated Knl1, which is followed by recruitment and binding of the BubR1:Bub3 complex. Interestingly, both Bub1 and

BubR1 bind to Bub3 through the same Bub3-binding or GLEBS domain. This is followed by Bub1-dependent recruitment of the Mad1:C-Mad2 complex, which catalyzes the formation of the Cdc20:C-Mad2 complex, thereby inhibiting Cdc20 [103].

Mad2 adopts two different conformational states. O-Mad2, or “open”-Mad2, is the unbound, cytosolic open conformation of the protein. When Mad2 binds to Mad1 or

Cdc20, two β-sheets move across the face of the protein to create the closed conformation (C-Mad2), with Mad1 or Cdc20 trapped within this fold. This creates either the Mad1:C-Mad2 complex or the Cdc20:C-Mad2 complex [103]. The process of binding to and inhibiting Cdc20 begins with the Bub1-dependent recruitment of a Mad1:C-Mad2 complex to kinetochores. Mad2 is able to dimerize, and as such, kinetochore-bound

Mad1:C-Mad2 recruits more O-Mad2 to kinetochores. When O-Mad2 dimerizes with the

Mad1:C-Mad2 complex, it is able to capture Cdc20, at which point it converts to C-Mad2 and subsequently forms the Cdc20:C-Mad2 complex [104-106]. Cdc20:C-Mad2 then recruits more O-Mad2, and catalyzes the formation of more Cdc20:C-Mad2 complexes, thereby amplifying the SAC signal [107].

Silencing the SAC:

As long as O-Mad2 is being recruited to kinetochores, the SAC signal will be maintained. As such, loss of Mad1:C-Mad2 from kinetochores halts Mad2 conversion and thus MCC formation, thereby extinguishing the checkpoint signal [104]. Two mechanisms contribute to dissociation of Mad1:C-Mad2 from kinetochores. The first is

20 mediated by dynein, a minus-end directed motor protein. Dynein localizes to kinetochores, and when microtubules make attachments, dynein is able to strip Mad1:C-

Mad2 from kinetochores by walking the complex towards the spindle poles. BubR1 is also removed by this mechanism [108, 109]. Interestingly, dynein does not localize to kinetochores in all eukaryotes, suggesting alternative mechanisms for SAC silencing

[110]. It is attractive to suggest that microtubule binding alone displaces the Mad1:C-

Mad2 complex. However, Mad1 can be recruited to attached kinetochores by Mps1 targeting, arguing against this hypothesis [111]. The second mechanism that contributes to the dissociation of Mad1:C-Mad2 from kinetochores is the most widely conserved, and involves the dephosphorylation of checkpoint proteins by PP1 (protein phosphatase 1)

[112]. PP1 associates with kinetochores through Knl1, and preventing this interaction results in sustained checkpoint signaling [113]. PP1 functions to remove the Bub1:Bub3 complex, and as a result Mad1:C-Mad2 [102].

Weakening of the SAC and Chromosomal Instability:

Complete loss of the SAC causes massive chromosome mis-segregation and cell death. However, weakening of the checkpoint can result in the mis-segregation of a few chromosome, and thus W-CIN. Many studies have been carried out to establish a causal relationship between mutations in genes encoding SAC proteins, and W-CIN. For example, several studies have linked mutations in the BubR1 gene to mosaic variegated aneuploidy, a condition that is characterized by a strong predisposition to cancer [114-

117]. One study identified mutations in the Bub1 and BubR1 genes in a panel of 19 aneuploid colorectal cancer cell lines [118], and in a screen of 49 W-CIN positive gastric cancer tissues and five gastric cancer cell lines, another study found that the Mad2 gene was mutated in 44.9% of the gastric tissues, as well as in one of the cell lines. Here, over-expression of mutated Mad2 in HeLa cells resulted in chromosome mis- segregation, implicating Mad2 in W-CIN development [119]. Further, a study

21 demonstrated that deletion of one Mad2 allele resulted in W-CIN in the human colon carcinoma cell line Hct-116 [120]. Despite these findings, many other attempts have failed to identify mutations in SAC proteins. For example, very few mutations were identified in Bub3, BubR1 and Bub1 genes in glioblastoma, breast, lung, bladder and thyroid cancers [121, 122]. Bub1 gene mutations are also rare in hematological cancers, head and neck cancers, and renal tumors [123-125]. Rather than gene mutation, what appears to be more common in cells with impaired SAC activity is over-expression of

SAC proteins. For example, in 181 gastric cancer samples examined, 50.3% showed

BubR1 over-expression, which correlated with aneuploidy, tumor invasiveness, metastasis likelihood and poor prognosis [126]. It was also found to be associated with

W-CIN in bladder cancer and clear cell kidney carcinomas [127, 128]. Similarly, Bub1 over-expression was found in breast tumor samples, salivary gland tumors, and gastric cancers [129-131]. Mad2 over-expression has been found in a number of cancers, including Familial Adenomatous Polyposis colorectal adenomas, advanced differentiated thyroid carcinomas, salivary duct carcinomas, and lung cancers [132-135].

Clinical Significance of W-CIN

Not only is W-CIN a characteristic of almost all cancers, it is also associated with poor patient prognosis [8], and thus has clinical relevance. The negative association between W-CIN and patient prognosis is likely driven by the tumor hererogeneity that results from W-CIN. This heterogeneity provides tumor cells with the opportunity to adapt to environmental stresses like chemotherapeutics, and thus may facilitate multidrug resistance, making treatment challenging. For these reasons, identifying the molecular mechanisms that most consistently generate W-CIN in cancers is paramount.

Once identified, this information may be exploited, and W-CIN can be used as a target for anti-cancer therapuetics. To highlight the importance of studying W-CIN with the goal

22 of identifying ways to exploit this common characteristic, this section focuses on the functional and clinical consequences of W-CIN, as well as current methods for diagnosing W-CIN in clinical settings, and how W-CIN may be utilized for anti-cancer therapeutics.

Functional Consequences of W-CIN:

The loss or gain of a whole chromosome can have profound effects on the genome of a cell, and can contribute to tumorigenesis through changes in gene dosage, including loss of heterozygosity (LOH) of tumor suppressor genes and/or amplification of oncogenes. Further, W-CIN results in tumor heterogeneity, which provides the tumor with the opportunity to adapt to environmental pressures such as chemotherapeutics, facilitating drug resistance.

The two-hit hypothesis, which was developed from studies of retinoblastoma in children [136], proposed that two “hits” to the retinoblastoma gene were required for cancer development. For the inherited form of this cancer, the first hit is a mutation in one of the two alleles in the germ line, and the second hit emerges in the other allele sometime during somatic cell division. For the sporadic form of retinoblastoma, mutations in both alleles arise during somatic cell division [15]. Tumor suppressor genes negatively regulate cell growth, thereby preventing uncontrolled proliferation. Inactivation of the first allele of a tumor suppressor gene often does not change the phenotype of the cell, whereas inactivation of the second allele can confer a growth advantage, contributing to tumorigenesis. LOH of tumor suppressor genes, defined as the loss of an entire gene and the surrounding region, can occur if the chromosome mis-segregation event results in the loss of a chromosome containing these genes. Thus, one functional consequence of W-CIN is an elevated rate of LOH, or the first hit of the two-hit hypothesis, which accelerates the potential for tumor suppressor gene inactivation [15].

Gaining a whole chromosome during chromosome mis-segregation can also result in increased expression of genes that promote tumorigenesis, or amplification of oncogenes. One example of this is in certain breast and ovarian cancers. Here, chromosome mis-segregation events can result in the loss of the heterochromatic X chromosome and a subsequent gain of the active X chromosome. The X chromosome contains both tumor suppressor and cancer-promoting genes, and studies have reported over-expression of these cancer-promoting genes in tumors, including BRCA1- associated ovarian tumors [137-139]. Further, in another study, a significant proportion of sporadic basal-like tumors and BRCA1 null tumors showed a complete loss of the heterochromatic X chromosome and the gain of several active X chromosomes, accompanied by the overexpression of a small, distinct subset of X-linked genes, suggesting that dysregulated X chromosomal genes contribute to tumor development

[140]. Thus, gaining whole chromosomes through W-CIN can amplify the expression of tumor-promoting or oncogenes, thereby contributing to tumorigenesis.

In addition to promoting tumorigenesis through LOH and/or gene amplification,

W-CIN also provides tumors with the opportunity to adapt to environmental stresses, including chemotherapeutics, allowing for the possibility of drug resistance and subsequently cancer reoccurrence. Several studies have confirmed this phenomenon.

For example, one study treated a panel of 18 W-CIN positive and 9 W-CIN negative colorectal cancer cell lines with a small molecule library of 160 kinase inhibitors, and found that the W-CIN positive cell lines were significantly more resistant to the inhibitors

[141]. Additionally, another study showed that in the OV01 ovarian cancer clinical trial, a high level of W-CIN was associated with taxane resistance [142]. In a study where puromycin resistant subclones were isolated from three human colon and breast cancer cell lines (MDA 231, SW 480, and HT 29 cell lines), a comparison of the karyotype of the subclones to the parental line showed that the drug-resistant subclones differed from the

24 parental line by six chromosomal alterations, whereas the average random subclone differed from the parental line by only 0.45 chromosomal alterations [143], suggesting that W-CIN was responsible for the development of drug resistance. Finally, in untransformed human mammary epithelial cells, cells surviving paclitaxel treatment showed higher basal and paclitaxel-induced chromosome mis-segregation as compared to controls, indicating that paclitaxel resistance is related to increased W-CIN in these cells [144]. Taken together, W-CIN has several functional consequences, including initiating changes in gene dosage, that likely contribute to tumorigenesis, as well as have a negative impact on patient outcome through the facilitation of chemotherapeutic resistance.

Clinical Consequences:

Table 1.2 summarizes the clinical outcome of several W-CIN positive cancers.

The most thoroughly studied cancers include lung cancer, colon cancer, and breast cancer.

In a study analyzing the prognostic importance of W-CIN in adenocarcinoma of the lung, 39.7% of the specimens studied were W-CIN positive, and the overall disease- free 5 year survival rate was 46.9% for W-CIN positive patients as compared to 71.0% for W-CIN negative patients. The difference in the 5 year overall survival rate was even more striking, at only 68.7% for W-CIN positive patients as compared to 93.5% for W-

CIN negative patients [145]. Several other studies have also correlated W-CIN with additional poor prognostic factors, like metastasis [146-148].

W-CIN is present in about 65-70% of colorectal cancers, and in a meta-analysis aimed at estimating the prognostic significance of W-CIN, the authors found that W-CIN is associated with poorer prognosis in terms of overall survival and progression-free survival, and could actually stratify colorectal cancer patients further after standard pathological staging. This study used both flow cytometry and image cytometry to

25 quantify aneuploidy status as a measure for W-CIN, and found that poorer survival was consistent, regardless of the patients’ ethnic background, location, or whether or not the patient received an adjuvant treatment [148].

In breast cancer, quantification of W-CIN by a variety of techniques, including

FISH, flow and image cytometry, similarly correlated with poor patient prognosis. In one study, W-CIN and prognostic factors were analyzed using 31 breast cancers and 5 benign breast lesions, and W-CIN was significantly higher in the breast cancers as compared to the benign lesions. Further, W-CIN showed a significant correlation with lymph node metastasis, as well as estrogen receptor negativity, suggesting that W-CIN status may be useful for predicting the aggressiveness of breast cancers [149]. In another study that used three independent measures for W-CIN, increased W-CIN was also correlated with poor prognosis in estrogen receptor positive breast cancers [150].

Clinical Diagnosis of W-CIN:

Studies in colon cancer demonstrated that W-CIN is capable of predicting whether or not a patient would relapse, independently of the information gained from the stage of the tumor [151]. This indicates that measuring W-CIN status in the clinic could provide valuable prognostic information. However, the methods used to measure W-CIN are inconsistent, sometimes technically difficult, and overall fall short of truly measuring chromosome mis-segregation events. A true measure of W-CIN involves evaluating chromosome mis-segregation by determining cell-to-cell variability in chromosome number, as well as assessing the rate at which chromosomes are gained and lost.

Currently, determination of W-CIN is based upon only one component of the phenomenon, such as aneuploidy status or tumor heterogeneity status, from which W-

CIN is inferred. Evaluating tumors for a change in chromosome number as well as for the rate at which that change occurs is often neither time nor cost effective. Measuring

26 only one aspect of W-CIN, however, has the potential to mask a more complex relationship between W-CIN and patient prognosis [8].

One method used for evaluating both changes in chromosome number as well as the rate at which that change occurs is by fixing tumor cells during anaphase, and evaluating each cell for evidence of chromosome mis-segregation, thereby capturing the dynamic nature of W-CIN. Experimental evidence shows that the most common indicators of chromosome mis-segregation are lagging chromosome and chromatin bridges [52]. This method has been used to assess the specific contribution of W-CIN to the prognosis of patients with diffuse large B-cell lymphoma (DLBCL). Here, 54 samples from cases of de novo DLBCL were formalin-fixed, paraffin-embedded, and stained with hematoxylin and eosin, and all anaphase cells were scored for evidence of chromosome mis-segregation. Further, radiologic imaging and bone marrow biopsy were used to score patients based on overall survival, progression-free survival, and requirement for treatment. Using these methods, they found that a two-fold increase in the frequency of chromosome mis-segregation led to a 24% decrease in overall survival and 48% decrease in relapse-free survival after treatment, concluding that increased rates of chromosome mis-segregation in DLBCL correlate with inferior outcome and poor patient prognosis [152]. This method for evaluating W-CIN is useful in cancers with a high mitotic index; however, this method may be challenging for slower growing tumors in which isolating and scoring anaphase-stage cells may be difficult.

Fluorescence in situ hybridization (FISH) is a common method for assessing W-

CIN status in tumors. This method uses fluorescently labeled DNA probes to quantify variations in chromosome copy number per cell, and can assess the chromosomal state of hundreds of cells at a time, but does not allow for evaluation of the rate of chromosome mis-segregation. This method has proven effective for analysis of W-CIN in lung cancer [145-147], breast cancer [149], and oral squamous cell carcinomas [153],

27 among others. However, this method has some limitations, including that tumors must be fixed, sectioned, and stained for evaluation. Thus, FISH is labor intensive, and therefore has limited clinical potential.

Another commonly used method for assessing W-CIN is flow cytometry. Here, cells must be labeled with a fluorochrome that is expected to stain DNA stoichiometrically, and cellular DNA content is reported using a laser-based electronic detection apparatus. This method can rapidly and accurately estimate the DNA content of single cells that are part of a larger population of cells, such as a tumor. The W-CIN status can be determined from both the aneuploidy status and heterogeneity of the population evaluated. Once again, this method is limited by the fact that it does not measure the rate of change of chromosome content [154]. A method similar to flow cytometery is DNA image cytometry, whereby cells are stained with a fluorochrome and evaluated by microscopy for increased DNA content. Like FISH, this method is labor intensive and thus has limited clinical potential.

More recently, a technique called comparative genomic hybridization (CGH) has been used, and allows for the detection of genomic variations as small as 5 to 10 megabases in individual cells as compared to control cells. Here, the DNA content of an individual sample cell and a control cell must be amplified, labeled with fluorophores of different colors, denatured, then hybridized at a one-to-one ratio to a metaphase spread originating from the same material as the control cell. The DNA samples will bind to the location from which they originated. Areas that show a higher intensity for the sample

DNA indicate a gain of material in the sample at that region, and vice versa [155].

Numerical chromosomal aberrations can be quantified, and heterogeneity can be determined by comparing CGH results from multiple cells. However, this method is expensive, time consuming, and cannot be easily used for high-throughput analysis.

Taken together, there are several methods currently available for assessing W-CIN

28 status in the clinic, each with its own set of limitations. Developing a consistent, accurate diagnostic method for determining the W-CIN status of patient tumors that is both practical and economical has the potential to contribute valuable information to predicted patient outcomes.

Exploitation of W-CIN for Anti-Cancer Therapies:

The low level of chromosome mis-segregation that is characteristic of W-CIN positive cancer cells is advantageous for continued proliferation and survival. However, severe aneuploidy is often not tolerated, and results in cell death. This opens up a therapeutic window, whereby the level of W-CIN in W-CIN positive cancer cells could be increased to a level that would force the cells into apoptosis. One potential therapeutic target that acts through this mechanism is the minus end-directed kinesin HSET. HSET is a kinesin-14 family member, and in normal cells is required for bundling microtubules

[156]. In cells containing supernumerary centrosomes, HSET is required for centrosome clustering [157]. Studies show that depleting HSET from non-transformed cells by RNAi has no effect, but depleting HSET from cancer cells containing extra centrosomes prevents the cell from clustering these centrosomes, causing multi-polar cell division and subsequently cell death. For example, in the W-CIN positive mouse neuroblastoma cell line N1E-115, depletion of HSET increased multi-polar mitotic spindles by 88%, and reduced cell viability by greater than 90%. Similarly, in the human breast cancer cell line

MDA-231, depletion of HSET increased multi-polar spindles by 45% and reduced cell viability by approximately 50%. In cancer cells that do not harbor supernumerary centrosomes, the effect of HSET depletion was insignificant [157]. Thus, HSET is a promising therapeutic target for W-CIN positive cancers containing supernumerary centrosomes.

Another study demonstrated that reducing the levels of checkpoint proteins

BubR1 and Mps1 greatly increased cancer cell sensitivity to the chemotherapeutic

Taxol. Taxol acts by stabilizing microtubules, preventing them from disassembling. This prevents spindle formation during mitosis and activates the mitotic checkpoint, and prolonged activation of the mitotic checkpoint results in apoptosis. In two W-CIN positive cancer cells lines (U2OS cells and HeLa cells), partial, independent depletion of mitotic checkpoint proteins BubR1 and Mps1 had no significant effect on cell viability. Similarly,

Taxol treatment alone had only marginal effects on cell viability. However, partial depletion of either BubR1 or Mps1 in combination with low, clinically relevant doses of

Taxol dramatically reduced cell viability by 2- to 10-fold, which also correlated with an increase in the amount and severity of chromosome segregation errors [158]. Therefore, using small molecule inhibitors to inhibit these or other checkpoint proteins, in combination with current therapeutics like Taxol, may prevent mitotic checkpoint activation, leading to severe aneuploidy and tumor cell death.

In addition to increasing the severity of chromosome mis-segregation as a therapeutic strategy, the opposite approach may hold some promise. W-CIN results in tumor heterogeneity, which allows tumor cells to adapt to environmental pressures, including chemotherapeutics. Therefore, preventing chromosome mis-segregation may limit the cell’s ability to adapt, allowing chemotherapeutics to be more effective. For example, treating tumors with a compound that forces microtubule de-polymerization may suppress the incidence of lagging chromosomes in W-CIN positive cancers, in a manner similar to the way over-expression of the microtubule-depolymerizing kinesins

Kif2b and MCAK do [99]. This would reduce chromosome mis-segregation, and limit W-

CIN.

Conclusions

The exploitation of W-CIN as a therapeutic target is in part limited by our current lack of understanding of the mechanisms that drive W-CIN in tumors. As mentioned,

30 disrupting the mechanisms that govern chromosome segregation can induce W-CIN, but whether these mechanisms are driving forces for transformation remains unknown.

Further, many “CIN genes” have been identified (Table 1.1), but it is not known if mutations or changes in the level of gene expression for any of these genes are initiating events for cancer development. Additionally, W-CIN is extremely prevalent in cancers, and its association with poor patient prognosis makes W-CIN an important clinical factor.

As such, there is an obvious need for research efforts to be put toward elucidating critical drivers of W-CIN in cancers.

One very important component that is often overlooked when discussing the mechanisms that govern chromosome segregation is the region of chromatin that many of those mechanisms are associated with, the centromere. The chromatin architecture of centromeres (discussed in detail in Chapter 2) must be maintained so that centromeres retain their ability to contribute to protecting cells against W-CIN. For example, the kinetochore assembles at centromeres, and without functional centromeres, kinetochore assembly is disrupted. Disrupted kinetochore assembly can result in issues with kinetochore-microtubule attachment, which can contribute to W-CIN, as discussed above. Similarly, because MCC proteins assemble at unattached kinetochores, disrupted kinetochores may prevent proper SAC signaling, allowing cells to exit mitosis with mal-attachments. Further, abnormalities in centromeric chromatin may weaken sister chromatid cohesion, allowing for precocious separation of sister chromatids.

Chapters 4 and 5 reveal Sp1 as an important factor for maintaining centromere identity.

The following chapter discusses centromere biology, and highlights the importance of maintaining centromere identity for proper chromosome segregation.

Figure 1.1: Acquired Capabilities of Cancer (Reprinted from [159] with permission Elsevier).

Figure 1.2: Genomic Instability. Most broadly definied, genomic instability is the failure of a cell to pass a complete and intact copy of its genome onto its daughter cells. Genomic aberations can occur on several different levels, including at the nucleotide level, the sub-chromosomal level (also referred to as structural/segmental CIN or S- CIN), and with changes in the number of whole chromosomes per cell (also referred to as numerical/whole CIN or W-CIN).

Table 1.1: Proteins Associated with W-CIN (Reprinted from [160] with permission from Elsevier).

Protein Alteration Putative mechanism(s) Reference APC Depletion, mutation Checkpoint defects, merotely [98, 161-166]

Aurora A Overexpression Centrosome amplification, cytokinesis [167, 168] failure Aurora B Depletion, drug inhibition Checkpoint defects, merotely [90, 92, 169]

β-catenin Mutation Dysregulation of cell-cycle proteins, [165] merotely BRCA1 Mutation Dysregulation of cell-cycle proteins, [170] merotely BRCA2 Mutation Dysregulation of cell-cycle proteins [170] Bub1 Heterozygous knockout, Checkpoint defects [118, 171] hypomorph, mutation Bub3 Heterozygous knockout Checkpoint defects [172, 173] BubR1 Knockout, mutation Checkpoint defect [114, 174- 176] CAML Knockout Cytokinesis failure, merotely [177] hCdc4/FBXW7 Depletion, knockout Dysregulation of cell-cycle proteins, [178] merotely Cdc20 Mutation Checkpoint defects [179] CENP-E Depletion, knockout Checkpoint defects, merotely [180-183] CENP-F Depletion Checkpoint defects, merotely [184, 185] CENP-H Overexpression Cytokinesis failure, merotely [186] CLASP Depletion Merotely [181, 187] Conductin/AXIN2 Overexpression Checkpoint defects, dysregulation of [166] cell-cycle proteins Cyclin E Overexpression Centrosome amplification, dysregulation [178, 188] of cell-cycle proteins, merotely EB1 Depletion Merotely [164] ECRG2 Depletion Centrosome amplification, checkpoint [189] defects, dysregulation of cell-cycle proteins Eg5 Overexpression Cytokinesis failure [190] FoxM1 Depletion, knockout Dysregulation of cell-cycle proteins [185] Hec1–NDC80 Antibody inhibition, mutation, Cytokinesis failure, merotely [191-193] complex overexpression Hice-1 Depletion Cytokinesis failure, merotely [194] Id1 Overexpression Cytokinesis failure [195] Kif2a Depletion, with MCAK depletion Merotely [196] Kif2b Depletion Merotely [99] Kif4 Knockout Centrosome amplification, merotely [197] Kruppel-like Knockout Centrosome amplification, chromosome [198] factor 4 breakage Mad1 Heterozygous knockout Checkpoint defects [199] Mad2 Depletion, heterozygous Checkpoint defects, merotely [120, 200- knockout, knockout, 203] overexpression MCAK Depletion Merotely [99, 204, 205] MCT-1 Overexpression Merotely [206] Mdm2 Overexpression Dysregulation of cell-cycle proteins [207] MdmX Knockout Centrosome amplification, cytokinesis [208] failure multipolar anaphases

Mps1 Mutation Checkpoint defects, merotely [209] p53 Knockout Dysregulation of cell-cycle proteins [206] PRP4 Depletion Checkpoint defects, merotely [210] Rad21/SCC1 Mutation Cohesion defects [71] Rae1 Heterozygous knockout Checkpoint defects [173] RanBP1 Depletion Merotely [211] Rb Depletion Centrosome amplification, dysregulation [202] of mitosis proteins causing overactivation of checkpoint REST Mutation Dysregulation of mitosis proteins [212] causing checkpoint defects SCC3 Mutation Cohesion defects [71] Securin Knockout, overexpression Cohesion defects [213, 214] Separase Knockout, overexpression Cohesion defects, cytokinesis failure [215, 216] SMC1 Depletion, mutation Cohesion defects, cytokinesis failure [71] SMC3 (cohesin Mutation Cohesion defects [71] subunit) Sgo1 Depletion Cohesion defects, cytokinesis failure [217] Sgo2/tripin Depletion Cohesion defects, merotely [218] TMAP/CKAP2 Depletion Merotely [219] Topoisomerase II Drug inhibition Catenation, merotely [220] Von Hippel Depletion Checkpoint defects [221] Lindau

Table 1.2: Clinical relevance of W-CIN (Reprinted from [8] with permission from John Wiley and Sons of EMBO Reports).

Method for Measuring W- Cancer Type Associated Outcome Reference CIN Lung Cancer Poor prognosis (OS and FISH (n = 63) [145] DFS) FISH (n = 47) Poor prognosis (OS) [147] FISH (n = 50) Poor prognosis (OS) [146]

12-gene signature (n = 647) Poor prognosis (OS) [151] W-CIN70 signature (n = 62) Poor clinical outcome [148] Breast Cancer SSI (n = 890) Poor prognosis (OS) [222] SNP (n = 313) Poor prognosis (MFS) [223] Poor prognosis (DFS and 12-gene signature (n = 469) [224] RFS) W-CIN70 signature (n = Poor clinical outcome [148] 1866) FISH (n = 31) Lymph node metastasis [149] Myelodysplastic FISH (n = 65) Poor prognosis (DFS) [225] syndrome Endocrine CGH (n = 62) Metastasis [226] pancreatic tumors Colon cancer Recurrence of colon 12-gene signature (n = 92) cancer [227] Flow cytometry/image Poor prognosis cytometry (n = 10126) Ovarian cancer 12-gene signature (n = 124) Poor prognosis (RFS) [151] Endometrial cancer SNP (n = 31) Poor prognosis (OS) [228] Synovial sarcoma CGH (n = 22) Poor prognosis (OS) [229] Oral cancer Poor prognosis (OS and FISH (n = 77) [153] (SCCs) DFS) Regional tumour FISH (n = 20) [230] outgrowth Diffuse large B-cell Anaphase segregation Poor prognosis (RFS) [152] lymphoma errors (n = 54) OS: Overall survival, DFS: Disease free survival, MFS: metastasis-free survival, RFS: relapse- free survival, SCC: squamous cell carcinoma, SSI: Stem line scatter index, SNP: Single nucleotide polymorphisms.

Chapter 2: Centromere Biology

Introduction

Our work discussed in Chapters 4 and 5 demonstrate that Sp1 is an important factor for maintaining centromere identity. Sp1 regulates centromere epigenetics by maintaining centromeric histone modifications, transcription through the core centromere, and proper levels of important centromere-associated proteins like centromeric protein A (CENP-A), the histone H3 variant that defines centromeres, and centromeric protein C (CENP-C). To place our findings in the context of the current knowledge in the field, this chapter focuses on centromere biology and the importance of maintaining centromere identity for proper chromosome segregation.

The centromere is a specialized region of chromatin that is critically important for faithful segregation of chromosomes during mitosis. Centromeres serve two major functions. First, the centromeric chromatin acts as a platform for assembly of the kinetochore. The kinetochore is a proteinaceous structure that serves as the interface between the mitotic chromatin and the mitotic spindle microtubules (Figure 2.1 and

Figure 2.2). Inner kinetochore proteins are constitutively associated with centromeric

DNA, and create a network of 16 proteins called the constitutive-centromere-associated network, or CCAN (Figure 2.2). CCAN proteins direct the assembly of outer kinetochore proteins when a cell enters mitosis, and these outer kinetochore proteins make connections with spindle microtubules. Thus, centromeres must maintain a specific chromatin architecture such that CCAN proteins recognize and assemble onto one and only one locus per chromosome. If this fails, resulting in more than one kinetochore per chromosome, that chromosome could be pulled in different directions during anaphase, resulting in fragmentation. Second, centromeres bind sister chromatid cohesion molecules, keeping sister chromatids together until they separate and move to opposite poles of the cell during anaphase. Defects in sister chromatid cohesion can result in premature chromosome segregation, and aneuploidy. Maintaining an active and

38 functioning centromere is essential for accurate chromosome segregation, as defects can lead to disruption or loss of genetic material during cell division, and pathological disease.

In 1882, the centromere was first observed using light microscopy simply as the attachment site for the mitotic spindle [231]. Since then, much effort has been directed toward understanding requirements for centromere identity and centromere function in many different species. Surprisingly, despite the importance and conservation of this region of chromatin, centromeres are highly divergent not only across species, but also on different chromosomes within a single organism [232]. The simplest centromere is the point centromere found in the budding yeast Saccharomyces cerevisiae. The point centromere is sequence specific, and exists on centromeric DNA that is organized into three domains, Centromere DNA Elements I, II, and III. A single nucleosome containing the centromere-specific histone H3 variant cenH3 wraps approximately 125 base pairs of centromeric DNA, and assembles onto Centromere DNA Element II [231, 233, 234].

More complexed centromeres in plants and animals contain several cenH3-containing nucleosomes, over a region of repetitive chromatin that spans several hundred kilobases to megabases in length, and are thus called regional centromeres [235-237].

Interestingly, there is a dramatic lack of homology between the repetitious centromeric sequences from different organisms [238, 239]. For example, Schizosaccharomyces pombe centromeres consist of a pair of repeated sequence arrays that are arranged in an inverted repeat around a central core sequence that is not conserved [240]. Oryza sativa (rice) centromeres contain a 155 base pair or 165 base pair satellite repeat called

CentO, as well as the centromere-specific retrotransposon CRR (Centromeric

Retrotransposon of Rice) that intermingles with CentO repeats, and spans 60 kilobases to 2 megabases, depending on the chromosome [241-243]. Drosophila melanogaster centromeres consist of short, 5 base pair repetitive sequences (AATAT and AAGAG)

39 interspersed with transposable elements [244], while chicken centromeres contain several hundred kilobases of repetitive arrays, where the repeat unit of each centromere is specific to each particular chromosome [245]. Mouse and human centromeres are composed of satellite DNA sequences. In mice, two types of repetitive DNA sequences are associated with centromeres, including major satellite repeats and minor satellite repeats. The major satellite sequences are located in the pericentromeric region, and encompass 6 megabases of 234 base pair repeat units [246]. The minor satellites encompass approximately 6 kilobases of 120 base pair repeat units [247]. Finally, human centromeres contain from 15,000 to more than 30,000 copies of a 171 base pair repeat, collectively called α-satellite (AS) DNA arrays, that span anywhere from 300 base pairs to several megabases of DNA [183]. Interestingly, although the majority of human centromeres form on AS DNA arrays, the rare neocentromere forms on genomic

DNA completely absent of AS DNA arrays, indicating that the underlying DNA sequence does not specify centromere formation [248]. Thus, although centromeres are necessary for faithful chromosome segregation, and although all species studied contain centromeres, this important chromatin region assembles on diverse types of DNA sequences rather than one that is evolutionarily conserved.

Human Centromeric DNA

Α-satellite (AS) DNA is thought to make up 2% to 3% of the human genome, with approximately 1 million copies of the 171 base pair repeat per diploid genome [249]. The highly repetitive nature of AS DNA arrays has made sequencing the human centromere extremely challenging. Currently, only a portion of the most distal AS DNA arrays that make up human centromeres have been annotated, and only on a few chromosomes.

Other AS sequences have been identified, but have not been placed on a specific chromosome [250]. Human AS DNA is classified into two groups. The first group,

40 monomeric AS DNA, includes DNA containing several copies of the basic 171 base pair unit. The second group is higher-order (HOR) AS DNA, in which a block of multiple 171 base pair units form a larger repeat unit [251-253]. Although a consensus sequence for the monomeric 171 base pair sequence has been described (Figure 2.3), there is a great deal of variation between monomeric sequences. For example, monomeric AS DNA sequences typically show identities of 70% to 90% to each other. Further, monomeric

AS DNA sequences within a HOR array typically share about 95% identity with the monomeric AS DNA sequence at the same position in another HOR array. It appears as though HOR arrays are mostly found at core centromeres, and monomers are found in the distal regions, although a lack of complete centromere sequence information makes confirming this theory difficult [253, 254]. This does indicate, however, that the HOR arrays are more important than the monomeric AS DNA sequences, and where studied, have been found to be most critical for centromere function. For example, studies using the HOR array sequences known to be present at the primary constriction site of chromosome X, called DXZ1, indicate that DXZ1 alone, and not the monomeric AS sequences flanking DXZ1, is sufficient for accurate chromosome segregation [255].

Currently, there is no evidence for direct involvement of monomeric AS DNA sequences in centromere identity and function.

Pericentromeric heterochromatin lacking AS DNA surrounds mammalian centromeres. This region of heterochromatin is thought to have several functions, including sister chromatid cohesion, and recruitment of centromere-specific proteins to centromeres [256]. To date, no pericentromeric region has been genetically manipulated

(deleted, reduced, or amplified), so the exact function of this region in mammals is still unclear. Like centromeric DNA, pericentromeric heterochromatin DNA sequence is not conserved between species, or even between chromosomes within the same species

[257]. Also like centromeric DNA, pericentromeric heterochromatin is made up of tandem

41 repeats of satellite DNA, including satellites I, II, and III. These satellite repeats are much shorter than AS DNA, at approximately 5 base pairs [257].

Defining Centromeres: Centromeric Protein A (CENP-A)

Our work in Chapters 4 and 5 indicate that Sp1 regulates CENP-A deposition, as depletion of Sp1 by RNAi results in a decrease in CENP-A at centromeres. With the exception of budding yeast Saccharomyces cerevisiae, centromeres are not defined by

DNA sequence. Instead, they are defined epigenetically, by the presence of nucleosomes containing the histone H3-like variant CENP-A (also referred to as CenH3 in some species). CENP-A containing nucleosomes are interspersed between canonical histone H3-containing nucleosomes at core centromeres. The requirement for CENP-A is best exemplified in the case of neocentromeres, centromeres formed on genomic

DNA devoid of the AS DNA arrays found at typical centromeres. Despite the lack of AS

DNA, neocentromeres contain CENP-A and form functional kinetochores. CENP-A homologs have been identified in every active centromere studied, in both single-celled and multicellular organisms [239, 258]. This histone H3-varient directs the recruitment of

CCAN proteins required for kinetochore formation, and thus for attachment of chromosomes to the mitotic spindle during mitosis [259]. As such, CENP-A is the most upstream component and single most important protein for centromere identity.

Disruption in the amount of CENP-A protein at centromeres results in disrupted CCAN assembly and kinetochore formation, disrupted attachment of spindle microtubules, and chromosome segregation errors.

The human CENP-A Gene and Protein:

CENP-A is a 17 kilodalton protein that belongs to the histone H3 family, and is a histone H3-like variant found exclusively at the core centromere. Here, it replaces conventional histone H3 within the nucleosome. The CENP-A gene is located on

42 chromosome 2 at position 2p23.3 (26,987,157-27,023,935 forward strand), from which it produces 6 transcripts, including CENP-A-001 through CENP-A-006. Of these transcripts, only CENP-A-001 (Isoform 1) and CENP-A-006 (Isoform 2) are protein coding transcripts. CENP-A-005 produces a protein product that is subjected to nonsense mediated decay, and all other transcripts are non-coding [260].

The CENP-A-001 gene is 1452 base pairs in length, and produces a 140 amino acid protein, while the CENP-A-006 gene is 1299 base pairs in length, and produces a

114 amino acid protein through alternative splicing [260]. CENP-A-006 lacks amino acids 71-96 [261]. CENP-A-001 is the functioning CENP-A molecule in the cell, while the function of CENP-A-006 is not yet clear.

The transcribed CENP-A-001 protein (hereafter referred to as CENP-A) has several secondary structures, including four helices and two beta strands [262]. The

CENP-A amino acid sequence shares significant homology with residues 48-135 in the

C-terminus of histone H3. This region, which contains the histone fold domain, shares

60% identity and 75% similarity to the CENP-A amino acid sequence [263]. Within

CENP-A, this H3-like domain allows for CENP-A to incorporate into the nucleosome in place of histone H3. In addition, CENP-A contains a unique region required for targeting the histone variant to the centromere, called the CENP-A targeting domain (CATD)

[264]. This region spans amino acids 75-116 [260], and contains the first loop and second alpha-helix of the protein. Replacement of the corresponding region of H3 with the CATD is sufficient to direct H3 to centromeres [259]. CENP-A mRNA and protein levels are maximal at the end of S phase [265]. Prior to assembly at centromeres, newly expressed CENP-A is sequestered for most of the cell cycle (late S-phase, G2, and most of mitosis) in a complex that contains its partner, H4, and its chaperone, HJURP

(Holliday junction recognition protein) [266].

The CENP-A Nucleosome:

Genomic DNA is wrapped around a canonical histone octamer, containing two of each of histones H2A, H2B, H3, and H4. Each octameric unit contains an H3/H4 tetramer, flanked by H2A/H2B dimers. DNA wraps around the histone octamer 1.7 times, or about 146 base pairs, in a left-handed orientation, a characteristic thought to be important for transcription, DNA replication, and DNA repair [267]. At centromeres,

CENP-A replaces canonical histone H3 in a large fraction of the nucleosomes.

Interestingly, the structure of CENP-A containing nucleosomes is currently a hotly debated topic, as several atypical nucleosome arrangements have been proposed. For example, in budding yeast, one study proposes that the H2A/H2B dimer is replaced with a non-histone protein, Scm3 (HJURP in mammalian cells), creating a hexameric nucleosome composed of two copies of the yeast CENP-A homolog Cse4, histone H4, and Scm3 [268]. Another study proposed that Cse4 containing nucleosomes form a hemisome, with only one copy of each of Cse4, H4, H2A, and H2B [269]. In both cases, it is thought that centromeric DNA may wrap around the budding yeast centromeric nucleosomes in a right-handed orientation, providing a specialized region of chromatin for centromere identification and kinetochore protein assembly [267]. This same hemisome model has been proposed in both Drosophila melanogaster [270], and in human cells [271]. Using atomic force microscopy, the latter study showed that CENP-A containing nucleosomes are one half the height of canonical octameric nucleosomes.

Here, bulk chromatin octameric nucleosomes ranged in height from 2.5-4.5 nanometers, and 75% of CENP-A containing nucleosomes ranged in height from 1.4-2.0 nanometers.

This study also used immunoelectron microscopy, a technique that permits mapping histones on chromatin at single molecule resolution [272], to support their claim that

CENP-A containing nucleosomes contain only one copy of the CENP-A and H2B molecules [271]. Many other groups contest that CENP-A containing nucleosomes are

44 octasomes, like canonical nucleosomes. In vitro studies, including bacterially expressed nucleosomes reconstituted with palindromic DNA designed from human AS sequences, when crystalized, appear to reveal a histone octamer [273-275]. In vivo studies using a photobleaching-assisted copy-number counting technique [276] in HeLa cells stably expressing a CENP-A-YFP fusion protein supports these claims. This system, which allows direct visual analysis of CENP-A-YFP stoichiometry in native-assembled nucleosomes, showed that a majority of the complexes contained two molecules of

CENP-A-YFP [275]. More recently, it has been proposed that CENP-A containing nucleosomes oscillate between octameric and tetrameric forms during cell-cycle progression in both human cells and budding yeast [277, 278]. In human cells, it appears as though CENP-A containing nucleosomes are tetrameric at early G1, convert to octamers at the G1 to S phase boundary, and then revert back to tetramers after replication. The tetrameric structure is then maintained for the remainder of the cell cycle

[277]. Although this theory remains to be substantiated by other groups, it could explain the variability in CENP-A containing nucleosome structure observed experimentally.

Crystal structures of octameric CENP-A containing nucleosomes reveal other distinct differences between those and canonical histone H3-containing nucleosomes that may be vital for centromere identity and function. The CATD makes the CENP-A/H4 tetramer more rigid than the H3/H4 tetramer [264]. It also causes the CENP-A/CENP-A protein interface to be rotated as compared to the H3/H3 interface [262]. Further, arrays of CENP-A containing nucleosomes are 30% more compact than arrays of canonical histone H3-containing nucleosomes [279], but with looser DNA at the entry and exit sites at the nucleosome boundaries [280]. Finally, post-translational modification of CENP-A can influence the conformation of CENP-A containing arrays. In most cases, NRMT (N- terminal RCC1 methyltransferase) catalyzes the tri-methylation of glycine 1. Further,

Serine 16 and 18 are phosphorylated, and these phosphorylated residues form a salt-

45 bridged secondary structure within the N-terminus of the protein. This secondary structure allows for intermolecular and intramolecular interactions that influence the conformation of CENP-A containing nucleosome arrays [281]. CENP-A is also phosphorylated on serine 7, which is required for stabilization of CENP-A nucleosomes and for the interaction between CENP-A and the centromere associated protein CENP-C

[282]. Recently, it was shown that CENP-A monoubiquitination on lysine (K) 124 is required for the interaction between CENP-A and the CENP-A specific chromatin assembly factor HJURP. This monoubiquitination is mediated by CUL4A-RBX1-COPS8

E3 ligase [259]. Therefore, although the true composition of CENP-A containing nucleosomes is still unclear and currently under intense investigation, it is clear that

CENP-A containing nucleosomes are distinct from canonical histone H3-containing nucleosomes, and that this distinction likely confers a specific identity to centromeric chromatin.

CENP-A Deposition at Centromeres:

Canonical nucleosomes are assembled in S phase, as DNA is replicated. Newly synthesized histones form complexes with chromatin assembly factors, which mediate histone deposition onto new DNA and the assembly of DNA into nucleosomes [283].

Conversely, in vertebrate cells, new CENP-A molecules are deposited onto DNA after the cell exits mitosis, in early G1, when CDK1 activity has declined [284]. Thus, when cellular DNA is replicated, CENP-A containing nucleosomes are diluted across both strands of DNA, and cells proceed through mitosis with only half the maximal number of

CENP-A containing nucleosomes [285]. Surprisingly, this is not true for all species. For example, during S phase, budding yeast first remove all existing Cse4 (the CENP-A homolog) before incorporating new Cse4 into centromeric DNA [286]. To date, several factors have been identified as regulators of CENP-A deposition, the major players of which include HJURP, the Mis18 complex, CENP-C, and CENP-I. Tight regulation of

CENP-A deposition is required, as deletion of CENP-A results in a mitotic lethal phenotype [284], and over-incorporation or mis-incorporation of CENP-A molecules is associated with chromosome segregation errors and cancer [287, 288].

Holliday junction recognition protein (HJURP):

HJURP was first identified in 2007 through a genome-wide expression profile analysis of non–small cell lung cancer (NSCLC) tissues, as a result of its having 5-fold or higher expression in cancer cells as compared to normal lung cells in more than 50% of the samples examined. In this study, HJURP was found to associate with hMSH5 and

NBS1 in cancer cells, and also bound to a synthetic nucleotide with Holliday junction structure, hence the name. Depletion of HJURP by RNAi resulted in genomic instability, suggesting for the first time that HJURP is an indispensable factor for maintaining chromosomal stability [289]. Shortly after, HJURP was co-purified with CENP-A from chromosome-depleted extracts. This study showed that HJURP is required for CENP-A localization to centromeres, as HJURP depletion by RNAi reduced the intensity of

CENP-A staining at centromeres. Further, adding the CATD to a canonical histone H3 molecule (H3CATD) induced an interaction between H3CATD and HJURP that was otherwise nonexistent [290]. Several studies have followed that confirm HJURP as the centromere-specific assembly factor that directs the incorporation of CENP-A into nucleosomes [291-294], including a study that showed that targeting of HJURP to non- centromeric locations results in CENP-A incorporation [292]. A recent study showed that the CENP-A/H4 tetramer is bound to HJURP for a majority of the cell cycle, including late S phase, G2, and mitosis [266].

HJURP interacts with the CENP-A/H4 tetramer through its CENP-A binding domain (CBD), located within the first 80 amino acids of the HJURP protein. Within this region, HJURP contains a TLTY box (TLTYETPQ in humans), a novel amino acid sequence that is highly conserved across species and is required for the HJURP-CENP-

A interaction, as deletion of the TLTY box alone abrogates HJURP binding to CENP-A.

The HJURP CBD interacts with the CATD of CENP-A to stimulate CENP-A/H4 tetramer deposition [291]. The CENP-A/H4-HJURP complex localizes to centromeres in G1, consistent with new CENP-A incorporation [294]. Although HJURP was first identified as a protein that binds to Holliday junction structures, this DNA structure has not been identified as a requirement for CENP-A incorporation at centromeres [265].

The phosphorylation status of HJURP may regulate its association with centromeric DNA. Multiple serine residues are phosphorylated, including serine (S) 123,

S140, S382, S412, S448, S472, S486, S557, S559, S595, and S686. Currently, the function of HJURP phosphorylation is only partially understood. One study showed that when HJURP is not associated with centromeres, it is most highly phosphorylated at

S123, S412, and S557, suggesting these phosphorylation sites are important for preventing HJURP from localizing and thus loading CENP-A outside of G1 [266]. Another study showed that mutating S412, S448, and S472 resulted in premature loading of

CENP-A [295], and thus must function to regulate the timing of HJURP association with centromeres as well.

The Mis18 Complex:

The Mis18 complex consists of 5 proteins, including Mis18α, Mis18β,

Mis18BP1KNL2, RbAp48, and RbAp46 [265]. This complex localizes to centromeres in late anaphase and is required for recruitment of HJURP, as depletion of Mis18α and

Mis18BP1KNL2 prevents localization [292]. Further, depletion of any one member of this complex prevents incorporation of new CENP-A [296, 297]. In mice, Mis18α deficiency leads to mislocalization of CENP-A in blastocysts, resulting in early embryonic lethality from severe chromosome mis-segregation [298]. Interestingly, despite the clear relationship between the Mis18 complex, HJURP, and CENP-A, no physical interaction between the Mis18 complex and HJURP, or the Mis18 complex and CENP-A, has been

48 observed [265]. Mis18 complex localization may also be regulated by CDK activity. One study showed that from cell cycle phase S through mitosis, when CDK activity is high,

Mis18BP1KNL2 is unable to localize to centromeres. However, the sharp decrease in CDK activity at the exit of mitosis changes the phosphorylation status of Mis18BP1KNL2, allowing it to be recruited [299]. Then, in G1, centromere associated Plk1 binds to and phosphorylates Mis18BP1KNL2, also to promote its localization [300]. This CDK and Plk1- dependent regulation of Mis18BP1KNL2 localization may help to restrict CENP-A incorporation to G1.

Current models suggest that the Mis18 complex may function to recruit chromatin modifying factors to centromeres to create or maintain chromatin architecture permissible for CENP-A deposition. For example, a 2007 study showed that treating

Mis18 complex-depleted cells with the histone deacetylase (HDAC) inhibitor trichostatin

A (TSA) suppressed the loss of CENP-A at centromeres, suggesting that histone acetylation is required for CENP-A deposition [296]. In support of this, RbAp46 was previously shown to bind to histone H4 and the histone acetyltransferase Hat1 [301].

Additionally, another study showed that promoting centromeric histone acetylation by targeting histone acetyltransferases p300 and PCAF to synthetic AS DNA enhanced

CENP-A assembly. This CENP-A assembly required HJURP, but did not require the

Mis18 complex [302]. Further, Mis18BP1KNL2 contains a SANT domain (Swi3-Ada2-

NCoR-TFIIIB), a domain found in a variety of different chromatin remodelers [303], including histone acetyltransferases [304]. These data argue that the Mis18 complex may play a role in maintaining the histone acetylation status required for CENP-A deposition. This is perplexing, as centromeric nucleosomes are known to lack H3 and

H4 acetylations typically found in euchromatin (H3K9 and H4K5, K8, K12 and K16)

[104].

In addition to a possible role in the maintenance of centromeric histone acetylation, the Mis18 complex may also affect histone methylation and DNA methylation at centromeres. In Mis18α-depleted mouse embryonic fibroblasts (MEFs), methylation levels of histone H3 lysine 9 (H3K9) and histone H3 lysine 4 (H3K4) were decreased. Further, recruitment of the histone methyltransferase Suv39h1 to centromeres, which has been shown to tri-methylate H3K9, was also reduced in Mis18α- depleted MEFs. DNA methyltransferases DNMT3A and DNMT3B were shown to interact with Mis18α in mice, and depletion of Mis18α results in a decrease in DNA methylation

[298]. Therefore, although the exact function of the Mis18 complex at centromeres is not clear, it is likely that this complex plays some role in centromeric chromatin maintenance vital for proper incorporation of CENP-A through recruitment of HJURP.

CENP-C and CENP-I:

CENP-C is one of several components of the CCAN, the network of proteins that associates with centromeric chromatin throughout the cell cycle. CENP-C is an important upstream factor for CENP-A deposition, as it binds to centromeric chromatin and recruits the Mis18 complex to this region. More specifically, the C-terminus of CENP-C interacts with the SANT domain of Mis18BP1KNL2, and knockdown of CENP-C results in decreased CENP-A at centromeres [305]. This has been demonstrated in several species, including Xenopus egg extracts [306], Drosophila melanogaster [307], and mouse cells [305]. Interestingly though, CENP-C appears to be recruited to centromeres through an interaction with CENP-A itself [308]. This suggests that CENP-C may be important for incorporation of CENP-A at existing centromeres, but not de novo incorporation of CENP-A. In support of this, targeting CENP-C to non-centromeric regions is not sufficient for CENP-A incorporation [309]. A recent study showed that

CENP-I, also a component of the CCAN, can recruit Mis18BP1KNL2 to synthetic AS DNA.

In this study, depletion of CENP-C had no effect on the ability of CENP-I to recruit

Mis18BP1KNL2, and vice versa [310].

Other DNA Binding Proteins at Centromeres

In Chapters 4 and 5, we show by chromatin immunoprecipitation that Sp1 binds to both centromeres and pericentromeres during mitosis. Here, we review other proteins known to bind to centromeres, as well as how they contribute to preserving centromere identity and function.

Centromeric Protein B (CENP-B):

Of the proteins that bind to centromeric DNA, only one protein, centromeric protein B (CENP-B), has been identified as requiring a specific sequencing for binding.

CENP-B is a highly conserved protein that binds to a 17 base pair sequence known as a

CENP-B box (CTTCGTTGGAAACGGGA) [311]. In human centromeres, CENP-B boxes appear in every other monomeric AS sequence in higher-order (HOR) AS arrays, but not in clusters of AS monomers that do not form HOR repeats [312]. CENP-B is an 80 kilodalton protein with a helix-loop-helix DNA-binding domain occupying the N-terminal

125 amino acids of the protein [313]. CENP-B also contains a dimerization domain in the

C-terminus, suggesting a model by which one molecule of CENP-B binds to HOR AS array and dimerizes with other DNA-bound CENP-B molecules to influence centromeric nucleosome positioning, and to establish higher order chromatin structures [314]. In studies using mammalian artificial chromosomes, CENP-B boxes are required for de novo centromere formation and assembly of some centromere associated proteins, including CENP-A [315]. However, mouse CENP-B knockout studies reveal that functional kinetochores are able to form in the absence of CENP-B [316]. Further, neocentromeres that lack AS DNA are also devoid of CENP-B boxes and thus CENP-B protein, yet form functional kinetochores [317], and CENP-B appears to be absent from

51 the Y chromosome all together [318]. A recent study showed that the amino-terminal tail of CENP-A binds to CENP-B, and that together, CENP-A and CENP-B are required for maintaining normal levels of CENP-C, an important chromatin and kinetochore assembly factor. This study also found that depletion of CENP-B yields a higher rate of chromosome mis-segregation [319]. Thus, a complete understanding of the function of this sequence-specific centromeric DNA binding protein is still needed.

Centromeric Protein C (CENP-C):

CENP-C was one of the first centromere proteins identified [320]. It is a component of the CCAN, and thus constitutively localizes to centromeres throughout the cell cycle. The human CENP-C protein consists of 943 amino acids and contains several regions that are important for binding to AS DNA, as well as to other centromere and kinetochore proteins.

The N-terminal region of CENP-C, consisting of amino acids 1-71, interacts with

Mis12, a component the KMN network of proteins (including the Knl1 complex, the

Mis12 complex, and the Ndc80 complex) of the outer kinetochore (Figure 2.2) [321].

The KMN network is essential for kinetochore-microtubule interactions and is thus vitally important for chromosome segregation. In HeLa cells, expression of amino acids 1-71

(CENP-C1-71) competed with endogenous CENP-C, and prevented recruitment of Mis12 to kinetochores, as well as other KMN network proteins. Further, CENP-C1-71 expression prevented proper alignment of chromosomes along the metaphase plate, premature mitotic exit, and chromosomal instability [322]. As such, CENP-C serves as a bridge between CENP-A and the CCAN at the inner kinetochore, and the outer kinetochore and spindle microtubules.

Amino acids 189 to approximately 400 are proline, glutamate, serine, and threonine rich, and are thus referred to as a “PEST rich” domain. Recently, this domain was found to be required for the interaction between CENP-C and the CENP-H-I-K-M

52 complex, a complex of centromere associated proteins consisting of CENP-H, CENP-I,

CENP-K and CENP-M (Figure 2.2). The CENP-H-I-K-M complex appears to be a central complex in CCAN organization, as it interacts with several other CCAN proteins, including the CENP-T-W-S-X heterotetramer, also required for recruitment of the KMN network proteins (Figure 2.2) [323, 324]. Although amino acids 426-537 are only conserved among mammals, this portion of CENP-C is required for centromere localization, as it binds centromere AS DNA as well as CENP-A. More specifically, amino acids 482-508 include many positively charged amino acids that are likely to bind

DNA, whereas amino acids 509-535 bind to CENP-A [325]. Amino acids 426-551, which contain a sequence homologous to the HP1 hinge domain, were also shown to bind AS- derived long non-coding RNAs (lncRNA) [326].

In contrast to the central domain, the C-terminus of CENP-C contains two distinct domains that are highly conserved from yeast to mammals. The first domain, located from amino acids 737–759, is called the Mif2p homology domain II (or the CENP-C domain). This domain is present in all homologs of CENP-C, and also binds AS DNA and centromeric nucleosomes. The second domain, located from amino acids 890 –

943, is called the Mif2p homology domain III, and is required for CENP-C dimerization and/or oligomerization [327]. Both of these regions are required for the interaction between CENP-C and the Mis18 complex, an important upstream factor for CENP-A deposition (Figure 2.2) [305].

Most recently, CENP-C was shown to influence the shape and dynamics of

CENP-A containing nucleosomes. Using purified proteins and AS DNA sequences in vitro, Black and colleagues showed that in the absence of the CENP-C central domain

(CENP-CCD), which binds CENP-A, the H2A/H2B dimers within CENP-A containing nucleosomes are approximately 5 angstroms farther apart than when reconstituted with the CENP-CCD. This increased distance between the H2A/H2B dimers is predicted to

53 weaken the physical connection between the H2A subunit on one face of the nucleosome, and the H4 subunit on the opposite face of the nucleosome. Thus, CENP-C functions to protect both the shape and the rigidity of CENP-A containing nucleosomes.

This study also showed that CENP-C binding affects how tightly DNA is wrapped around

CENP-A containing nucleosomes [328].

The CENP-T-W-S-X Heterotetramer:

The CENP-T-W-S-X heterotetramer is composed of four CCAN proteins, CENP-

T, CENP-W, CENP-S, and CENP-X. This heterotetramer was discovered as two separate complexes, including the CENP-T-W complex and the CENP-S-X complex.

CENP-T-W was first shown to be vital for kinetochore formation and function, and a detailed structural analysis of this complex revealed that CENP-T and CENP-W both contain histone fold domains, suggesting DNA-binding properties. Indeed, digested centromeric DNA immunoprecipitated with both CENP-T and CENP-W, and mutational analysis revealed that the histone fold domains are required for centromere targeting.

Further, this complex was shown to associate with histone H3-containing nucleosomes, but not with CENP-A containing nucleosomes at centromeres [329]. A short time later,

CENP-X was identified as an interacting partner with CENP-S, another CCAN protein containing a histone fold domain [330]. A detailed structural analysis of these complexes has revealed that CENP-T-W forms a dimer highly homologous to the structure of the histone H2A–H2B or H3–H4 dimers within the nucleosome, as does the CENP-S-X complex. When these complexes were combined in equimolar ratios in solution, CENP-

S-X and CENP-T-W co-migrated and formed a single stoichiometric complex, suggesting that both dimers interact to form a heterotetramer. Further, the CENP-T-W-S-

X complex was shown to bind to approximately 100 base pairs of DNA as a (CENP-T-W-

S-X)2 structure, and although this complex is predicted to bend centromeric DNA in a manner similar to canonical histones, it is believed that the CENP-T-W-S-X complex

54 induces positive supercoils rather than negative coils, adding specificity to centromeric chromatin [331]. In this study, depletion of either the CENP-T-W or the CENP-S-X complex prevented formation of a functional kinetochore, suggesting that the CENP-T-

W-S-X heterotetramer contributes to the specialized chromatin architecture of this region

[331, 332].

In addition to epigenetically distinguishing the centromere from other genomic loci, the CENP-T-W-S-X complex contributes to faithful chromosome segregation by recruiting the KMN network to kinetochores. More specifically, the N-terminal tail of

CENP-T interacts with KMN network component Ndc80, a key microtubule binding component of the outer kinetochore [333]. The CENP-T-W-S-X complex is interspersed between CENP-A containing and histone H3-containing nucleosomes at core centromeres.

The Centrochromatin: Centromere-Specific Histone Modifications

In addition to the presence of CENP-A, it is becoming more and more evident that other centromere-specific histone modifications are just as important for centromere identity. The centromeric chromatin maintains a unique bivalent signature of histone modifications that are neither heterochromatic, nor euchromatic. Thus, because this region is neither typical heterochromatin nor typical euchromatin, it has been dubbed

“centrochromatin” to reflect the uniqueness of the blended histone marks. While studies over the past fifteen years have shed light on the characteristics of the centrochromatin, it continues to be an area that is difficult to study because of the repetitiveness of the AS

DNA sequences at centromeres. To circumvent this problem, many groups have examined extended chromatin fibers probed with specific antibodies to identify histone modifications present in CENP-A containing regions. We have used chromatin immunoprecipitation experiments to evaluate the effect of Sp1 depletion on histone

55 modifications at centromeres, and have found that Sp1 regulates several of them. Thus, this section focuses on our current understanding of histone modifications at core centromeres, as well as the enzymes that may maintain them.

Histone Modifications at the Core Centromere:

As a result of its repetitive nature, centromeric DNA was long thought to be heterochromatic. Then, a landmark 2004 study identified histone modifications typically associated with active transcription in CENP-A containing regions. This study showed that centromeric histone H3-containing nucleosomes are di-methylated on lysine 4

(H3K4me2) [334]. H3K4me2 is often enriched at promoters of active genes, as well as at promoters of genes that are primed for future gene expression during cell development.

H3K4me2 can also exist in large domains over both the gene promoter and the gene body, indicating that it may be a mechanism for fine-tuning gene expression [335, 336].

A later study showed that H3K4me2 is required for HJURP-dependent incorporation of

CENP-A, and also identified H3K36me2 and H3K36me3 at centromeres, both modifications associated with active transcription [337]. Recently, H4K20me1 was identified in CENP-A containing nucleosomes, and was found to be important for kinetochore assembly by promoting localization of CENP-T, a component of the CENP-

T-W-S-X complex [338]. In other areas of the genome, H4K20me1 is necessary for active elongation by facilitating the release of RNA Polymerase II [339]. Additionally, monoubiquitinated H2B (H2Bub1), a modification associated with active transcription of many genes, is present at centromeres during the G2-M transition [340]. Interestingly however, H3K4me3, a prominent histone mark associated with active genes, is absent from centromeres, as are histone H3 and H4 acetylations typically found in euchromatin

(H3K9 and H4K5, K8, K12 and K16) [334]. Thus, although many euchromatic marks are maintained at centromeres, other prominent marks are conspicuously absent.

Despite an incomplete set of euchromatic marks, centromeres are not heterochromatic, as they mostly lack heterochromatic modifications like H3K9me2 and

H3K9me3 [334], although some studies show that a small fraction of H3 histones are tri- methylated on lysine 9 within the core centromere [302, 341]. Heavily heterochromatinizing centromeric DNA is devastating for centromere function. One study showed that targeting tTs (a powerful transcriptional repressor) to centromeres on human artificial chromosomes resulted in the accumulation H3K9me3 and HP1, a classic component of heterochromatin, and prevented kinetochore protein assembly

[342]. Heterochromatin appears to be restricted to the pericentromeric region, as the surrounding pericentromeric DNA is enriched for H3K9me2/3, H3K27me2/3, and

H4K20me3 (Figure 2.4) [343-345].

Histone Modifiers:

While much effort has been directed toward identifying the enzymes responsible for adding, maintaining, or removing modifications from histones in genomic DNA over the years, the specific enzymes responsible for maintaining core centromere modifications specifically, have not been identified. It is attractive to assume that the same enzymes are responsible, but this remains to be determined. For example, the

H3K4 methylases Set1 and Ash2 are required for inheritance of active transcriptional states in other genomic areas during mitosis [346], but have not been identified as being required for H3K4me2 at core centromeres. Similarly, methylases Set2 and WHSC1 methylate H3K36, but have not been observed at centromeres [347]. Some histone modifiers, including p300/CBP, HDAC1, and DNMT1, have been shown to localize to metaphase centromeres, but the function of these factors at centromeres is not yet known [348]. At pericentromeres, Suv39h is responsible for H3K9me3, and also indirectly regulates H4K20me3 (Figure 2.4) [344, 349].

Transcription Through Core Centromeres

Centromeres are actively transcribed in several different organisms, including

Saccharomyces cerevisiae [350], Schizosaccharomyces pombe [351], rice [352], mouse

[353], tammar wallaby [354], and humans [326, 355-358]. While this region was once thought to be transcriptionally inactive in human cells [359], a landmark study in 2004 identified histone modifications associated with active transcription [334]. Since then, much effort has been directed toward understanding the significance of the centrochromatin, including the significance of transcription through the core centromere.

Accumulating evidence reveals that transcription through the core centromere in human cells produces lncRNAs. Core transcription and the lncRNA transcripts are required for the incorporation of new CENP-A molecules, as well as kinetochore assembly and stability, making transcription a critical process in maintaining centromere identity and function [360].

Evidence for Transcription at Core Centromeres in Human Cells:

One of the first studies that examined transcription through core centromeres in human cells evaluated transcription at neocentromeres, active centromeres that form on genomic DNA instead of AS DNA. This study showed that neocentromere formation on genomic DNA had no effect on the expression of the genes within the region of neocentromere kinetochore assembly, despite the requirement for major chromatin remodeling at these sites [358]. Thus, although transcription through core centromeres formed on AS DNA had not yet been evaluated, this study demonstrated that despite the unique bivalent signature of centromeric chromatin that isn’t necessarily “active”, it is permissive to transcription machinery. Another study using the 10q25 neocentromere on the Mardel (10) chromosome showed that RNA transcribed from an L1 retrotransposon that resides within the CENP-A bound chromatin of this neocentromere (called FL-L1b) was directly incorporated into the neocentromeric chromatin, and knockdown of this

RNA transcript resulted in a reduction in CENP-A binding [356]. A later study showed that single-stranded RNA is directly transcribed from AS DNA, binds CENP-C, and is required for the association of kinetochore proteins with centromeres. Here, the authors used RNA-FISH with centromere specific AS RNA probes to show that AS-derived RNA is enriched at the nucleolus, along with centromere proteins CENP-C and INCENP.

Electrophoretic mobility-shift assays using radioactive-labeled RNA probes confirmed that the AS-derived RNA binds CENP-C, and RNAse treatment abolished CENP-C localization to centromeres by immunofluorescence [326]. A recent study biochemically purified AS-derived lncRNA and used northern blotting to reveal that HeLa cells possess a unique centromeric RNA species that migrates at approximately 1.3 kilobases [361].

Transcription Factors at Centromeres in Human Cells:

In 2012, an important study identified RNA Polymerase II as the polymerase responsible for transcribing core centromeres. By immunofluorescence, Chan, et al.

[355] showed that RNA Polymerase II localizes to mitotic centromeres but not interphase centromeres in HeLa cells, as early as prometaphase. Using the same techniques, they also identified CTDP1 (carboxy-terminal domain, RNA polymerase II, polypeptide A phosphatase, subunit 1), a transcription factor specific to RNA Polymerase II required for transcription elongation, as well as SSRP1 (structure-specific recognition protein 1), another key component of the RNA Polymerase II machinery. SSRP1 is a component of the FACT complex (facilitates chromatin transcription), a complex previously identified as being required for CENP-A targeting to centromeres, along with the ATP-dependent chromatin remodeling factor CHD1 [362]. TBP (TATA-box binding protein) was also shown to localize to CENP-A containing regions [361].

Timing of Core Centromere Transcription:

To determine whether RNA Polymerase II actively transcribes at centromeres during mitosis, Chan, et al. used a transcription assay with Fluorescein-12-UTP to label

59 actively transcribed RNA, and found that discrete Fluorescein signals were visible at the kinetochore during mitosis, but not during interphase [355]. Quantitative real-time reverse transcriptase PCR (qRT-PCR) for AS-derived lncRNA confirmed mitotic transcription of core centromeres, and inhibition of RNA Polymerase II with α-amanitin resulted in a 68% reduction in these transcripts, as well as reduced CENP-C localization and chromosome segregation errors [355].

The exact timing of centromeric transcription is under debate. Many studies point to mitosis, and more specifically metaphase, as the point at which lncRNAs are transcribed. For example, RNA Polymerase II was not visible at centromeres in prophase cells, whereas a small subset of prometaphase cells showed localization by immunofluorescence. Localization was robust by metaphase [355]. However, another study indicates that centromeres are not transcribed until early G1. This study evaluated chromatin fibers from HeLa cells at different stages of the cell cycle for RNA Polymerase

II and TBP binding by immunofluorescence. They were not able to detect these proteins at CENP-A containing regions until after mitosis, in early G1 [361]. Although it is widely accepted that centromeres are only transcribed during a short period of time between the onset of mitosis and late G1, the exact timing is still unclear.

Human Artificial Chromosomes (HACs):

The generation of human artificial chromosomes (HACs) has been an important tool for studying human centromeres. The centromere and kinetochore structure within

HACs mimic those within endogenous chromosomes, and are able to be manipulated through protein targeting without affecting endogenous centromeres [363]. The most commonly used HAC for studying centromeres is the synthetic alphoidtetO array. The basic repeating unit of the alphoidtetO array includes an AS monomer from chromosome

17, followed by a CENP-B box, then a second consensus sequence AS monomer. In place of the CENP-B box in the AS monomer consensus sequence is a tetracycline

60 operator (tetO). This cassette was cloned head to tail to create an 8-mer, subjected to rolling circle amplification, and subsequently transfected into yeast with a YAC/BAC backbone. Recombination within the yeast generated a HAC with 1.1 megabases of AS

DNA [364, 365]. The tetO sites are recognized by a tetracycline repressor (tetR) protein in a tetracycline-inducible manner. This allows for specific and inducible targeting of tetR-fusion proteins to the AS sequences on HACs, providing a powerful tool for studying centromere structure and dynamics.

HACs have been useful for determining the impact of certain histone modifications on centromeric transcription. In one study, centromeric H3K4me2 was depleted from the HAC centromere by targeting LSD1, an H3K4me2-specific demethylase, to the alphoidtetO array. Loss of H3K4me2 resulted in a decrease in transcription of the alphoidtetO array, as well as a decrease in HJURP recruitment and subsequent loss of CENP-A. The decrease in H3K4me2 was also paralleled by a decrease in H3K36me2 [337]. These data imply that centromeric transcription is an important upstream regulator in CENP-A deposition. Further, a later study showed that acetylating centromeric chromatin results in an increase in centromeric transcription, and a similar loss of CENP-A. Here, the authors tethered the p65 subunit of the NF-kB transcription factor complex (NF-kB-p65) to the alphoidtetO array by creating an NF-kB- p65-tetR fusion protein. This resulted in increased H3K9 acetylation, and a 10 fold increase in transcription of the alphoidtetO arrays. They also tethered the herpes virus transcriptional activator protein VP16 to the alphoidtetO arrays using the same technique, which resulted in similar histone modifications and a 150-fold increase in transcription at the HAC core centromere. This increase in transcription was paralleled by a dramatic loss of CENP-A at the HAC centromere due to a defect in CENP-A loading [366]. Taken together, these two studies reinforce the notion that dysregulation of centromeric

61 transcription, whether it be an increase in transcription or a decrease, has detrimental effects on centromere identity and function.

A Functional Requirement for Transcription at Core Centromeres:

In recent years, it has become clear that transcription through the core centromere is important for centromere function. Disrupting core centromere transcription has resulted in several different defects, including reduction or loss of kinetochore proteins, aberrant mitoses, and a decrease in CENP-A. However, the exact function of transcription is still unclear.

Some suggest that the act of transcription alone facilitates CENP-A deposition, as the general process of transcription requires chromatin and nucleosome remodeling.

Passage of RNA Polymerase II through the region may initiate nucleosome remodeling that allows for HJURP-dependent deposition of CENP-A. Although this theory has not been substantiated, the presence of the FACT complex at centromeres allows for the possibility. During gene transcription, FACT aids in nucleosome disassembly prior to the passage of RNA Polymerase II, and then reassembles nucleosomes behind RNA

Polymerase II, contributing to efficient transcription [367]. The FACT complex has been linked to CENP-A in several species, including humans. Although it has not been implicated specifically in CENP-A deposition in human cells, one study showed that knockdown of FACT subunit SSRP1 resulted in a decrease in CENP-A [362]. In yeast,

FACT is capable of removing CENP-A from non-centromeric loci [368], and in flies,

CENP-A assembly requires RNA Polymerase II-mediated transcription of the core centromere, and that transcription is dependent on FACT [369]. Thus, the movement of

RNA Polymerase II through the centromere, and the accompanying chromatin modifications required, may facilitate CENP-A deposition independently of the transcripts that result.

Studies do show that AS-derived lncRNA transcripts plays an important role in centromere function. Some centromere proteins have been shown to directly bind to lncRNA, and this RNA binding is required for their centromeric localization. For example,

CENP-C contains a domain that is homologous to the HP1 (heterochromatin protein 1) hinge region (CENP-C amino acids 426–551), a region shown to have RNA-binding activity. This CENP-C domain binds AS-derived lncRNA by electrophoretic mobility-shift assay, and RNAse treatment abolishes CENP-C localization to centromeres by immunofluorescence [326]. This indicates that the centromere-derived transcript is required for CENP-C localization. As mentioned, CENP-C is a critical protein for maintaining centromere identity and function, as it is required for CENP-A deposition through the recruitment of the Mis18 complex, as well as for recruitment of outer kinetochore proteins. Thus, AS-derived lncRNA transcripts are indispensable for localization of these proteins as well. The same study showed that INCENP (inner centromere protein) and Survivin, both components of the CPC (chromosomal passenger complex), fail to localize to metaphase centromeres in cells treated with

RNAseA, suggesting that an RNA component is required. Further, introduction of endogenous AS RNA restored localization of CENP-C and INCENP. Survivin was not evaluated [326]. Another study found that AS-derived lncRNA binds to Aurora B kinase, also a component of the CPC, and is required for its localization to metaphase centromeres [370].

A recent study showed that inhibition of transcription results in a decrease in

CENP-A and HJURP recruitment to centromeres. Computational RNA-binding prediction algorithms suggest that both HJURP and CENP-A may bind RNA. To test if AS-derived lncRNAs bind to these proteins, HJURP and CENP-A were immunoprecipitated from both a soluble fraction and a chromatin fraction of HeLa cells. Then, RNA was purified from the immunoprecipitated proteins, and evaluated for AS-derived lncRNAs by

63 northern blot. Results showed that a 1.3 kilobase RNA species binds to CENP-A in both the soluble and the chromatin fraction, but only binds HJURP in the soluble fraction.

Thus, lncRNA may bind to the HJURP/CENP-A complex prior to CENP-A loading, and may also bind CENP-A that has been incorporated into the centrochromatin. Abolishing

AS-derived lncRNAs by RNAi prevented recruitment of both proteins to centromeres

[361].

The RNA transcripts themselves may hybridize to the centrochromatin from which they are transcribed, forming an RNA-DNA hybrid called an R-loop. An R-loop is a three-stranded nucleic acid structure generated by the act of transcription, and is formed by the RNA-DNA hybrid plus the displaced DNA strand identical to the RNA molecule

[371]. Generally, these structures pose a threat to genomic integrity, as they cause an increase in spontaneous mutagenicity. R-loops are also a substrate for DNA modifying enzymes like AID (activation induced cytidine deaminase), an enzyme known to cause

DNA double-strand-breaks during antibody gene diversification [372]. At centromeres, the repair of double-strand-breaks caused by R-loops could result in repair-coupled chromatin remodeling that facilitates CENP-A deposition [373]. Evidence for this has only been observed in Xenopus, where CENP-A assembly is modulated by the presence or absence of active DNA repair machinery at centromeres [374].

Dysregulation of Centromeric Transcription:

Many studies have shown that loss of AS-derived lncRNAs result in aberrant mitoses. In one study using HeLa cells, knockdown of AS-derived transcripts using antisense RNA/DNA chimeric oligonucleotides, which degrade annealed RNAs by endogenous RNase H activity, resulted in nuclei with an abnormal morphology.

Immunofluroescence analysis revealed that the abnormal nuclei resulted from the inability of spindle microtubules to make connections with chromosomes, and as a result, chromosomes failed to align along the metaphase plate [370]. In another study,

64 depletion of AS-derived lncRNAs by RNAi resulted in 42.2% of cells having multipolar spindles and lagging chromosomes [361].

While it is evident that depletion of AS-derived lncRNAs results in mitotic defects, the same appears to be true for an increase in expression of these lncRNAs. The tammar wallaby is a mammalian model system useful for studying centromeres because their centromeres are small (about 420 kilobases) and well characterized. One component of the tammar wallaby centromere is the kangaroo endogenous retrovirus

(KERV) long terminal repeat (kLTR), a region of centromeric chromatin that produces

RNA transcripts ranging in size from 34 to 42 nucleotides. These RNAs are termed centromere repeat-associated short interacting RNAs, or crasiRNAs. When crasiRNAs were over-expressed by transfection of kLTR sequences, cells failed to load newly transcribed CENP-A molecules at centromeres. Further, a range of mitotic defects resulted, including chromatin fragments and lagging chromosomes [375]. Studies in

HACs showed a similar effect when transcription of HAC centromeres were upregulated.

Acetylating centromeric chromatin resulted in an increase in centromeric transcription, and a similar loss of CENP-A [366]. Finally, ectopically expressing AS DNA in human mammary epithelial cells resulted in increased abnormal mitoses, including anaphase bridges, lagging chromosomes, and disorganization of chromosomes during metaphase

[376].

Conclusions

Centromere identity must be carefully maintained throughout the cell cycle so that the CCAN and the kinetochore complex properly associate and assemble onto one and only one locus per chromosome to avoid errors in chromosome segregation. As discussed above, it is clear that CENP-A is the most important factor for centromere identity. Deletion of CENP-A results in a mitotic lethal phenotype [284], and over-

65 incorporation, under-incorporation, or mis-incorporation of CENP-A molecules is associated with chromosome segregation errors and cancer [287, 288]. This likely results from disrupted CCAN assembly and kinetochore formation, and subsequent spindle attachment errors. Current research suggest that CENP-A deposition depends on CCAN components CENP-C and CENP-I. These CCAN components recruit the

Mis18 complex to centromeres, which in turn recruits HJURP to facilitate the exchange of canonical histone H3 for CENP-A, thereby maintaining CENP-A protein levels and centromere identity.

Although the exact function of transcription through core centromeres is still unclear, recent work has indicated that transcription is an important regulatory component for CENP-A deposition. As discussed above, the transcription-coupled chromatin remodeling associated with core transcription may itself assist in CENP-A deposition. Additionally or alternatively, the lncRNA transcripts that result from core transcription bind to centromere associated proteins, including CENP-C, HJURP, and

CENP-A, and act to recruit and/or stabilize these proteins at centromeres. Further, lncRNAs may bind to centromeres to create R-loops, and the chromatin remodeling associated with repair of R-loop-induced DNA double-strand-breaks could facilitate

CENP-A deposition. Despite its emerging importance, very little is known about what regulates transcription through core centromeres. Our work discussed in Chapters 4 and

5 demonstrates that Sp1 is a negative regulator of core transcription and is therefore an important upstream factor for CENP-A deposition. Sp1 likely regulates core transcription by maintaining histone modifications at this region, as these are also disrupted in Sp1- depleted cells. Further, our work shows that Sp1 depletion causes a decrease in CENP-

C binding to centromeres, as well as a decrease in CENP-A deposition. Taken together, these data indicate that Sp1 is an important component of centromere biology, and as such contributes to faithful chromosome segregation.

Figure 2.1: Overview of Mitotic Chromatin. As a cell enters mitosis, chromatin condenses to form mitotic chromosomes. Sister chromatids remain joined at centromeres with the help of sister-chromatid-cohesin complexes that are loaded during DNA synthesis. Centromeres are defined by CENP-A-containing nucleosomes and are flanked on either side by pericentromeric heterochromatin. The constitutive-centromere- associated network of proteins (CCAN) associates with CENP-A-containing regions throughout the cell cycle. During mitosis, the CCAN recruits outer kinetochore proteins, which interact with spindle microtubules emanating from centrosomes at opposite poles of the cells

Figure 2.2: The Constitutive-Centromere-Associated Network (CCAN) and the Kinetochore. The CCAN encompasses a network of 16 proteins that localize to centromeres throughout the cell cycle and serve as a bridge between the centromeric chromatin and the outer kinetochore and spindle microtubules. CCAN components can be broken down into five groups, including CENP-C, the CENP-L-N complex, the CENP- H-I-K-M complex, the CENP‑O‑P‑Q‑U‑R complex, and the CENP‑T‑W‑S‑X complex. CENP-C is a key kinetochore assembly factor. It is recruited to centromeres by CENP-A and is also required for CENP-A loading at centromeres through recruitment of the Mis18 complex. CENP-C also recruits the CENP-H-I-K-M complex to centromeres, as well as outer kinetochore component Mis12. CENP-H-I-K-M interacts with several CCAN proteins, including CENP-T-W-S-X and CENP-C, and thus may function to stabilize these proteins. CENP-L-N interacts with both CENP-C and the CENP-H-I-K-M complex. The function of the CENP-O-P-Q-U-R is still unclear, as disruption of this complex by protein knockdown of CENP-O had no measurable effect on kinetochore assembly or function [377]. CENP-T-W-S-X binds centromeric DNA and is thought to induce positive supercoiling. It also interacts with KMN network protein Ndc80. The KMN network of the outer kinetochore makes attachments with spindle microtubules. Each centromere contains 20 to 25 kinetochores and attaches to the same number of microtubules.

Figure 2.3: Human A-satellite Consensus Sequence. The human α-satellite consensus sequence is 171 base pairs in length, with a high adenine and thymine content (62%) [378].

Figure 2.4: Histone Modifications at the Centrochromatin. Histone H3 is di- methylated on lysine 4, and di- and tri-methylated on lysine 36. These modifications are known to be associated with active transcription. CENP-A is tri-methylated on glycine 1, catalyzed by NRMT. Serine 7 is phosphorylated, which is important for stabilization of the CENP-A nucleosome, as well as the interaction between CENP-A and CENP-C. CENP-A is also phosphorylated on serine 16 and 18, modifications that also influence the stability of the CENP-A nucleosome as well as the conformation of CENP-A containing nucleosome arrays. Finally, lysine 124 is monoubiquitinated by the CUL4A- RBX1-COPS8 E3 ubiquitin ligase, required for the interaction between CENP-A and its assembly factor HJURP. The histone H4 within CENP-A containing nucleosomes is methylated on lysine 20, and promotes localization of CENP-T. H2B is monoubiquitinated (H2Bub1), although it is not known if this modification is restricted to CENP-A- or H3-containing nucleosomes. H2Bub1 is also associated with active transcription in other areas of the genome. Pericentromeric heterochromatin contains typical heterochromatin modifications, some of which are catalyzed by Suv39h1, including H3K9me3 and H4K20me3. Centromchromatin generally lacks histone H3 and H4 acetylation. p300/CBP, HDAC1, and DNMT1 have all been shown to localize to centromeres during mitosis, but the targets of these modifiers have not been identified.

Chapter 3: Specificity Protein 1

Introduction

Specificity Protein 1 (Sp1) was identified in the early 1980’s as a factor that has the ability to alter the activity of RNA Polymerase II so that it can recognize the SV40 promoter and activate transcription of SV40 genes [379, 380]. In the last almost four decades, hundreds of studies have focused on attempting to gain a complete understanding of the transcriptional functions of Sp1. Initial studies suggested that Sp1 was a general transcription factor required for basal transcription of housekeeping genes. This notion was supported by the fact that the GC-rich Sp1 binding sequence 5’-

(G/T)GGGCGG(G/A)(G/A)(G/T)-3’ is present in the promoters of many housekeeping genes [381] and that Sp1 was found to be required for the assembly of the RNA

Polymerase II preinitiation complex (discussed below) [382, 383]. It is now clear that Sp1 is not just a general transcription factor required for basal transcription but also has promoter-specific functions that include activating or repressing transcription of specific genes with TATA-less or TATA-containing promoters. Here, Sp1 functions to recruit transcriptional co-regulators such as chromatin remodeling complexes to promoters to influence gene transcription [382].

In recent years, we have focused efforts on elucidating non-transcriptional roles for Sp1, adding another layer of complexity to a protein with a diversity of functions. The most well characterized non-transcriptional role is its facilitation of DNA double-strand- break repair. Our studies have demonstrated that Sp1 is required for repair, as depletion of Sp1 by RNAi inhibits repair at site-specific DNA double-strand-breaks. Further, we have shown that Sp1 is phosphorylated by the DNA-damage sensory kinase ATM

(ataxia telangiectasia mutated) in response to DNA double-strand-breaks and is rapidly recruited to regions immediately adjacent to sites of DNA damage [384]. Within this context, Sp1 appears to recruit the histone acetyltransferase p300 to break sites to facilitate chromatin remodeling required for repair (Guo, Beishline, Azizkhan-Clifford,

72 unpublished). In addition to its role in DNA double-strand-break repair, Sp1 also localizes to centrosomes and regulates centrosome number as depletion of Sp1 by RNAi results in centrosome amplification. Further, a deletion mutant of Sp1 lacking amino acids 1-182 does not localize to centrosomes or rescue the supernumerary centrosome phenotype in Sp1-depleted cells. Because this portion of the Sp1 protein has almost

100% transcriptional activity on its own, these data suggest that regulation of centrosome number and function by Sp1 is non-transcriptional [7]. More recently, we have shown that amino acids 1-182 alone can localize to centrosomes and rescue the centrosome defects observed in Sp1-depleted cells, further supporting this notion

(Flashner, Sowash, Azizkhan-Clifford, unpublished).

In Chapters 4 and 5, I discuss a novel function for Sp1 in preserving centromere identity and function by regulating transcription through the core centromere during mitosis, likely through the recruitment of chromatin modifiers. As such, this chapter will introduce the Sp1 protein and review what is currently known about Sp1 during mitosis.

It will also discuss chromatid modifiers known to interact with Sp1 at gene promoters to influence transcription; these modifiers may be recruited by Sp1 to centromeres to maintain the architecture of the centrochromatin.

The Specificity Protein/Krüppel-like Factor (SP/KLF) Transcription Factor Family

Sp1 is a member of the Specificity Protein/Krüppel-like Factor (SP/KLF) transcription factor family [385]. In addition to Sp1, the Sp/KLF transcription factor family contains Sp2-4, Sp5-9, and a number of KLF transcription factors [386-388]. This family is united by the highly conserved Cys2His2-type zinc-finger, DNA-binding domain that consists of three adjacent zinc fingers close to the C-terminus of the proteins, which shares more than 65% sequence identity throughout [385]. Sp and KLF factors differ in the DNA sequences with which they preferentially bind. Sp factors bind with high affinity

73 to GC-boxes (GGGGCGGGG), whereas KLF factors bind to the related GT/CACCC-box

(GGTGTGGGG) [389]. Of these family members, Sp1-4 are most similar. In addition to the DNA-binding domain, Sp1-4 also each contain transactivation domains that are serine/threonine-rich and are flanked by glutamine rich regions [388]. These and other domain features of Sp1-4 are illustrated in Figure 3.1. Sp1 and Sp3 are ubiquitously expressed in mammalian cells and are the most thoroughly studied. Sp2 is also expressed ubiquitously [390], but Sp2 does not bind to GC-boxes, significantly differentiating it from the other Sp proteins. Sp4 expression is mostly limited to the brain

[391]. Because the data in Chapters 4 and 5 describe a novel function for Sp1, this section will focus solely on Sp1.

The Sp1 Gene and Protein(s)

The Sp1 gene is located on chromosome 12 at position q13.1 (position

53380195 in the current GRCh38.p2 assembly) and consists of 6 exons and 5 introns spanning 36 kilobases. There are three known isoforms of the Sp1 protein. Isoform 1, also known as Sp1a, is translated from a 2355 base pair transcript, creating a 785 amino acids protein that is 80.9 kilodaltons in the unmodified form. Isoform 2 (Sp1b) lacks the first 7 amino acids found in isoform 1, contains an alternative exon in the 5′ coding region, and is created using a second ATG start codon found at position +21 from the start codon used by isoform 1. The unmodified form of Sp1b is 778 amino acids and 80 kilodaltons. A third isoform, Sp1c, is created through alternative splicing and results in a protein that lacks amino acids 54 through 101. The unmodified form of Sp1c is 737 amino acids and 75.8 kilodaltons [392, 393]. Sp1a is the primary form of the protein and is ubiquitously expressed in all mammalian tissues. While Sp1a is likely one of the most well-characterized transcription factors to date, much less is known about Sp1b and

Sp1c. One group showed that Sp1c is also ubiquitously expressed, although at very low

74 levels as compared to Sp1a [393]. The Sp1 protein is highly unstructured, with the exception of the DNA-binding domain. Sp1 is also highly post-translationally modified

(discussed in more detail below). These two features allow the Sp1 protein to have multiple interacting partners, contributing to its functional diversity.

Sp1 and Gene Regulation

The promoters of genes contain a variety of different DNA elements that are recognized by both general and specific transcription factors. These transcription factors bind to promoter elements and function to control gene expression. It is predicted that there are over 12,000 Sp1 binding sites within the genome, indicating that Sp1 transcriptionally regulates a vast amount of genes that are involved with almost all cellular processes [385]. Sp1 has been shown to act as a general transcription factor as well as both a transcriptional activator and transcriptional repressor.

Sp1 as a General Transcription Factor:

General transcription factors mark promoters that require RNA Polymerase II for gene transcription. To begin transcription, the TFIID protein complex binds to a short sequence of adenine and thymine nucleotides, called the TATA box, through the TFIID component TBP (TATA-binding protein). It is thought that binding of the TFIID complex causes a topological distortion in the DNA, which serves as a physical indicator of promoter activation. From there, other general transcription factors, as well as RNA

Polymerase II, assemble to form the transcription-initiation complex. However, not all gene promoters contain TATA-boxes and require alternative means for recruiting and binding the TFIID complex. As a general transcription factor, Sp1 binds to Sp1 binding sites within gene promoters and interacts with components of the TFIID complex to recruit the complex to DNA. Sp1 interacts directly with TBP as well as TAFII130 (TBP- associated factor) and TAFII55 to stimulate transcription initiation [394, 395]. Sp1 also

75 assists in the stable assembly of TFIIB and TFIIE [383] and may interact with TFIIA, although it is not clear if this interaction is direct [396].

In addition to general transcription factors, eukaryotic genes bind other proteins to influence gene transcription, including mediator proteins, transcriptional activators and repressors, and chromatin modifying enzymes. Mediator proteins act as a bridge between general transcription factors and transcriptional activators and repressors that may be thousands of nucleotides away from the transcription-initiation complex. In some contexts, Sp1 interacts with the mediator protein CRSP (cofactor required for Sp1), as with the synergistic activation by Sp1 and SREBP-1a (sterol regulatory element-binding protein-1a) [397, 398]. Sp1 as a transcriptional activator and repressor is discussed below.

Sp1 as a Specific Transcription Factor:

In addition to aiding in the assembly of the preinitiation complex at RNA

Polymerase II-dependent gene promoters, Sp1 acts as a transcriptional activator and transcriptional repressor at specific gene control regions. Transcriptional activators and repressors function to direct local alterations in chromatin structure to stimulate or repress transcription, respectively. This occurs through alterations in histone modifications and nucleosome architecture, which is accomplished though attracting histone modifying enzymes, chromatin remodeling complexes, and histone chaperones to DNA.

Histone acetyltransferases function by transferring acetyl groups to lysine residues on histones, neutralizing the positive charge normally present on lysines. This reduces the affinity between the histones and the negatively charged DNA, making the

DNA more accessible for transcription factors, thereby accelerating the rate of transcription. As a transcriptional activator, Sp1 binds directly to histone acetyltransferases like p300 and CBP (CREB-binding protein) and positions them at

76 promoters to stimulate transcription. For example, Sp1 recruits p300 to the promoter of the 12(S)-lipoxygenase gene to enhance its expression [399]. Similarly, in some contexts, Sp1 interacts with p300 to activate the p21 promoter and the p27 promoter

[400, 401] and stimulates the histone acetyltransferase activity of CBP in vitro [402].

Histone deacetylases and DNA methyltransferases function to repress gene activity. Histone deacetylases remove acetyl groups from lysines, allowing for DNA to wrap more tightly around nucleosomes, thereby preventing access by transcription factors. DNA methyltransferases catalyze the transfer of methyl groups onto DNA, and when this occurs in promoter regions, DNA methylation typically acts to repress gene expression. Sp1 is known to bind to both histone deacetylases and DNA methyltransferases to negatively affect gene expression. For example, amino acids 619-

785 of Sp1 interact with HDAC1 (histone deacetylase 1) to facilitate repression of the p21 gene. Here, HDAC1 competes with p53 for Sp1 binding, and as such, Sp1 acts as a mediator of both positive (when associated with p53 or p300, as mentioned above) and negative (when associated with HDAC1) p21 expression, depending on the cellular context [403-405]. Sp1 is also required for HDAC1-dependent repression of SSeCKS

(Src-suppressed C kinase substrate) as transcriptional repression of SSeCKS correlates with increased binding of Sp1 and increased recruitment of HDAC1 to the SSeCKS promoter [406]. Another study using the S-phase-specific mouse thymidine-kinase (TK) promoter as a model system showed that HDAC1 interacts with the C-terminus of Sp1 to repress transcription of TK expression. Here, HDAC1 competes with transcription factor

E2F1 for Sp1 binding; the E2F1-Sp1 interaction relieves HDAC1-mediated repression

[407]. Further, in normal human somatic cells, Sp1 is tightly associated with hTERT

(human telomerase reverse transcriptase) gene promoters and recruits HDAC2 (histone deacetylase 2) to transcriptionally silence hTERT expression [408]. In HeLa cells, one study demonstrated that Sp1 recruits the DNA methyltransferase DMNT1 to the MAZ

(Myc-associated zinc-finger protein) promoter to repress transcription of the MAZ gene.

Here, treatment with the methylation inhibitor 5-azacytidine reversed Sp1-dependent

MAZ repression [409]. Finally, Sp1 interacts with a complex of proteins including

HDAC1, HDAC2, and RbAp48 at the hLHR (human luteinizing hormone receptor) promoter to silence transcription of hLHR [410].

ATP-dependent chromatin remodeling complexes can move, eject, or restructure nucleosomes in order to influence gene expression. Sp1 is known to interact with several of the components of the nucleosome remodeling complex SWI/SNF

(SWItch/Sucrose Non-Fermentable) [382]. In particular, at the MMP-2 (matrix metalloproteinase 2) promoter, Sp1 interacts with BRG1, a component of the SWI/SNF complex, for constitutive expression of MMP-2 in tumor cells [411]. Further, an in vitro transcription assay showed that Sp1-dependent transcription required SWI/SNF component PBAF (polybromo- and BAF-containing complex) [412]. Finally, Sp1 interacts with and recruits SWI/SNF1 to human β-globin-gene promoters to activate transcription

[413, 414]. In addition to the SWI/SNF complex, Sp1 interacts with HMGA1 and HMGA2

(High Mobility Group A) proteins to enhance gene transcription. HMGA proteins can both positively and negatively regulate gene transcription and do so by binding to the minor groove of DNA, resulting in a change in the DNA structure, as well as by binding to a variety of transcription factors. Sp1 was found to associate with HMGA1 at the promoter of the IGF-I receptor gene to stimulate transcription [415] and interacts with HMGA2 at the hTERT promoter, thereby stimulating hTERT expression in HeLa cells [416]. Overall,

Sp1 not only acts as a general transcription factor but also influences transcription of specific genes by interacting with a multitude of activators and repressors at gene control regions.

Post-Translational Modifications (PTMs) for Sp1

Sp1 is a highly post-translationally modified protein. The unstructured nature of

Sp1 allows for many different interacting partners, and as such, the PTMs of Sp1 likely confer specificity to those interactions. Sp1 is known to be phosphorylated, O-linked glycosylated, acetylated, SUMOylated, and ubiquitinylated. For a complete list of Sp1’s

PTMs, including the enzyme responsible for the PTM and the consequence of the modification (when known), refer to Table 3.1. Mitosis-specific PTMs will be reviewed in this section.

An early study aimed at examining phosphorylation of the 5-amino-acid-linker sequence separating the zinc fingers in proteins containing Cys2His2-type zinc-finger,

DNA-binding domains showed that the linker sequence in Sp1 is phosphorylated during mitosis as a potential mechanism for mitotic inactivation of the protein. The phosphorylated residues were identified as threonine 651 and threonine 681 [417]. While

PKC-ζ (protein kinase C-ζ) has been shown to phosphorylate Sp1 at these residues in other contexts, mitotic-specific phosphorylation of Sp1 by PKC-ζ has not been demonstrated [418]. Sp1 was shown to be phosphorylated at serine 59 but the cell cycle timing of this particular phosphorylation is controversial. One study demonstrated that serine 59 is phosphorylated by CDK2/Cyclin A in late G1 and S phases of the cell cycle to enhance Sp1-mediated transcription [419]. Another study, however, showed that Sp1 is phosphorylated at this residue during mitosis; in this study, the authors monitored phosphorylation of this residue in resting and dividing T lymphocytes using an antibody developed specifically against phospho-serine 59. They found that the phospho-serine-

59 signal is diminished in interphase cells but not in mitotic cells, suggesting that it is a mitotic phosphorylation. They also showed that the phospho-serine-59 signal could only be partly decreased by treatment with the CDK2 inhibitor roscovitine, suggesting that another kinase is also responsible for phosphorylation of this residue. Interestingly though, over-expression of the mitotic kinase CDK1/Cyclin B1 did not increase

79 phosphorylation status at serine 59, even when wild-type Sp1 was over-expressed [420].

Sp1 is phosphorylated by JNK1 (c-Jun NH(2)-terminal kinase 1) at threonines 278 and

739 during mitosis and also on threonine 739 by CDK1/Cyclin B1 during mitosis, as discussed below.

Sp1 and Mitosis

Studies have shown that during mitosis, as chromatin condenses into chromosomes, most transcription factors dissociate from chromatin and become dispersed throughout the cell. By telophase, these transcription factors are largely restored so that they are available for DNA binding as the chromatin decondenses at the exit of mitosis. Examples of transcription factors that follow this pattern include Oct-1,

Fos-family transcription factors, and TFIIB [421, 422]. Other transcription factors have been shown to remain associated with mitotic chromatin throughout mitosis, including the transcription factor Runx1 (Runt-related transcription factor 1) [423] and AP2

(activating protein 2) [421]. As such, it is difficult to predict the localization pattern of a particular transcription factor without careful evaluation. Our studies focus on the behavior of Sp1 during mitosis. Thus, this section summarizes how Sp1 is described during mitosis in the current literature.

A handful of studies have shown that Sp1 is ejected from chromatin during mitosis. The first study, published in 1995, demonstrated that in mitotic extracts from

HeLa cells Sp1 showed an approximately 5-fold reduction in specific DNA-binding activity, which was accompanied by a change in phosphorylation status [421]. Then, in

2006, another group analyzed the spatial distribution of Sp1 in MCF-7 cells during mitosis and similarly found that Sp1 translocates away from chromatin during prometaphase and returns by telophase. This study additionally showed that Sp1 co- localizes with F-actin microfilaments in the cytoplasm, likely to retain organization and to

80 stabilize the protein during cell division [424]. Importantly, a later study revealed that Sp1 is phosphorylated by JNK1 (c-Jun NH(2)-terminal kinase 1) at two residues during mitosis, including threonine 278 and 739, and inhibition of JNK1 activity resulted in the ubiquitination and degradation of Sp1 [425]. Threonine 739 is close to the Sp1 DNA- binding domain and thus likely facilitates Sp1’s eviction from the chromatin by sterically interfering with zinc-finger binding. The interaction between JNK1 and Sp1 is facilitated by Hsp90 (heat shock protein 90), which binds Sp1 primarily during mitosis [426]. A later study demonstrated that phosphorylation on threonine 739 also functions to stabilize

Sp1 by preventing the interaction between Sp1 and the E3 ubiquitin ligase RNF4 [427].

Interestingly, Sp1 is also phosphorylated on threonine 739 by CDK1/Cyclin B1 during mitosis to suppress DNA binding. In this study, mutation of threonine 739 to alanine resulted in Sp1 remaining bound to chromatin. Mass spectrometry also revealed that

Sp1 interacts with F-actin and myosin during mitosis, supporting earlier studies. This study also showed that Sp1 is dephosphorylated by PP2A (protein phosphatase 2A) upon mitotic exit to restore Sp1 DNA binding affinity [420, 428]. In contrast to what is reported in the literature, Chapters 4 and 5 describe our studies showing that a small pool of Sp1 remains at centromeres during mitosis and is involved with maintaining centromeric architecture through transcriptional regulation at this region.

Sp1 and Whole Chromosomal Instability (W-CIN)

The cell employs several surveillance systems to ensure that its genomic content is passed to its daughter cells with little genomic variation. As discussed in Chapter 1 and 2, there are several mechanisms at centromeres and centrosomes that function to ensure proper chromosome segregation. These mechanisms can be disrupted in a number of different ways, resulting in chromosome segregation errors that can lead to whole chromosomal instability (W-CIN), aneuploidy, and cancer. The data presented in

Chapters 4 and 5 reveal a previously uncharacterized role for Sp1 in preserving chromosomal stability. Other known links between Sp1 and W-CIN are discussed below.

Sp1 and Centrosome Regulation:

In a previous study, we demonstrated that Sp1 depletion results in phenotypes that are consistent with W-CIN. Here, NHDF (normal human diploid fibroblast) cells depleted of Sp1 by RNAi showed a significant increase in the number of multi-nucleated cells as compared to control cells. Sp1 depletion also resulted in mis-alignment of metaphase chromosomes along the metaphase plate. In the W-CIN-negative mammary epithelial cell line MCF 10A, Sp1 depletion resulted in aneuploidy as Sp1-depleted cells showed a statistically significant increase in chromosome number as compared to control cells. The lack of tetraploid karyotypes suggested that this was not due to cytokinesis defects but rather from errors in chromosome segregation [7]. In addition to

W-CIN, these studies also demonstrated that Sp1 depletion results in centrosome amplification. As discussed in Chapter 1, centrosome amplification can lead to improper attachment of microtubules to kinetochores and if not corrected, chromosome segregation errors. Further, this study showed that Sp1 localizes to centrosomes, and a deletion mutant of Sp1 lacking amino acids 1-182 did not localize to centrosomes or rescue the supernumerary centrosome phenotype in Sp1-depleted cells. Because this portion of the Sp1 protein has almost 100% transcriptional activity, these data suggest that regulation of centrosome number and function by Sp1 is non-transcriptional. More recently, we have found that a fragment of Sp1 comprising amino acids 1-182 can on its own localize to centrosomes and rescue the supernumerary centrosome phenotype

(Flashner, Sowash, Azizkhan-Clifford, unpublished). Taken together, it is clear that Sp1 maintains faithful chromosome segregation through centrosome regulation.

Sp1 and Survivin:

Survivin is a component of the CPC (chromosomal passenger complex) and functions to correct mal-attached microtubules through Aurora B kinase activity. This mitotic function for Survivin is evolutionally conserved and severe chromosome segregation defects have been found in Survivin-mutant yeast strains [429-432],

Caenorhabditis elegans in which Survivin was knocked down by RNAi [433, 434], and

Survivin-deficient mice [435]. Within this complex, Survivin dictates CPC localization by targeting it to centromeres through two aspartic acid residues in the BIR domain

(discussed below) and is thus vital for CPC function [436, 437].

In addition to functioning as a mitotic regulator, Survivin is also a member of the family of inhibitor of apoptosis proteins (IAPs) containing one baculovirus IAP repeat

(BIR) domain [438]. IAPs function to inhibit apoptosis by directly binding to caspases, which prevents caspase cleavage and activation [439]. Survivin is dramatically over- expressed in many tumors and fetal tissues [440], and the National Cancer Institute's cancer drug-screening program found that Survivin is expressed in all 60 human tumor lines that were tested with the highest levels in breast and lung cancer cells [441, 442].

Conversely, Survivin is expressed at very low levels or undetectable in normal healthy tissues. Over-expression of Survivin has been shown to inhibit apoptosis both through the extrinsic and the intrinsic apoptotic pathways. Here, Survivin binds to caspases 3, 7, and 9, disrupting the caspase-cleavage cascade required for apoptosis [443]. Further,

Survivin was shown to inhibit cytochrome C and caspase 8 cleavage activity [444]. As such, in addition to being an important mitotic regulator, Survivin is also an integral component for apoptosis inhibition and cancer cell survival. Sp1 transcriptionally regulates Survivin gene expression. The Survivin promoter contains several Sp1 binding sites, and as such, perturbations in Sp1 expression affect Survivin-dependent apoptotic signaling [445].

Although it has not been carefully evaluated, changes in Sp1 protein levels likely also affect mitosis in a Survivin-dependent manner. Over-expression of Sp1 in tumors is correlated with increased Survivin expression [446, 447]. In a stomach adenocarcinoma cell line, over-expression of Survivin resulted in a decrease in cells that underwent mitotic slippage and subsequent cell death in response to treatment with the microtubule destabilizing drug monastrol. Here, over-expression of Survivin sustained mitotic arrest, potentially through increased CPC activity, thereby increasing viability of the cells and contributing to cell survival and continued proliferation [448]. Further studies are needed to evaluate the impact of Sp1-dependent changes in Survivin levels and CPC activity during mitosis.

Sp1 and Cancer

In their 2000 paper entitled “Hallmarks of Cancer”, Hanahan and Weinberg described eight major “hallmarks” or characteristics that allow a transformed cell to survive, proliferate, and disseminate. These include sustained proliferative signaling, replicative immortality, resistance to cell death and avoidance of immune destruction, induction of angiogenesis, invasion and metastasis, and de-regulation of cellular energetics (Figure 1.1) [159]. Sp1 regulates genes whose products contribute to each of these characteristics, indicating that Sp1 could, and likely does, play a substantial role in cell transformation. A comprehensive list of cancer-related genes regulated by Sp1 can be found in Table 3.2. In support of this, Sp1 is over-expressed in several different types of cancers, including breast, gastric, pancreatic, lung, brain, and thyroid cancers, and increased Sp1 protein levels correlate with poor patient survival in several of these cancers [449-453]. Interestingly, Sp1 both activates and suppresses a number of oncogenes and tumor suppressor genes, respectively, and regulates both pro-survival genes and pro-apoptotic genes, underscoring its complexity as a transcription factor and

84 a contributor to tumorigenesis. Targeting Sp1 as an anti-cancer therapeutic is attractive because multiple pro-oncogenic pathways and genes would be affected [454]. However, the opposing functions of Sp1 also makes targeting Sp1 challenging, and a more complete mechanistic understanding of Sp1’s activities is required. We recently published an in-depth review addressing the role of Sp1 in cancer, including a thorough description of Sp1 target genes from each of the “hallmarks of cancer” [388].

Conclusions

Although Sp1 is likely the most thoroughly studied transcription factor, a complete understanding of the diversity its functions is still lacking. Sp1 regulates a vast number of genes involved with almost all cellular processes. Further, it is dysregulated in many types of cancers and is associated with poor prognosis. In addition to its role as a general and gene-specific transcription factor, we have characterized non-transcriptional roles for Sp1, including its role in DNA double-strand-break repair and centrosome regulation. By regulating centrosome number and function, Sp1 contributes to faithful chromosome segregation, thereby preserving chromosome stability. While Sp1 appears to be a promising target for anti-cancer therapeutics, a more complete understanding of its diverse functions is required.

Chapters 4 and 5 describe a novel function for Sp1 during mitosis. The current literature describes Sp1 as being evicted from chromatin during mitosis and stabilized in the cytoplasm (discussed above). Our work shows that a small pool or Sp1 remains associated with centromeres and functions to preserve centrochromatin architecture and centromeric transcription. This, in turn, allows for proper localization and assembly of centromere-associated proteins that are necessary for faithful chromosome segregation.

Therefore, in addition to regulating centrosome number and function, Sp1 contributes to

85 maintaining chromosomal stability at centromeres, further diversifying the contributions of Sp1 to normal cell biology.

Figure 3.1: Domain Features of Sp1, Sp2, Sp3, and Sp4 (reprinted from [388] with permission from John Wiley and Sons and The FEBS Journal). The most significant features that unite Sp1–Sp4 include the highly conserved Cys2His2 zinc-finger, DNA- binding domain as well as transactivation domains (TAD) indicated above. Sp1, Sp3, and Sp4 contain two TADs that are serine/threonine rich, and are flanked by regions that are glutamine rich. By contrast, Sp2 only contains one TAD. The zinc-finger, DNA- binding domain consists of three adjacent zinc fingers, which share more than 65% sequence identity throughout the Sp/KLF transcription factor family. Each protein contains a Buttonhead (Btd) domain just N-terminal to the DNA-binding domain within a highly charged region of the proteins. The Btd domain is a conserved stretch of 11 amino acids originally identified in the Drosophila melanogaster Sp1 homolog Buttonhead. Deletion of this domain reduces in vitro activity of Sp1 [387]. Sp1 also contains an N-terminal inhibitory domain as well as a C-terminal multimerization domain, which allows for formation of Sp1 tetramers, multiple stacked tetramers, and synergistic transcriptional activation [385, 388].

Table 3.1: Sp1 Post-Translational Modifications (reprinted from [388] with permission from John Wiley and Sons and The FEBS Journal).

Residue Modification Enzyme Function References

Phosphorylation; N– Ser2 Unknown Unknown [455, 456] acetylserine Ser7 Phosphorylation Unknown Stability [455-457] SUMOylation; [425, 427, 457, Lys16 RNF4 Stability ubiquitinylation 458] Ser56 Phosphorylation ATM/ATR Unknown [459] CDK2; ERK1/2; Stability, [419, 420, 457, Ser59 Phosphorylation PP2A DNA binding 460, 461] Ser101 Phosphorylation ATM/ATR Unknown [459, 462, 463] b

Ser111-114 Glycosylation OGT Unknown Unpublished c

Asp183 Cleavage Casp3 Unknown Unpublished a

Ser220 Phosphorylation DNA-PK Transcription [464] Thr278 Phosphorylation JNK1; ERK1/2 Stability [425] b

Ser301 Glycosylation OGT Unknown Unpublished a

Thr355 Phosphorylation ERK1/2; JNK1 Transcription [465] Thr453 Phosphorylation ERK1/2 Transcription [466-468] Ser491 Glycosylation OGT Transcription [469, 470] b

Ser535/540 Glycosylation OGT Unknown Unpublished Asp584 Cleavage Casp3 Unknown [471] Phosphorylation; Ser612 OGT Localization [472] glycosylation Phosphorylation; Thr640 OGT Localization [472] glycosylation Phosphorylation; [472] Ser641 PKCζ; OGT Transcription glycosylation [473] Thr651 Phosphorylation PKCζ [474] a

Thr668 Phosphorylation CKII; PPI; PKCζ DNA binding [472, 475, 476] Ser670 Phosphorylation PKCζ Unknown [476] Thr681 Phosphorylation PP2A; PKCζ Unknown [420, 476] Ser698 Glycosylation Unknown Localization [472] Ser702 Glycosylation OGT Unknown [472] Lys703 Acetylation P300; HDAC1 Unknown [399, 477-479] Ser728 Phosphorylation GSK3β Degradation [480] Ser732 Phosphorylation GSK3β Degradation [480] Transcription, [425, 427, 466, Thr739 Phosphorylation Erk1/2, JNK1 stability 480] a Amino acid residue number is based on the full length 785 amino acid Sp1 protein. b Unpublished data from in vitro glycosylation assays (Beishline, Azizkhan-Clifford). c Unpulished data (Torabi, Azizkhan-Clifford). Paranthetic Numbers indicate amino acid number in original Sp1 sequence which lacks amino acids 1–89.

Table 3.2: Cancer-related genes regulated by Sp1 (reprinted from [388] with permission from John Wiley and Sons and The FEBS Journal). Gene References Sustained proliferation/immortality hTERT/hTERC [408, 416, 481-491] p53/MDM2 [492-497] p16 [498-500] p21 [501-504] IGF1R [415, 505-513] EGFR [514-519] EGF [520] FGF [521-524] IGF [525-530] Apoptosis Survivin [446, 482, 483, 531-537] Trail–R2 [538-540] Bcl–2 [477, 541] TRAIL [542] MCL–1 [534, 543] XIAP [544] Bak [477, 545] FasL [446, 546-548] Angiogenesis VEGF [467, 549-569] TSP–1 [570] PDGF [466, 571-575] uPA [576-578] DNA damage/stress response Brca1 [579, 580] ATM [581-583] MDC1 [584] Cdc25B [585, 586] RECQ4 [587] PARP [588, 589] XRCC1 [590, 591] BLM [592] CHEK2 [593, 594] XPC [595] XPB [596, 597] XPD [596] DDB1/2 [598] DNA–PK/Ku70/80 [599] XRCC5 [600] ERCC6 (CSB) [601] WRN [602] Invasion and metastasis MMP9 [468, 603] MT1–MMP [604, 605] RECK [468, 606-608] E–cadherin [609-614] Integrin α5 [615-617] MMP2 [411, 452, 603, 618-621]

Chapter 4: Transcription Factor Sp1 Regulates the Centrochromatin Landscape and Centromeric Transcription During Mitosis

Introduction

The centromere is a unique region of chromatin that is vitally important for proper segregation of chromosomes during mitosis and meiosis. In most eukaryotes, the centromere is defined epigenetically by the presence of nucleosomes containing the histone H3 variant CENP-A interspersed between canonical H3-containing nucleosomes

[320, 373, 622-626]. The centromere is vitally important for proper segregation of chromosomes because it serves as a platform for assembly of the multi-protein kinetochore complex, the proteinaceous interface between the mitotic spindle and the chromosomes. Proper assembly and function of kinetochores allows for error-free attachment of spindle microtubules to these structures and faithful chromosome segregation [627, 628]. Defects in chromosome segregation can lead to chromosome instability and thus can have significant negative consequences to the health of the organism [52, 629].

Sp1 is a ubiquitously expressed transcription factor that regulates genes involved with a variety of cellular processes, including cell proliferation, DNA repair, apoptosis, and senescence [386, 388]. Sp1 binds to the GC-rich Sp consensus sequence 5’-

(G/T)GGGCGG(G/A)(G/A)(G/T)-3’ in both proximal and distal gene promoters through its highly conserved Cys2His2-type zinc-finger, DNA-binding domain and functions to recruit general transcription factors and transcriptional co-regulators to promotors to influence transcription [382, 388, 630-633]. Sp1 is also important for maintaining chromosomal stability in human cells. Depletion of Sp1 by RNAi results in abnormal chromosome alignment along the metaphase plate, accumulation of micronucleated cells, and aneuploidy, all of which are phenotypes consistent with whole chromosomal instability (W-CIN) [7]. Based on our previous work showing that depletion of Sp1 by

RNAi results in chromosome mis-segregation [7], we sought to gain a more detailed understanding of the localization and function of Sp1 during mitosis. In this section, we

91 show for the first time that Sp1 localizes and binds to centromeres of mitotic chromosomes where it functions to regulate the transcription of α-satellite (AS)-derived lncRNAs from core centromeres. Localization does not require the Sp1 DNA-binding domain but is dependent on ATM-kinase activity. Loss of Sp1 results in disrupted centrochromatin, including changes in histone modifications, and decreases in centromere-associated proteins including CENP-C and CENP-A. These data implicate

Sp1, independent of its sequence-specific DNA-binding domain, as a factor that regulates AS-derived lncRNAs and maintains the structure, function, and identity of centromeres during mitosis, thereby preserving chromosomal stability.

Sp1 Localizes to Centromeres in Mitotic Cells

During mitosis, as the chromatin condenses into chromosomes, most transcription factors become dissociated from chromatin and are dispersed throughout the cell. However, others stay bound to chromatin. Therefore, it is difficult to predict the localization pattern of a particular transcription factor without careful evaluation. To gain a more detailed understanding of the localization and function of Sp1 during mitosis, we used the non-tumorigenic mammalian breast epithelial cell line MCF 10A to prepare chromosome spreads. Based on the overlap of Sp1 signal with that of CENP-A, immunofluorescence analysis (IF) showed that Sp1 was detected at centromeres during mitosis (Figures 4.1A and 4.1D). To verify that this was not an observation unique to

MCF 10A cells, centromere localization of Sp1 was confirmed in chromosome spreads obtained from the osteosarcoma cell line Saos-2 as well as mouse embryonic fibroblasts

(MEF) (Figures 4.1B, and 4.1C and E, respectively). We next examined Sp1 localization at different stages of mitosis (Figure 4.2) by fixing and staining untreated normal human diploid fibroblasts (NHDF) at various stages of mitosis. At interphase, there was no apparent localization of Sp1 to the centromeric region. Pro-metaphase was the earliest

92 point at which Sp1 was detected at centromeres, and by anaphase, much of the Sp1 localized to centromeres was dissociated. Finally, by telophase, no Sp1 was detected at centromeres. Thus, Sp1 exhibited a dynamic localization pattern at centromeres in mitotic cells.

Sp1 Binds to Centromeres and Pericentromeres in MCF 10A Cells

The kinetochore is a multilayered structure composed of over 100 different proteins that assembles onto core centromeres. Several of these proteins bind directly to the centrochromatin, including CENP-A, CENP-B, CENP-C, CENP-N, and the CENP-T-

W-S-X complex [634, 635]. To determine if Sp1 binds directly to core centromeres, chromatin immunoprecipitation (ChIP) experiments were performed using MCF 10A cells. Sp1 binding to core centromeres was confirmed using SimpleChIP® Human α-

Satellite Repeat Primers. ChIP with a CENP-A antibody was performed as a positive control (Figure 4.3A). Careful examination of IF images revealed that Sp1 appeared to be localizing to both CENP-A containing regions as well as surrounding regions that were devoid of CENP-A (Figure 4.1D). To determine if Sp1 was also bound to surrounding pericentromeric heterochromatin, we designed primers that specifically amplified similar pericentromeric heterochromatin regions on chromosomes 9, 20, 21, and 22 (Table 4.1). Here, ChIP with a CENP-A antibody was performed as a negative control (Figure 4.3B). All results are presented as fold binding relative to the negative control, IgG. Results showed that Sp1 binds to AS arrays and pericentromeric heterochromatin in MCF 10A cells.

Sp1 Binding to Centromeres is Not Dependent on the Sp1 Zinc-Finger, DNA-

Binding Domain

The Sp1 DNA-binding domain consists of three Cys2His2-type zinc fingers spanning amino acids 626-708, which bind directly to the very GC-rich consensus sequence 5’-(G/T)GGGCGG(G/A)(G/A)(G/T)-3’ [385]. The mammalian centromere is composed of several megabases of 171-bp AT-rich repetitive sequences, and does not appear to contain the Sp1 consensus sequence (Figure 2.3). Although a complete DNA sequence for centromeres is lacking, it seems unlikely that Sp1 localization and binding to centromeres is mediated by the zinc-finger domain. To test this, we prepared chromosome spreads from MEFs expressing full length Sp1 and from a MEF cell line that expressed a truncated form of Sp1, consisting of the 65 kilodalton intact N-terminal domain, and lacking the DNA-binding domain region in the C-terminus (hereafter referred to as Sp1-65) (Figure 4.4B and described in [636]). IF analysis of these spreads revealed that full length Sp1 (Figure 4.1C) and Sp1-65 localized to centromeres (Figure

4.4A), suggesting that Sp1 binding to centromeres does not require the zinc-finger,

DNA-binding domain. To more specifically determine the region of Sp1 that mediates localization to centromeres, we transfected MCF 10A cells with a plasmid expressing amino acids 1-182 of Sp1 (hereafter referred to as Sp11-182). IF analysis of chromosome spreads prepared from this cell line revealed that endogenous Sp1 and exogenous Sp11-

182 co-localized in a pattern consistent with centromeric localization (Figure 4.4C), confirming that the Sp1 DNA-binding domain was not required for localization of Sp1 to centromeres.

ATM Activity Is Required for Rapid Sp1 Localization to Centromeres

Several studies have reported that Sp1 is phosphorylated during mitosis. The c-

Jun NH2-terminal kinase 1 (JNK1) phosphorylates Sp1 at threonine 278 and threonine

739 [426, 637]. Additionally, Sp1 is a target of CDK1/Cyclin B1, which also phosphorylates Sp1 at threonine 739 [428]. These phosphorylation events decrease the

DNA binding affinity of Sp1 during mitosis as well as stabilize Sp1 in the cytoplasm by protecting it from RNF4-mediated degradation [421, 428, 638].

Previously, we have shown that Sp1 is phosphorylated by the serine/threonine protein kinase ATM on serine 101 in response to DNA double-strand-breaks [384, 462].

Sp1 is also phosphorylated by ATM on serine 101 in response to viral infection [418,

459]. ATM was recently shown to be phosphorylated and activated by Aurora B kinase during mitosis in a DNA damage-independent manner, resulting in the phosphorylation and activation of other centromere associated proteins, including Bub1 and Mad1 [639].

Therefore, we sought to determine if ATM activity is required for Sp1 localization to centromeres during mitosis. Here, we treated MCF 10A cells with the ATM-specific inhibitor KU-55933 [640] for 2 or 16 hours before preparing chromosome spreads. IF analysis revealed that 2 hours of KU-55933 treatment prevented Sp1 from localizing to centromeres (Figure 4.5A), suggesting that ATM activity was required for rapid localization. Sp1 re-appeared at centromeres after 16 hours (Figure 4.5A) when KU-

55933 was still active based on lack of phosphorylation of Chk2 as threonine 68 (Figure

4.5B). To determine if serine 101 was the residue of Sp1 that was phosphorylated during mitosis, we used immunoblotting to probe interphase and mitotic lysates with a phospho- serine 101 specific antibody (pSp1101) using untreated or Adriamycin treated cells as negative and positive controls, respectively (Figure 4.5C). Our results showed that within the context of mitosis, serine 101 remains unphosphorylated; however, it should be noted that there are several additional SQ/TQ sites clustered in Sp11-182, which are potential targets of ATM phosphorylation during mitosis.

Sp1 Knockdown Results in a Decrease in CENP-A at Centromeres

A commonly used technique for deducing the function of a particular protein is to eliminate that protein from a cell and examine the consequences. To better understand

95 the role of Sp1 during mitosis, we transduced MCF 10A cells with a control shRNA or an shRNA targeting the Sp1 transcript and then prepared chromosome spreads.

Immunoblotting analysis was used to verify Sp1 protein knockdown (Figure 4.6B, left panel). IF analysis revealed two predominant phenotypes. First, metaphase stage cells lacking Sp1 showed a slight but significant increase in distance between the CENP-A signals on paired sister chromatids (measured using ImageJ), suggesting a centromeric or cohesion defect (Figure 4.6A, 4.6B right panel, and 4.6C). Second, the CENP-A signal intensity was significantly decreased by 15.9% in Sp1 knockdown cells, at 5181 ±

86.05 rfu (relative fluorescent units) as compared to 6163 ± 99.58 rfu in control cells (rfu calculated as described in [641]), suggesting that Sp1 knockdown resulted in a decrease in CENP-A binding to centromeres (Figure 4.6D). To confirm our imaging data, ChIP experiments were performed using MCF 10A cells transduced with control shRNA or shRNA targeting the Sp1 transcript. Antibodies against CENP-A and H3 were used for immunoprecipitation and binding to core centromeres was determined using

SimpleChIP® Human α-Satellite Repeat Primers. Comparison of the ratio of AS signal from CENP-A to H3 ChIP showed that in Sp1-depleted cells there was a 14.5% decrease in CENP-A binding to core centromeres as compared to control cells (Figure

4.6E), corroborating the 15.9% decrease that we observed by IF.

Decreased CENP-A at Centromeres is Not Due to Reduced Expression of CENP-A

CENP-A, like other histone proteins, is an exceptionally stable protein [642, 643].

Because of this, it was unlikely that the decrease in CENP-A protein levels at centromeres in Sp1-depleted cells was a result of Sp1-mediated transcriptional down- regulation or normal CENP-A protein turnover. However, to explore this possibility, we first evaluated the distal and proximal CENP-A promoter regions for Sp1 binding sites.

We found two Sp consensus sequences 2583 base pairs and 173 base pairs upstream

96 from the transcription start site (Figure 4.7A). ChIP experiments confirmed that Sp1 binds to both regions, with more binding at the -173 site (Figure 4.7B, Table 4.1 for primers). Next, we transduced MCF 10A cells with control shRNA or shRNA targeting the Sp1 transcript exactly as performed for the experiments in Figure 4.6 and evaluated protein lysates for CENP-A protein levels. Protein densitometry measurements from three separate experiments revealed no difference in total CENP-A protein levels between Sp1-depleted and control cells (Figure 4.7C). Additionally, we evaluated CENP-

A mRNA transcript levels in both control and Sp1-depleted cells and found no significant difference in transcript levels (Figure 4.7D, Table 4.1 for primers). Thus, although Sp1 binds to two Sp consensus sequences in the CENP-A promoter region, Sp1 does not appear to significantly alter CENP-A protein or mRNA levels within the timeframe of the experiments performed. These results showed that the decrease in CENP-A at centromeres in Sp1-depleted cells was not due to reduced expression of CENP-A.

Sp1 Knockdown Results in a Decrease in CENP-C at Centromeres

CENP-C is an important upstream factor for CENP-A deposition. CENP-C is one of several components of the CCAN, the network of proteins that associates with centromeric chromatin throughout the cell cycle. CENP-C is an important upstream factor for CENP-A deposition because it binds to centromeric chromatin and recruits the

Mis18 complex to this region. The Mis18 complex then recruits HJURP, the chaperone that facilitates the exchange of histone H3 for CENP-A. Knockdown of CENP-C results in decreased CENP-A levels at centromeres [305], which has been demonstrated in

Xenopus egg extracts [306], Drosophila [307], and mouse cells [305]. In an attempt to gain a better understanding of why Sp1 knockdown resulted in a decrease in CENP-A at centromeres, we used ChIP to evaluate CENP-C binding to centromeres in MCF 10A cells transduced with a control shRNA or an shRNA targeting the Sp1 transcript (Table

4.1 for primers). Here, IgG was used as a negative control and binding of histone H2B to centromeres was used as a positive control. Two separate CENP-C antibodies were used. Results showed that Sp1 depletion resulted in a significant reduction in CENP-C binding to centromeres, likely contributing to the decrease in CENP-A at centromeres through a reduction in Mis18 complex and HJURP recruitment (Figure 4.8A). To confirm that the decrease in CENP-C at centromeres was not due to a decrease in total CENP-C protein levels in Sp1-depleted cells, we used immunoblotting to evaluate total cell lysates for CENP-C. Our results showed that depletion of Sp1 had no effect on total

CENP-C levels (Figure 4.8B).

Sp1 Contributes to Regulation of α-Satellite-Derived Long Non-Coding RNAs

For many years, the core centromere was thought to be a purely heterochromatic region and thus transcriptionally inactive. However, several recent studies have shown that the core centromere is actively transcribed in a number of different species and that centromeric transcripts are important for maintaining the structure and function of centromeres and kinetochores [350, 352, 355, 358, 644]. Several studies have identified centromere-derived transcripts, including those in yeast [350, 368], rice [352, 645], mice

[353], and humans [326, 355-358]. Consistent with these studies are those that have identified transcriptionally permissive epigenetic modifications at centromeres, including

H3K4me1/2, H2Bub1, and H3K36me2 [337, 340, 646]. Although a complete understanding of the purpose of RNA Polymerase II-dependent transcription through this region, as well as the importance of centromere-derived transcripts, is still lacking, studies continue to demonstrate that this phenomenon plays a major role in maintaining active centromeres.

In human cells, long non-coding RNAs (lncRNAs) derived from AS DNA are transcribed by RNA Polymerase II during mitosis, in conjunction with other general

98 transcription factors, including RNA polymerase II subunit A C-terminal domain phosphatase (CTDP1), structure-specific recognition protein 1 (SSRP1), and TATA-box binding protein (TBP) [355, 361]. These AS-derived lncRNAs have been shown to be present at mitotic centromeres to maintain the centrochromatin architecture as well as to facilitate the localization of centromere and kinetochore proteins, including CENP-A,

CENP-C, Aurora B kinase, INCENP, and Survivin [326, 353, 361, 647]. In budding yeast and HeLa cells, both increases and decreases in centromeric transcription lead to chromosome segregation errors and W-CIN, a characteristic of many solid tumors and hematological malignancies [8, 350, 355]. However, how transcription of AS-derived lncRNAs is regulated is poorly understood.

Because Sp1 is a transcription factor, we wanted to determine if Sp1 plays a role in the transcription of AS-derived lncRNAs at centromeres during mitosis. We transduced MCF 10A cells with a control shRNA or shRNA targeting the Sp1 transcript

(Figure 4.9C) and evaluated each cell line for AS-derived lncRNAs by quantitative RT-

PCR (Table 4.1 for primers). Additionally, we transduced MCF 10A cells with retrovirus containing an Sp1 expression construct to over-express Sp1 in these cells (Figure 4.9D).

Our results show that in response to Sp1 protein knockdown, AS-derived lncRNAs were increased approximately 3.5 fold (Figure 4.9A). For the Sp1 over-expressing MCF 10A cell line, AS-derived lncRNAs were decreased by 50% (Figure 4.9B). Here, because the

AS sequences are highly repetitive, we chose to use two additional highly repetitive sequences as controls. First, we used 18S ribosomal RNA, an RNA polymerase I transcript that should not be impacted by changes in Sp1 protein levels [648, 649], and transcripts from Alu elements, a family of small interspersed elements (SINEs) widely distributed throughout the genome (Table 4.1 for primers). The transcription of Alu elements depends on both RNA Polymerase II and RNA Polymerase III, and because of their wide genomic distribution, changes in gene expression of individual proteins does

99 not influence expression of total Alu transcript levels [650-654]. Thus, Alu transcript levels should also remain unchanged in response to changes in total Sp1 protein levels.

Indeed, our results show that in response to both Sp1 protein knockdown and Sp1 protein over-expression, both 18S and Alu transcript levels remained unchanged.

In support of these data, we used SimpleChIP® Human α-Satellite Repeat

Primers to perform a ChIP experiment in MCF 10A cells that were transduced with control shRNA or shRNA targeting the Sp1 transcript and evaluated these cell lines for binding of RNA Polymerase II, phosphorylated on serine5 (RNA Polymerase IIpS5), at centromeres. RNA Polymeriase II becomes phorphorylated on serine 5 as the RNA

Polymerase II pre-initiation complex transitions from pre-initiation to elongation.

Therefore, RNA Polymerase IIpS5 is a measure of active transcription [655]. Our results showed an approximately 2-fold increase in RNA Polymerase IIpS5 binding at centromeres in Sp1-depleted cells, supporting an increase in transcription of AS-derived lncRNAs under these conditions (Figure 4.10A). To confirm that the increase in RNA

Polymerase IIpS5 at centromeres was not due to an increase in total RNA Polymerase

IIpS5 protein levels in Sp1-depleted cells, we used immunoblotting to evaluate total cell lysates for RNA Polymerase IIpS5. Our results showed that depletion of Sp1 had no effect on total RNA Polymerase IIpS5 levels (Figure 4.10B). Thus, changes in total Sp1 protein levels significantly alter AS-derived lncRNA transcription at centromeres.

Over-Expression of Sp1 Results in a Decrease in CENP-A at Centromeres

We observed an increase in transcription through the core centromere in Sp1- depleted cells as well as a decrease in CENP-A binding at centromeres. Further, Sp1 over-expression resulted in a decrease in centromeric transcription. Studies have indicated that a perturbation in centromeric transcription in either direction prevents proper CENP-A deposition, resulting in a decrease in CENP-A binding. To determine if

100 over-expression of Sp1, and thus decreased centromeric transcription, prevents proper

CENP-A deposition at centromeres, we used ChIP to evaluate CENP-A binding at centromeres in MCF 10A cells transduced with a control retrovirus and a retrovirus containing an Sp1 expression construct to over-express Sp1. Antibodies against CENP-

A and H3 were used for immunoprecipitation, and binding to core centromeres was determined using SimpleChIP® Human α-Satellite Repeat Primers. Immunoblotting was used to confirm over-expression of Sp1 and to ensure that over-expression of Sp1 had no effect on total CENP-A protein levels (Figure 4.11B). Comparison of the ratio of AS signal from CENP-A to H3 ChIP showed that in Sp1 over-expressing cells there was a slight but significant decrease in CENP-A at centromeres (Figure 4.11A).

Sp1 Depletion Disrupts the Centrochromatin Landscape

Within the context of transcriptional regulation, Sp1 directly or indirectly recruits several different factors to chromatin to modulate the chromatin landscape. Some of these factors include p300, HDAC1 and HDAC2, DNMT1, and the SWI/SNF complex

[388, 408, 411, 498, 656, 657]. Therefore, we sought to determine if the increase in AS- derived transcripts in Sp1-depleted cells was associated with changes in centrochromatin epigenetics, specifically in histone modifications. To address this, MCF

10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript, followed by ChIP to evaluate changes in monoubiquitinated H2B (H2Bub1), a modification shown to be required for proper transcription of AS-derived lncRNAs [340].

Our results show that upon Sp1 depletion, there was a significant increase in the

H2Bub1 to H2B ratio at centromeres as compared to our control cells (Figure 4.12A).

We also used ChIP to evaluate H3K4me2 levels at centromeres, a modification associated with open but not active chromatin [335], as well as H3K36me2 levels at centromeres, associated with active transcription [337]. Interestingly, our results

101 revealed a significant decrease in the H3K4me2 to H3 ratios (Figure 4.12B) and the

H3K36me2 to H3 ratios (Figure 4.12C), in Sp1-depleted cells as compared to control.

Further, because HDAC1 has been shown to interact with Sp1 to facilitate deacetylation of H3K9 [658], we used ChIP to evaluated H3K9Ac levels at centromeres and found that

Sp1 depletion resulted in a decrease in H3K9Ac levels at centromeres (Figure 4.12D).

For these ChIP studies, binding was determined using SimpleChIP® Human α-Satellite

Repeat Primers. Finally, studies have shown that changes in the pericentromeric heterochromatin can similarly affect transcription of core centromeres [659]. Because our studies showed that Sp1 binds to both core centromeres and pericentromeres

(Figure 4.3B), we evaluated the prominent heterochromatic marks H3K9me2/3 at pericentromeres by ChIP. Our results showed that Sp1 depletion had no effect on this modification at pericentromeres (Figure 4.12E, Table 4.1 for primers). We used immunoblotting to confirm that Sp1 depletion alone had no effect on any of these histone modifications on a global level (Figure 4.13). This suggests that the changes in epigenetic modifications upon Sp1 depletion may be specific to centromeres.

Conclusions

With these experiments, we showed that transcription factor Sp1 localizes and binds to centromeres and pericentromeres during mitosis and that this interaction is not dependent upon the Sp1 zinc-finger, DNA-binding domain. Rapid localization required

ATM activity, and knockdown of Sp1 protein resulted in a decrease in CENP-A and

CENP-C at centromeres and a potential centromeric cohesion defect. Finally, knockdown of Sp1 protein disrupted the centrochromatin landscape. In a previous report

[7], we demonstrated that Sp1 depletion resulted in chromosome mis-segregation and

W-CIN. Taken together with the data presented in this chapter, we believe that although transcription factor Sp1 likely maintains chromosomal stability through numerous

102 mechanisms, maintaining centromere identity by regulating CENP-A levels, centromeric transcription, and the centrochromatin landscape are likely contributing factors. The implications of these findings are discussed in Chapter 6.

103

B B

D E

Figure 4.1: Sp1 Localizes to Centromeres in Mitotic Cells. (A) Chromosome spreads were prepared using MCF 10A cells treated with colcemid for 4 hours. Spreads where stained with antibodies against CENP-A and Sp1 and with DAPI for visualization of chromatin. Sp1 localized to CENP-A containing regions in metaphase stage cells. Inset is (D). (B) Chromosome spreads were prepared using a human osteosarcoma cell line Saos-2 treated with colcemid for 4 hours. Spreads where stained with antibodies against CENP-A and Sp1 and with DAPI for visualization of chromatin. Sp1 localized to CENP-A containing regions in metaphase Saos-2 cells. (C) Chromosome spreads were prepared using Sp1+/+ mouse embryonic fibroblasts treated with colcemid for 4 hours. Spreads

104 were stained with antibodies against CREST and Sp1 and with DAPI for visualization of chromatin. Sp1 localized to centromeric regions in MEF cells. (D) Inset from (A) suggests that Sp1 localizes to both centromeric and pericentromeric regions. (E) Inset from (C), showing Sp1 (green) localized to CENP-A (red) containing regions of a metaphase chromosome.

105

Figure 4.2: Temporal Localization of Sp1 at Centromeres. Untreated, normal human diploid fibroblasts (NHDF) were fixed and stained using antibodies against Sp1 and CENP-A and with DAPI for visualization of chromatin. Cells at different stages of mitosis were examined for Sp1 localization at centromeres. Results showed that at interphase, there was no apparent localization of Sp1 to centromeres. Pro-metaphase was the earliest point at which Sp1 was detected at centromeres, and by anaphase, much of the Sp1 localized to the centromeres was dissociated. Finally, by telophase, no Sp1 was detected at centromeres.

106

A B

Figure 4.3: Transcription Factor Sp1 Localizes and Binds to Centromeres and Pericentromeres in Mitotic Cells. (A) Untreated MCF 10A cells were collected, and SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect Sp1 binding to AS arrays (A), or for Sp1 binding to pericentromeric heterochromatin regions (B), as described in methods. Normal rabbit IgG and CENP-A were used as controls. Immunoprecipitated chromatin was assessed by SYBR green quantitative PCR. All CT values were normalized to mock IgG control to show fold increase in Sp1 binding over control. Results confirmed Sp1 binding to core centromere and pericentromeric regions in untreated MCF 10A cells. * p<0.04.

107

Figure 4.4: Sp1 Binding to Centromeres is Not Dependent on the Sp1 Zinc-Finger, DNA-Binding Domain. (A) Chromosome spreads were prepared using Sp1-/- MEFs. This cell line is referred to as Sp1-/- because the Sp1 gene is truncated, resulting in the expression of a truncated form of Sp1, consisting of the 65 kDa intact N-terminal domain and lacking the DNA-binding domain region in the C-terminus (Sp1-65). The truncated Sp1 protein expressed in Sp1-/- MEFs localized to centromeres. (B) Western blot analysis of Sp1+/+ and Sp1-/- total cell lysates probed for Sp1. Actin was used as a loading control. Sp1+/+ cell lysates show wild-type Sp1 protein at approximately 100 kilodalton as expected, whereas Sp1-/- cell lysates show truncated 65 kilodalton N- terminal domain portion of Sp1 protein, as described in [636]. (C) Chromosome spreads were prepared using MCF 10A cells expressing LXSN-Sp11-182-HA then stained using antibodies against Sp1 (green), HA (red), and DAPI (blue). Sp1 and HA localized in a pattern consistent with centromeric localization, suggesting LXSN- Sp11-182-HA localized to centromeres.

108

B C

Figure 4.5: ATM Activity Is Required for Rapid Sp1 Localization to Centromeres. (A) Chromosome spreads were prepared using MCF 10A cells treated with vehicle (DMSO) or 10 μM KU-55933 for 2 or 16 hours. Inhibition of ATM with KU-55933 prevented Sp1 localization after 2 hours of treatment but not after 16 hours of treatment. (B) Western blot analysis of MCF 10A cell lysates after 1, 5, and 15 hours of KU-55933 treatment plus 1 additional hour of KU-55933 treatment with 100 μM H2O2, for a total of 2, 6, and 16 hours of KU-55933 treatment. Western blot was probed for phospho-Chk2 as an indicator of ATM activity. ATM activity was inhibited by KU-55933 at all time points. (C) Western blot analysis of interphase and mitotic MCF 10A cell lysates from 2 separate mechanical shake-off experiments were probed for pSp1101, total Sp1, and nucleolin as a loading control. Untreated MCF 10A cell lysates were used as a negative control (-), and Adriamycin treated MCF 10A cell lysates were used as a positive control (+). Sp1 was not phosphorylated on serine 101 during mitosis.

109

C D E

Figure 4.6: Sp1 Knockdown Results in a Decrease in CENP-A at Centromeres. (A) Chromosome spreads were prepared using MCF 10A cells transduced with control shRNA or shRNA targeting the Sp1 transcript and then stained using antibodies against Sp1 (green), CENP-A (red), and with DAPI (blue) for chromatin visualization. *background fluorescence. (B) Left: Western blot analysis of MCF 10A cell lysates used in (A). Nucleolin was used as a loading control. Right: Insets from (A). (C) The distance between CENP-A signals from (A) were measured using ImageJ and plotted. For control shRNA, n = 399. For Sp1 shRNA, n = 389. p = 0.0008. (D) The CENP-A signal intensities from (A) were quantified using ImageJ and plotted. For control shRNA, n = 259. For Sp1 shRNA, n = 279. p < 0.0001. (E) MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to evaluate binding of histone H3 and CENP-A to AS arrays, as described in methods. Normal rabbit IgG was used as a negative control. Immunoprecipitated chromatin was assessed by SYBR green quantitative PCR. CENP- A to H3 ratios were determined using raw CT values, and those ratios were normalized

110 to mock IgG control. Results showed a decrease in the CENP-A to H3 ratio in Sp1- depleted cells. n = 2, p = 0.01.

111

Figure 4.7: Sp1 Binds to CENP-A Promoter, but Does Not Influence CENP-A mRNA or Protein Levels Upon Knockdown. (A) Schematic of the CENP-A distal and proximal promoter regions showing Sp1 consensus sequences at 173 and 2583 nucleotides upstream from the transcriptional start site. (B) ChIP was performed using MCF 10A cells transduced with control shRNA or shRNA targeting the Sp1 transcript. Immunoprecipitated chromatin was evaluated for binding at the -173 locus and the -2583 locus using primers described in Table 4.1. Immunoprecipitated chromatin was assessed by SYBR green quantitative PCR. All CT values were normalized to mock IgG control to show fold increase in Sp1 binding over control. Results confirm Sp1 binding to both the -173 and the -2583 loci. n=3, *p<0.04, **p<0.01. (C) MCF 10A cells were transduced with control shRNA or shRNA targeting Sp1 transcript. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein knockdown and CENP- A levels to evaluate the effect of Sp1 knockdown on CENP-A protein. Left panel: Representative western blot. Right panel: CENP-A densitometry for all experimental replicates was plotted. Results show that Sp1 depletion did not significantly alter total CENP-A protein levels. n=3. (D) RNA was extracted from MCF 10A cells transduced with control shRNA or shRNA targeting Sp1 transcript. RNA extraction was performed using the Qiagen RNeasy® Mini Kit per manufacturer’s instructions. RNA was reverse transcribed using the qScriptTM cDNA Supermix, and CENP-A transcript levels were evaluated using primers described in methods. Results show that CENP-A transcript levels were not significantly altered in Sp1-depleted cells. n=3.

112

A B

Figure 4.8: Sp1 Knockdown Results in a Decrease in CENP-C at Centromeres. (A) MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect CENP-C binding to AS arrays, as described in methods. Normal rabbit IgG was used as a negative control, and histone H2B was used as a positive control. Immunoprecipitated chromatin was assessed by SYBR green quantitative PCR. Results showed a decrease in CENP-C binding to centromeres in Sp1-depleted cells. 1CENP-C antibody was purchased from abcam (ab193666). 2CENP-C antibody was a gracious gift from Dr. Andrea Musacchio [321]. n = 3. (B) MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein depletion and for changes in CENP-C protein levels. Nucleolin was used as a loading control. Results showed Sp1 depletion had no effect on total CENP-C protein levels.

113

A B

C D

Figure 4.9: Sp1 Contributes to Regulation of α-Satellite-Derived Long Non-Coding RNAs. RNA was extracted from MCF 10A cells transduced with control shRNA or shRNA targeting the Sp1 transcript (A), or from MCF 10A cells transduced with empty pLXSN vector or pLXSN-Sp1-HA over-expression construct (B). RNA extraction was performed using the Qiagen RNeasy® Mini Kit. RNA was reverse transcribed using the qScriptTM cDNA Supermix, and transcripts were analyzed by quantitative PCR. 18S ribosomal RNA and Alu transcripts were used as controls. All results were normalized to nucleolin transcripts. Depletion of Sp1 resulted in an approximately 3.5 fold increase in AS-derived transcripts as compared to control, while Sp1 over-expression resulted in a 50% decrease. n = 9. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein knockdown (C) and pLXSN-Sp1-HA over-expression (D). Nucleolin was used as a loading control.

114

A B

Figure 4.10: Sp1 Depletion Results in an Increase in RNA Polymerase IIpS5 Binding at Centromeres. (A) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect RNA Polymerase IIpS5 binding at AS arrays. Results showed an increase in RNA Polymerase IIpS5 binding in Sp1-depleted cells. (B) MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein depletion, and for changes in RNA Polymerase IIpS5. Nucleolin was used as a loading control. Results showed Sp1 depletion had no effect on RNA Polymerase llpS5 protein levels.

115

A B

Figure 4.11: Over-Expression of Sp1 Results in a Decrease in CENP-A at Centromeres (A) MCF 10A cells transduced with a control retrovirus (pLXSN empty vector) or a retrovirus containing an Sp1 expression construct to over-express Sp1 (pLXSN-Sp1-HA). Antibodies against CENP-A and H3 were used for immunoprecipitation, and binding to core centromeres was determined using SimpleChIP® Human α-Satellite Repeat Primers. Results showed a slight but significant decrease in CENP-A binding at centromeres in Sp1 over-expressing cells. (B) MCF 10A cells were transduced with control retrovirus (pLXSN empty vector) or a retrovirus containing an Sp1 expression construct to over-express Sp1 (pLXSN-Sp1-HA). Protein lysates were collected and immunoblotting was used to evaluate lysates for total Sp1 protein levels, exogenous Sp1 protein levels (HA), and total CENP-A levels. A-tubulin was used as a loading control. Results confirmed over-expression and that over- expression of Sp1 had no effect on total CENP-A protein levels.

116

A B

C D

Figure 4.12: Sp1 Depletion Disrupts the Centrochromatin Landscape. (A) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and SimpleChIP® Human α- Satellite Repeat Primers and ChIP were used to detect H2Bub1 and H2B binding at AS arrays. Results showed an increase in the H2Bub1 to H2B ratio at centromeres in Sp1- depleted cells. (B) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect H3K4me2 and H3 binding at AS arrays. Results showed a decrease in H3K4me2 to H3 ratio in Sp1-depleted cells. (C) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect H3K36me2 and H3 binding at AS arrays. Results showed a decrease in H3K36me2 to H3 ratio in Sp1-depleted cells. (D) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and SimpleChIP® Human α-Satellite Repeat Primers and ChIP were used to detect H3K9Ac and H3 binding at AS arrays. Results showed a

117 decrease in H3K9Ac to H3 ratio in Sp1-depleted cells. (E) MCF 10A cells were transduced with control or Sp1 targeted shRNA, and ChIP was performed for H3K9me2/3 and H3 binding at the pericentromeric heterochromatin region. Results showed that Sp1 depletion had no effect on the H3K9me2/3 to H3 ratio.

118

A B C

D E

Figure 4.13: Control Western Blots. MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein depletion and for changes in H2Bub1 (A), H3K4me2 (B), H3K36me2 (C), H3K9Ac (D), or H3K9me2/3 (E). α-tubulin was used as a loading control. Results showed Sp1 depletion had no effect on total H2Bub1, H2K4me2, H3K36me2, H3K9Ac, or H3K9me2/3 protein levels.

119

Table 4.1: Primers Used in Described Experiments.

Oligonucleotides Sequence

SimpleChIP® Human α-Satellite Repeat Primers (Cell α-Satellites ChIP - DNA Signaling 4486)

Pericentromeric Heterochromatin Regions forward 5’-AGCTCGGGCAAAGAGTTCAA-3’, on Chromosomes 9, 20, 21, and 22 reverse 5’-TTTGGGGGAAATTCCTGCA-3’

forward 5’-AGTCGCTGGATTCGGGTTTT-3’, CENP-A -173 Locus reverse 5’-CTCACATGAGCCGGTGCTT-3’

forward 5’-AATAGTGCAAGACCCGGTTG-3’, CENP-A -2583 Locus reverse 5’-AGGTGGGGAGTTTTAGTTTGA

forward 5’-CTTAGGCGCTTCCTCCCATC-3’, CENP-A Transcripts reverse 5’-AGAGGTGTGTGCTCTTCTGA-3’

forward 5’-CATCACAAAGAAGTTTCTGAGAATGCTTC-3’ α-Satellite Transcripts - RNA reverse 5’-TGCATTCAACTCACAGAGTTGAACCTTCC-3’

forward 5’-GTAACCCGTTGAACCCCATT-3’, 18S Control Transcripts reverse 5’-CCATCCAATCGGTAGTAGCG-3’

forward 5'-CATGGTGAAACCCCGTCTCTA-3' Alu Control Transcripts reverse 5'-GCCTCAGCCTCCCGAGTAG-3'

forward 5'-CTCGCGAAGGCAGGTAAAAAT-3' C23 Control Transcripts reverse 5'-CAGCAGCCTTCTTGCCTTTC-3'

120

Chapter 5: Sp11-182 May Be Sufficient for Functional Centromeres

121

Introduction

In addition to being regulated by various post-translational modifications, Sp1 is also regulated through proteolytic cleavage. For example, when cellular glucose levels are low, Sp1 is specifically rapidly degraded by the proteasome, likely in an effort to conserve cellular resources by down-regulating Sp1-mediated transcription [660]. Using an in vitro system to further analyze this degradation, it was determined that Sp1 contains a proteolytic cleavage site between leucine 56 and 57 and that this cleavage is required for proteasome-dependent degradation [661]. Sp1 has also been shown to be cleaved by caspases during apoptosis. This was first observed during retinoid-induced apoptosis, in which Sp1 was specifically cleaved, likely by caspase 2 [662]. Further, in the human Burkitt lymphoma cell line BL60, Sp1 is cleaved by caspase 3 during anti-

IgM-induced apoptosis. This study identified a caspase 3 cleavage site (AQPQAGR) directly after aspartic acid 584 of Sp1b (or 590 of Sp1a), suggesting that cleavage of

Sp1 plays a role in anti-IgM-induced apoptosis [471].

We previously demonstrated that Sp1 is cleaved by caspase 3 at aspartic acid

183 of Sp1a during DNA damage- or TRAIL-induced apoptosis. Cleavage at aspartic acid 183 produces a 70 kilodalton fragment comprising the DNA-binding domain and trans-activation domains (herein referred to as Sp1183-785), as well as a 30 kilodalton fragment containing the N-terminal 182 amino acids (herein referred to as Sp11-182).

Although it is not clear why Sp1 is cleaved, our data suggest that this cleavage is actively involved with apoptosis; mutating aspartic acid 183 to alanine protects cells against ultraviolet-induced apoptosis (Torabi, Azizkhan-Clifford, unpublished). It is attractive to speculate that genes are differentially regulated by cleaved Sp1183-785 as compared to full-length Sp1. Although there is currently no evidence to suggest that this is the case, we continue to investigate this possibility.

122

Interestingly, new evidence indicates that Sp1 may be cleaved at aspartic acid

183 to stabilize the Sp1183-785 cleavage product. Studies have shown that SUMOylation of

Sp1 on lysine 16 facilitates Sp1 degradation through the recruitment of RNF4, a SUMO- targeted ubiquitin ligase (STUbL). RNF4 interacts with the C-terminus of Sp1, at amino acids 619 – 785 [427]. Therefore, both the N-terminus (lysine 16) and the C-terminus of

Sp1 (amino acids 619 – 785) are required for RNF4-dependent degradation. If Sp1 is involved with transcriptional induction of apoptosis, cleavage of Sp1 at aspartic acid 183 may function to prevent RNF4-mediated degradation by physically separating the N- terminus of the protein from the C-terminus of the protein, thereby allowing the C- terminus of the protein to function as a transcriptional regulator. To determine if this is the case, we created three cell lines expressing exogenous full-length Sp1 (pLXSN-Sp1-

HA), Sp1 in which aspartic acid 183 is mutated to alanine (pLXSN-Sp1-D183A-HA), and cleavage product Sp1183-785 (pLXSN-Sp1183-785-HA). These cell lines were treated with the

DNA damaging agent Adriamycin to induce apoptosis, and the stability of each of the exogenously expressed proteins was monitored over time. Results show that at 12 and

24 hours post treatment, Sp1-D183A, which cannot be cleaved, was more quickly degraded as compared to full-length Sp1 and Sp1183-785. Full-length Sp1 and Sp1183-785 showed comparable stability at both time points (Figure 5.1). These data suggest that cleavage of Sp1 at aspartic acid 183 stabilizes the Sp1183-785 cleavage product, potentially allowing for Sp1183-785-dependent transcriptional regulation during apoptosis.

In addition to a likely pro-apoptotic role for the Sp1183-785 cleavage product, we have demonstrated that the N-terminal cleavage product, Sp11-182, functions independently from the C-terminal cleavage product in a number of cellular pathways, including DNA double-strand-break repair and centrosome regulation (both discussed below). Further, Sp11-183 localizes to centromeres and rescues centromere defects observed in Sp1-depleted cells. As such, this chapter reviews the established functions

123 for Sp11-182, as well as presents evidence indicating that Sp11-182 may be sufficient for maintaining centromere identity and function.

Sp11-182 Is Sufficient for DNA Double-Strand-Break Repair

Recently, we showed that Sp1 is required for efficient site-specific DNA double- strand-break repair. In response to ionizing radiation-induced DNA damage, Sp1 is phosphorylated by the DNA damage sensory kinase ATM at serine 101 and co-localizes with γH2AX (i.e. histone H2AX phosphorylated at serine 139 at sites of DNA damage)

[384]. H2AX is a variant of histone H2A that is phosphorylated early by damage sensory kinases in response to DNA damage and is thus commonly used as a DNA damage marker. We demonstrated that depletion of Sp1 by RNAi inhibits repair of DNA double- strand-breaks, similar to the effect of depletion of the MRN complex component Nbs1, demonstrating an important role for Sp1 in break repair. Interestingly, Sp11-182 alone localized to sites of DNA damage and appeared to be sufficient for rescuing the repair defect observed in Sp1-depleted cells [384].

Current studies indicate that Sp1 modulates the chromatin surrounding break sites to allow for efficient repair. Depletion of Sp1 by RNAi resulted in loss of H4K18 acetylation and an increase in H4K16 acetylation at break sites; these modifications are essential for double-strand-break repair by non-homologous end joining (NHEJ). This

Sp1-mediated chromatin modulation was accomplished, at least in part, through Sp1- dependent recruitment of the histone acetyltransferase p300 and HDACs to break sites.

Remarkably, expression of Sp11-183 rescued the changes in histone modifications observed in Sp1-depleted cells and was sufficient for recruitment of p300 (Guo,

Beishline, Azizkhan-Clifford, unpublished).

Sp11-182 Is Sufficient for Centrosome Regulation

124

Centrosome amplification can lead to improper attachment of microtubules to kinetochores, and if not properly corrected, chromosome segregation errors. We previously showed that Sp1 localized to centrosomes and regulated centrosome number, as depletion of Sp1 by RNAi resulted in centrosome amplification. Sp1183-785 did not localize to centrosomes or rescue the supernumerary centrosome phenotype in Sp1- depleted cells [7]. Recently, new data showed that Sp11-182 can localize to centrosomes and rescue the centrosome defects observed in Sp1-depleted cells (Flashner, Sowash,

Azizkhan-Clifford, unpublished). The function of Sp1 at centrosomes is still not clear and requires further investigation.

Sp11-182 Rescues Centromere Distance Phenotype

We have presented data showing that Sp11-182 localized to centromeres (Figure

4.4C). To determine if Sp11-182 is functional at centromeres, we created four cell lines, including MCF 10A cells transduced with: control shRNA with pLXSN empty expression vector (cell line A), shRNA targeting the Sp1 transcript with pLXSN empty expression vector (cell line B), control shRNA with pLXSN-SP11-182-HA over-expression vector (cell line C), and shRNA targeting the Sp1 transcript with pLXSN-SP11-182-HA over-expression vector (cell line D). Immunoblotting was used to verify Sp1 protein knockdown and/or

Sp11-182 over-expression in each cell line (Figure 5.2A). We next prepared chromosome spreads and evaluated metaphase stage cells for the centromere distance phenotype previously observed in Sp1-depleted cells (Figures 4.6A, 4.6B, and 4.6C). As expected, we observed a small but significant increase in the distance between CENP-A signals in cell line B as compared to cell line A. Over-expression of Sp11-182 alone did not affect the distance between CENP-A signals (cell line C compared to cell line A). Remarkably,

Sp11-182 rescued the CENP-A distance phenotype observed in Sp1-depleted cells (cell

125 line D compared to cell line A) (Figure 5.2B and 5.2C). Thus, Sp11-182 localized to centromeres and was sufficient for the function of Sp1 at this region.

Sp11-182 Rescues CENP-A Intensity Phenotype by Immunofluorescence Imaging

In addition to causing an increased distance between CENP-A signals, Sp1 depletion also resulted in a decrease in CENP-A at centromeres (Figures 4.6D and

4.6E). To determine if Sp11-182 is capable of rescuing this phenotype, the intensity of the centromeric CENP-A signals from the cell lines in Figure 5.2A were quantified using

ImageJ. As expected, Sp1 depletion resulted in a decrease in CENP-A signal (cell line B compared to cell line A). Over-expression of Sp11-182 alone did not affect the intensity of the CENP-A signals (cell line C compared to cell line A). Remarkably, Sp11-182 rescued the CENP-A signal intensity phenotype observed in Sp1-depleted cells (cell line D compared to cell line A) (Figure 5.3). Thus, Sp11-182 localized to centromeres and was sufficient for function of Sp1 at this region.

Sp11-182 Partially Rescues CENP-A Binding by Chromatin Immunoprecipitation

To confirm the imaging data presented in Figures 5.2 and 5.3, chromatin immunoprecipitation experiments were performed using cell lines A, B and D. Antibodies against CENP-A and H3 were used for immunoprecipitation, and binding to core centromeres was determined using SimpleChIP® Human α-Satellite Repeat Primers.

Comparison of the ratio of AS signal from CENP-A to H3 ChIP showed that there was a decrease in CENP-A binding to core centromeres in cell line B as compared to cell line

A. Further, the decrease in CENP-A binding at centromeres was rescued by expression of Sp11-182 in cell line D as compared to cell line B (Figure 5.4). These data are preliminary, and experiments need to be repeated to confirm rescue and establish significance.

126

Conclusion

Within the context of DNA double-strand-break repair, Sp11-182 was capable of recruiting p300 to break sites to facilitate the chromatin remodeling required for efficient repair. Further, although the function of Sp1 at centrosomes is still unclear, Sp11-182 localized to centrosomes and rescued the centrosomal defects observed in Sp1- depleted cells. With the data presented in Chapters 4 and 5, we have shown that Sp11-

182 localized to centromeres (Figure 4.4) and partially rescued centromeric defects observed in Sp1-depleted cells. Implications of these findings are discussed in Chapter

127

HA Protein Levels, 10 μM Adriamycin

Full-length Sp1

Sp1-D183A 183-785 Sp1

Tubulin Control Tubulin

- α to

Normalized Levels, Protein HA UT 12 hours 24 hours 48 hours

Figure 5.1: Cleavage of Sp1 at Aspartic Acid 183 Stabilizes Sp1183-785. MCF 10A cells were transduced with full-length Sp1 (pLXSN-Sp1-HA), Sp1 in which aspartic acid is mutated to alanine (pLXSN-Sp1-D183A-HA), and the C-terminal cleavage product (pLXSN-Sp1183-785-HA). Then, cells were treated with 10 μM Adraimycin to induce DNA damage, and total cell lysates were collected at 12, 24, and 48 hours post treatment. The signal intensity of the HA tag from each cell line was quantified using ImageJ and normalized to α-tubulin loading control. Results suggested that non-cleavage Sp1 was less stable than full-length Sp1 and Sp1183-785.

128

A B

Figure 5.2: Sp11-182 Rescues Centromere Distance Phenotype. (A) MCF 10A cells were transduced with: control shRNA with pLXSN empty vector (cell line A), shRNA targeting the Sp1 transcript with pLXSN empty vector (cell line B), control shRNA with pLXSN-SP11-182-HA over-expression vector (cell line C), and shRNA targeting the Sp1 transcription with pLXSN-SP11-182-HA over-expression vector (cell line D). Immunoblotting was used to verify Sp1 protein knockdown and/or Sp11-182 over- expression in each cell line. (B) The distance between CENP-A signals from (A) were measured using ImageJ and plotted. (C) Representative images from each cell line from (A).

129

Figure 5.3: Sp11-182 Rescues CENP-A Intensity Phenotype by Immunofluorescence Imaging. Intensity of the centromeric CENP-A signals from the cell lines in Figure 5.2A were quantified using ImageJ. As expected, Sp1 depletion resulted in a decrease in CENP-A signal (cell line A compared to cell line B). Over-expression of Sp11-182 alone did not affect the intensity of the CENP-A signals (cell line A compared to cell line C). Remarkably, Sp11-182 rescued the CENP-A signal intensity phenotype observed in Sp1- depleted cells (cell line A compared to cell line D).

130

Figure 5.4: Sp11-182 Partially Rescues CENP-A Binding by Chromatin Immunoprecipitation. Chromatin immunoprecipitation experiments were performed using cell lines A, B, and D. Antibodies against CENP-A and H3 were used for immunoprecipitation, and binding to core centromeres was determined using SimpleChIP® Human α-Satellite Repeat Primers. Comparison of the ratio of AS signal from CENP-A to H3 ChIP show that there was a decrease in CENP-A binding to core centromeres in cell line B as compared to cell line A. Further, the decrease in CENP-A binding at centromeres was rescued by expression of Sp11-182 in cell line D as compared to cell line B. Normal rabbit IgG was used as a negative control.

131

Chapter 6: Discussion and Future Directions

132

Introduction

The experiments presented in Chapter 4 demonstrate that transcription factor

Sp1 localizes and binds to centromeres and pericentromeres during mitosis and that this interaction is not dependent upon the Sp1 zinc-finger, DNA-binding domain. This localization requires ATM activity, and knockdown of Sp1 protein results in a decrease in

CENP-A and CENP-C at centromeres, as well as a potential centromeric cohesion defect. Finally, knockdown of Sp1 protein significantly changes histone modifications and the centrochromatin landscape. In a previous report [7], we demonstrated that Sp1 depletion results in chromosome mis-segregation and W-CIN. Taken together with this newly identified function for Sp1, we believe that although transcription factor Sp1 likely maintains chromosomal stability through numerous mechanisms, maintaining centromere identity by regulating CENP-A deposition, centromeric transcription, and the centrochromatin landscape are likely contributing factors. The implications of these findings are discussed below.

Within the context of DNA double-strand-break repair, the first 182 amino acids of Sp1 (Sp11-182) are capable of recruiting p300 to break sites to facilitate the chromatin remodeling required for efficient repair (Guo, Beishline, Azizkhan-Clifford, unpublished).

Further, although the function of Sp1 at centrosomes is still unclear, Sp11-182 localizes to centrosomes and rescues the centrosomal defects observed in Sp1-depleted cells

(Flashner, Sowash, Azizkhan-Clifford, unpublished). With the data presented in Chapter

5, we shown that Sp11-182 localizes to centromeres and rescues some of the centromeric defects observed in Sp1-depleted cells. The implications of these findings are also discussed below.

Discussion of Presented Results

133

Transcription Factor Sp1 Regulates the Centrochromatin Landscape and Centromeric

Transcription During Mitosis:

We show by immunofluorescence that Sp1 is absent from centromeres during interphase but transiently localizes and binds to centromeres during mitosis. More specifically, Sp1 localizes to all CENP-A containing regions by pro-metaphase, partially dissociating by anaphase, and completely dissociating by late anaphase or early telophase (Figure 4.1 and Figure 4.2); the timing is consistent with transcription through the core centromere [355]. Our cytological data is supported by ChIP experiments where we confirm that Sp1 binds to both AS DNA as well as surrounding pericentromeric heterochromatin (Figure 4.3A and 4.3B). Sp1 localizes to centromeres in all cell types examined, including several different human cell types and in mouse cells, illustrating that this phenomenon is conserved across different cell lines and may be conserved across species (Figure 4.1). This level of conservation indicates that Sp1 has an important function at centromeres during mitosis.

The localization of Sp1 to centromeres is not dependent on its zinc-finger, DNA- binding domain, which, within the context of gene transcription, binds to GC-rich Sp consensus sequences (Figure 4.4). This observation is consistent with the properties of the core centromere, composed of AT-rich repetitive AS arrays (Figure 2.3). Although it is not yet clear how Sp1 localizes to centromeres during mitosis, our data suggest that

Sp1 is likely recruited as a component of a complex, where complex formation requires only Sp11-182 as this fragment alone is capable of localizing to centromeres. Additionally,

ATM activity is required for rapid Sp1 localization, at least initially (Figure 4.5). ATM is best known for its role as a sensor for detecting DNA double-strand-breaks and subsequent activation of the DNA damage response [663]. In humans, loss of ATM causes Ataxia-Telangiectasia (A-T). Lymphoblastoid cells from A-T patients have defective spindle checkpoints as well as increased incidence of aneuploidy, suggesting a

134 role for ATM during mitosis [664-666]. It was recently shown that ATM is phosphorylated and activated by Aurora B kinase in a DNA damage-independent manner during mitosis, resulting in the phosphorylation and activation of other centromere associated proteins, including Bub1 and Mad1 [639]. We have not identified Sp1 as a direct target of ATM during mitosis, nor have we identified a phospho-residue required for Sp1 localization to centromeres. We confirmed that during mitosis Sp1 is not phosphorylated at serine 101

(Figure 4.5C), an ATM-dependent modification that we and others have shown to be important for the repair of DNA double-strand-breaks [459, 462]. As ATM is an important signaling kinase in this and other cellular contexts, it will be interesting to determine if

Sp1 is a direct target of ATM during mitosis or if ATM targets another signaling factor(s) required for Sp1 localization.

Cells depleted of Sp1 protein by RNAi show a decrease in CENP-A at centromeres and a potential cohesion defect (Figure 4.6). When cellular DNA is replicated, CENP-A containing nucleosomes are diluted across both strands of DNA, and cells proceed through mitosis with only half the maximal number of CENP-A containing nucleosomes [285]. Then, new CENP-A is incorporated into centrochromatin in early G1, after CDK1 activity declines [667], a process vitally important for maintaining centromere identity and function. This is exemplified in studies demonstrating that depletion of CENP-A by RNAi results in centromere inactivation and chromosome mis- segregation [668, 669]. Further, CENP-A over-expression is associated with several different types of cancer, including hepatocellular carcinoma [670], colorectal cancer

[287], and lung adenocarcinoma [671]. These data suggest that CENP-A protein levels must be carefully regulated to maintain centromere identity in normally dividing cells.

Upon Sp1 depletion, there is a 15.9% decrease in CENP-A signal intensity at centromeres in Sp1-depleted cells as quantified using ImageJ (Figure 4.6D). These cytological data are supported by ChIP experiments showing a similar decrease in

135

CENP-A binding to core centromeres in response to Sp1 depletion (Figure 4.6E).

Although we identified two Sp binding sites within the distal and proximal promoter of the

CENP-A gene, and confirmed that Sp1 binds to both of these sites (Figure 4.7A and

4.7B), our CENP-A protein and CENP-A gene transcript studies in Sp1-depleted cells fail to show a significant decrease in total CENP-A protein levels and total CENP-A gene transcript levels, respectively, as compared to controls (Figure 4.7C and 4.7D). This confirms that the decrease in CENP-A at centromeres is not due to decreased CENP-A gene transcription as a result of Sp1 depletion within the time frame of our experiments.

Additionally, CENP-A is an exceptionally stable protein, and thus the decrease at centromeres is unlikely due to normal protein turnover [642, 643]. Therefore, Sp1 maintains CENP-A protein levels at centromeres, emerging as an important factor for preserving centromere identity and function.

Recently, studies have shown that precise regulation of centromere-associated transcription is required for proper CENP-A deposition in G1 [337, 366]. Current models for CENP-A nucleosome assembly suggest that the centromere binding protein CENP-C recruits components of the Mis18 complex to centromeres, which in turn recruits the

CENP-A chaperone protein HJURP to facilitate the exchange of H3 for CENP-A [306].

Core centromere-derived transcripts interact directly with CENP-C, HJURP, and CENP-

A, and loss of these transcripts prevents centromeric recruitment of these proteins. This results in loss of centromere identity, at least in part due to a decrease in CENP-A deposition, and chromosome mis-segregation [306, 361]. Studies have also shown that an increase in centromeric transcription can result in decreased CENP-A deposition at the centromere. Using human artificial chromosomes (HACs), one group showed that forcing a more open HAC centrochromatin conformation through histone acetylation resulted in an increase in centromeric transcription, coupled with a decrease in CENP-A loading and loss of kinetochore function [366]. In the tammer wallaby, hypermorphic

136 expression of centromeric RNAs also led to decreased CENP-A loading at centromeres

[375], and in budding yeast, both increases and decreases in centromeric-derived transcripts results in chromosome mis-segregation [350]. These combined data suggest that precise regulation of centromere-derived transcription is required for proper CENP-A loading, centromere identity, and kinetochore function, and perturbations in transcript levels in either direction has detrimental effects on chromosome segregation. Here, we show that depletion of Sp1 by RNAi results in an increase in AS-derived lncRNAs and a concomitant increased binding of active RNA Polymerase IIpS5 at centromeres (Figure

4.9A and 4.9C, and Figure 4.10, respectively). Over-expression of Sp1 has the opposite effect on transcription, resulting in a decrease in AS derived lncRNAs and a similar decrease in CENP-A binding at centromeres (Figure 4.9B and 4.9D, and Figure 4.11, respectively). Thus, Sp1 acts as a negative regulator of AS-derived lncRNAs during mitosis. As such, the increase in AS-derived lncRNAs in Sp1-depleted cells likely contributes to decreased CENP-C binding to centromeres (Figure 4.8) resulting in decreased CENP-A loading at centromeres (Figure 4.6E), implicating Sp1 as a critically important factor for regulating centromeric transcripts and thus downstream loading of

CENP-A molecules in G1. This decrease in CENP-A at centromeres likely contributes to chromosome segregation errors in Sp1-depleted cells [7].

Centromeric histone modifications, or centrochromatin architecture, is also significantly altered in Sp1-depleted cells (Figure 4.12). During gene transcription, Sp1 collaborates with and recruits chromatin modifiers and remodelers to gene promoters.

Here, these enzymes modify histones to alter the chromatin structure, resulting in both positive and negative changes to gene transcription, depending on the context [388].

Thus, it is likely that Sp1 is acting to recruit enzymes to centromeres to down-regulate centromeric transcription. In other areas of the genome, Sp1 and HDAC1/2 interact to down-regulate genes, including SSeCKS (Src-suppressed C kinase substrate), hLHR

137

(human luteinizing hormone receptor), and hTERT (telomerase reverse transcriptase)

[406, 410, 481]. Further, Sp1 interacts with DNMT1 to down-regulate expression of the

MAZ (Myc-associated zinc-finger protein) gene [409]. HDAC1, HDAC2 (as a component of the mSin3 complex), and DNMT1 have all been shown to localize to metaphase centromeres. With the exception of mSin3 (discussed below), the function of these factors at centromeres is not yet determined [348]. We predict that Sp1 interacts with or recruits these or other chromatin remodeling/modifying factors to mitotic centromeres to regulate histone modifications and subsequently influence transcription of core centromeres.

Our current understanding of the centrochromatin architecture indicates that the centrochromatin maintains a unique bivalent signature consisting of both heterochromatic modifications and transcription-coupled modifications. For example, the centrochromatin is enriched for H3K4me2 and H3K36me2, both modifications associated with transcription. H3K4me2 is often enriched at promoters of active genes, as well as at promoters of genes that are primed for future gene expression during cell development. H3K4me2 can also exist in large domains over both the gene promoter and the gene body, indicating that it may be a mechanism for fine-tuning gene expression [335]. H3K36me2 is highly enriched 3’ to open reading frames, and methylation of H3K36 is carried out by Set2, a conserved histone methyltransferase that specifically associates with elongating RNA Polymerase II [672, 673]. Although it is widely accepted that H3K36me2 is associated with active transcription, the mechanisms behind H3K36me2-dependent activation remain largely obscure [674-676]. Additionally,

H2Bub1, a modification associated with active transcription of many genes, is associated with centromeres during the G2-M transition [340]. Conversely, the centrochromatin does not display other histone modifications typically associated with active transcription, including H3K4me3, acetylated H3K9, and acetylated H4K5, H4K8, H4K12, and H4K16

138

[334], indicating that the pattern of modification at this region is less clear and may not necessarily conform to “active” or “inactive” marks. We show that some of these histone modifications are disrupted in Sp1-depleted cells, supporting the notion that Sp1 may interact with chromatin remodeling/modifying factors to maintain centrochromatin epigenetics during mitosis. We observe an increase in the H2Bub1 to H2B ratio at centromeres (Figure 4.12A) and a decrease in the H3K4me2 to H3 ratio and H3K36me2 to H3 ratio at centromeres (Figure 4.12B and Figure 4.12C, respectively) in Sp1- depleted cells. The increase in the ratio of H2Bub1 to H2B may contribute to the increase in centromeric transcripts as this modification is associated with active transcription. The decrease in the ratio of H3K4me2 to H3 and H3K36me2 to H3 is surprising as this would typically indicate more tightly packed centrochromatin in the absence of Sp1 and thus a decrease in core centromere transcription. As our data shows the opposite effect, we believe these data highlight the complexity of histone modifications such as H3K4me2 and H3K36me2 at the centrochromatin.

In an attempt to explain the increase in transcription in Sp1-depleted cells despite changes in histone modifications that might indicate the opposite effect, we evaluated centromeres for acetylated histones, and found a significant decrease in histone 3 acetylated on lysine 9 (H3K9Ac) in Sp1-depleted cells (Figure 4.12D). Because centromeres are hypo-acetylated in normal cells, this specific change may not have a significant impact on centromere function. However, how changes in centrochromatin histone modifications affect transcription at centromeres has not been fully elucidated nor has how changes in some histone modifications may affect others. More information on the acetylation status of centromeric histones in Sp1-depleted cells will help to determine if Sp1 negatively regulates centromeric transcription though histone deacetylation.

139

The surrounding pericentromeric DNA is enriched for heterochromatic marks, including H3K9me2/3, H3K27me2/3, and H4K20me3 [343-345]. We did not see a change in the pericentromeric mark H3K9me2/3 (Figure 4.12E). This is consistent with

Sp1 functioning primarily at the core centromere rather than at the surrounding pericentromeric region. However, because disrupted pericentromeric heterochromatin can affect centromere architecture and function [677], we plan to more thoroughly analyze pericentromeric heterochromatin marks in the context of Sp1 depletion.

Taken together, the data presented in Chapter 4 indicate that Sp1 likely interacts with chromatin remodeling/modifying factors to modulate histone modifications and the centrochromatin architecture at centromeres, subsequently regulating transcription through this region. The resulting lncRNA transcripts maintain centromere identity and function by regulating the recruitment of centromere-associated proteins including

CENP-C and CENP-A, both of which are vital for maintaining centromere identity.

Disrupted centromere identity contributes to chromosome segregation errors and W-CIN

(see our model in Figure 6.1). A more complete understanding of centrochromatin epigenetics, as well as what factors influence centrochromatin architecture, is required to confirm our model.

Sp11-182 May Be Sufficient for Functional Centromeres:

As discussed above, we hypothesize that Sp1 regulates the centrochromatin histone code, subsequently acting as a negative regulator of core centromeric transcription, thereby maintaining centromere identity and function. This is likely accomplished through Sp1-dependent recruitment of chromatin modifiers and/or remodelers to centromeres, although a specific enzyme(s) has not been identified. The data presented in Chapter 5 indicate that the N-terminal 182 amino acids may be sufficient for function of Sp1 at this region. This information may help us identify centromere-specific Sp1 interacting partners because it potentially limits candidate

140 enzymes to those that interact with Sp1 at its N-terminus. For example, HDAC1 has been shown to localize to centromeres, although its function at this region has not been characterized [348]. HDAC1 also interacts with Sp1 at promoters of a number of genes to deacetylate histones and down-regulate transcription [403-407], making HDAC1 an attractive candidate for an Sp1-dependent chromatin modifier at centromeres. However,

HDAC1 reportedly interacts with the C-terminus of Sp1 in these cases and therefore may not interact with Sp11-182 directly [403]. This does not rule out the possibility that Sp1 may interact with HDAC1 indirectly as a component of a larger complex of proteins where complex formation only requires amino acids 1-182. HDAC2, for example, interacts with Sp1 transactivation domains and therefore may interact with Sp11-182.

HDAC2 is often found in complex with HDAC1, including in three major co-repressor complexes, Sin3, NuRD (nucleosome remodeling and deacetylation), and CoREST (co- repressor for element-1-silencing transcription factor) [678]. Interestingly, in fission yeast, the Sin3 co-repressor Pst1p was shown to localize to centromeres and is required for maintaining low levels of acetylation at this region. Pst1p mutant cells show centromeric and chromosome segregation defects, including hyperacetylation, lagging chromosomes, and defects in sister chromatid cohesion [679]. The mammalian Sin3 complex, mSin3, includes HDAC1, HDAC2, and RbAp48, among other proteins. Sp1 was shown to direct the localization of the mSin3 complex to the hLHR promoter through its interaction with HDAC2 [410]. Further, although the mSin3 complex has not been identified as being targeted to centromeres, deletion of the mSin3 component mSds3 results in defects in pericentromeric histone modifications and chromosome segregation errors [680]. Thus, although it is unlikely that Sp1 interacts with HDAC1 directly at centromeres, it may do so indirectly, possibly through HDAC2 and the mSin3 complex.

Taken together, the data presented in Chapter 5 provide valuable information that can

141 be used for identifying potential Sp1 interacting partners at centromeres in an effort to gain a full understanding of its function at centromeres during mitosis.

Current Model:

To ensure that chromosomes segregate faithfully during mitosis, CCAN proteins and kinetochore complexes must assemble properly at centromeres for correct attachment of chromosomes to spindle microtubules. Centromeres are identified by

CENP-A containing nucleosomes, and defects in CENP-A deposition can result in disrupted CCAN and kinetochore assembly. CENP-A deposition is dependent on lncRNAs transcribed from core centromeres, and transcription through core centromeres is regulated by the centrochromatin architecture. The data presented in Chapters 4 and

5 suggest that Sp1 regulates the centrochromatin architecture during mitosis, thereby regulating centromeric transcription, subsequent assembly of the CCAN protein CENP-

C, and finally deposition of CENP-A at centromeres. This identifies Sp1 as an important upstream factor for centromere identity and function. This information is summarized in our model in Figure 6.1.

Future Directions

Although the data presented in this body of work makes a strong case for Sp1- dependent regulation of centromeric transcription, and therefore downstream deposition of CENP-A for maintaining centromere identity and function, there are significant unknowns in our model. Further, additional information is required to conclusively link the function of Sp1 at centromeres to chromosome instability as well as to implicate this function of Sp1 in cell transformation and cancer development. This section discusses future directions for this area of study, with the goal of filling these gaps and gaining a more complete understanding of the function of Sp1 at centromeres during mitosis.

Identifying Sp1 Interacting Partners at Centromeres:

142

Our studies indicate that Sp1 is a negative regulator of centromeric transcription during mitosis. We show that depletion of Sp1 results in a dramatic increase in centromere-derived lncRNAs (Figure 4.9A and 4.9C) as well as an increase in RNA

Polymerase II, phosphorylated on serine 5, an indicator of active transcription (Figure

4.10). Further, over-expression of Sp1 has the opposite effect, decreasing lncRNA expression (Figure 4.9B and 4.9D). It is well established that centromeres are transcribed during mitosis, although the timing of this transcription is slightly controversial (see review in Chapter 2). Some studies suggest that transcription of core centromeres peaks during metaphase, the stage at which Sp1 localizes to centromeres.

Thus, though Sp1 appears to be acting as a negative regulator of transcription, it does not completely repress transcription, but may fine-tune expression, preventing over- expression and reduced expression of lncRNAs, both of which are detrimental to proper mitoses.

How Sp1 negatively regulates transcription at core centromeres is not clear. Sp1 is most commonly known as a transcriptional activator. However, there are examples of

Sp1 acting as a transcriptional repressor at other locations within the genome by recruiting histone deacetylases and DNA methyltransferases to gene promoters

(discussed in Chapter 3). To understand Sp1 as a negative regulator of centromeric transcription, it will be necessary to identify the Sp1-dependent enzyme or enzymes required for this regulation. One potential candidate is the mammalian transcriptional co- repressor mSin3. As discussed above, the Sin3 co-repressor Pst1p was shown to localize to centromeres in fission yeast and is required for maintaining low levels of acetylation at this region [679]. In mammalian cells, deletion of the mSin3 component mSds3 results in failure of pericentric histones to be deacetylated, thereby preventing the cascade of histone modification events required for the establishment of a functional pericentric heterochromatin structure [680]. mSin3 may act at both centromeres and

143 pericentromeres in an Sp1-dependent manner and merits further investigation.

Alternatively, immunoprecipitation of Sp1 from an exclusively mitotic population of cells, coupled with mass spectrometry, could identify chromatin remodelers or modifiers that interact with Sp1 during mitosis and could therefore be candidate enzymes for regulators of Sp1-dependent histone modifications at centromeres. Once candidate interacting partners are identified, it must be shown that these factors localize to centromeres in a manner that is both spatially and temporally consistent with being recruited by Sp1 and also show that Sp1 knockdown prevents localization of these proteins. An experiment designed such that a candidate protein is artificially tethered to centromeres in the absence of Sp1, perhaps through creating a fusion protein with the CENP-B DNA- binding domain, may add strength to these data, with the anticipation that artificial tethering will rescue phenotypes observed in Sp1-depleted cells.

Linking Histone Modifications to Transcription:

Although Sp1 depletion results in an increase in centromeric transcription, the alterations in histone modifications observed in Sp1-depleted cells do not align with transcriptional upregulation. For example, H3K4me2 and H3K36me2 are both histone modifications associated with active transcription. In Sp1-depleted cells, however, both of these modifications are decreased, which is what might be expected if Sp1 were acting as a positive regulator of centromeric transcription. Continuing to evaluate histone modifications in Sp1-depleted cells with the goal of identifying the aberrant modification(s) that contributes to up-regulation of centromeric transcription may help identify the enzymes responsible for normal maintenance of these modifications. For example, centromeres are hypo-acetylated, and an increase in histone acetylation in response to Sp1 depletion may explain the increase in transcription. Although we observed a decrease in H3K9Ac, investigating other acetylated residues by ChIP, such as those known to be low or absent at centromeres (see Figure 2.4), may provide

144 valuable information to this end. Studies in human artificial chromosomes have shown that increased acetylation results in an increase in core transcription and a coupled decrease in CENP-A assembly [366]. Further, studies in fission yeast have shown that treatment with the histone deacetylase inhibitor Trichostatin A (TSA) increases histone

H3 and H4 acetylation at centromeres and causes chromosome segregation defects

[681]. Within the context of DNA damage, Sp1 depletion results in an increase in

H4K16Ac at DNA double-strand-breaks, and may have a similar effect at centromeres.

Thus, a closer look at the acetylation status of centromeres in response to Sp1 depletion may help to determine the mechanism behind negative regulation of centromeric transcription by Sp1 as well as aid in identifying the Sp1-dependent enzymes responsible for maintaining the centrochromatin.

Demonstrating a Decrease in CENP-A Over Time:

Our model suggests that Sp1 is required for downstream loading of new CENP-A molecules at centromeres after cells exit mitosis (Figure 6.1). Therefore, after Sp1 has been depleted, cells must continue to cycle through mitosis and into G1 in order to show the CENP-A phenotype. As such, for each subsequent cell cycle, centromeres will show a progressive decrease in CENP-A, as the cell continuously fails to load new CENP-A molecules. For experiments in which CENP-A levels were evaluated in Sp1-depleted cells, samples were collected 96 hours post infection. Although we have not specifically evaluated the cell cycle after lentivirus infection, it is likely that within this timeframe, cells cycle a very limited number of times. To support our model, it will be important to evaluate CENP-A levels at centromeres over an extended timeframe to see this expected progressive decrease in CENP-A at centromeres. Such an experiment will confirm our hypothesis that Sp1 is required for loading of new CENP-A molecules in G1.

Completing the Sp11-182 Story:

145

The data presented in Chapter 5 indicate that Sp11-182 may be capable of rescuing some of the centromere phenotypes observed in Sp1-depleted cells. These phenotypes include the centromere distance phenotype (Figure 5.2) as well as the decrease in CENP-A at centromeres (Figure 5.3 and 5.4). These data indicate that the first 182 amino acids may be sufficient for function. Though meaningful, much work is needed to make these data relevant. First, the experiment presented in Figure 5.4 requires repeating in order to establish significance. Next, the first 182 amino acids of

Sp1 do not contain the sequence-specific DNA-binding domain. Therefore, how full length or Sp11-182 localizes to centromeres remains undetermined. One could postulate that Sp1 is recruited to this region as part of a larger complex, where complex formation requires only amino acids 1-182. Co-immunoprecipitation experiments using Sp11-182, coupled with mass spectrometry may provide clues as to how this occurs. Next, any centromere-specific chromatin modifiers or remodeling complexes identified as being

Sp1-dependent should interact with and be recruited to centromeres by Sp11-182, either directly or indirectly. To conclusively determine that Sp11-182 is functional at centromeres, it will be important to show that Sp11-182 is capable of rescuing the changes in centromeric transcription observed in Sp1-depleted cells as well as binding of CENP-C and all aberrant histone modifications that result from Sp1 knockdown.

Linking the Function of Sp1 at Centromeres to Chromosomal Instability:

Throughout this body of work, we have linked the function of Sp1 at centromeres to W-CIN. We previously demonstrated that Sp1 depletion results in phenotypes consistent with W-CIN [7]. Here, we showed that Sp1 depletion results in decreased

CENP-A loading at centromeres (Figure 4.6), which has been shown in other studies to cause segregation defects and W-CIN (discussed in Chapter 2). However, we have not directly shown an Sp1 depletion-dependent inactivation of centromeres as a result of decreased CENP-A loading and subsequent chromosome segregation errors. To

146 accomplish this, we would need to measure centromeric function directly within the context of Sp1 knockdown. This could be done by evaluating kinetochore assembly and the attachment of microtubules to kinetochores. Kinetochore assembly is often measured by evaluating the localization of outer kinetochore proteins using immunofluorescence, as the outer kinetochore assembles during mitosis. Defects in kinetochore assembly in Sp1-depleted cells would be identifiable by changes in the staining pattern of these proteins as compared to control cells. Further, attachment of microtubules to kinetochores could similarly be measured by immunofluorescence. Cells in which end-on kinetochore-microtubule attachments are disrupted would show fewer attachments as compared to control cells. W-CIN could be measured by fixing control and Sp1-depleted cells in anaphase and evaluating those cells for evidence of chromosome mis-segregation. To confirm that these defects are caused by loss of Sp1 specifically at centromeres, and not from changes in Sp1-dependent transcription of other proteins, we would need to show that Sp11-182, or a DNA binding-deficient mutant of Sp1, is capable of rescuing kinetochore assembly errors, microtubule attachment errors, and W-CIN. Finally, as both increases and decreases in transcription of core centromeres is associated with chromosome segregation errors and W-CIN, a study correlating the rate of transcription of core centromeres in W-CIN positive and W-CIN negative cell lines with total Sp1 protein levels may provide important data in support of our model.

Identifying Post-Translational Modifications Required for Localization:

Sp1 is phosphorylated on several residues during mitosis, including serine 59

(controversially; see Chapter 3), threonine 278, threonine 651, threonine 681, and threonine 739. Threonine 651, 681, and 739 phosphorylation events likely decrease the

DNA-binding affinity of Sp1 and stabilize Sp1 by preventing ubiquitination and degradation (discussed in Chapter 3). Interestingly, our data show that ATM kinase

147 activity is required for rapid Sp1 recruitment to centromeres, as specific inhibition of ATM by the small molecule inhibitor KU-55933 prevents localization (Figure 4.5). ATM is best known for its role in DNA double-strand-break repair, where it phosphorylates a variety of proteins involved in the DNA damage response [682]. Recently, ATM was shown to be activated by Aurora B kinase in a DNA-damage independent context during mitosis

(discussed above). ATM preferentially phosphorylates substrates on serine or threonine residues that are followed by glutamine residues, called SQ/TQ, or S/TQ motifs [683].

Sp1 has 15 SQ/TQ motifs in total, 7 of which are immediately preceded by a hydrophobic residue, which is the preferred ATM phosphorylation site (Table 6.1) [683].

None of the residues identified as being phosphorylated during mitosis are SQ/TQ sites and thus are not likely phosphorylated by ATM. Of the 15 SQ/TQ sites, only serine 56 and serine 101 have been identified as being phosphorylated by ATM (Table 3.1), and our data shows that Sp1 is not phosphorylated on serine 101 during mitosis (Figure 4.5).

Although serine 56 is phosphorylated by ATM, we have not evaluated mitotic cells for this post-translational modification. Because ATM directly phosphorylates other centromere-associated proteins for localization to centromeres, it is possible that ATM phosphorylates Sp1 at serine 56 for recruitment to centromeres. This hypothesis has not yet been tested nor have we evaluated any other SQ/TQ sites, although this could be accomplished through mutagenesis. Additionally, although the data in Figure 4.5 are compelling, an experiment showing that ATM depletion by RNAi abrogates Sp1 localization to centromeres, or the same study in ATM-null cells, would strengthen our conclusion.

Effects of Sp1 Knockdown on the Chromosomal Passenger Complex:

The Chromosomal Passenger Complex (CPC) is a complex of proteins consisting of Aurora B kinase, INCENP (inner centromere protein), Borealin, and

Survivin. During early stages of mitosis, the CPC localizes to chromosome arms and

148 centromeres as Aurora B kinase together with Plk1 phosphorylates arm sister- chromatid-cohesin (SCC) complexes to allow for their dissociation from chromosomes.

As the cell progresses into metaphase, localization is restricted to centromeres, where the CPC is thought to assist in recruiting other centromere-associated proteins (like SAC proteins [87] and Sgo1 [88]) as well as to assist in destabilizing incorrectly attached microtubules. When CPC function is compromised, a cell fails to detach incorrectly attached microtubules, resulting in an increased persistence of merotelic attachments and chromosome segregation errors. Failed CPC function also likely affects SCC at both chromosome arms and at centromeres. Sp1 transcriptionally regulates Survivin gene expression, and although it has not been specifically evaluated, changes in Sp1 protein levels likely also affect mitosis in a Survivin-dependent manner. Additional studies are needed to evaluate the impact of Sp1-dependent changes in Survivin levels on CPC activity during mitosis.

Understanding the Centromeric Cohesion Defect in Sp1-depleted Cells:

One apparent phenotype upon Sp1 depletion is an increased distance between

CENP-A signals as analyzed by IF (Figure 4.6C and Figure 5.2). These data suggest that Sp1 may be an important factor for maintaining centromeric sister chromatid cohesion at centromeres. In an attempt to begin to understand the relationship between

Sp1 and SCC complexes, we have evaluated Scc1, Smc1, and Smc3 protein levels in

Sp1-depleted cells to determine if Sp1 depletion results in a decrease in total cellular levels of SCC complex components. If so, this could explain the decrease in sister chromatin cohesion observed upon Sp1 depletion. Our data from a single experiment suggests that Scc1, Smc1, and Smc3 protein levels do not change significantly in Sp1- depleted cells (Figure 6.2A). SA1/SA2 levels have not been evaluated. Additional experiments evaluating the effect of Sp1 depletion on SCC complex components is

149 necessary to conclusively determine if Sp1 depletion affects SCC complex component protein levels and thus SCC itself.

SCC complexes are known to play a role in transcriptional regulation by enabling the DNA-binding protein CTCF (CCCTC-binding factor) to protect gene promoters from distal enhancers, thereby acting as a transcriptional insulator [684]. Additionally, SCC complexes are also required for efficient repair of DNA double stranded breaks [685,

686]. Further, SCC complexes are thought to maintain the connection between the two centrioles of a centrosome, and our previous work shows that Sp1 depletion results in an increase in the distance between centrioles as measured by IF [7], a phenotype similar to that seen at centromeres in Sp1-depleted cells. Because Sp1 is an integral component of both transcriptional regulation and efficient DNA double-strand-break repair, and because Sp1 localizes to both centrosomes and centromeres, it is possible that Sp1 regulates the localization and/or function of SCC complexes through a physical interaction. To begin to evaluate this possibility, we isolated mitotic cells by mechanical shake-off (Figure 6.2B) and used co-immunoprecipitation (Co-IP) to determine if Sp1 interacts with Smc3 during mitosis. Our data show that Sp1 does not pull down Smc3, nor does Smc3 pull down Sp1 (Figure 6.2C). It would be compelling to determine if Sp1 interacts with other SCC complex components during mitosis.

Much about how SCC complexes are loaded and established at mammalian centromeres remains unclear. However, studies in yeast have indicated that the epigenetic state of centromeric and pericentromeric chromatin may play an important role. Fission yeast Swi6 (HP1) is required for heterochromatin formation at pericentromeres, as well as promoting sister chromatin cohesion, as Swi6 deficiency specifically disrupts cohesion at centromeres but not along chromosome arms [687-689].

Similarly, Sp1-dependent centrochromatin regulation may be required for proper establishment and/or maintenance of centromeric cohesion. The changes in histone

150 modifications observed in Sp1-depleted cells may prevent proper loading and/or establishment of sister chromatid cohesion and merits further investigation.

During mitosis, sister chromatid cohesion is maintained at centromeres largely in part by phosphatase activity of the Sgo1-PP2A complex. Chromosome arm SCC complexes are removed at the beginning of mitosis by Plk1 and Aurora B kinase- mediated phosphorylation; Sgo1-PP2A complexes maintain a hypo-phosphorylated state at centromeres, preventing premature SCC dissociation [70]. Centromeric localization of

Sgo1 depends on histone H2A phosphorylation mediated by kinase Bub1 [690], as well as CPC localization [88]. As such, evaluating cells for Sp1 depletion-dependent changes

Bub1 kinase localization, histone H2B phosphorylation status, and/or Sgo1 localization and function may reveal defects that contribute to the sister chromatid cohesion phenotype observed in Sp1-depleted cells.

151

Figure 6.1: Model. During mitosis, Sp1 localizes to centromeres and pericentromeres in a sequence independent manner. The mechanism by which Sp1 localizes to centromeres and pericentromeres is not known. Because binding at this region is not sequence dependent, Sp1 is likely either recruited to this region by an unknown factor (A,1a) or is recruited as a component of a complex that has yet to be identified (A,1b). At centromeres, Sp1 likely recruits chromatin modifiers or remodelers (A, 2) to maintain the histone modifications (A, 3) necessary for RNA Polymerase II-dependent transcription of core centromeres (B, 4). The lncRNA transcripts generated from RNA Polymerase II-dependent transcription (B, 5) ensure proper assembly of CCAN proteins like CENP-C (B,6), which is required for recruitment of the Mis18 complex (C, 7). The Mis18 complex recruits the CENP-A chaperone HJURP to centromeres (C, 8), which facilitates the exchange of histone H3 for CENP-A (D, 9), thereby maintain proper levels of CENP-A at centromeres, and thus centromere identity, preventing the chromosome mis-segregation events that lead to W-CIN (D, 10).

152

A B

Figure 6.2: Understanding the Centromeric Cohesion Defect in Sp1-depleted Cells. (A) MCF 10A cells were transduced with control shRNA or shRNA targeting the Sp1 transcript. Protein lysates were collected and evaluated for Sp1 protein levels to confirm protein knockdown, and for changes in Scc1, Smc1, and Smc3. Nucleolin was used as a loading control. Results show that Sp1 depletion has no effect on Scc1, Smc1, or Smc3 protein levels. (B) Mitotic cells were isolated by mechanical shake-off. Both the mitotic population and the remaining interphase population of cells were lysed, and protein lysates were evaluated for cyclins to ensure the mitotic population of cells contained no interphase cells. (C) Mitotic lysates from (B) were used for co-immunoprecipitation, using Sp1 Ab581, Sp1 conjugated beads (Santa Cruz sc-59 AC) and Smc3 for pull-down. Immunoprecipitated lysates were analyzed by immunoblotting using Sp1 Ab581 and Smc3. Here, normal rabbit IgG was used as a negative control. Results show that Sp1 does not co-immunoprecipitate with Smc3 in the mitotic population and vice versa.

153

Table 6.1: Sp1 SQ/TQ Sites.

Residue with Preceding and Proceeding Sp1 SQ/TQ Site Amino Acids Serine 36 Phenylalanine-Serine-Glutamine Serine 56 Glutamic Acid-Serine-Glutamine Serine 81 Asparagine-Serine-Glutamine Serine 85 Proline-Serine-Glutamine Threonine 98 Alanine-Threonine-Glutamine Serine 101 Leucine-Serine-Glutamine Threonine 250 Glutamine-Threonine-Glutamine Serine 281 Serine-Serine-Glutamine Serine 291 Glycine-Serine-Glutamine Serine 296 Glycine-Serine-Glutamine Serine 313 Serine-Serine-Glutamine Serine 351 Asparagine-Serine-Glutamine Threonine 394 Glutamine-Threonine-Glutamine Threonine 427 Threonine-Threonine-Glutamine Serine 431 Isoleucine-Serine-Glutamine

Residues highlighted in green are preferred ATM sites; the serine or threonine is directly preceded by a hydrophobic residue.

154

Chapter 7: Experimental Procedures

155

Cell Lines and Culture Conditions: MCF 10As (ATCC) were cultured in Dulbecco's

Modified of Eagle's Medium/Ham's F-12 50:50 Mix (Cellgrow, Mediatech, Inc) supplemented with 5% fetal bovine serum (FBS; Gemini), 10 μg recombinant human epidermal growth factor (EGF; PeproTech), 250 ug hydrocortisone (Sigma), 5 mg insulin from bovine pancreas (Sigma), 50 μg Cholera Toxin from Vibrio cholera (Sigma), 100 mg penicillin, and 60 mg streptomycin (Pen-Strep; Sigma). Sp1+/+ and Sp1-/- mouse embryonic fibroblasts (MEFs) have been previously described [636] and were kindly provided by Dr. G.Suske. MEFs, Normal Human Diploid Fibroblasts (NHDFs; ATCC), and human osteosarcoma cell line Saos-2 (ATCC), were cultured in Dulbecco's

Modification of Eagle's Medium (DMEM; Cellgrow, Mediatech, Inc) supplemented with

10% FBS. The retroviral packaging cell line 293-GPG (VSV-G) was maintained in

DMEM supplemented with 10% heat-inactivated FBS, Pen-Strep, 1 μg/mL tetracycline, 2

μg/mL puromycin, and 0.3 mg/mL G418. During production of lentiviruses, 293T cells were maintained in DMEM supplemented with 10% heat-inactivated FBS and Pen-Strep.

All cells were maintained in a 37ᵒC humidified atmosphere with 5% CO2.

Plasmids and Viral Infections: All vectors, including pLKO-shRNA control vector, pLKO-shRNA vector for Sp1, pLXSN empty vector, pLXSN-Sp1-HA, and pLXSN-Sp11-

182-HA (also known as pLXSN-Sp1-damage response domain), pLXSN-Sp1183-785-HA, and pLXSN-Sp1-D183A-HA were previously described [384]. Viral packaging vectors, pCMV-VSV-G, pRSV-Rev, and pMDLg/pRRE, were generously provided by M. Reginato

(Drexel University College of Medicine, Philadelphia, PA). For lentivirus production, 293T cells were co-transfected for 6 h with 6 μg of pLKO vectors containing the shRNA sequences, along with 2 μg of each viral packaging component. Virus was collected 48 and 72 hours post-transfection, and stored at -80ᵒC. For retrovirus production, 293-GPG cells were transfected with 10 μg pLXSN empty vector, pLXSN-Sp1-HA, or pLXSN-Sp11-

156

182-HA plasmid using Gendrill transfection reagent (Bamagin) per manufacturer’s instructions. Virus was collected on days 6, 7, and 8 post-transfection, and stored at -

80ᵒC.

Transduction of MCF 10A Cells: For protein knockdown or expression vectors: MCF

10A cells were plated and treated with pLKO-shRNA control lentivirus, pLKO-Sp1 shRNA lentivirus, pLXSN empty vector retrovirus, pLXSN-Sp11-182-HA retrovirus, pLXSN-

Sp1183-785-HA retrovirus, or pLXSN-Sp1-D183A-HA retrovirus at a concentration of 4 μL of virus per cell plated. A complete medium change was performed 16 hours post infection, and cells were used for experiments 96 hours post infection. For rescue experiments: MCF 10A cells were first treated with retroviral vectors as described above.

24 hours after post infection, cells were infected with lentivirus as described above. Cells were used for experiments 72 hours post lentivirus infection.

DNA Spreads and Immunofluorescence: Cells were treated with 100 mg/mL Colcemid

(Sigma) for 4 hours, harvested, and pelleted at a concentration of 80,000 cells/mL. Then, cell pellets were resuspended in hypotonic solution (10 mM Hepes pH 7.3; 2% FBS; 30 mM Glycerol; 1.0 mM CaCl2; 0.8 mM MgCl2) and incubated at 4ᵒC for 15 minutes. DNA spreads were prepared using a Shandon Cytospin 3 (2000 rpm for 20 minutes at room temperature), and fixed in 100% methanol for 30 minutes at -20ᵒC, rehydrated in acetone for 30 seconds at -20ᵒC, then dried at room temperature. For immunofluorescence: dried DNA spreads were rehydrated in 1xPBS with 0.01% sodium azide for 5 minutes. Slides were washed 3 times for 1 minute each in TTB solution (1.0 mM Triethanolamine:HCl pH 8.5; 0.2 mM Na-EDTA; 25 mM NaCl; 0.1% Triton-X 100;

0.1% BSA, Gemini), then incubated with primary antibodies diluted in TTB solution for 30 minutes at 37ᵒC. Slides were washed for 2, 5, then 3 minutes with 1xKB solution (10 mM

157

Tris:HCl pH 7.7; 0.15 M NaCl; 0.1% BSA), then incubated with secondary antibodies diluted in 1xKB solution for 30 minutes at 37ᵒC. Slides were washed with 1xKB solution for 5 minutes, then with 1xPBS solution for 5 minutes. Coverslips were applied using

VectaMountTM AQ (Vector Laboratories, Inc). Images were collected using an Olympus

AX-70 compound microscope, and iVisionTM Scientific Image Process software by

BioVision Technologies. For KU-55933 treatment: MCF 10A cells were incubated with

KU-55933 (Tocris) or DMSO (Sigma) for indicated time. Primary antibodies: Sp1 H225

(Santa Cruz sc-14027, 1:200), CENP-A (Abcam ab13939, 1:200), CREST

(ImmunoVision HCT-0100, 1:200), HA (Cell Signaling 2367, 1:500). Secondary antibodies: Donkey anti-Rabbit IgG (H+L) Secondary Antibody, Alexa Fluor® 488 conjugate (A21206, 1:1000), Donkey anti-Mouse IgG (H+L) Secondary Antibody, Alexa

Fluor® 594 conjugate (A21203, 1:1000). DNA stain: DAPI (Sigma).

Western Blotting: For Western blot analysis, cells were directly lysed in 2x SDS sample buffer (12.5 mMTris pH 6.8; 20% glycerol; 4% SDS). Proteins were separated by traditional SDS-PAGE, transferred to a polyvinylidene difluoride (PVDF) membrane, and analyzed by immunoblotting with the following antibodies, diluted in TBST with 5% BSA:

Sp1 Ab581, (1:500 [381]), Sp1p101 (1:500, [384]), Sp1 H225 (Santa Cruz sc-14027,

1:200), Sp1 pre-conjugated beads (Santa Cruz sc-59 AC), CENP-A (Abcam ab13939,

1:1000), CENP-C (Abcam ab196666, 1:1000), H2Bub1 (Cell Signaling 5546, 1:1000),

H3K36me2 (Abcam 9049, 1:1000), H3K9me2/3 (Cell Signaling 5327, 1:1000), H3K4me2

(NeoBioLab A2356, 1:500), H3K9Ac (Cell Signaling 9649, 1:1000), HA (Cell Signaling

3724, 1:1000), pChk2 Thr68 (Cell Signaling 2661,1:1000), Chk2 (Cell Signaling 3440,

1:1000), Cyclin-B1 (Santa Cruz sc-245, 1:500), Cyclin-D (Santa Cruz sc-753, 1:500),

Cyclin-E (Santa Cruz sc-198, 1:500), Cyclin-A (Santa Cruz sc-239, 1:500), Scc1 (Abcam ab992, 1:500), Smc1 (Bethyl A300-055A, 1:500), Smc3 (Bethyl A300-060A, 1:500), actin

158

(Santa Cruz sc-1615, 1:2500) α-tubulin (GeneTex GTX27291, 1:3000), Nucleolin (Santa

Cruz sc-8031, 1:500)

Chromatin Immunoprecipitation: Cells were cross-linked with 1% formaldehyde while rocking for 15 minutes. Reaction was quenched by the addition of 125 mM glycine for 5 minutes, cells were then scraped and collected in 1xPBS with protease and phosphatase inhibitors. Cell pellets were lysed in 1 mL cell lysis buffer (5 mM PIPES, pH

8.0; 85 mM KCl; 0.5% NP-40; plus protease and phosphatase inhibitors), then in 1 mL nuclei lysis buffer (50 mM Tris-HCl pH 8.0; 10 mM EDTA; 1% SDS; plus protease and phosphatase inhibitors). Chromatin was sheared by sonication using a Branson sonicator at 50% duty for a total of 30 seconds at pulsed intervals. Protein was quantified using a standard bicinchoninic acid (BCA) assay. A total of 5% of the lysate was reserved, and genomic DNA was isolated as input. Equal amounts of the remaining lysate was aliquoted and used for immunoprecipitation with CENP-A (Abcam ab13939),

CENP-C (Abcam 196666 and a gift from Dr. Andrea Musacchio, Ph.D [321]), H3 (Cell

Signaling 2650), H2Bub1 (Cell Signaling 5546), H2B (Millipore 07-371), H3K4me2

(NeoBioLab A2356), H3K36me2 (Abcam 9049), H3K9me2/3 (Cell Signaling 5327),

H3K9Ac (Cell Signaling 9649), RNA Polymerase II pCTD Ser5 (Abcam ab5095) and IgG

(Abcam ab46540) antibodies. Antibodies were pre-incubated with protein A/G agarose beads (Santa Cruz sc-2003). All beads were pre-blocked with 1 mg/mL low IgG BSA and

1 mg/mL tRNA (Sigma) prior to immunoprecipitation. Immunoprecipitated chromatin was isolated overnight, then beads were washed 4 times in high salt wash buffer (50 mM

HEPES pH 7.9; 250 mM NaCl; 1 mM EDTA; 0.1% SDS; 1% Triton X-100; 0.1%

Deoxycholate; plus protease and phosphatase inhibitors). DNA was treated with 10%

Chelex resin (BioRad), 10 mg/mL RNAse A (Qiagen), and 20 μg/mL Proteinase K

(Ambion), then collected. Input chromatin was treated with 10% Chelex resin, 10 mg/mL

159

RNAse A, and 20 ug/mL Proteinase K, then precipitated using 100% ethanol. Dry pellets were resuspended clean water. Analysis of protein binding at specific locations was assessed by SYBR green (Bio-Rad) quantitative PCR (qPCR) using a Bio-Rad CFX-96 real-time PCR detection system. Refer to Table 1 for primer information.

Co-Immunoprecipitation: Mitotic MCF 10A cells were isolated by mechanical shake-off as described in [691]. Then, cells were lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4;

1% NP-40; 120 mM NaCl; 0.25% Deoxycholic Acid; 1 mM EDTA; plus protease and phosphatase inhibitors). Cell lysates were pre-cleared by incubation with a 50% protein

A/G agarose bead slurry (Santa Cruz sc-2003) for 2 hours with rotation at 4C. Pre- cleared samples were centrifuged, and cell lysate supernatants were transferred to a fresh tube. After quantifying lysates by BCA assay, a portion of the total sample was set aside to be used as the input control, then equal amounts of the sample were incubated with the appropriate antibody for 2 hours with rotation at 4C to allow antibody-antigen complex formation. Antibody concentrations used were based on manufacturer’s recommendation. Then, a 60% protein A/G agarose bead slurry was added to each sample, and incubated for 4 hours with rotation at 4C to allow the antibody-antigen complex to conjugate to the beads. Beads were washed twice with a wash buffer (50 mM Tris-HCl, pH 7.4; 1% NP-40; 1% Triton-X; 250 mM NaCl; 0.25% Deoxycholic Acid;

1 mM EDTA; plus protease and phosphatase inhibitors), then resuspended in 70 μL of

2x SDS sample buffer (12.5 mMTris pH 6.8; 20% glycerol; 4% SDS). Samples were boiled at 100C for 5 minutes, then cooled at 4C for 5 minutes, then loaded onto an SDS-

PAGE gel for analysis by western blot, along with a 10% input control. Antibodies used:

Sp1 Ab581, Sp1 pre-conjugated beads (Santa Cruz sc-59 AC), Smc3 (Bethyl A300-

060A) and IgG (Abcam ab46540).

160

RNA Isolation, cDNA Synthesis and Quantitative PCR: RNA was extracted from MCF

10A cells using the Qiagen RNeasy® Mini Kit, per manufacturer’s instructions. RNA was reverse transcribed using the qScriptTM cDNA Supermix (Quanta Biosciences). PCR was performed using SYBR green (Bio-Rad) quantitative PCR (qPCR) using a Bio-Rad CFX-

96 real-time PCR detection system. Refer to Table 1 for primer information.

Sp1183-785 Stabilization Experiment: MCF 10A cells were plated and infected with pLXSN empty vector retrovirus, pLXSN-Sp11-182-HA retrovirus, pLXSN-Sp1183-785-HA retrovirus, or pLXSN-Sp1-D183A-HA retrovirus as described above. Treated each cell line with 10 μM Adriamycin, then collected cell lysates in 2x SDS sample buffer at 12,

24, and 48 hours post treatment. Proteins were separated by traditional SDS-PAGE, transferred to a polyvinylidene difluoride (PVDF) membrane, and analyzed by immunoblotting with the following antibodies, diluted in TBST with 5% BSA: Sp1 Ab581,

HA, and α-tubulin. The HA signal and the α-tubulin signal from each cell line were quantified using ImageJ, and the change in HA signal over time was determine as a ratio of HA to α-tubulin control. Then, results were normalized to ratios from untreated controls.

161

List of References

1. Langie, S.A., et al., Causes of genome instability: the effect of low dose chemical exposures in modern society. Carcinogenesis, 2015. 36 Suppl 1: p. S61-88.

2. Holland, A.J. and D.W. Cleveland, Losing balance: the origin and impact of aneuploidy in cancer. EMBO Rep, 2012. 13(6): p. 501-14.

3. Pfau, S.J. and A. Amon, Chromosomal instability and aneuploidy in cancer: from yeast to man. EMBO Rep, 2012. 13(6): p. 515-27.

4. Drake, J.W., et al., Rates of spontaneous mutation. Genetics, 1998. 148(4): p. 1667-86.

5. Hanahan, D. and R.A. Weinberg, Hallmarks of cancer: the next generation. Cell, 2011. 144(5): p. 646-74.

6. Aguilera, A. and T. Garcia-Muse, Causes of genome instability. Annu Rev Genet, 2013. 47: p. 1-32.

7. Astrinidis, A., et al., The transcription factor SP1 regulates centriole function and chromosomal stability through a functional interaction with the mammalian target of rapamycin/raptor complex. Genes, chromosomes & cancer, 2010. 49(3): p. 282-97.

8. McGranahan, N., et al., Cancer chromosomal instability: therapeutic and diagnostic challenges. EMBO Rep, 2012. 13(6): p. 528-38.

9. Perry, S.E., et al., Maternal Child Nursing Care. 2014: Elsevier Health Sciences.

10. Holland, A.J. and D.W. Cleveland, Boveri revisited: chromosomal instability, aneuploidy and tumorigenesis. Nat Rev Mol Cell Biol, 2009. 10(7): p. 478-87.

11. Weaver, B.A. and D.W. Cleveland, Aneuploidy: instigator and inhibitor of tumorigenesis. Cancer Res, 2007. 67(21): p. 10103-5.

12. Paulsson, K. and B. Johansson, High hyperdiploid childhood acute lymphoblastic leukemia. Genes Chromosomes Cancer, 2009. 48(8): p. 637-60.

13. Kaneko, Y. and A.G. Knudson, Mechanism and relevance of ploidy in neuroblastoma. Genes Chromosomes Cancer, 2000. 29(2): p. 89-95.

162

14. Duesberg, P. and R. Li, Multistep carcinogenesis: a chain reaction of aneuploidizations. Cell Cycle, 2003. 2(3): p. 202-10.

15. Michor, F., et al., Can chromosomal instability initiate tumorigenesis? Semin Cancer Biol, 2005. 15(1): p. 43-9.

16. Zimonjic, D., et al., Derivation of human tumor cells in vitro without widespread genomic instability. Cancer Res, 2001. 61(24): p. 8838-44.

17. Conduit, P.T., A. Wainman, and J.W. Raff, Centrosome function and assembly in animal cells. Nat Rev Mol Cell Biol, 2015. 16(10): p. 611-24.

18. Bettencourt-Dias, M., Q&A: Who needs a centrosome? BMC Biol, 2013. 11: p. 28.

19. O'Connell, C.B. and A.L. Khodjakov, Cooperative mechanisms of mitotic spindle formation. J Cell Sci, 2007. 120(Pt 10): p. 1717-22.

20. Saunders, W., Centrosomal amplification and spindle multipolarity in cancer cells. Semin Cancer Biol, 2005. 15(1): p. 25-32.

21. Sato, N., et al., Centrosome abnormalities in pancreatic ductal carcinoma. Clin Cancer Res, 1999. 5(5): p. 963-70.

22. Pihan, G.A., et al., Centrosome defects can account for cellular and genetic changes that characterize prostate cancer progression. Cancer Res, 2001. 61(5): p. 2212-9.

23. Ghadimi, B.M., et al., Centrosome amplification and instability occurs exclusively in aneuploid, but not in diploid colorectal cancer cell lines, and correlates with numerical chromosomal aberrations. Genes Chromosomes Cancer, 2000. 27(2): p. 183-90.

24. Kuo, K.K., et al., Centrosome abnormalities in human carcinomas of the gallbladder and intrahepatic and extrahepatic bile ducts. Hepatology, 2000. 31(1): p. 59-64.

25. Haruki, N., et al., Persistent increase in chromosome instability in lung cancer: possible indirect involvement of p53 inactivation. Am J Pathol, 2001. 159(4): p. 1345-52.

26. Ganem, N.J., S.A. Godinho, and D. Pellman, A mechanism linking extra centrosomes to chromosomal instability. Nature, 2009. 460(7252): p. 278-82.

27. Fu, J., I.M. Hagan, and D.M. Glover, The centrosome and its duplication cycle. Cold Spring Harb Perspect Biol, 2015. 7(2): p. a015800.

28. Feldman, J.L. and S.F.C.B. University of California, Deconstructing Cell Architecture: Exploring Centriole Structure, Function, and Position in the Green Alga Chlamydomonas Reinhardtii. 2008: University of California, San Francisco.

163

29. Schockel, L., et al., Cleavage of cohesin rings coordinates the separation of centrioles and chromatids. Nat Cell Biol, 2011. 13(8): p. 966-72.

30. Oliveira, R.A. and K. Nasmyth, Cohesin cleavage is insufficient for centriole disengagement in Drosophila. Curr Biol, 2013. 23(14): p. R601-3.

31. Matsuo, K., et al., Kendrin is a novel substrate for separase involved in the licensing of centriole duplication. Curr Biol, 2012. 22(10): p. 915-21.

32. Nigg, E.A., Centrosome duplication: of rules and licenses. Trends Cell Biol, 2007. 17(5): p. 215-21.

33. Firat-Karalar, E.N. and T. Stearns, The centriole duplication cycle. Philos Trans R Soc Lond B Biol Sci, 2014. 369(1650).

34. Korzeniewski, N., et al., Daughter centriole elongation is controlled by proteolysis. Mol Biol Cell, 2010. 21(22): p. 3942-51.

35. Gudi, R., et al., Centrobin-mediated regulation of the centrosomal protein 4.1- associated protein (CPAP) level limits centriole length during elongation stage. J Biol Chem, 2015. 290(11): p. 6890-902.

36. Schmidt, T.I., et al., Control of centriole length by CPAP and CP110. Curr Biol, 2009. 19(12): p. 1005-11.

37. Mardin, B.R., et al., Plk1 controls the Nek2A-PP1gamma antagonism in centrosome disjunction. Curr Biol, 2011. 21(13): p. 1145-51.

38. Smith, E., et al., Differential control of Eg5-dependent centrosome separation by Plk1 and Cdk1. EMBO J, 2011. 30(11): p. 2233-45.

39. Habedanck, R., et al., The Polo kinase Plk4 functions in centriole duplication. Nat Cell Biol, 2005. 7(11): p. 1140-6.

40. Kleylein-Sohn, J., et al., Plk4-induced centriole biogenesis in human cells. Dev Cell, 2007. 13(2): p. 190-202.

41. Cunha-Ferreira, I., et al., The SCF/Slimb ubiquitin ligase limits centrosome amplification through degradation of SAK/PLK4. Curr Biol, 2009. 19(1): p. 43-9.

42. Godinho, S.A. and D. Pellman, Causes and consequences of centrosome abnormalities in cancer. Philos Trans R Soc Lond B Biol Sci, 2014. 369(1650).

43. Fukasawa, K., et al., Abnormal centrosome amplification in the absence of p53. Science, 1996. 271(5256): p. 1744-7.

44. Li, J., et al., USP33 regulates centrosome biogenesis via deubiquitination of the centriolar protein CP110. Nature, 2013. 495(7440): p. 255-9.

45. Loncarek, J., et al., Control of daughter centriole formation by the pericentriolar material. Nat Cell Biol, 2008. 10(3): p. 322-8.

164

46. Starita, L.M., et al., BRCA1-dependent ubiquitination of gamma-tubulin regulates centrosome number. Mol Cell Biol, 2004. 24(19): p. 8457-66.

47. Lingle, W.L., et al., Centrosome hypertrophy in human breast tumors: implications for genomic stability and cell polarity. Proc Natl Acad Sci U S A, 1998. 95(6): p. 2950-5.

48. Pihan, G.A., et al., Centrosome defects and genetic instability in malignant tumors. Cancer Res, 1998. 58(17): p. 3974-85.

49. Hsu, L.C., et al., Centrosome abnormalities in ovarian cancer. Int J Cancer, 2005. 113(5): p. 746-51.

50. Kramer, A., K. Neben, and A.D. Ho, Centrosome aberrations in hematological malignancies. Cell Biol Int, 2005. 29(5): p. 375-83.

51. Giehl, M., et al., Centrosome aberrations in chronic myeloid leukemia correlate with stage of disease and chromosomal instability. Leukemia, 2005. 19(7): p. 1192-7.

52. Thompson, S.L. and D.A. Compton, Examining the link between chromosomal instability and aneuploidy in human cells. J Cell Biol, 2008. 180(4): p. 665-72.

53. Chan, J.Y., A clinical overview of centrosome amplification in human cancers. Int J Biol Sci, 2011. 7(8): p. 1122-44.

54. Melby, T.E., et al., The symmetrical structure of structural maintenance of chromosomes (SMC) and MukB proteins: long, antiparallel coiled coils, folded at a flexible hinge. J Cell Biol, 1998. 142(6): p. 1595-604.

55. Peters, J.M., A. Tedeschi, and J. Schmitz, The cohesin complex and its roles in chromosome biology. Genes Dev, 2008. 22(22): p. 3089-114.

56. Haering, C.H., et al., Molecular architecture of SMC proteins and the yeast cohesin complex. Mol Cell, 2002. 9(4): p. 773-88.

57. Losada, A., et al., Identification and characterization of SA/Scc3p subunits in the Xenopus and human cohesin complexes. J Cell Biol, 2000. 150(3): p. 405-16.

58. Sumara, I., et al., Characterization of vertebrate cohesin complexes and their regulation in prophase. J Cell Biol, 2000. 151(4): p. 749-62.

59. Prieto, I., et al., Mammalian STAG3 is a cohesin specific to sister chromatid arms in meiosis I. Nat Cell Biol, 2001. 3(8): p. 761-6.

60. Revenkova, E., et al., Cohesin SMC1 beta is required for meiotic chromosome dynamics, sister chromatid cohesion and DNA recombination. Nat Cell Biol, 2004. 6(6): p. 555-62.

61. Glynn, E.F., et al., Genome-wide mapping of the cohesin complex in the yeast Saccharomyces cerevisiae. PLoS Biol, 2004. 2(9): p. E259.

165

62. Laloraya, S., V. Guacci, and D. Koshland, Chromosomal addresses of the cohesin component Mcd1p. J Cell Biol, 2000. 151(5): p. 1047-56.

63. Watrin, E., et al., Human Scc4 is required for cohesin binding to chromatin, sister-chromatid cohesion, and mitotic progression. Curr Biol, 2006. 16(9): p. 863-74.

64. Gruber, S., et al., Evidence that loading of cohesin onto chromosomes involves opening of its SMC hinge. Cell, 2006. 127(3): p. 523-37.

65. Zhang, N. and D. Pati, Handcuff for sisters: a new model for sister chromatid cohesion. Cell Cycle, 2009. 8(3): p. 399-402.

66. Huang, C.E., M. Milutinovich, and D. Koshland, Rings, bracelet or snaps: fashionable alternatives for Smc complexes. Philos Trans R Soc Lond B Biol Sci, 2005. 360(1455): p. 537-42.

67. Nasmyth, K., How might cohesin hold sister chromatids together? Philos Trans R Soc Lond B Biol Sci, 2005. 360(1455): p. 483-96.

68. Zhang, N., et al., A handcuff model for the cohesin complex. J Cell Biol, 2008. 183(6): p. 1019-31.

69. Hauf, S., et al., Dissociation of cohesin from chromosome arms and loss of arm cohesion during early mitosis depends on phosphorylation of SA2. PLoS Biol, 2005. 3(3): p. e69.

70. McGuinness, B.E., et al., Shugoshin prevents dissociation of cohesin from centromeres during mitosis in vertebrate cells. PLoS Biol, 2005. 3(3): p. e86.

71. Barber, T.D., et al., Chromatid cohesion defects may underlie chromosome instability in human colorectal cancers. Proc Natl Acad Sci U S A, 2008. 105(9): p. 3443-8.

72. Mehes, K., P. Kajtar, and G. Kosztolanyi, Association of nonsyndromic Wilms tumor with premature centromere division (PCD). Am J Med Genet, 2002. 112(2): p. 215-6.

73. Kajii, T., et al., Cancer-prone syndrome of mosaic variegated aneuploidy and total premature chromatid separation: report of five infants. Am J Med Genet, 2001. 104(1): p. 57-64.

74. Plaja, A., et al., Variegated aneuploidy related to premature centromere division (PCD) is expressed in vivo and is a cancer-prone disease. Am J Med Genet, 2001. 98(3): p. 216-23.

75. Rao, N.M., et al., Premature separation of centromere and aneuploidy: an indicator of high risk in unaffected individuals from familial breast cancer families? Eur J Cancer Prev, 1996. 5(5): p. 343-50.

166

76. German, J., Roberts' syndrome. I. Cytological evidence for a disturbance in chromatid pairing. Clin Genet, 1979. 16(6): p. 441-7.

77. Mehes, K. and E.M. Buhler, Premature centromere division: a possible manifestation of chromosome instability. Am J Med Genet, 1995. 56(1): p. 76-9.

78. Moorhead, P.S. and A. Heyman, Chromosome studies of patients with Alzheimer disease. Am J Med Genet, 1983. 14(3): p. 545-56.

79. Spremo-Potparevic, B., et al., Analysis of premature centromere division (PCD) of the X chromosome in Alzheimer patients through the cell cycle. Exp Gerontol, 2004. 39(5): p. 849-54.

80. Scappaticci, S., et al., Chromosome abnormalities in tuberous sclerosis. Hum Genet, 1988. 79(2): p. 151-6.

81. Thompson, S.L. and D.A. Compton, Chromosome missegregation in human cells arises through specific types of kinetochore-microtubule attachment errors. Proc Natl Acad Sci U S A, 2011. 108(44): p. 17974-8.

82. Kelly, A.E. and H. Funabiki, Correcting aberrant kinetochore microtubule attachments: an Aurora B-centric view. Curr Opin Cell Biol, 2009. 21(1): p. 51-8.

83. Salmon, E.D., et al., Merotelic kinetochores in mammalian tissue cells. Philos Trans R Soc Lond B Biol Sci, 2005. 360(1455): p. 553-68.

84. Cimini, D., L.A. Cameron, and E.D. Salmon, Anaphase spindle mechanics prevent mis-segregation of merotelically oriented chromosomes. Curr Biol, 2004. 14(23): p. 2149-55.

85. Khodjakov, A., et al., Chromosome fragments possessing only one kinetochore can congress to the spindle equator. J Cell Biol, 1997. 136(2): p. 229-40.

86. Cimini, D., Merotelic kinetochore orientation, aneuploidy, and cancer. Biochim Biophys Acta, 2008. 1786(1): p. 32-40.

87. Lens, S.M., et al., Survivin is required for a sustained spindle checkpoint arrest in response to lack of tension. EMBO J, 2003. 22(12): p. 2934-47.

88. Kawashima, S.A., et al., Shugoshin enables tension-generating attachment of kinetochores by loading Aurora to centromeres. Genes Dev, 2007. 21(4): p. 420- 35.

89. Tanaka, T.U., et al., Evidence that the Ipl1-Sli15 (Aurora kinase-INCENP) complex promotes chromosome bi-orientation by altering kinetochore-spindle pole connections. Cell, 2002. 108(3): p. 317-29.

90. Cimini, D., et al., Aurora kinase promotes turnover of kinetochore microtubules to reduce chromosome segregation errors. Curr Biol, 2006. 16(17): p. 1711-8.

167

91. Lampson, M.A., et al., Correcting improper chromosome-spindle attachments during cell division. Nat Cell Biol, 2004. 6(3): p. 232-7.

92. Knowlton, A.L., W. Lan, and P.T. Stukenberg, Aurora B is enriched at merotelic attachment sites, where it regulates MCAK. Curr Biol, 2006. 16(17): p. 1705-10.

93. Waters, J.C., et al., Localization of Mad2 to kinetochores depends on microtubule attachment, not tension. J Cell Biol, 1998. 141(5): p. 1181-91.

94. Bolton, M.A., et al., Aurora B kinase exists in a complex with survivin and INCENP and its kinase activity is stimulated by survivin binding and phosphorylation. Mol Biol Cell, 2002. 13(9): p. 3064-77.

95. Reing, J.E., S.M. Gollin, and W.S. Saunders, The occurrence of chromosome segregational defects is an intrinsic and heritable property of oral squamous cell carcinoma cell lines. Cancer Genet Cytogenet, 2004. 150(1): p. 57-61.

96. Saunders, W.S., et al., Chromosomal instability and cytoskeletal defects in oral cancer cells. Proc Natl Acad Sci U S A, 2000. 97(1): p. 303-8.

97. Wolf, K.W., M. Mentzel, and A.S. Mendoza, DNA-containing cytoplasmic bridges in a human breast cancer cell line, MX-1: morphological markers of a highly mobile cell type? J Submicrosc Cytol Pathol, 1996. 28(3): p. 369-73.

98. Bakhoum, S.F., G. Genovese, and D.A. Compton, Deviant kinetochore microtubule dynamics underlie chromosomal instability. Curr Biol, 2009. 19(22): p. 1937-42.

99. Bakhoum, S.F., et al., Genome stability is ensured by temporal control of kinetochore-microtubule dynamics. Nat Cell Biol, 2009. 11(1): p. 27-35.

100. Vleugel, M., et al., Evolution and function of the mitotic checkpoint. Dev Cell, 2012. 23(2): p. 239-50.

101. Shepperd, L.A., et al., Phosphodependent recruitment of Bub1 and Bub3 to Spc7/KNL1 by Mph1 kinase maintains the spindle checkpoint. Curr Biol, 2012. 22(10): p. 891-9.

102. London, N., et al., Phosphoregulation of Spc105 by Mps1 and PP1 regulates Bub1 localization to kinetochores. Curr Biol, 2012. 22(10): p. 900-6.

103. Lara-Gonzalez, P., F.G. Westhorpe, and S.S. Taylor, The spindle assembly checkpoint. Curr Biol, 2012. 22(22): p. R966-80.

104. London, N. and S. Biggins, Signalling dynamics in the spindle checkpoint response. Nat Rev Mol Cell Biol, 2014. 15(11): p. 736-47.

105. Foley, E.A. and T.M. Kapoor, Microtubule attachment and spindle assembly checkpoint signalling at the kinetochore. Nat Rev Mol Cell Biol, 2013. 14(1): p. 25-37.

168

106. Musacchio, A., The Molecular Biology of Spindle Assembly Checkpoint Signaling Dynamics. Curr Biol, 2015. 25(20): p. R1002-18.

107. Musacchio, A. and E.D. Salmon, The spindle-assembly checkpoint in space and time. Nat Rev Mol Cell Biol, 2007. 8(5): p. 379-93.

108. Howell, B.J., et al., Cytoplasmic dynein/dynactin drives kinetochore protein transport to the spindle poles and has a role in mitotic spindle checkpoint inactivation. J Cell Biol, 2001. 155(7): p. 1159-72.

109. Barisic, M. and S. Geley, Spindly switch controls anaphase: spindly and RZZ functions in chromosome attachment and mitotic checkpoint control. Cell Cycle, 2011. 10(3): p. 449-56.

110. Gassmann, R., et al., Removal of Spindly from microtubule-attached kinetochores controls spindle checkpoint silencing in human cells. Genes Dev, 2010. 24(9): p. 957-71.

111. Kops, G.J. and J.V. Shah, Connecting up and clearing out: how kinetochore attachment silences the spindle assembly checkpoint. Chromosoma, 2012. 121(5): p. 509-25.

112. Pinsky, B.A., C.R. Nelson, and S. Biggins, Protein phosphatase 1 regulates exit from the spindle checkpoint in budding yeast. Curr Biol, 2009. 19(14): p. 1182-7.

113. Liu, D., et al., Regulated targeting of protein phosphatase 1 to the outer kinetochore by KNL1 opposes Aurora B kinase. J Cell Biol, 2010. 188(6): p. 809- 20.

114. Hanks, S., et al., Constitutional aneuploidy and cancer predisposition caused by biallelic mutations in BUB1B. Nat Genet, 2004. 36(11): p. 1159-61.

115. Suijkerbuijk, S.J. and G.J. Kops, Preventing aneuploidy: the contribution of mitotic checkpoint proteins. Biochim Biophys Acta, 2008. 1786(1): p. 24-31.

116. Suijkerbuijk, S.J., et al., Molecular causes for BUBR1 dysfunction in the human cancer predisposition syndrome mosaic variegated aneuploidy. Cancer Res, 2010. 70(12): p. 4891-900.

117. Yen, T.J. and G.D. Kao, Mitotic checkpoint, aneuploidy and cancer. Adv Exp Med Biol, 2005. 570: p. 477-99.

118. Cahill, D.P., et al., Mutations of mitotic checkpoint genes in human cancers. Nature, 1998. 392(6673): p. 300-3.

119. Kim, H.S., et al., Frequent mutations of human Mad2, but not Bub1, in gastric cancers cause defective mitotic spindle checkpoint. Mutat Res, 2005. 578(1-2): p. 187-201.

120. Michel, L.S., et al., MAD2 haplo-insufficiency causes premature anaphase and chromosome instability in mammalian cells. Nature, 2001. 409(6818): p. 355-9.

169

121. Ouyang, B., et al., Mechanisms of aneuploidy in thyroid cancer cell lines and tissues: evidence for mitotic checkpoint dysfunction without mutations in BUB1 and BUBR1. Clin Endocrinol (Oxf), 2002. 56(3): p. 341-50.

122. Myrie, K.A., et al., Mutation and expression analysis of human BUB1 and BUB1B in aneuploid breast cancer cell lines. Cancer Lett, 2000. 152(2): p. 193-9.

123. Lin, S.F., et al., Expression of hBUB1 in acute myeloid leukemia. Leuk Lymphoma, 2002. 43(2): p. 385-91.

124. Yamaguchi, K., et al., Mutation analysis of hBUB1 in aneuploid HNSCC and lung cancer cell lines. Cancer Lett, 1999. 139(2): p. 183-7.

125. Shichiri, M., et al., Genetic and epigenetic inactivation of mitotic checkpoint genes hBUB1 and hBUBR1 and their relationship to survival. Cancer Res, 2002. 62(1): p. 13-7.

126. Ando, K., et al., High expression of BUBR1 is one of the factors for inducing DNA aneuploidy and progression in gastric cancer. Cancer Sci, 2010. 101(3): p. 639- 45.

127. Yamamoto, Y., et al., Overexpression of BUBR1 is associated with chromosomal instability in bladder cancer. Cancer Genet Cytogenet, 2007. 174(1): p. 42-7.

128. Pinto, M., et al., Overexpression of the mitotic checkpoint genes BUB1 and BUBR1 is associated with genomic complexity in clear cell kidney carcinomas. Cell Oncol, 2008. 30(5): p. 389-95.

129. Bieche, I., et al., Expression analysis of mitotic spindle checkpoint genes in breast carcinoma: role of NDC80/HEC1 in early breast tumorigenicity, and a two- gene signature for aneuploidy. Mol Cancer, 2011. 10: p. 23.

130. Shigeishi, H., et al., Correlation of human Bub1 expression with tumor- proliferating activity in salivary gland tumors. Oncol Rep, 2006. 15(4): p. 933-8.

131. Grabsch, H., et al., Overexpression of the mitotic checkpoint genes BUB1, BUBR1, and BUB3 in gastric cancer--association with tumour cell proliferation. J Pathol, 2003. 200(1): p. 16-22.

132. Abal, M., et al., APC inactivation associates with abnormal mitosis completion and concomitant BUB1B/MAD2L1 up-regulation. Gastroenterology, 2007. 132(7): p. 2448-58.

133. Wada, N., et al., Overexpression of the mitotic spindle assembly checkpoint genes hBUB1, hBUBR1 and hMAD2 in thyroid carcinomas with aggressive nature. Anticancer Res, 2008. 28(1A): p. 139-44.

134. Ko, Y.H., et al., Expression of mitotic checkpoint proteins BUB1B and MAD2L1 in salivary duct carcinomas. J Oral Pathol Med, 2010. 39(4): p. 349-55.

170

135. Kato, T., et al., Overexpression of MAD2 predicts clinical outcome in primary lung cancer patients. Lung Cancer, 2011. 74(1): p. 124-31.

136. Knudson, A.G., Jr., Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A, 1971. 68(4): p. 820-3.

137. Richardson, A.L., et al., X chromosomal abnormalities in basal-like human breast cancer. Cancer Cell, 2006. 9(2): p. 121-32.

138. Jazaeri, A.A., et al., Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers. J Natl Cancer Inst, 2002. 94(13): p. 990-1000.

139. Spatz, A., C. Borg, and J. Feunteun, X-chromosome genetics and human cancer. Nat Rev Cancer, 2004. 4(8): p. 617-29.

140. Carrel, L. and H.F. Willard, X-inactivation profile reveals extensive variability in X- linked gene expression in females. Nature, 2005. 434(7031): p. 400-4.

141. Lee, A.J., et al., Chromosomal instability confers intrinsic multidrug resistance. Cancer Res, 2011. 71(5): p. 1858-70.

142. Swanton, C., et al., Chromosomal instability determines taxane response. Proc Natl Acad Sci U S A, 2009. 106(21): p. 8671-6.

143. Li, R., et al., Chromosomal alterations cause the high rates and wide ranges of drug resistance in cancer cells. Cancer Genet Cytogenet, 2005. 163(1): p. 44-56.

144. Bouchet, B.P., et al., Paclitaxel resistance in untransformed human mammary epithelial cells is associated with an aneuploidy-prone phenotype. Br J Cancer, 2007. 97(9): p. 1218-24.

145. Choi, C.M., et al., Chromosomal instability is a risk factor for poor prognosis of adenocarcinoma of the lung: Fluorescence in situ hybridization analysis of paraffin-embedded tissue from Korean patients. Lung Cancer, 2009. 64(1): p. 66- 70.

146. Nakamura, H., et al., Chromosomal instability detected by fluorescence in situ hybridization in surgical specimens of non-small cell lung cancer is associated with poor survival. Clin Cancer Res, 2003. 9(6): p. 2294-9.

147. Yoo, J.W., et al., The relationship between the presence of chromosomal instability and prognosis of squamous cell carcinoma of the lung: fluorescence in situ hybridization analysis of paraffin-embedded tissue from 47 Korean patients. J Korean Med Sci, 2010. 25(6): p. 863-7.

148. Carter, S.L., et al., A signature of chromosomal instability inferred from gene expression profiles predicts clinical outcome in multiple human cancers. Nat Genet, 2006. 38(9): p. 1043-8.

171

149. Takami, S., et al., Chromosomal instability detected by fluorescence in situ hybridization in Japanese breast cancer patients. Clin Chim Acta, 2001. 308(1-2): p. 127-31.

150. Roylance, R., et al., Relationship of extreme chromosomal instability with long- term survival in a retrospective analysis of primary breast cancer. Cancer Epidemiol Biomarkers Prev, 2011. 20(10): p. 2183-94.

151. Mettu, R.K., et al., A 12-gene genomic instability signature predicts clinical outcomes in multiple cancer types. Int J Biol Markers, 2010. 25(4): p. 219-28.

152. Bakhoum, S.F., et al., Chromosomal instability substantiates poor prognosis in patients with diffuse large B-cell lymphoma. Clin Cancer Res, 2011. 17(24): p. 7704-11.

153. Sato, H., et al., Prognostic utility of chromosomal instability detected by fluorescence in situ hybridization in fine-needle aspirates from oral squamous cell carcinomas. BMC Cancer, 2010. 10: p. 182.

154. Darzynkiewicz, Z., H.D. Halicka, and H. Zhao, Analysis of cellular DNA content by flow and laser scanning cytometry. Adv Exp Med Biol, 2010. 676: p. 137-47.

155. Fiegler, H., et al., High resolution array-CGH analysis of single cells. Nucleic Acids Res, 2007. 35(3): p. e15.

156. Karabay, A. and R.A. Walker, Identification of microtubule binding sites in the Ncd tail domain. Biochemistry, 1999. 38(6): p. 1838-49.

157. Kwon, M., et al., Mechanisms to suppress multipolar divisions in cancer cells with extra centrosomes. Genes Dev, 2008. 22(16): p. 2189-203.

158. Janssen, A., G.J. Kops, and R.H. Medema, Elevating the frequency of chromosome mis-segregation as a strategy to kill tumor cells. Proc Natl Acad Sci U S A, 2009. 106(45): p. 19108-13.

159. Hanahan, D. and R.A. Weinberg, The hallmarks of cancer. Cell, 2000. 100(1): p. 57-70.

160. Thompson, S.L., S.F. Bakhoum, and D.A. Compton, Mechanisms of chromosomal instability. Curr Biol, 2010. 20(6): p. R285-95.

161. Tighe, A., et al., Aneuploid colon cancer cells have a robust spindle checkpoint. EMBO Rep, 2001. 2(7): p. 609-14.

162. Green, R.A. and K.B. Kaplan, Chromosome instability in colorectal tumor cells is associated with defects in microtubule plus-end attachments caused by a dominant mutation in APC. J Cell Biol, 2003. 163(5): p. 949-61.

163. Fodde, R., et al., Mutations in the APC tumour suppressor gene cause chromosomal instability. Nat Cell Biol, 2001. 3(4): p. 433-8.

172

164. Draviam, V.M., et al., Misorientation and reduced stretching of aligned sister kinetochores promote chromosome missegregation in EB1- or APC-depleted cells. EMBO J, 2006. 25(12): p. 2814-27.

165. Aoki, K., et al., Chromosomal instability by beta-catenin/TCF transcription in APC or beta-catenin mutant cells. Oncogene, 2007. 26(24): p. 3511-20.

166. Hadjihannas, M.V., et al., Aberrant Wnt/beta-catenin signaling can induce chromosomal instability in colon cancer. Proc Natl Acad Sci U S A, 2006. 103(28): p. 10747-52.

167. Lentini, L., et al., Simultaneous Aurora-A/STK15 overexpression and centrosome amplification induce chromosomal instability in tumour cells with a MIN phenotype. BMC Cancer, 2007. 7: p. 212.

168. Nishida, N., et al., High copy amplification of the Aurora-A gene is associated with chromosomal instability phenotype in human colorectal cancers. Cancer Biol Ther, 2007. 6(4): p. 525-33.

169. 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, 2003. 161(2): p. 281-94.

170. Joukov, V., et al., The BRCA1/BARD1 heterodimer modulates ran-dependent mitotic spindle assembly. Cell, 2006. 127(3): p. 539-52.

171. Jeganathan, K., et al., Bub1 mediates cell death in response to chromosome missegregation and acts to suppress spontaneous tumorigenesis. J Cell Biol, 2007. 179(2): p. 255-67.

172. Kalitsis, P., et al., Bub3 gene disruption in mice reveals essential mitotic spindle checkpoint function during early embryogenesis. Genes Dev, 2000. 14(18): p. 2277-82.

173. Babu, J.R., et al., Rae1 is an essential mitotic checkpoint regulator that cooperates with Bub3 to prevent chromosome missegregation. J Cell Biol, 2003. 160(3): p. 341-53.

174. Matsuura, S., et al., Monoallelic BUB1B mutations and defective mitotic-spindle checkpoint in seven families with premature chromatid separation (PCS) syndrome. Am J Med Genet A, 2006. 140(4): p. 358-67.

175. Wang, Q., et al., BUBR1 deficiency results in abnormal megakaryopoiesis. Blood, 2004. 103(4): p. 1278-85.

176. Baker, D.J., et al., BubR1 insufficiency causes early onset of aging-associated phenotypes and infertility in mice. Nat Genet, 2004. 36(7): p. 744-9.

177. Liu, Y., et al., CAML loss causes anaphase failure and chromosome missegregation. Cell Cycle, 2009. 8(6): p. 940-9.

173

178. Rajagopalan, H., et al., Inactivation of hCDC4 can cause chromosomal instability. Nature, 2004. 428(6978): p. 77-81.

179. Li, M., et al., Loss of spindle assembly checkpoint-mediated inhibition of Cdc20 promotes tumorigenesis in mice. J Cell Biol, 2009. 185(6): p. 983-94.

180. Weaver, B.A., et al., Aneuploidy acts both oncogenically and as a tumor suppressor. Cancer Cell, 2007. 11(1): p. 25-36.

181. Maffini, S., et al., Motor-independent targeting of CLASPs to kinetochores by CENP-E promotes microtubule turnover and poleward flux. Curr Biol, 2009. 19(18): p. 1566-72.

182. Yao, X., et al., CENP-E forms a link between attachment of spindle microtubules to kinetochores and the mitotic checkpoint. Nat Cell Biol, 2000. 2(8): p. 484-91.

183. Weaver, B.A., et al., Centromere-associated protein-E is essential for the mammalian mitotic checkpoint to prevent aneuploidy due to single chromosome loss. J Cell Biol, 2003. 162(4): p. 551-63.

184. Holt, S.V., et al., Silencing Cenp-F weakens centromeric cohesion, prevents chromosome alignment and activates the spindle checkpoint. J Cell Sci, 2005. 118(Pt 20): p. 4889-900.

185. Laoukili, J., et al., FoxM1 is required for execution of the mitotic programme and chromosome stability. Nat Cell Biol, 2005. 7(2): p. 126-36.

186. Tomonaga, T., et al., Centromere protein H is up-regulated in primary human colorectal cancer and its overexpression induces aneuploidy. Cancer Res, 2005. 65(11): p. 4683-9.

187. Pereira, A.L., et al., Mammalian CLASP1 and CLASP2 cooperate to ensure mitotic fidelity by regulating spindle and kinetochore function. Mol Biol Cell, 2006. 17(10): p. 4526-42.

188. Spruck, C.H., K.A. Won, and S.I. Reed, Deregulated cyclin E induces chromosome instability. Nature, 1999. 401(6750): p. 297-300.

189. Cheng, X., et al., ECRG2 disruption leads to centrosome amplification and spindle checkpoint defects contributing chromosome instability. J Biol Chem, 2008. 283(9): p. 5888-98.

190. Castillo, A., et al., Overexpression of Eg5 causes genomic instability and tumor formation in mice. Cancer Res, 2007. 67(21): p. 10138-47.

191. DeLuca, J.G., et al., Kinetochore microtubule dynamics and attachment stability are regulated by Hec1. Cell, 2006. 127(5): p. 969-82.

192. Diaz-Rodriguez, E., et al., Hec1 overexpression hyperactivates the mitotic checkpoint and induces tumor formation in vivo. Proc Natl Acad Sci U S A, 2008. 105(43): p. 16719-24.

174

193. Du, J., et al., The mitotic checkpoint kinase NEK2A regulates kinetochore microtubule attachment stability. Oncogene, 2008. 27(29): p. 4107-14.

194. Wu, G., et al., Hice1, a novel microtubule-associated protein required for maintenance of spindle integrity and chromosomal stability in human cells. Mol Cell Biol, 2008. 28(11): p. 3652-62.

195. Wang, X., et al., Id-1 promotes chromosomal instability through modification of APC/C activity during mitosis in response to microtubule disruption. Oncogene, 2008. 27(32): p. 4456-66.

196. Ganem, N.J., K. Upton, and D.A. Compton, Efficient mitosis in human cells lacking poleward microtubule flux. Curr Biol, 2005. 15(20): p. 1827-32.

197. Mazumdar, M., et al., Tumor formation via loss of a molecular motor protein. Curr Biol, 2006. 16(15): p. 1559-64.

198. Hagos, E.G., et al., Mouse embryonic fibroblasts null for the Kruppel-like factor 4 gene are genetically unstable. Oncogene, 2009. 28(9): p. 1197-205.

199. Iwanaga, Y., et al., Heterozygous deletion of mitotic arrest-deficient protein 1 (MAD1) increases the incidence of tumors in mice. Cancer Res, 2007. 67(1): p. 160-6.

200. Dobles, M., et al., Chromosome missegregation and apoptosis in mice lacking the mitotic checkpoint protein Mad2. Cell, 2000. 101(6): p. 635-45.

201. Sotillo, R., et al., Mad2 overexpression promotes aneuploidy and tumorigenesis in mice. Cancer Cell, 2007. 11(1): p. 9-23.

202. Hernando, E., et al., Rb inactivation promotes genomic instability by uncoupling cell cycle progression from mitotic control. Nature, 2004. 430(7001): p. 797-802.

203. Burds, A.A., A.S. Lutum, and P.K. Sorger, Generating chromosome instability through the simultaneous deletion of Mad2 and p53. Proc Natl Acad Sci U S A, 2005. 102(32): p. 11296-301.

204. Maney, T., et al., Mitotic centromere-associated kinesin is important for anaphase chromosome segregation. J Cell Biol, 1998. 142(3): p. 787-801.

205. Kline-Smith, S.L., et al., Depletion of centromeric MCAK leads to chromosome congression and segregation defects due to improper kinetochore attachments. Mol Biol Cell, 2004. 15(3): p. 1146-59.

206. Kasiappan, R., et al., Loss of p53 and MCT-1 overexpression synergistically promote chromosome instability and tumorigenicity. Mol Cancer Res, 2009. 7(4): p. 536-48.

207. Wang, P., et al., Elevated Mdm2 expression induces chromosomal instability and confers a survival and growth advantage to B cells. Oncogene, 2008. 27(11): p. 1590-8.

175

208. Matijasevic, Z., et al., MdmX regulates transformation and chromosomal stability in p53-deficient cells. Cell Cycle, 2008. 7(19): p. 2967-73.

209. Jelluma, N., et al., Mps1 phosphorylates Borealin to control Aurora B activity and chromosome alignment. Cell, 2008. 132(2): p. 233-46.

210. Montembault, E., et al., PRP4 is a spindle assembly checkpoint protein required for MPS1, MAD1, and MAD2 localization to the kinetochores. J Cell Biol, 2007. 179(4): p. 601-9.

211. Tedeschi, A., et al., RANBP1 localizes a subset of mitotic regulatory factors on spindle microtubules and regulates chromosome segregation in human cells. J Cell Sci, 2007. 120(Pt 21): p. 3748-61.

212. Guardavaccaro, D., et al., Control of chromosome stability by the beta-TrCP- REST-Mad2 axis. Nature, 2008. 452(7185): p. 365-9.

213. Jallepalli, P.V., et al., Securin is required for chromosomal stability in human cells. Cell, 2001. 105(4): p. 445-57.

214. Yu, R., et al., Overexpressed pituitary tumor-transforming gene causes aneuploidy in live human cells. Endocrinology, 2003. 144(11): p. 4991-8.

215. Wirth, K.G., et al., Separase: a universal trigger for sister chromatid disjunction but not chromosome cycle progression. J Cell Biol, 2006. 172(6): p. 847-60.

216. Zhang, N., et al., Overexpression of Separase induces aneuploidy and mammary tumorigenesis. Proc Natl Acad Sci U S A, 2008. 105(35): p. 13033-8.

217. Iwaizumi, M., et al., Human Sgo1 downregulation leads to chromosomal instability in colorectal cancer. Gut, 2009. 58(2): p. 249-60.

218. Huang, H., et al., Tripin/hSgo2 recruits MCAK to the inner centromere to correct defective kinetochore attachments. J Cell Biol, 2007. 177(3): p. 413-24.

219. Hong, K.U., et al., TMAP/CKAP2 is essential for proper chromosome segregation. Cell Cycle, 2009. 8(2): p. 314-24.

220. Cimini, D., et al., Topoisomerase II inhibition in mitosis produces numerical and structural chromosomal aberrations in human fibroblasts. Cytogenet Cell Genet, 1997. 76(1-2): p. 61-7.

221. Thoma, C.R., et al., VHL loss causes spindle misorientation and chromosome instability. Nat Cell Biol, 2009. 11(8): p. 994-1001.

222. Kronenwett, U., et al., Improved grading of breast adenocarcinomas based on genomic instability. Cancer Res, 2004. 64(3): p. 904-9.

223. Smid, M., et al., Patterns and incidence of chromosomal instability and their prognostic relevance in breast cancer subtypes. Breast Cancer Res Treat, 2011. 128(1): p. 23-30.

176

224. Habermann, J.K., et al., The gene expression signature of genomic instability in breast cancer is an independent predictor of clinical outcome. Int J Cancer, 2009. 124(7): p. 1552-64.

225. Heilig, C.E., et al., Chromosomal instability correlates with poor outcome in patients with myelodysplastic syndromes irrespectively of the cytogenetic risk group. J Cell Mol Med, 2010. 14(4): p. 895-902.

226. Jonkers, Y.M., et al., Chromosomal instability predicts metastatic disease in patients with insulinomas. Endocr Relat Cancer, 2005. 12(2): p. 435-47.

227. Walther, A., R. Houlston, and I. Tomlinson, Association between chromosomal instability and prognosis in colorectal cancer: a meta-analysis. Gut, 2008. 57(7): p. 941-50.

228. Murayama-Hosokawa, S., et al., Genome-wide single-nucleotide polymorphism arrays in endometrial carcinomas associate extensive chromosomal instability with poor prognosis and unveil frequent chromosomal imbalances involved in the PI3-kinase pathway. Oncogene, 2010. 29(13): p. 1897-908.

229. Nakagawa, Y., et al., Chromosomal and genetic imbalances in synovial sarcoma detected by conventional and microarray comparative genomic hybridization. J Cancer Res Clin Oncol, 2006. 132(7): p. 444-50.

230. Bergshoeff, V.E., et al., Chromosome instability in resection margins predicts recurrence of oral squamous cell carcinoma. J Pathol, 2008. 215(3): p. 347-8.

231. McKinley, K.L. and I.M. Cheeseman, The molecular basis for centromere identity and function. Nat Rev Mol Cell Biol, 2016. 17(1): p. 16-29.

232. Choo, K.H., Domain organization at the centromere and neocentromere. Dev Cell, 2001. 1(2): p. 165-77.

233. Furuyama, S. and S. Biggins, Centromere identity is specified by a single centromeric nucleosome in budding yeast. Proc Natl Acad Sci U S A, 2007. 104(37): p. 14706-11.

234. Steiner, F.A. and S. Henikoff, Diversity in the organization of centromeric chromatin. Curr Opin Genet Dev, 2015. 31: p. 28-35.

235. Pluta, A.F., et al., The centromere: hub of chromosomal activities. Science, 1995. 270(5242): p. 1591-4.

236. Meraldi, P., et al., Phylogenetic and structural analysis of centromeric DNA and kinetochore proteins. Genome Biol, 2006. 7(3): p. R23.

237. Plohl, M., N. Mestrovic, and B. Mravinac, Centromere identity from the DNA point of view. Chromosoma, 2014. 123(4): p. 313-25.

238. Willard, H.F., Centromeres of mammalian chromosomes. Trends Genet, 1990. 6(12): p. 410-6.

177

239. Malik, H.S. and S. Henikoff, Major evolutionary transitions in centromere complexity. Cell, 2009. 138(6): p. 1067-82.

240. Cleveland, D.W., Y. Mao, and K.F. Sullivan, Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell, 2003. 112(4): p. 407-21.

241. Cheng, Z., et al., Functional rice centromeres are marked by a satellite repeat and a centromere-specific retrotransposon. Plant Cell, 2002. 14(8): p. 1691-704.

242. Yan, H. and J. Jiang, Rice as a model for centromere and heterochromatin research. Chromosome Res, 2007. 15(1): p. 77-84.

243. Wang, G., X. Zhang, and W. Jin, An overview of plant centromeres. J Genet Genomics, 2009. 36(9): p. 529-37.

244. Sun, X., J. Wahlstrom, and G. Karpen, Molecular structure of a functional Drosophila centromere. Cell, 1997. 91(7): p. 1007-19.

245. Shang, W.H., et al., Chickens possess centromeres with both extended tandem repeats and short non-tandem-repetitive sequences. Genome Res, 2010. 20(9): p. 1219-28.

246. Guenatri, M., et al., Mouse centric and pericentric satellite repeats form distinct functional heterochromatin. J Cell Biol, 2004. 166(4): p. 493-505.

247. Choo, K.H., Centromere DNA dynamics: latent centromeres and neocentromere formation. Am J Hum Genet, 1997. 61(6): p. 1225-33.

248. du Sart, D., et al., A functional neo-centromere formed through activation of a latent human centromere and consisting of non-alpha-satellite DNA. Nat Genet, 1997. 16(2): p. 144-53.

249. Hayden, K.E., et al., Sequences associated with centromere competency in the human genome. Mol Cell Biol, 2013. 33(4): p. 763-72.

250. Benson, D.A., et al., GenBank. Nucleic Acids Res, 2014. 42(Database issue): p. D32-7.

251. Willard, H.F., Chromosome-specific organization of human alpha satellite DNA. Am J Hum Genet, 1985. 37(3): p. 524-32.

252. Alexandrov, I., et al., Alpha-satellite DNA of primates: old and new families. Chromosoma, 2001. 110(4): p. 253-66.

253. Koga, A., et al., Evolutionary origin of higher-order repeat structure in alpha- satellite DNA of primate centromeres. DNA Res, 2014. 21(4): p. 407-15.

254. Rudd, M.K. and H.F. Willard, Analysis of the centromeric regions of the human genome assembly. Trends Genet, 2004. 20(11): p. 529-33.

255. Schueler, M.G., et al., Genomic and genetic definition of a functional human centromere. Science, 2001. 294(5540): p. 109-15.

178

256. Grewal, S.I. and S. Jia, Heterochromatin revisited. Nat Rev Genet, 2007. 8(1): p. 35-46.

257. Dejardin, J., Switching between Epigenetic States at Pericentromeric Heterochromatin. Trends Genet, 2015. 31(11): p. 661-72.

258. Black, B.E. and E.A. Bassett, The histone variant CENP-A and centromere specification. Curr Opin Cell Biol, 2008. 20(1): p. 91-100.

259. Niikura, Y., et al., CENP-A K124 Ubiquitylation Is Required for CENP-A Deposition at the Centromere. Dev Cell, 2015. 32(5): p. 589-603.

260. Kersey, P.J., et al., Ensembl Genomes 2016: more genomes, more complexity. Nucleic Acids Res, 2016. 44(D1): p. D574-80.

261. Gerhard, D.S., et al., The status, quality, and expansion of the NIH full-length cDNA project: the Mammalian Gene Collection (MGC). Genome Res, 2004. 14(10B): p. 2121-7.

262. Sekulic, N., et al., The structure of (CENP-A-H4)(2) reveals physical features that mark centromeres. Nature, 2010. 467(7313): p. 347-51.

263. Sullivan, K.F., M. Hechenberger, and K. Masri, Human CENP-A contains a histone H3 related histone fold domain that is required for targeting to the centromere. J Cell Biol, 1994. 127(3): p. 581-92.

264. Black, B.E., et al., Structural determinants for generating centromeric chromatin. Nature, 2004. 430(6999): p. 578-82.

265. Stellfox, M.E., A.O. Bailey, and D.R. Foltz, Putting CENP-A in its place. Cell Mol Life Sci, 2013. 70(3): p. 387-406.

266. Bailey, A.O., et al., Identification of the Post-translational Modifications Present in Centromeric Chromatin. Mol Cell Proteomics, 2016. 15(3): p. 918-31.

267. Furuyama, T. and S. Henikoff, Centromeric nucleosomes induce positive DNA supercoils. Cell, 2009. 138(1): p. 104-13.

268. Mizuguchi, G., et al., Nonhistone Scm3 and histones CenH3-H4 assemble the core of centromere-specific nucleosomes. Cell, 2007. 129(6): p. 1153-64.

269. Dalal, Y., et al., Structure, dynamics, and evolution of centromeric nucleosomes. Proc Natl Acad Sci U S A, 2007. 104(41): p. 15974-81.

270. Dalal, Y., et al., Tetrameric structure of centromeric nucleosomes in interphase Drosophila cells. PLoS Biol, 2007. 5(8): p. e218.

271. Dimitriadis, E.K., et al., Tetrameric organization of vertebrate centromeric nucleosomes. Proc Natl Acad Sci U S A, 2010. 107(47): p. 20317-22.

272. Frado, L.L., et al., Mapping of histone H5 sites on nucleosomes using immunoelectron microscopy. J Biol Chem, 1983. 258(19): p. 11984-90.

179

273. Tachiwana, H., et al., Crystal structure of the human centromeric nucleosome containing CENP-A. Nature, 2011. 476(7359): p. 232-5.

274. Hasson, D., et al., The octamer is the major form of CENP-A nucleosomes at human centromeres. Nat Struct Mol Biol, 2013. 20(6): p. 687-95.

275. Padeganeh, A., et al., Octameric CENP-A nucleosomes are present at human centromeres throughout the cell cycle. Curr Biol, 2013. 23(9): p. 764-9.

276. Jain, A., et al., Probing cellular protein complexes using single-molecule pull- down. Nature, 2011. 473(7348): p. 484-8.

277. Bui, M., et al., Cell-cycle-dependent structural transitions in the human CENP-A nucleosome in vivo. Cell, 2012. 150(2): p. 317-26.

278. Shivaraju, M., et al., Cell-cycle-coupled structural oscillation of centromeric nucleosomes in yeast. Cell, 2012. 150(2): p. 304-16.

279. Geiss, C.P., et al., CENP-A arrays are more condensed than canonical arrays at low ionic strength. Biophys J, 2014. 106(4): p. 875-82.

280. Panchenko, T., et al., Replacement of histone H3 with CENP-A directs global nucleosome array condensation and loosening of nucleosome superhelical termini. Proc Natl Acad Sci U S A, 2011. 108(40): p. 16588-93.

281. Bailey, A.O., et al., Posttranslational modification of CENP-A influences the conformation of centromeric chromatin. Proc Natl Acad Sci U S A, 2013. 110(29): p. 11827-32.

282. Goutte-Gattat, D., et al., Phosphorylation of the CENP-A amino-terminus in mitotic centromeric chromatin is required for kinetochore function. Proc Natl Acad Sci U S A, 2013. 110(21): p. 8579-84.

283. Krude, T. and C. Keller, Chromatin assembly during S phase: contributions from histone deposition, DNA replication and the cell division cycle. Cell Mol Life Sci, 2001. 58(5-6): p. 665-72.

284. Black, B.E., et al., Centromere identity maintained by nucleosomes assembled with histone H3 containing the CENP-A targeting domain. Mol Cell, 2007. 25(2): p. 309-22.

285. Fukagawa, T. and W.C. Earnshaw, The centromere: chromatin foundation for the kinetochore machinery. Dev Cell, 2014. 30(5): p. 496-508.

286. Wisniewski, J., et al., Imaging the fate of histone Cse4 reveals de novo replacement in S phase and subsequent stable residence at centromeres. Elife, 2014. 3: p. e02203.

287. Tomonaga, T., et al., Overexpression and mistargeting of centromere protein-A in human primary colorectal cancer. Cancer research, 2003. 63(13): p. 3511-6.

180

288. McGovern, S.L., et al., Centromere protein-A, an essential centromere protein, is a prognostic marker for relapse in estrogen receptor-positive breast cancer. Breast Cancer Res, 2012. 14(3): p. R72.

289. Kato, T., et al., Activation of Holliday junction recognizing protein involved in the chromosomal stability and immortality of cancer cells. Cancer Res, 2007. 67(18): p. 8544-53.

290. Foltz, D.R., et al., Centromere-specific assembly of CENP-a nucleosomes is mediated by HJURP. Cell, 2009. 137(3): p. 472-84.

291. Shuaib, M., et al., HJURP binds CENP-A via a highly conserved N-terminal domain and mediates its deposition at centromeres. Proc Natl Acad Sci U S A, 2010. 107(4): p. 1349-54.

292. Barnhart, M.C., et al., HJURP is a CENP-A chromatin assembly factor sufficient to form a functional de novo kinetochore. J Cell Biol, 2011. 194(2): p. 229-43.

293. Mishra, P.K., et al., Misregulation of Scm3p/HJURP causes chromosome instability in Saccharomyces cerevisiae and human cells. PLoS Genet, 2011. 7(9): p. e1002303.

294. Dunleavy, E.M., et al., HJURP is a cell-cycle-dependent maintenance and deposition factor of CENP-A at centromeres. Cell, 2009. 137(3): p. 485-97.

295. Muller, S., et al., Phosphorylation and DNA binding of HJURP determine its centromeric recruitment and function in CenH3(CENP-A) loading. Cell Rep, 2014. 8(1): p. 190-203.

296. Fujita, Y., et al., Priming of centromere for CENP-A recruitment by human hMis18alpha, hMis18beta, and M18BP1. Dev Cell, 2007. 12(1): p. 17-30.

297. Hayashi, T., et al., Mis16 and Mis18 are required for CENP-A loading and histone deacetylation at centromeres. Cell, 2004. 118(6): p. 715-29.

298. Kim, I.S., et al., Roles of Mis18alpha in epigenetic regulation of centromeric chromatin and CENP-A loading. Mol Cell, 2012. 46(3): p. 260-73.

299. Silva, M.C., et al., Cdk activity couples epigenetic centromere inheritance to cell cycle progression. Dev Cell, 2012. 22(1): p. 52-63.

300. McKinley, K.L. and I.M. Cheeseman, Polo-like kinase 1 licenses CENP-A deposition at centromeres. Cell, 2014. 158(2): p. 397-411.

301. Verreault, A., et al., Nucleosome assembly by a complex of CAF-1 and acetylated histones H3/H4. Cell, 1996. 87(1): p. 95-104.

302. Ohzeki, J., et al., Breaking the HAC Barrier: histone H3K9 acetyl/methyl balance regulates CENP-A assembly. EMBO J, 2012. 31(10): p. 2391-402.

181

303. Zhang, D., C.J. Martyniuk, and V.L. Trudeau, SANTA domain: a novel conserved protein module in Eukaryota with potential involvement in chromatin regulation. Bioinformatics, 2006. 22(20): p. 2459-62.

304. Boyer, L.A., et al., Essential role for the SANT domain in the functioning of multiple chromatin remodeling enzymes. Mol Cell, 2002. 10(4): p. 935-42.

305. Dambacher, S., et al., CENP-C facilitates the recruitment of M18BP1 to centromeric chromatin. Nucleus, 2012. 3(1): p. 101-10.

306. Moree, B., et al., CENP-C recruits M18BP1 to centromeres to promote CENP-A chromatin assembly. J Cell Biol, 2011. 194(6): p. 855-71.

307. Orr, B. and C.E. Sunkel, Drosophila CENP-C is essential for centromere identity. Chromosoma, 2011. 120(1): p. 83-96.

308. Carroll, C.W., K.J. Milks, and A.F. Straight, Dual recognition of CENP-A nucleosomes is required for centromere assembly. J Cell Biol, 2010. 189(7): p. 1143-55.

309. Gascoigne, K.E., et al., Induced ectopic kinetochore assembly bypasses the requirement for CENP-A nucleosomes. Cell, 2011. 145(3): p. 410-22.

310. Shono, N., et al., CENP-C and CENP-I are key connecting factors for kinetochore and CENP-A assembly. J Cell Sci, 2015. 128(24): p. 4572-87.

311. Masumoto, H., et al., A human centromere antigen (CENP-B) interacts with a short specific sequence in alphoid DNA, a human centromeric satellite. J Cell Biol, 1989. 109(5): p. 1963-73.

312. Ikeno, M., H. Masumoto, and T. Okazaki, Distribution of CENP-B boxes reflected in CREST centromere antigenic sites on long-range alpha-satellite DNA arrays of human chromosome 21. Hum Mol Genet, 1994. 3(8): p. 1245-57.

313. Yoda, K., et al., A human centromere protein, CENP-B, has a DNA binding domain containing four potential alpha helices at the NH2 terminus, which is separable from dimerizing activity. J Cell Biol, 1992. 119(6): p. 1413-27.

314. Yoda, K., et al., In vitro assembly of the CENP-B/alpha-satellite DNA/core histone complex: CENP-B causes nucleosome positioning. Genes Cells, 1998. 3(8): p. 533-48.

315. Ohzeki, J., et al., CENP-B box is required for de novo centromere chromatin assembly on human alphoid DNA. J Cell Biol, 2002. 159(5): p. 765-75.

316. Hudson, D.F., et al., Centromere protein B null mice are mitotically and meiotically normal but have lower body and testis weights. J Cell Biol, 1998. 141(2): p. 309-19.

182

317. Voullaire, L.E., et al., A functional marker centromere with no detectable alpha- satellite, satellite III, or CENP-B protein: activation of a latent centromere? Am J Hum Genet, 1993. 52(6): p. 1153-63.

318. Earnshaw, W.C., et al., Molecular cloning of cDNA for CENP-B, the major human centromere autoantigen. J Cell Biol, 1987. 104(4): p. 817-29.

319. Fachinetti, D., et al., DNA Sequence-Specific Binding of CENP-B Enhances the Fidelity of Human Centromere Function. Dev Cell, 2015. 33(3): p. 314-27.

320. Earnshaw, W.C. and N. Rothfield, Identification of a family of human centromere proteins using autoimmune sera from patients with scleroderma. Chromosoma, 1985. 91(3-4): p. 313-21.

321. Klare, K., et al., CENP-C is a blueprint for constitutive centromere-associated network assembly within human kinetochores. J Cell Biol, 2015. 210(1): p. 11-22.

322. Screpanti, E., et al., Direct binding of Cenp-C to the Mis12 complex joins the inner and outer kinetochore. Curr Biol, 2011. 21(5): p. 391-8.

323. Schleiffer, A., et al., CENP-T proteins are conserved centromere receptors of the Ndc80 complex. Nat Cell Biol, 2012. 14(6): p. 604-13.

324. Basilico, F., et al., The pseudo GTPase CENP-M drives human kinetochore assembly. Elife, 2014. 3: p. e02978.

325. Kato, H., et al., A conserved mechanism for centromeric nucleosome recognition by centromere protein CENP-C. Science, 2013. 340(6136): p. 1110-3.

326. Wong, L.H., et al., Centromere RNA is a key component for the assembly of nucleoproteins at the nucleolus and centromere. Genome Res, 2007. 17(8): p. 1146-60.

327. Trazzi, S., et al., The C-terminal domain of CENP-C displays multiple and critical functions for mammalian centromere formation. PLoS One, 2009. 4(6): p. e5832.

328. Falk, S.J., et al., Chromosomes. CENP-C reshapes and stabilizes CENP-A nucleosomes at the centromere. Science, 2015. 348(6235): p. 699-703.

329. Hori, T., et al., CCAN makes multiple contacts with centromeric DNA to provide distinct pathways to the outer kinetochore. Cell, 2008. 135(6): p. 1039-52.

330. Amano, M., et al., The CENP-S complex is essential for the stable assembly of outer kinetochore structure. J Cell Biol, 2009. 186(2): p. 173-82.

331. Takeuchi, K., et al., The centromeric nucleosome-like CENP-T-W-S-X complex induces positive supercoils into DNA. Nucleic Acids Res, 2014. 42(3): p. 1644- 55.

332. Nishino, T., et al., CENP-T-W-S-X forms a unique centromeric chromatin structure with a histone-like fold. Cell, 2012. 148(3): p. 487-501.

183

333. Nishino, T., et al., CENP-T provides a structural platform for outer kinetochore assembly. EMBO J, 2013. 32(3): p. 424-36.

334. Sullivan, B.A. and G.H. Karpen, Centromeric chromatin exhibits a histone modification pattern that is distinct from both euchromatin and heterochromatin. Nature structural & molecular biology, 2004. 11(11): p. 1076-83.

335. Pekowska, A., et al., A unique H3K4me2 profile marks tissue-specific gene regulation. Genome Res, 2010. 20(11): p. 1493-502.

336. Bernstein, B.E., et al., Methylation of histone H3 Lys 4 in coding regions of active genes. Proc Natl Acad Sci U S A, 2002. 99(13): p. 8695-700.

337. Bergmann, J.H., et al., Epigenetic engineering shows H3K4me2 is required for HJURP targeting and CENP-A assembly on a synthetic human kinetochore. EMBO J, 2011. 30(2): p. 328-40.

338. Hori, T., et al., Histone H4 Lys 20 monomethylation of the CENP-A nucleosome is essential for kinetochore assembly. Dev Cell, 2014. 29(6): p. 740-9.

339. Kapoor-Vazirani, P. and P.M. Vertino, A dual role for the histone methyltransferase PR-SET7/SETD8 and histone H4 lysine 20 monomethylation in the local regulation of RNA polymerase II pausing. J Biol Chem, 2014. 289(11): p. 7425-37.

340. Sadeghi, L., et al., Centromeric histone H2B monoubiquitination promotes noncoding transcription and chromatin integrity. Nat Struct Mol Biol, 2014. 21(3): p. 236-43.

341. Martins, N.M., et al., Epigenetic engineering shows that a human centromere resists silencing mediated by H3K27me3/K9me3. Mol Biol Cell, 2016. 27(1): p. 177-96.

342. Nakano, M., et al., Inactivation of a human kinetochore by specific targeting of chromatin modifiers. Dev Cell, 2008. 14(4): p. 507-22.

343. Rice, J.C., et al., Histone methyltransferases direct different degrees of methylation to define distinct chromatin domains. Molecular cell, 2003. 12(6): p. 1591-8.

344. Schotta, G., et al., A silencing pathway to induce H3-K9 and H4-K20 trimethylation at constitutive heterochromatin. Genes & development, 2004. 18(11): p. 1251-62.

345. Gopalakrishnan, S., et al., DNMT3B interacts with constitutive centromere protein CENP-C to modulate DNA methylation and the histone code at centromeric regions. Human molecular genetics, 2009. 18(17): p. 3178-93.

346. Muramoto, T., et al., Methylation of H3K4 Is required for inheritance of active transcriptional states. Curr Biol, 2010. 20(5): p. 397-406.

184

347. Mohan, M., H.M. Herz, and A. Shilatifard, SnapShot: Histone lysine methylase complexes. Cell, 2012. 149(2): p. 498-498 e1.

348. Craig, J.M., et al., Analysis of mammalian proteins involved in chromatin modification reveals new metaphase centromeric proteins and distinct chromosomal distribution patterns. Hum Mol Genet, 2003. 12(23): p. 3109-21.

349. Lehnertz, B., et al., Suv39h-mediated histone H3 lysine 9 methylation directs DNA methylation to major satellite repeats at pericentric heterochromatin. Curr Biol, 2003. 13(14): p. 1192-200.

350. Ohkuni, K. and K. Kitagawa, Endogenous transcription at the centromere facilitates centromere activity in budding yeast. Curr Biol, 2011. 21(20): p. 1695- 703.

351. Choi, E.S., et al., Identification of noncoding transcripts from within CENP-A chromatin at fission yeast centromeres. J Biol Chem, 2011. 286(26): p. 23600-7.

352. Yan, H., et al., Genomic and genetic characterization of rice Cen3 reveals extensive transcription and evolutionary implications of a complex centromere. Plant Cell, 2006. 18(9): p. 2123-33.

353. Bouzinba-Segard, H., A. Guais, and C. Francastel, Accumulation of small murine minor satellite transcripts leads to impaired centromeric architecture and function. Proc Natl Acad Sci U S A, 2006. 103(23): p. 8709-14.

354. Carone, D.M., et al., A new class of retroviral and satellite encoded small RNAs emanates from mammalian centromeres. Chromosoma, 2009. 118(1): p. 113-25.

355. Chan, F.L., et al., Active transcription and essential role of RNA polymerase II at the centromere during mitosis. Proc Natl Acad Sci U S A, 2012. 109(6): p. 1979- 84.

356. Chueh, A.C., et al., LINE retrotransposon RNA is an essential structural and functional epigenetic component of a core neocentromeric chromatin. PLoS Genet, 2009. 5(1): p. e1000354.

357. Lam, A.L., et al., Human centromeric chromatin is a dynamic chromosomal domain that can spread over noncentromeric DNA. Proc Natl Acad Sci U S A, 2006. 103(11): p. 4186-91.

358. Saffery, R., et al., Transcription within a functional human centromere. Mol Cell, 2003. 12(2): p. 509-16.

359. Eissenberg, J.C. and S.C. Elgin, The HP1 protein family: getting a grip on chromatin. Curr Opin Genet Dev, 2000. 10(2): p. 204-10.

360. Scott, K.C., Transcription and ncRNAs: at the cent(rome)re of kinetochore assembly and maintenance. Chromosome Res, 2013. 21(6-7): p. 643-51.

185

361. Quenet, D. and Y. Dalal, A long non-coding RNA is required for targeting centromeric protein A to the human centromere. Elife, 2014. 3: p. e03254.

362. Okada, M., et al., CENP-H-containing complex facilitates centromere deposition of CENP-A in cooperation with FACT and CHD1. Mol Biol Cell, 2009. 20(18): p. 3986-95.

363. Nakashima, H., et al., Assembly of additional heterochromatin distinct from centromere-kinetochore chromatin is required for de novo formation of human artificial chromosome. J Cell Sci, 2005. 118(Pt 24): p. 5885-98.

364. Kouprina, N., et al., A new generation of human artificial chromosomes for functional genomics and gene therapy. Cell Mol Life Sci, 2013. 70(7): p. 1135-48.

365. Bergmann, J.H., et al., HACking the centromere chromatin code: insights from human artificial chromosomes. Chromosome Res, 2012. 20(5): p. 505-19.

366. Bergmann, J.H., et al., Epigenetic engineering: histone H3K9 acetylation is compatible with kinetochore structure and function. J Cell Sci, 2012. 125(Pt 2): p. 411-21.

367. Williams, S.K. and J.K. Tyler, Transcriptional regulation by chromatin disassembly and reassembly. Curr Opin Genet Dev, 2007. 17(2): p. 88-93.

368. Choi, E.S., et al., Factors that promote H3 chromatin integrity during transcription prevent promiscuous deposition of CENP-A(Cnp1) in fission yeast. PLoS Genet, 2012. 8(9): p. e1002985.

369. Chen, C.C., et al., Establishment of Centromeric Chromatin by the CENP-A Assembly Factor CAL1 Requires FACT-Mediated Transcription. Dev Cell, 2015. 34(1): p. 73-84.

370. Ideue, T., et al., Involvement of satellite I noncoding RNA in regulation of chromosome segregation. Genes Cells, 2014. 19(6): p. 528-38.

371. Aguilera, A. and T. Garcia-Muse, R loops: from transcription byproducts to threats to genome stability. Mol Cell, 2012. 46(2): p. 115-24.

372. Dorsett, Y., et al., MicroRNA-155 suppresses activation-induced cytidine deaminase-mediated Myc-Igh translocation. Immunity, 2008. 28(5): p. 630-8.

373. Allshire, R.C. and G.H. Karpen, Epigenetic regulation of centromeric chromatin: old dogs, new tricks? Nature reviews Genetics, 2008. 9(12): p. 923-37.

374. Zeitlin, S.G., et al., Xenopus CENP-A assembly into chromatin requires base excision repair proteins. DNA Repair (Amst), 2005. 4(7): p. 760-72.

375. Carone, D.M., et al., Hypermorphic expression of centromeric retroelement- encoded small RNAs impairs CENP-A loading. Chromosome Res, 2013. 21(1): p. 49-62.

186

376. Zhu, Q., et al., BRCA1 tumour suppression occurs via heterochromatin-mediated silencing. Nature, 2011. 477(7363): p. 179-84.

377. McKinley, K.L., et al., The CENP-L-N Complex Forms a Critical Node in an Integrated Meshwork of Interactions at the Centromere-Kinetochore Interface. Mol Cell, 2015. 60(6): p. 886-98.

378. Vissel, B. and K.H. Choo, Human alpha satellite DNA--consensus sequence and conserved regions. Nucleic Acids Res, 1987. 15(16): p. 6751-2.

379. Dynan, W.S. and R. Tjian, Isolation of transcription factors that discriminate between different promoters recognized by RNA polymerase II. Cell, 1983. 32(3): p. 669-80.

380. Dynan, W.S. and R. Tjian, The promoter-specific transcription factor Sp1 binds to upstream sequences in the SV40 early promoter. Cell, 1983. 35(1): p. 79-87.

381. Lin, S.Y., et al., Cell cycle-regulated association of E2F1 and Sp1 is related to their functional interaction. Mol Cell Biol, 1996. 16(4): p. 1668-75.

382. Wierstra, I., Sp1: emerging roles--beyond constitutive activation of TATA-less housekeeping genes. Biochemical and biophysical research communications, 2008. 372(1): p. 1-13.

383. Choy, B. and M.R. Green, Eukaryotic activators function during multiple steps of preinitiation complex assembly. Nature, 1993. 366(6455): p. 531-6.

384. Beishline, K., et al., Sp1 facilitates DNA double-strand break repair through a nontranscriptional mechanism. Mol Cell Biol, 2012. 32(18): p. 3790-9.

385. Li, L. and J.R. Davie, The role of Sp1 and Sp3 in normal and cancer cell biology. Ann Anat, 2010. 192(5): p. 275-83.

386. Black, A.R., J.D. Black, and J. Azizkhan-Clifford, Sp1 and kruppel-like factor family of transcription factors in cell growth regulation and cancer. Journal of cellular physiology, 2001. 188(2): p. 143-60.

387. Bouwman, P. and S. Philipsen, Regulation of the activity of Sp1-related transcription factors. Mol Cell Endocrinol, 2002. 195(1-2): p. 27-38.

388. Beishline, K. and J. Azizkhan-Clifford, Sp1 and the 'hallmarks of cancer'. FEBS J, 2015. 282(2): p. 224-58.

389. Suske, G., The Sp-family of transcription factors. Gene, 1999. 238(2): p. 291- 300.

390. Kingsley, C. and A. Winoto, Cloning of GT box-binding proteins: a novel Sp1 multigene family regulating T-cell receptor gene expression. Mol Cell Biol, 1992. 12(10): p. 4251-61.

187

391. Hagen, G., et al., Cloning by recognition site screening of two novel GT box binding proteins: a family of Sp1 related genes. Nucleic Acids Res, 1992. 20(21): p. 5519-25.

392. UniProt, C., UniProt: a hub for protein information. Nucleic Acids Res, 2015. 43(Database issue): p. D204-12.

393. Infantino, V., et al., Identification of a novel Sp1 splice variant as a strong transcriptional activator. Biochem Biophys Res Commun, 2011. 412(1): p. 86-91.

394. Emili, A., J. Greenblatt, and C.J. Ingles, Species-specific interaction of the glutamine-rich activation domains of Sp1 with the TATA box-binding protein. Mol Cell Biol, 1994. 14(3): p. 1582-93.

395. Chiang, C.M. and R.G. Roeder, Cloning of an intrinsic human TFIID subunit that interacts with multiple transcriptional activators. Science, 1995. 267(5197): p. 531-6.

396. De Cesare, D., et al., Transcriptional control in male germ cells: general factor TFIIA participates in CREM-dependent gene activation. Mol Endocrinol, 2003. 17(12): p. 2554-65.

397. Ryu, S., et al., The transcriptional cofactor complex CRSP is required for activity of the enhancer-binding protein Sp1. Nature, 1999. 397(6718): p. 446-50.

398. Naar, A.M., et al., Chromatin, TAFs, and a novel multiprotein coactivator are required for synergistic activation by Sp1 and SREBP-1a in vitro. Genes Dev, 1998. 12(19): p. 3020-31.

399. Hung, J.J., Y.T. Wang, and W.C. Chang, Sp1 deacetylation induced by phorbol ester recruits p300 to activate 12(S)-lipoxygenase gene transcription. Mol Cell Biol, 2006. 26(5): p. 1770-85.

400. Xiao, H., T. Hasegawa, and K. Isobe, p300 collaborates with Sp1 and Sp3 in p21(waf1/cip1) promoter activation induced by histone deacetylase inhibitor. J Biol Chem, 2000. 275(2): p. 1371-6.

401. Gizard, F., et al., Progesterone inhibits human breast cancer cell growth through transcriptional upregulation of the cyclin-dependent kinase inhibitor p27Kip1 gene. FEBS Lett, 2005. 579(25): p. 5535-41.

402. Soutoglou, E., et al., Transcription factor-dependent regulation of CBP and P/CAF histone acetyltransferase activity. EMBO J, 2001. 20(8): p. 1984-92.

403. Lagger, G., et al., The tumor suppressor p53 and histone deacetylase 1 are antagonistic regulators of the cyclin-dependent kinase inhibitor p21/WAF1/CIP1 gene. Mol Cell Biol, 2003. 23(8): p. 2669-79.

404. Gartel, A.L. and A.L. Tyner, Transcriptional regulation of the p21((WAF1/CIP1)) gene. Exp Cell Res, 1999. 246(2): p. 280-9.

188

405. Gartel, A.L. and S.K. Radhakrishnan, Lost in transcription: p21 repression, mechanisms, and consequences. Cancer Res, 2005. 65(10): p. 3980-5.

406. Bu, Y. and I.H. Gelman, v-Src-mediated down-regulation of SSeCKS metastasis suppressor gene promoter by the recruitment of HDAC1 into a USF1-Sp1-Sp3 complex. J Biol Chem, 2007. 282(37): p. 26725-39.

407. Doetzlhofer, A., et al., Histone deacetylase 1 can repress transcription by binding to Sp1. Mol Cell Biol, 1999. 19(8): p. 5504-11.

408. Won, J., J. Yim, and T.K. Kim, Sp1 and Sp3 recruit histone deacetylase to repress transcription of human telomerase reverse transcriptase (hTERT) promoter in normal human somatic cells. J Biol Chem, 2002. 277(41): p. 38230- 8.

409. Song, J., et al., Independent repression of a GC-rich housekeeping gene by Sp1 and MAZ involves the same cis-elements. J Biol Chem, 2001. 276(23): p. 19897- 904.

410. Zhang, Y. and M.L. Dufau, Silencing of transcription of the human luteinizing hormone receptor gene by histone deacetylase-mSin3A complex. J Biol Chem, 2002. 277(36): p. 33431-8.

411. Ma, Z., et al., Brg-1 is required for maximal transcription of the human matrix metalloproteinase-2 gene. J Biol Chem, 2004. 279(44): p. 46326-34.

412. Lemon, B., et al., Selectivity of chromatin-remodelling cofactors for ligand- activated transcription. Nature, 2001. 414(6866): p. 924-8.

413. Kadam, S., et al., Functional selectivity of recombinant mammalian SWI/SNF subunits. Genes Dev, 2000. 14(19): p. 2441-51.

414. Kadam, S. and B.M. Emerson, Transcriptional specificity of human SWI/SNF BRG1 and BRM chromatin remodeling complexes. Mol Cell, 2003. 11(2): p. 377- 89.

415. Aiello, A., et al., HMGA1 protein is a positive regulator of the insulin-like growth factor-I receptor gene. Eur J Cancer, 2010. 46(10): p. 1919-26.

416. Li, A.Y., et al., High-mobility group A2 protein modulates hTERT transcription to promote tumorigenesis. Mol Cell Biol, 2011. 31(13): p. 2605-17.

417. Dovat, S., et al., A common mechanism for mitotic inactivation of C2H2 zinc finger DNA-binding domains. Genes Dev, 2002. 16(23): p. 2985-90.

418. Tan, N.Y. and L.M. Khachigian, Sp1 phosphorylation and its regulation of gene transcription. Mol Cell Biol, 2009. 29(10): p. 2483-8.

419. Fojas de Borja, P., et al., Cyclin A-CDK phosphorylates Sp1 and enhances Sp1- mediated transcription. EMBO J, 2001. 20(20): p. 5737-47.

189

420. Vicart, A., et al., Increased chromatin association of Sp1 in interphase cells by PP2A-mediated dephosphorylations. J Mol Biol, 2006. 364(5): p. 897-908.

421. Martinez-Balbas, M.A., et al., Displacement of sequence-specific transcription factors from mitotic chromatin. Cell, 1995. 83(1): p. 29-38.

422. Chen, D., et al., TBP dynamics in living human cells: constitutive association of TBP with mitotic chromosomes. Mol Biol Cell, 2002. 13(1): p. 276-84.

423. Zaidi, S.K., et al., Mitotic partitioning and selective reorganization of tissue- specific transcription factors in progeny cells. Proc Natl Acad Sci U S A, 2003. 100(25): p. 14852-7.

424. He, S. and J.R. Davie, Sp1 and Sp3 foci distribution throughout mitosis. J Cell Sci, 2006. 119(Pt 6): p. 1063-70.

425. Chuang, J.Y., et al., Phosphorylation by c-Jun NH2-terminal kinase 1 regulates the stability of transcription factor Sp1 during mitosis. Mol Biol Cell, 2008. 19(3): p. 1139-51.

426. Wang, S.A., et al., Heat shock protein 90 is important for Sp1 stability during mitosis. J Mol Biol, 2009. 387(5): p. 1106-19.

427. Wang, Y.T., et al., Interplay of posttranslational modifications in Sp1 mediates Sp1 stability during cell cycle progression. J Mol Biol, 2011. 414(1): p. 1-14.

428. Chuang, J.Y., et al., Sp1 phosphorylation by cyclin-dependent kinase 1/cyclin B1 represses its DNA-binding activity during mitosis in cancer cells. Oncogene, 2012. 31(47): p. 4946-59.

429. Uren, A.G., et al., Role for yeast inhibitor of apoptosis (IAP)-like proteins in cell division. Proc Natl Acad Sci U S A, 1999. 96(18): p. 10170-5.

430. Rajagopalan, S. and M.K. Balasubramanian, S. pombe Pbh1p: an inhibitor of apoptosis domain containing protein is essential for chromosome segregation. FEBS Lett, 1999. 460(1): p. 187-90.

431. Li, F., et al., Cell division regulation by BIR1, a member of the inhibitor of apoptosis family in yeast. J Biol Chem, 2000. 275(10): p. 6707-11.

432. Petersen, J. and I.M. Hagan, S. pombe aurora kinase/survivin is required for chromosome condensation and the spindle checkpoint attachment response. Curr Biol, 2003. 13(7): p. 590-7.

433. Fraser, A.G., et al., Caenorhabditis elegans inhibitor of apoptosis protein (IAP) homologue BIR-1 plays a conserved role in cytokinesis. Curr Biol, 1999. 9(6): p. 292-301.

434. Speliotes, E.K., et al., The survivin-like C. elegans BIR-1 protein acts with the Aurora-like kinase AIR-2 to affect chromosomes and the spindle midzone. Mol Cell, 2000. 6(2): p. 211-23.

190

435. Uren, A.G., et al., Survivin and the inner centromere protein INCENP show similar cell-cycle localization and gene knockout phenotype. Curr Biol, 2000. 10(21): p. 1319-28.

436. Lens, S.M., G. Vader, and R.H. Medema, The case for Survivin as mitotic regulator. Curr Opin Cell Biol, 2006. 18(6): p. 616-22.

437. Vader, G., et al., Survivin mediates targeting of the chromosomal passenger complex to the centromere and midbody. EMBO Rep, 2006. 7(1): p. 85-92.

438. Ambrosini, G., C. Adida, and D.C. Altieri, A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma. Nat Med, 1997. 3(8): p. 917-21.

439. Brady, S.W., et al., PI3K-independent mTOR activation promotes lapatinib resistance and IAP expression that can be effectively reversed by mTOR and Hsp90 inhibition. Cancer Biol Ther, 2015. 16(3): p. 402-11.

440. Hunter, A.M., E.C. LaCasse, and R.G. Korneluk, The inhibitors of apoptosis (IAPs) as cancer targets. Apoptosis, 2007. 12(9): p. 1543-68.

441. Luk, S.U., et al., The BIRC6 gene as a novel target for therapy of prostate cancer: dual targeting of inhibitors of apoptosis. Oncotarget, 2014. 5(16): p. 6896-908.

442. Kusner, L.L., et al., Survivin as a potential mediator to support autoreactive cell survival in myasthenia gravis: a human and animal model study. PLoS One, 2014. 9(7): p. e102231.

443. Chandele, A., et al., Upregulation of survivin in G2/M cells and inhibition of caspase 9 activity enhances resistance in staurosporine-induced apoptosis. Neoplasia, 2004. 6(1): p. 29-40.

444. Tamm, I., et al., IAP-family protein survivin inhibits caspase activity and apoptosis induced by Fas (CD95), Bax, caspases, and anticancer drugs. Cancer Res, 1998. 58(23): p. 5315-20.

445. Xu, R., et al., Sp1 and Sp3 regulate basal transcription of the survivin gene. Biochem Biophys Res Commun, 2007. 356(1): p. 286-92.

446. Asanuma, K., et al., Survivin enhances Fas ligand expression via up-regulation of specificity protein 1-mediated gene transcription in colon cancer cells. J Immunol, 2004. 172(6): p. 3922-9.

447. Wu, J., et al., Molecular mechanism of inhibition of survivin transcription by the GC-rich sequence-selective DNA binding antitumor agent, hedamycin: evidence of survivin down-regulation associated with drug sensitivity. J Biol Chem, 2005. 280(10): p. 9745-51.

448. Asraf, H., et al., Mitotic slippage and expression of survivin are linked to differential sensitivity of human cancer cell-lines to the Kinesin-5 inhibitor monastrol. PLoS One, 2015. 10(6): p. e0129255.

191

449. Chiefari, E., et al., Increased expression of AP2 and Sp1 transcription factors in human thyroid tumors: a role in NIS expression regulation? BMC Cancer, 2002. 2: p. 35.

450. Wang, L., et al., Transcription factor Sp1 expression is a significant predictor of survival in human gastric cancer. Clin Cancer Res, 2003. 9(17): p. 6371-80.

451. Jiang, N.Y., et al., Sp1, a new biomarker that identifies a subset of aggressive pancreatic ductal adenocarcinoma. Cancer Epidemiol Biomarkers Prev, 2008. 17(7): p. 1648-52.

452. Guan, H., et al., Sp1 is upregulated in human glioma, promotes MMP-2-mediated cell invasion and predicts poor clinical outcome. Int J Cancer, 2012. 130(3): p. 593-601.

453. Wang, X.B., et al., [Expression and prognostic value of transcriptional factor sp1 in breast cancer]. Ai Zheng, 2007. 26(9): p. 996-1000.

454. Safe, S., et al., Transcription factor Sp1, also known as specificity protein 1 as a therapeutic target. Expert Opin Ther Targets, 2014. 18(7): p. 759-69.

455. Rigbolt, K.T., et al., System-wide temporal characterization of the proteome and phosphoproteome of human embryonic stem cell differentiation. Sci Signal, 2011. 4(164): p. rs3.

456. Olsen, J.V., et al., Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis. Sci Signal, 2010. 3(104): p. ra3.

457. Spengler, M.L., L.W. Guo, and M.G. Brattain, Phosphorylation mediates Sp1 coupled activities of proteolytic processing, desumoylation and degradation. Cell Cycle, 2008. 7(5): p. 623-30.

458. Spengler, M.L. and M.G. Brattain, Sumoylation inhibits cleavage of Sp1 N- terminal negative regulatory domain and inhibits Sp1-dependent transcription. J Biol Chem, 2006. 281(9): p. 5567-74.

459. Iwahori, S., et al., Identification of phosphorylation sites on transcription factor Sp1 in response to DNA damage and its accumulation at damaged sites. Cell Signal, 2008. 20(10): p. 1795-803.

460. Kim, H.S. and I.K. Lim, Phosphorylated extracellular signal-regulated protein kinases 1 and 2 phosphorylate Sp1 on serine 59 and regulate cellular senescence via transcription of p21Sdi1/Cip1/Waf1. J Biol Chem, 2009. 284(23): p. 15475-86.

461. Beausoleil, S.A., et al., A probability-based approach for high-throughput protein phosphorylation analysis and site localization. Nat Biotechnol, 2006. 24(10): p. 1285-92.

462. Olofsson, B.A., et al., Phosphorylation of Sp1 in response to DNA damage by ataxia telangiectasia-mutated kinase. Mol Cancer Res, 2007. 5(12): p. 1319-30.

192

463. Iwahori, S., et al., Enhanced phosphorylation of transcription factor sp1 in response to herpes simplex virus type 1 infection is dependent on the ataxia telangiectasia-mutated protein. J Virol, 2007. 81(18): p. 9653-64.

464. Chun, R.F., et al., Modulation of Sp1 phosphorylation by human immunodeficiency virus type 1 Tat. J Virol, 1998. 72(4): p. 2615-29.

465. Zheng, X.L., et al., Epidermal growth factor induction of apolipoprotein A-I is mediated by the Ras-MAP kinase cascade and Sp1. J Biol Chem, 2001. 276(17): p. 13822-9.

466. Bonello, M.R. and L.M. Khachigian, Fibroblast growth factor-2 represses platelet- derived growth factor receptor-alpha (PDGFR-alpha) transcription via ERK1/2- dependent Sp1 phosphorylation and an atypical cis-acting element in the proximal PDGFR-alpha promoter. J Biol Chem, 2004. 279(4): p. 2377-82.

467. Milanini-Mongiat, J., J. Pouyssegur, and G. Pages, Identification of two Sp1 phosphorylation sites for p42/p44 mitogen-activated protein kinases: their implication in vascular endothelial growth factor gene transcription. J Biol Chem, 2002. 277(23): p. 20631-9.

468. Hsu, M.C., H.C. Chang, and W.C. Hung, HER-2/neu represses the metastasis suppressor RECK via ERK and Sp transcription factors to promote cell invasion. J Biol Chem, 2006. 281(8): p. 4718-25.

469. Roos, M.D., et al., O glycosylation of an Sp1-derived peptide blocks known Sp1 protein interactions. Mol Cell Biol, 1997. 17(11): p. 6472-80.

470. Yang, X., et al., O-linkage of N-acetylglucosamine to Sp1 activation domain inhibits its transcriptional capability. Proc Natl Acad Sci U S A, 2001. 98(12): p. 6611-6.

471. Rickers, A., et al., Cleavage of transcription factor SP1 by caspases during anti- IgM-induced B-cell apoptosis. Eur J Biochem, 1999. 261(1): p. 269-74.

472. Majumdar, G., et al., Insulin dynamically regulates calmodulin gene expression by sequential o-glycosylation and phosphorylation of sp1 and its subcellular compartmentalization in liver cells. J Biol Chem, 2006. 281(6): p. 3642-50.

473. Zhang, Y., M. Liao, and M.L. Dufau, Phosphatidylinositol 3-kinase/protein kinase Czeta-induced phosphorylation of Sp1 and p107 repressor release have a critical role in histone deacetylase inhibitor-mediated derepression [corrected] of transcription of the luteinizing hormone receptor gene. Mol Cell Biol, 2006. 26(18): p. 6748-61.

474. Dephoure, N., et al., A quantitative atlas of mitotic phosphorylation. Proc Natl Acad Sci U S A, 2008. 105(31): p. 10762-7.

475. Armstrong, S.A., et al., Casein kinase II-mediated phosphorylation of the C terminus of Sp1 decreases its DNA binding activity. J Biol Chem, 1997. 272(21): p. 13489-95.

193

476. Tan, N.Y., et al., Angiotensin II-inducible platelet-derived growth factor-D transcription requires specific Ser/Thr residues in the second zinc finger region of Sp1. Circ Res, 2008. 102(4): p. e38-51.

477. Waby, J.S., et al., Sp1 acetylation is associated with loss of DNA binding at promoters associated with cell cycle arrest and cell death in a colon cell line. Mol Cancer, 2010. 9: p. 275.

478. Suzuki, T., et al., Regulation of interaction of the acetyltransferase region of p300 and the DNA-binding domain of Sp1 on and through DNA binding. Genes Cells, 2000. 5(1): p. 29-41.

479. Kang, J.E., et al., Histone deacetylase-1 represses transcription by interacting with zinc-fingers and interfering with the DNA binding activity of Sp1. Cell Physiol Biochem, 2005. 16(1-3): p. 23-30.

480. Wei, S., et al., Thiazolidinediones mimic glucose starvation in facilitating Sp1 degradation through the up-regulation of beta-transducin repeat-containing protein. Mol Pharmacol, 2009. 76(1): p. 47-57.

481. Hou, M., et al., The histone deacetylase inhibitor trichostatin A derepresses the telomerase reverse transcriptase (hTERT) gene in human cells. Exp Cell Res, 2002. 274(1): p. 25-34.

482. Endoh, T., et al., Survivin enhances telomerase activity via up-regulation of specificity protein 1- and c-Myc-mediated human telomerase reverse transcriptase gene transcription. Exp Cell Res, 2005. 305(2): p. 300-11.

483. Furuya, M., et al., Interaction between survivin and aurora-B kinase plays an important role in survivin-mediated up-regulation of human telomerase reverse transcriptase expression. Int J Oncol, 2009. 34(4): p. 1061-8.

484. Kyo, S., et al., Sp1 cooperates with c-Myc to activate transcription of the human telomerase reverse transcriptase gene (hTERT). Nucleic Acids Res, 2000. 28(3): p. 669-77.

485. Liu, C., et al., The telomerase reverse transcriptase (hTERT) gene is a direct target of the histone methyltransferase SMYD3. Cancer Res, 2007. 67(6): p. 2626-31.

486. Marconett, C.N., et al., Indole-3-carbinol downregulation of telomerase gene expression requires the inhibition of estrogen receptor-alpha and Sp1 transcription factor interactions within the hTERT promoter and mediates the G1 cell cycle arrest of human breast cancer cells. Carcinogenesis, 2011. 32(9): p. 1315-23.

487. Oh, S.T., S. Kyo, and L.A. Laimins, Telomerase activation by human papillomavirus type 16 E6 protein: induction of human telomerase reverse transcriptase expression through Myc and GC-rich Sp1 binding sites. J Virol, 2001. 75(12): p. 5559-66.

194

488. Palumbo, S.L., S.W. Ebbinghaus, and L.H. Hurley, Formation of a unique end-to- end stacked pair of G-quadruplexes in the hTERT core promoter with implications for inhibition of telomerase by G-quadruplex-interactive ligands. J Am Chem Soc, 2009. 131(31): p. 10878-91.

489. Takakura, M., et al., Cloning of human telomerase catalytic subunit (hTERT) gene promoter and identification of proximal core promoter sequences essential for transcriptional activation in immortalized and cancer cells. Cancer Res, 1999. 59(3): p. 551-7.

490. Wick, M., D. Zubov, and G. Hagen, Genomic organization and promoter characterization of the gene encoding the human telomerase reverse transcriptase (hTERT). Gene, 1999. 232(1): p. 97-106.

491. Zhao, J.Q., et al., Activation of telomerase rna gene promoter activity by NF-Y, Sp1, and the retinoblastoma protein and repression by Sp3. Neoplasia, 2000. 2(6): p. 531-9.

492. Bond, G.L., et al., A single nucleotide polymorphism in the MDM2 promoter attenuates the p53 tumor suppressor pathway and accelerates tumor formation in humans. Cell, 2004. 119(5): p. 591-602.

493. Knappskog, S. and P.E. Lonning, Effects of the MDM2 promoter SNP285 and SNP309 on Sp1 transcription factor binding and cancer risk. Transcription, 2011. 2(5): p. 207-10.

494. Knappskog, S. and P.E. Lonning, MDM2 promoter SNP285 and SNP309; phylogeny and impact on cancer risk. Oncotarget, 2011. 2(3): p. 251-8.

495. Knappskog, S., et al., The MDM2 promoter SNP285C/309G haplotype diminishes Sp1 transcription factor binding and reduces risk for breast and ovarian cancer in Caucasians. Cancer Cell, 2011. 19(2): p. 273-82.

496. Schutte, M., et al., Exon expression arrays as a tool to identify new cancer genes. PLoS One, 2008. 3(8): p. e3007.

497. Yarden, R.I., et al., MDM2 SNP309 accelerates breast and ovarian carcinogenesis in BRCA1 and BRCA2 carriers of Jewish-Ashkenazi descent. Breast Cancer Res Treat, 2008. 111(3): p. 497-504.

498. Wang, X., et al., P300 plays a role in p16(INK4a) expression and cell cycle arrest. Oncogene, 2008. 27(13): p. 1894-904.

499. Wang, X., et al., The proximal GC-rich region of p16(INK4a) gene promoter plays a role in its transcriptional regulation. Mol Cell Biochem, 2007. 301(1-2): p. 259- 66.

500. Wu, J., et al., Sp1 is essential for p16 expression in human diploid fibroblasts during senescence. PLoS One, 2007. 2(1): p. e164.

195

501. Xu, W., et al., A novel evolutionarily conserved element is a general transcriptional repressor of p21WAF(1)/CIP(1). Cancer Res, 2012. 72(23): p. 6236-46.

502. Biggs, J.R., J.E. Kudlow, and A.S. Kraft, The role of the transcription factor Sp1 in regulating the expression of the WAF1/CIP1 gene in U937 leukemic cells. J Biol Chem, 1996. 271(2): p. 901-6.

503. Koutsodontis, G., A. Moustakas, and D. Kardassis, The role of Sp1 family members, the proximal GC-rich motifs, and the upstream enhancer region in the regulation of the human cell cycle inhibitor p21WAF-1/Cip1 gene promoter. Biochemistry, 2002. 41(42): p. 12771-84.

504. Kavurma, M.M. and L.M. Khachigian, Vascular smooth muscle cell-specific regulation of cyclin-dependent kinase inhibitor p21(WAF1/Cip1) transcription by Sp1 is mediated via distinct cis-acting positive and negative regulatory elements in the proximal p21(WAF1/Cip1) promoter. J Cell Biochem, 2004. 93(5): p. 904- 16.

505. Ohlsson, C., et al., p53 regulates insulin-like growth factor-I (IGF-I) receptor expression and IGF-I-induced tyrosine phosphorylation in an osteosarcoma cell line: interaction between p53 and Sp1. Endocrinology, 1998. 139(3): p. 1101-7.

506. Abramovitch, S., et al., BRCA1-Sp1 interactions in transcriptional regulation of the IGF-IR gene. FEBS Lett, 2003. 541(1-3): p. 149-54.

507. Werner, H., et al., Cloning and characterization of the proximal promoter region of the rat insulin-like growth factor I (IGF-I) receptor gene. Biochem Biophys Res Commun, 1990. 169(3): p. 1021-7.

508. Jensen, D.E., et al., Transcriptional regulation of the elastin gene by insulin-like growth factor-I involves disruption of Sp1 binding. Evidence for the role of Rb in mediating Sp1 binding in aortic smooth muscle cells. J Biol Chem, 1995. 270(12): p. 6555-63.

509. Kaytor, E.N., et al., Physiological concentrations of insulin promote binding of nuclear proteins to the insulin-like growth factor I gene. Endocrinology, 2001. 142(3): p. 1041-9.

510. Maor, S., et al., Estrogen receptor regulates insulin-like growth factor-I receptor gene expression in breast tumor cells: involvement of transcription factor Sp1. J Endocrinol, 2006. 191(3): p. 605-12.

511. Maor, S., et al., Elevated insulin-like growth factor-I receptor (IGF-IR) levels in primary breast tumors associated with BRCA1 mutations. Cancer Lett, 2007. 257(2): p. 236-43.

512. Maor, S.B., et al., BRCA1 suppresses insulin-like growth factor-I receptor promoter activity: potential interaction between BRCA1 and Sp1. Mol Genet Metab, 2000. 69(2): p. 130-6.

196

513. Shahrabani-Gargir, L., T.K. Pandita, and H. Werner, Ataxia-telangiectasia mutated gene controls insulin-like growth factor I receptor gene expression in a deoxyribonucleic acid damage response pathway via mechanisms involving zinc- finger transcription factors Sp1 and WT1. Endocrinology, 2004. 145(12): p. 5679- 87.

514. Haley, J., et al., The human EGF receptor gene: structure of the 110 kb locus and identification of sequences regulating its transcription. Oncogene Res, 1987. 1(4): p. 375-96.

515. Johnson, A.C., et al., Epidermal growth factor receptor gene promoter. Deletion analysis and identification of nuclear protein binding sites. J Biol Chem, 1988. 263(12): p. 5693-9.

516. Kageyama, R., G.T. Merlino, and I. Pastan, Epidermal growth factor (EGF) receptor gene transcription. Requirement for Sp1 and an EGF receptor-specific factor. J Biol Chem, 1988. 263(13): p. 6329-36.

517. Kitadai, Y., et al., The level of a transcription factor Sp1 is correlated with the expression of EGF receptor in human gastric carcinomas. Biochem Biophys Res Commun, 1992. 189(3): p. 1342-8.

518. Salvatori, L., et al., Oestrogens and selective oestrogen receptor (ER) modulators regulate EGF receptor gene expression through human ER alpha and beta subtypes via an Sp1 site. Oncogene, 2003. 22(31): p. 4875-81.

519. Carpentier, C., et al., Polymorphism in Sp1 recognition site of the EGF receptor gene promoter and risk of glioblastoma. Neurology, 2006. 67(5): p. 872-4.

520. Pascall, J.C. and K.D. Brown, Identification of a minimal promoter element of the mouse epidermal growth factor gene. Biochem J, 1997. 324 ( Pt 3): p. 869-75.

521. Tiesman, J. and A. Rizzino, Nucleotide sequence of the 5'-flanking region of the mouse k-FGF oncogene exhibits an alternating purine:pyrimidine motif with the potential to form Z-DNA. Gene, 1990. 96(2): p. 311-2.

522. Luster, T.A., et al., Effects of three Sp1 motifs on the transcription of the FGF-4 gene. Mol Reprod Dev, 2000. 57(1): p. 4-15.

523. Payson, R.A., M.A. Chotani, and I.M. Chiu, Regulation of a promoter of the fibroblast growth factor 1 gene in prostate and breast cancer cells. J Steroid Biochem Mol Biol, 1998. 66(3): p. 93-103.

524. Jimenez, S.K., et al., Transcriptional regulation of FGF-2 gene expression in cardiac myocytes. Cardiovasc Res, 2004. 62(3): p. 548-57.

525. Zhu, J.L., et al., Involvement of Sp1 in the transcriptional regulation of the rat insulin-like growth factor-1 gene. Mol Cell Endocrinol, 2000. 164(1-2): p. 205-18.

197

526. Cadoret, A., et al., Oncogene-induced up-regulation of Caco-2 cell proliferation involves IGF-II gene activation through a protein kinase C-mediated pathway. Oncogene, 1998. 17(7): p. 877-87.

527. Hyun, S.W., et al., Characterization of the P4 promoter region of the human insulin-like growth factor II gene. FEBS Lett, 1993. 332(1-2): p. 153-8.

528. Jiang, Y., et al., A high expression level of insulin-like growth factor I receptor is associated with increased expression of transcription factor Sp1 and regional lymph node metastasis of human gastric cancer. Clin Exp Metastasis, 2004. 21(8): p. 755-64.

529. Lee, Y.I., et al., The human hepatitis B virus transactivator X gene product regulates Sp1 mediated transcription of an insulin-like growth factor II promoter 4. Oncogene, 1998. 16(18): p. 2367-80.

530. Li, T., et al., Using DNA microarray to identify Sp1 as a transcriptional regulatory element of insulin-like growth factor 1 in cardiac muscle cells. Circ Res, 2003. 93(12): p. 1202-9.

531. Sun, Y., N.J. Giacalone, and B. Lu, Terameprocol (tetra-O-methyl nordihydroguaiaretic acid), an inhibitor of Sp1-mediated survivin transcription, induces radiosensitization in non-small cell lung carcinoma. J Thorac Oncol, 2011. 6(1): p. 8-14.

532. Chae, J.I., et al., Role of transcription factor Sp1 in the quercetin-mediated inhibitory effect on human malignant pleural mesothelioma. Int J Mol Med, 2012. 30(4): p. 835-41.

533. Kim, Y.H., et al., Quercetin augments TRAIL-induced apoptotic death: involvement of the ERK signal transduction pathway. Biochem Pharmacol, 2008. 75(10): p. 1946-58.

534. Lee, K.A., et al., The flavonoid resveratrol suppresses growth of human malignant pleural mesothelioma cells through direct inhibition of specificity protein 1. Int J Mol Med, 2012. 30(1): p. 21-7.

535. Lee, S.O., et al., Inactivation of the orphan nuclear receptor TR3/Nur77 inhibits pancreatic cancer cell and tumor growth. Cancer Res, 2010. 70(17): p. 6824-36.

536. Li, F. and D.C. Altieri, Transcriptional analysis of human survivin gene expression. Biochem J, 1999. 344 Pt 2: p. 305-11.

537. Li, F. and D.C. Altieri, The cancer antiapoptosis mouse survivin gene: characterization of locus and transcriptional requirements of basal and cell cycle- dependent expression. Cancer Res, 1999. 59(13): p. 3143-51.

538. Kim, J.Y., et al., Quercetin sensitizes human hepatoma cells to TRAIL-induced apoptosis via Sp1-mediated DR5 up-regulation and proteasome-mediated c- FLIPS down-regulation. J Cell Biochem, 2008. 105(6): p. 1386-98.

198

539. Kim, Y.H., et al., Sodium butyrate sensitizes TRAIL-mediated apoptosis by induction of transcription from the DR5 gene promoter through Sp1 sites in colon cancer cells. Carcinogenesis, 2004. 25(10): p. 1813-20.

540. Yoshida, T., et al., Promoter structure and transcription initiation sites of the human death receptor 5/TRAIL-R2 gene. FEBS Lett, 2001. 507(3): p. 381-5.

541. Duan, H., C.A. Heckman, and L.M. Boxer, Histone deacetylase inhibitors down- regulate bcl-2 expression and induce apoptosis in t(14;18) lymphomas. Mol Cell Biol, 2005. 25(5): p. 1608-19.

542. Azahri, N.S., et al., Sp1, acetylated histone-3 and p300 regulate TRAIL transcription: mechanisms of PDGF-BB-mediated VSMC proliferation and migration. J Cell Biochem, 2012. 113(8): p. 2597-606.

543. Lee, K.A., J.I. Chae, and J.H. Shim, Natural diterpenes from coffee, cafestol and kahweol induce apoptosis through regulation of specificity protein 1 expression in human malignant pleural mesothelioma. J Biomed Sci, 2012. 19: p. 60.

544. Lee, T.J., et al., Mithramycin A sensitizes cancer cells to TRAIL-mediated apoptosis by down-regulation of XIAP gene promoter through Sp1 sites. Mol Cancer Ther, 2006. 5(11): p. 2737-46.

545. Lee, K.E., et al., Effect of methanol extracts of Cnidium officinale Makino and Capsella bursa-pastoris on the apoptosis of HSC-2 human oral cancer cells. Exp Ther Med, 2013. 5(3): p. 789-792.

546. Kavurma, M.M., Y. Bobryshev, and L.M. Khachigian, Ets-1 positively regulates Fas ligand transcription via cooperative interactions with Sp1. J Biol Chem, 2002. 277(39): p. 36244-52.

547. Kavurma, M.M., et al., Sp1 phosphorylation regulates apoptosis via extracellular FasL-Fas engagement. J Biol Chem, 2001. 276(7): p. 4964-71.

548. Savickiene, J., et al., p21 (Waf1/Cip1) and FasL gene activation via Sp1 and NFkappaB is required for leukemia cell survival but not for cell death induced by diverse stimuli. Int J Biochem Cell Biol, 2005. 37(4): p. 784-96.

549. Pal, S., et al., Activation of Sp1-mediated vascular permeability factor/vascular endothelial growth factor transcription requires specific interaction with protein kinase C zeta. J Biol Chem, 1998. 273(41): p. 26277-80.

550. Jia, Z., et al., Molecular basis of the synergistic antiangiogenic activity of bevacizumab and mithramycin A. Cancer Res, 2007. 67(10): p. 4878-85.

551. Yao, J.C., et al., Association between expression of transcription factor Sp1 and increased vascular endothelial growth factor expression, advanced stage, and poor survival in patients with resected gastric cancer. Clin Cancer Res, 2004. 10(12 Pt 1): p. 4109-17.

199

552. Abdelrahim, M. and S. Safe, Cyclooxygenase-2 inhibitors decrease vascular endothelial growth factor expression in colon cancer cells by enhanced degradation of Sp1 and Sp4 proteins. Mol Pharmacol, 2005. 68(2): p. 317-29.

553. Abdelrahim, M., et al., Role of Sp proteins in regulation of vascular endothelial growth factor expression and proliferation of pancreatic cancer cells. Cancer Res, 2004. 64(18): p. 6740-9.

554. Eisermann, K., et al., Androgen up-regulates vascular endothelial growth factor expression in prostate cancer cells via an Sp1 binding site. Mol Cancer, 2013. 12: p. 7.

555. Finkenzeller, G., et al., Sp1 recognition sites in the proximal promoter of the human vascular endothelial growth factor gene are essential for platelet-derived growth factor-induced gene expression. Oncogene, 1997. 15(6): p. 669-76.

556. Kazi, A.A., J.M. Jones, and R.D. Koos, Chromatin immunoprecipitation analysis of gene expression in the rat uterus in vivo: estrogen-induced recruitment of both estrogen receptor alpha and hypoxia-inducible factor 1 to the vascular endothelial growth factor promoter. Mol Endocrinol, 2005. 19(8): p. 2006-19.

557. Koos, R.D., et al., New insight into the transcriptional regulation of vascular endothelial growth factor expression in the endometrium by estrogen and relaxin. Ann N Y Acad Sci, 2005. 1041: p. 233-47.

558. Milanini, J., et al., p42/p44 MAP kinase module plays a key role in the transcriptional regulation of the vascular endothelial growth factor gene in fibroblasts. J Biol Chem, 1998. 273(29): p. 18165-72.

559. Motoyama, K., et al., SREBP inhibits VEGF expression in human smooth muscle cells. Biochem Biophys Res Commun, 2006. 342(1): p. 354-60.

560. Mukhopadhyay, D., et al., The von Hippel-Lindau tumor suppressor gene product interacts with Sp1 to repress vascular endothelial growth factor promoter activity. Mol Cell Biol, 1997. 17(9): p. 5629-39.

561. Pore, N., et al., Sp1 is involved in Akt-mediated induction of VEGF expression through an HIF-1-independent mechanism. Mol Biol Cell, 2004. 15(11): p. 4841- 53.

562. Reisinger, K., R. Kaufmann, and J. Gille, Increased Sp1 phosphorylation as a mechanism of hepatocyte growth factor (HGF/SF)-induced vascular endothelial growth factor (VEGF/VPF) transcription. J Cell Sci, 2003. 116(Pt 2): p. 225-38.

563. Santra, M., et al., Ectopic decorin expression up-regulates VEGF expression in mouse cerebral endothelial cells via activation of the transcription factors Sp1, HIF1alpha, and Stat3. J Neurochem, 2008. 105(2): p. 324-37.

564. Schafer, G., et al., Oxidative stress regulates vascular endothelial growth factor- A gene transcription through Sp1- and Sp3-dependent activation of two proximal GC-rich promoter elements. J Biol Chem, 2003. 278(10): p. 8190-8.

200

565. Shi, Q., et al., Constitutive Sp1 activity is essential for differential constitutive expression of vascular endothelial growth factor in human pancreatic adenocarcinoma. Cancer Res, 2001. 61(10): p. 4143-54.

566. Stoner, M., et al., Estrogen regulation of vascular endothelial growth factor gene expression in ZR-75 breast cancer cells through interaction of estrogen receptor alpha and SP proteins. Oncogene, 2004. 23(5): p. 1052-63.

567. Tischer, E., et al., The human gene for vascular endothelial growth factor. Multiple protein forms are encoded through alternative exon splicing. J Biol Chem, 1991. 266(18): p. 11947-54.

568. Wei, D., et al., Celecoxib inhibits vascular endothelial growth factor expression in and reduces angiogenesis and metastasis of human pancreatic cancer via suppression of Sp1 transcription factor activity. Cancer Res, 2004. 64(6): p. 2030-8.

569. Yu, D.C., et al., Butyrate suppresses expression of neuropilin I in colorectal cell lines through inhibition of Sp1 transactivation. Mol Cancer, 2010. 9: p. 276.

570. Okamoto, M., et al., Up-regulation of thrombospondin-1 gene by epidermal growth factor and transforming growth factor beta in human cancer cells-- transcriptional activation and messenger RNA stabilization. Biochim Biophys Acta, 2002. 1574(1): p. 24-34.

571. Rafty, L.A. and L.M. Khachigian, von Hippel-Lindau tumor suppressor protein represses platelet-derived growth factor B-chain gene expression via the Sp1 binding element in the proximal PDGF-B promoter. J Cell Biochem, 2002. 85(3): p. 490-5.

572. Liu, M.Y., et al., Inducible platelet-derived growth factor D-chain expression by angiotensin II and hydrogen peroxide involves transcriptional regulation by Ets-1 and Sp1. Blood, 2006. 107(6): p. 2322-9.

573. Rafty, L.A. and L.M. Khachigian, Sp1 phosphorylation regulates inducible expression of platelet-derived growth factor B-chain gene via atypical protein kinase C-zeta. Nucleic Acids Res, 2001. 29(5): p. 1027-33.

574. Rafty, L.A., F.S. Santiago, and L.M. Khachigian, NF1/X represses PDGF A-chain transcription by interacting with Sp1 and antagonizing Sp1 occupancy of the promoter. EMBO J, 2002. 21(3): p. 334-43.

575. Santiago, F.S. and L.M. Khachigian, Ets-1 stimulates platelet-derived growth factor A-chain gene transcription and vascular smooth muscle cell growth via cooperative interactions with Sp1. Circ Res, 2004. 95(5): p. 479-87.

576. Belaguli, N.S., et al., GATA6 promotes colon cancer cell invasion by regulating urokinase plasminogen activator gene expression. Neoplasia, 2010. 12(11): p. 856-65.

201

577. Benasciutti, E., et al., MAPK and JNK transduction pathways can phosphorylate Sp1 to activate the uPA minimal promoter element and endogenous gene transcription. Blood, 2004. 104(1): p. 256-62.

578. Zannetti, A., et al., Coordinate up-regulation of Sp1 DNA-binding activity and urokinase receptor expression in breast carcinoma. Cancer Res, 2000. 60(6): p. 1546-51.

579. Hockings, J.K., et al., Involvement of a specificity proteins-binding element in regulation of basal and estrogen-induced transcription activity of the BRCA1 gene. Breast Cancer Res, 2008. 10(2): p. R29.

580. Maor, S., et al., Insulin-like growth factor-I controls BRCA1 gene expression through activation of transcription factor Sp1. Horm Metab Res, 2007. 39(3): p. 179-85.

581. Keating, K.E., et al., Transcriptional downregulation of ATM by EGF is defective in ataxia-telangiectasia cells expressing mutant protein. Oncogene, 2001. 20(32): p. 4281-90.

582. Truman, J.P., et al., Down-regulation of ATM protein sensitizes human prostate cancer cells to radiation-induced apoptosis. J Biol Chem, 2005. 280(24): p. 23262-72.

583. Gueven, N., et al., Site-directed mutagenesis of the ATM promoter: consequences for response to proliferation and ionizing radiation. Genes Chromosomes Cancer, 2003. 38(2): p. 157-67.

584. Bu, Y., et al., Sp1-mediated transcriptional regulation of NFBD1/MDC1 plays a critical role in DNA damage response pathway. Genes Cells, 2008. 13(1): p. 53- 66.

585. Dalvai, M., et al., Cdc25B is negatively regulated by p53 through Sp1 and NF-Y transcription factors. Oncogene, 2011. 30(19): p. 2282-8.

586. Dalvai, M., et al., Doxorubicin promotes transcriptional upregulation of Cdc25B in cancer cells by releasing Sp1 from the promoter. Oncogene, 2013. 32(42): p. 5123-8.

587. Sengupta, S., et al., Tumor suppressor p53 represses transcription of RECQ4 helicase. Oncogene, 2005. 24(10): p. 1738-48.

588. Oei, S.L., et al., Transcriptional regulation and autoregulation of the human gene for ADP-ribosyltransferase. Mol Cell Biochem, 1994. 138(1-2): p. 99-104.

589. Laniel, M.A., G.G. Poirier, and S.L. Guerin, A conserved initiator element on the mammalian poly(ADP-ribose) polymerase-1 promoters, in combination with flanking core elements, is necessary to obtain high transcriptional activity. Biochim Biophys Acta, 2004. 1679(1): p. 37-46.

202

590. Hao, B., et al., A novel T-77C polymorphism in DNA repair gene XRCC1 contributes to diminished promoter activity and increased risk of non-small cell lung cancer. Oncogene, 2006. 25(25): p. 3613-20.

591. Zhou, Z.Q. and C.A. Walter, Cloning and characterization of the promoter of baboon XRCC1, a gene involved in DNA strand-break repair. Somat Cell Mol Genet, 1998. 24(1): p. 23-39.

592. Iso, T., et al., Modulation of the expression of bloom helicase by estrogenic agents. Biol Pharm Bull, 2007. 30(2): p. 266-71.

593. Wang, H., et al., Chk2 down-regulation by promoter hypermethylation in human bulk gliomas. Life Sci, 2010. 86(5-6): p. 185-91.

594. Zhang, S., et al., A variant in the CHEK2 promoter at a methylation site relieves transcriptional repression and confers reduced risk of lung cancer. Carcinogenesis, 2010. 31(7): p. 1251-8.

595. Xu, X.S., et al., Histone deacetylases (HDACs) in XPC gene silencing and bladder cancer. J Hematol Oncol, 2011. 4: p. 17.

596. Jaitovich-Groisman, I., et al., Transcriptional regulation of the TFIIH transcription repair components XPB and XPD by the hepatitis B virus x protein in liver cells and transgenic liver tissue. J Biol Chem, 2001. 276(17): p. 14124-32.

597. Ma, L., et al., Molecular and functional analysis of the XPBC/ERCC-3 promoter: transcription activity is dependent on the integrity of an Sp1-binding site. Nucleic Acids Res, 1992. 20(2): p. 217-24.

598. Nichols, A.F., et al., Basal transcriptional regulation of human damage-specific DNA-binding protein genes DDB1 and DDB2 by Sp1, E2F, N-myc and NF1 elements. Nucleic Acids Res, 2003. 31(2): p. 562-9.

599. Hosoi, Y., et al., Up-regulation of DNA-dependent protein kinase activity and Sp1 in colorectal cancer. Int J Oncol, 2004. 25(2): p. 461-8.

600. Wang, S., et al., A novel variable number of tandem repeats (VNTR) polymorphism containing Sp1 binding elements in the promoter of XRCC5 is a risk factor for human bladder cancer. Mutat Res, 2008. 638(1-2): p. 26-36.

601. Lin, Z., et al., A variant of the Cockayne syndrome B gene ERCC6 confers risk of lung cancer. Hum Mutat, 2008. 29(1): p. 113-22.

602. Yamabe, Y., et al., Sp1-mediated transcription of the Werner helicase gene is modulated by Rb and p53. Mol Cell Biol, 1998. 18(11): p. 6191-200.

603. Hung, W.C., et al., Skp2 overexpression increases the expression of MMP-2 and MMP-9 and invasion of lung cancer cells. Cancer Lett, 2010. 288(2): p. 156-61.

604. Sroka, I.C., R.B. Nagle, and G.T. Bowden, Membrane-type 1 matrix metalloproteinase is regulated by sp1 through the differential activation of AKT,

203

JNK, and ERK pathways in human prostate tumor cells. Neoplasia, 2007. 9(5): p. 406-17.

605. Yun, S., et al., Transcription factor Sp1 phosphorylation induced by shear stress inhibits membrane type 1-matrix metalloproteinase expression in endothelium. J Biol Chem, 2002. 277(38): p. 34808-14.

606. Chang, H.C., L.T. Liu, and W.C. Hung, Involvement of histone deacetylation in ras-induced down-regulation of the metastasis suppressor RECK. Cell Signal, 2004. 16(6): p. 675-9.

607. Sasahara, R.M., C. Takahashi, and M. Noda, Involvement of the Sp1 site in ras- mediated downregulation of the RECK metastasis suppressor gene. Biochem Biophys Res Commun, 1999. 264(3): p. 668-75.

608. Sasahara, R.M., et al., Oncogene-mediated downregulation of RECK, a novel transformation suppressor gene. Braz J Med Biol Res, 1999. 32(7): p. 891-5.

609. Hsu, T.I., et al., Sp1 expression regulates lung tumor progression. Oncogene, 2012. 31(35): p. 3973-88.

610. Apt, D., et al., High Sp1/Sp3 ratios in epithelial cells during epithelial differentiation and cellular transformation correlate with the activation of the HPV- 16 promoter. Virology, 1996. 224(1): p. 281-91.

611. D'Costa, Z.J., et al., Transcriptional repression of E-cadherin by human papillomavirus type 16 E6. PLoS One, 2012. 7(11): p. e48954.

612. Faraldo, M.L., et al., Analysis of the E-cadherin and P-cadherin promoters in murine keratinocyte cell lines from different stages of mouse skin carcinogenesis. Mol Carcinog, 1997. 20(1): p. 33-47.

613. Graff, J.R., et al., Mapping patterns of CpG island methylation in normal and neoplastic cells implicates both upstream and downstream regions in de novo methylation. J Biol Chem, 1997. 272(35): p. 22322-9.

614. Liu, Y.N., et al., Regulatory mechanisms controlling human E-cadherin gene expression. Oncogene, 2005. 24(56): p. 8277-90.

615. Sisci, D., et al., 17beta-estradiol enhances alpha(5) integrin subunit gene expression through ERalpha-Sp1 interaction and reduces cell motility and invasion of ERalpha-positive breast cancer cells. Breast Cancer Res Treat, 2010. 124(1): p. 63-77.

616. Han, S. and J. Roman, COX-2 inhibitors suppress integrin alpha5 expression in human lung carcinoma cells through activation of Erk: involvement of Sp1 and AP-1 sites. Int J Cancer, 2005. 116(4): p. 536-46.

617. Nam, E.H., et al., ZEB2 upregulates integrin alpha5 expression through cooperation with Sp1 to induce invasion during epithelial-mesenchymal transition of human cancer cells. Carcinogenesis, 2012. 33(3): p. 563-71.

204

618. Kuo, L., et al., Src oncogene activates MMP-2 expression via the ERK/Sp1 pathway. J Cell Physiol, 2006. 207(3): p. 729-34.

619. Pan, M.R. and W.C. Hung, Nonsteroidal anti-inflammatory drugs inhibit matrix metalloproteinase-2 via suppression of the ERK/Sp1-mediated transcription. J Biol Chem, 2002. 277(36): p. 32775-80.

620. Qin, H., Y. Sun, and E.N. Benveniste, The transcription factors Sp1, Sp3, and AP-2 are required for constitutive matrix metalloproteinase-2 gene expression in astroglioma cells. J Biol Chem, 1999. 274(41): p. 29130-7.

621. Wang, C.H., H.C. Chang, and W.C. Hung, p16 inhibits matrix metalloproteinase- 2 expression via suppression of Sp1-mediated gene transcription. J Cell Physiol, 2006. 208(1): p. 246-52.

622. Foltz, D.R., et al., The human CENP-A centromeric nucleosome-associated complex. Nature cell biology, 2006. 8(5): p. 458-69.

623. Okamoto, Y., et al., A minimal CENP-A core is required for nucleation and maintenance of a functional human centromere. The EMBO journal, 2007. 26(5): p. 1279-91.

624. Vafa, O. and K.F. Sullivan, Chromatin containing CENP-A and alpha-satellite DNA is a major component of the inner kinetochore plate. Current biology : CB, 1997. 7(11): p. 897-900.

625. Warburton, P.E., et al., Immunolocalization of CENP-A suggests a distinct nucleosome structure at the inner kinetochore plate of active centromeres. Current biology : CB, 1997. 7(11): p. 901-4.

626. Van Hooser, A.A., et al., Specification of kinetochore-forming chromatin by the histone H3 variant CENP-A. Journal of cell science, 2001. 114(Pt 19): p. 3529- 42.

627. Cheeseman, I.M. and A. Desai, Molecular architecture of the kinetochore- microtubule interface. Nature reviews Molecular cell biology, 2008. 9(1): p. 33-46.

628. Chan, G.K., S.-T. Liu, and T.J. Yen, Kinetochore structure and function. Trends in cell biology, 2005. 15(11): p. 589-98.

629. Bakhoum, S.F. and D.A. Compton, Chromosomal instability and cancer: a complex relationship with therapeutic potential. J Clin Invest, 2012. 122(4): p. 1138-43.

630. Courey, A.J. and R. Tjian, Analysis of Sp1 in vivo reveals multiple transcriptional domains, including a novel glutamine-rich activation motif. Cell, 1988. 55(5): p. 887-98.

631. Courey, A.J., et al., Synergistic activation by the glutamine-rich domains of human transcription factor Sp1. Cell, 1989. 59(5): p. 827-36.

205

632. Pascal, E. and R. Tjian, Different activation domains of Sp1 govern formation of multimers and mediate transcriptional synergism. Genes & development, 1991. 5(9): p. 1646-56.

633. Li, L. and J.R. Davie, The role of Sp1 and Sp3 in normal and cancer cell biology. Annals of anatomy = Anatomischer Anzeiger : official organ of the Anatomische Gesellschaft, 2010. 192(5): p. 275-83.

634. Cheeseman, I.M., The kinetochore. Cold Spring Harb Perspect Biol, 2014. 6(7): p. a015826.

635. McAinsh, A.D. and P. Meraldi, The CCAN complex: linking centromere specification to control of kinetochore-microtubule dynamics. Semin Cell Dev Biol, 2011. 22(9): p. 946-52.

636. Marin, M., et al., Transcription factor Sp1 is essential for early embryonic development but dispensable for cell growth and differentiation. Cell, 1997. 89(4): p. 619-28.

637. Chuang, J.-Y., et al., Phosphorylation by c-Jun NH2-terminal kinase 1 regulates the stability of transcription factor Sp1 during mitosis. Molecular biology of the cell, 2008. 19(3): p. 1139-51.

638. Yang, H.-C., et al., Pin1-mediated Sp1 phosphorylation by CDK1 increases Sp1 stability and decreases its DNA-binding activity during mitosis. Nucleic acids research, 2014. 42(22): p. 13573-87.

639. Boohaker, R.J. and B. Xu, The versatile functions of ATM kinase. Biomed J, 2014. 37(1): p. 3-9.

640. Hickson, I., et al., Identification and characterization of a novel and specific inhibitor of the ataxia-telangiectasia mutated kinase ATM. Cancer Res, 2004. 64(24): p. 9152-9.

641. McCloy, R.A., et al., Partial inhibition of Cdk1 in G 2 phase overrides the SAC and decouples mitotic events. Cell Cycle, 2014. 13(9): p. 1400-12.

642. Shelby, R.D., K. Monier, and K.F. Sullivan, Chromatin assembly at kinetochores is uncoupled from DNA replication. J Cell Biol, 2000. 151(5): p. 1113-8.

643. Jansen, L.E., et al., Propagation of centromeric chromatin requires exit from mitosis. J Cell Biol, 2007. 176(6): p. 795-805.

644. Ferri, F., et al., Non-coding murine centromeric transcripts associate with and potentiate Aurora B kinase. Nucleic Acids Res, 2009. 37(15): p. 5071-80.

645. Yan, H., et al., Transcription and histone modifications in the recombination-free region spanning a rice centromere. Plant Cell, 2005. 17(12): p. 3227-38.

206

646. Hall, L.E., S.E. Mitchell, and R.J. O'Neill, Pericentric and centromeric transcription: a perfect balance required. Chromosome Res, 2012. 20(5): p. 535- 46.

647. Du, Y., C.N. Topp, and R.K. Dawe, DNA binding of centromere protein C (CENPC) is stabilized by single-stranded RNA. PLoS Genet, 2010. 6(2): p. e1000835.

648. Grummt, I., Life on a planet of its own: regulation of RNA polymerase I transcription in the nucleolus. Genes Dev, 2003. 17(14): p. 1691-702.

649. Kadonaga, J.T., K.A. Jones, and R. Tjian, Promoter-Specific Activation of Rna Polymerase-Ii Transcription by Sp1. Trends in Biochemical Sciences, 1986. 11(1): p. 20-23.

650. Rihani, A., et al., Effective Alu repeat based RT-Qpcr normalization in cancer cell perturbation experiments. PLoS One, 2013. 8(8): p. e71776.

651. Vossaert, L., et al., Reference loci for RT-qPCR analysis of differentiating human embryonic stem cells. BMC Mol Biol, 2013. 14: p. 21.

652. Hasler, J., T. Samuelsson, and K. Strub, Useful 'junk': Alu RNAs in the human transcriptome. Cell Mol Life Sci, 2007. 64(14): p. 1793-800.

653. Pandey, R. and M. Mukerji, From 'JUNK' to just unexplored noncoding knowledge: the case of transcribed Alus. Brief Funct Genomics, 2011. 10(5): p. 294-311.

654. Wang, C. and S. Huang, Nuclear function of Alus. Nucleus, 2014. 5(2): p. 131-7.

655. Phatnani, H.P. and A.L. Greenleaf, Phosphorylation and functions of the RNA polymerase II CTD. Genes Dev, 2006. 20(21): p. 2922-36.

656. Jin, W., et al., TIEG1 inhibits breast cancer invasion and metastasis by inhibition of epidermal growth factor receptor (EGFR) transcription and the EGFR signaling pathway. Mol Cell Biol, 2012. 32(1): p. 50-63.

657. Esteve, P.O., H.G. Chin, and S. Pradhan, Molecular mechanisms of transactivation and doxorubicin-mediated repression of survivin gene in cancer cells. J Biol Chem, 2007. 282(4): p. 2615-25.

658. Gui, C.Y., et al., Histone deacetylase (HDAC) inhibitor activation of p21WAF1 involves changes in promoter-associated proteins, including HDAC1. Proc Natl Acad Sci U S A, 2004. 101(5): p. 1241-6.

659. Carroll, C.W. and A.F. Straight, Centromere formation: from epigenetics to self- assembly. Trends in cell biology, 2006. 16(2): p. 70-8.

660. Han, I. and J.E. Kudlow, Reduced O glycosylation of Sp1 is associated with increased proteasome susceptibility. Mol Cell Biol, 1997. 17(5): p. 2550-8.

207

661. Su, K., et al., An N-terminal region of Sp1 targets its proteasome-dependent degradation in vitro. J Biol Chem, 1999. 274(21): p. 15194-202.

662. Piedrafita, F.J. and M. Pfahl, Retinoid-induced apoptosis and Sp1 cleavage occur independently of transcription and require caspase activation. Mol Cell Biol, 1997. 17(11): p. 6348-58.

663. Shiloh, Y. and Y. Ziv, The ATM protein kinase: regulating the cellular response to genotoxic stress, and more. Nat Rev Mol Cell Biol, 2013. 14(4): p. 197-210.

664. Shigeta, T., et al., Defective control of apoptosis and mitotic spindle checkpoint in heterozygous carriers of ATM mutations. Cancer Res, 1999. 59(11): p. 2602-7.

665. Takagi, M., et al., Defective control of apoptosis, radiosensitivity, and spindle checkpoint in ataxia telangiectasia. Cancer Res, 1998. 58(21): p. 4923-9.

666. Iourov, I.Y., et al., Increased chromosome instability dramatically disrupts neural genome integrity and mediates cerebellar degeneration in the ataxia- telangiectasia brain. Hum Mol Genet, 2009. 18(14): p. 2656-69.

667. Wang, J., et al., Mitotic regulator Mis18beta interacts with and specifies the centromeric assembly of molecular chaperone holliday junction recognition protein (HJURP). J Biol Chem, 2014. 289(12): p. 8326-36.

668. Howman, E.V., et al., Early disruption of centromeric chromatin organization in centromere protein A (Cenpa) null mice. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(3): p. 1148-53.

669. Oegema, K., et al., Functional analysis of kinetochore assembly in Caenorhabditis elegans. The Journal of cell biology, 2001. 153(6): p. 1209-26.

670. Li, Y., et al., ShRNA-targeted centromere protein A inhibits hepatocellular carcinoma growth. PloS one, 2011. 6(3): p. e17794.

671. Wu, Q., et al., Expression and prognostic significance of centromere protein A in human lung adenocarcinoma. Lung cancer (Amsterdam, Netherlands), 2012. 77(2): p. 407-14.

672. Li, B., M. Carey, and J.L. Workman, The role of chromatin during transcription. Cell, 2007. 128(4): p. 707-19.

673. Li, B., et al., Histone H3 lysine 36 dimethylation (H3K36me2) is sufficient to recruit the Rpd3s histone deacetylase complex and to repress spurious transcription. J Biol Chem, 2009. 284(12): p. 7970-6.

674. Suganuma, T. and J.L. Workman, Crosstalk among Histone Modifications. Cell, 2008. 135(4): p. 604-7.

675. Zhu, L., et al., ASH1L links histone H3 lysine 36 di-methylation to MLL leukemia. Cancer Discov, 2016.

208

676. Zentner, G.E. and S. Henikoff, Regulation of nucleosome dynamics by histone modifications. Nat Struct Mol Biol, 2013. 20(3): p. 259-66.

677. Robbins, A.R., et al., Inhibitors of histone deacetylases alter kinetochore assembly by disrupting pericentromeric heterochromatin. Cell Cycle, 2005. 4(5): p. 717-26.

678. Kelly, R.D. and S.M. Cowley, The physiological roles of histone deacetylase (HDAC) 1 and 2: complex co-stars with multiple leading parts. Biochem Soc Trans, 2013. 41(3): p. 741-9.

679. Silverstein, R.A., et al., A new role for the transcriptional corepressor SIN3; regulation of centromeres. Curr Biol, 2003. 13(1): p. 68-72.

680. David, G., et al., mSin3-associated protein, mSds3, is essential for pericentric heterochromatin formation and chromosome segregation in mammalian cells. Genes Dev, 2003. 17(19): p. 2396-405.

681. Ekwall, K., et al., Transient inhibition of histone deacetylation alters the structural and functional imprint at fission yeast centromeres. Cell, 1997. 91(7): p. 1021-32.

682. Shiloh, Y., The ATM-mediated DNA-damage response: taking shape. Trends Biochem Sci, 2006. 31(7): p. 402-10.

683. Traven, A. and J. Heierhorst, SQ/TQ cluster domains: concentrated ATM/ATR kinase phosphorylation site regions in DNA-damage-response proteins. Bioessays, 2005. 27(4): p. 397-407.

684. Wendt, K.S., et al., Cohesin mediates transcriptional insulation by CCCTC- binding factor. Nature, 2008. 451(7180): p. 796-801.

685. Kim, J.S., et al., Specific recruitment of human cohesin to laser-induced DNA damage. J Biol Chem, 2002. 277(47): p. 45149-53.

686. Potts, P.R., M.H. Porteus, and H. Yu, Human SMC5/6 complex promotes sister chromatid homologous recombination by recruiting the SMC1/3 cohesin complex to double-strand breaks. EMBO J, 2006. 25(14): p. 3377-88.

687. Bernard, P., et al., Requirement of heterochromatin for cohesion at centromeres. Science, 2001. 294(5551): p. 2539-42.

688. Nonaka, N., et al., Recruitment of cohesin to heterochromatic regions by Swi6/HP1 in fission yeast. Nat Cell Biol, 2002. 4(1): p. 89-93.

689. Bailis, J.M., et al., Hsk1-Dfp1 is required for heterochromatin-mediated cohesion at centromeres. Nat Cell Biol, 2003. 5(12): p. 1111-6.

690. Williams, S.J., A. Abrieu, and A. Losada, Bub1 targeting to centromeres is sufficient for Sgo1 recruitment in the absence of kinetochores. Chromosoma, 2016.

209

691. Schorl, C. and J.M. Sedivy, Analysis of cell cycle phases and progression in cultured mammalian cells. Methods, 2007. 41(2): p. 143-50.

210