University of Southern California Dissertations and Theses

Understanding the role of BRCA1 and Aurora A kinase in polyploidy development in ovarian carcinoma precursor cells

(USC Thesis Other)

Understanding the role of BRCA1 and Aurora A kinase in polyploidy development in ovarian carcinoma precursor cells

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

UNDERSTANDING THE ROLE OF BRCA1 AND AURORA A KINASE IN
POLYPLOIDY DEVELOPMENT IN OVARIAN CARCINOMA PRECURSOR CELLS

Christine M. Marion

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA

In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
GENETICS, MOLECULAR, AND CELLULAR BIOLOGY

August 2015

Copyright 2015 Christine M. Marion
2
Dedication

I dedicate this dissertation to my mother, the late Maria Rifici Marion who instilled a love
of learning at an early age and a tenacity with which to live my life from a professional
perspective. You have been the silent driving force for my scientific discovery since I
stepped foot in my first lab at the University of Chicago. I will forever be grateful.

3
Acknowledgements

First off, I would like to thank my father, Paul D. Marion, Jr. who has continually
supported me throughout my scholastic endeavors (both emotionally and financially!).
You were always there to listen to me when I hit a road block and your endless words of
encouragement have led me to what I feel is my greatest accomplishment, other than my
two children, of course.

I would also like to thank my brother, John Paul Marion, my sister-in-law, Colleen
Marion, and my niece and nephew, Lilly Marion, and Connor Marion, who supported me
throughout my doctorate and always exuded confidence in my capabilities. At some
point, I started to believe them!

To the rest of my extended New Jersey family, who have been with me in mind and spirit
despite 2500 miles of distance, especially my grandmothers, Elizabeth Marion and the
late Genevieve Rifici: thank you for your love, support and guidance. I will eternally be
grateful for our perfect, crazy family.

To all of my friends, near and far, especially Stacey Ford Lysik: thanks for listening to
me drone on about science and for all of the welcome distractions along the way.

I’d also like to thank my mentor, Dr. Louis Dubeau for granting me the opportunity to
work in his lab and teaching me to think like a scientist. I could not have asked for a
kinder, more available and more understanding PI. To my committee members, Dr.
4
Florence Hofman, Dr. Michael Press, and Dr. Axel Schonthal: thank you for your
suggestions over the years. You have given me invaluable advice and it has been a
pleasure having you on my committee.

To the members of the Dubeau lab, both past and present, thanks for your guidance,
support, and endless discussions, especially Vanessa Yu, PhD, Jennifer Yeh, PhD, Ying
Liu, MD PhD, and Theresa Austria.

Most importantly, to my husband, Jeff Enden: words can never express the gratitude I
have for you. You have truly been my rock without which I could have never dreamed to
complete such a scholastic endeavor. Thank you for your continual support and for
letting me know that anything is possible, even amidst two kids and crazy work
schedules.

Finally to my two magnificent sons, Anthony Leo and Maxwell Joseph: I hope that I can
inspire you to follow your dream, just like my parents taught me. When it comes down
to it, this is really all for you. Thank you for motivating me to be better.

5

Table of Contents

Dedication 2

Acknowledgements 3

List of figures 7

Chapter 1: Introduction 8
1.1 Ovarian Cancer as a Disease 8
1.1.1 Background 8
1.1.2 Molecular Abnormalities Associated with High-Grade 10
Serous Ovarian Carcinomas
1.2 The Role of BRCA1 in Ovarian Carcinogenesis 12
1.2.1 Background 12
1.2.2 BRCA1 as a Ubiquitin Ligase 13
1.2.3 BRCA1 as a Classical Tumor Suppressor Gene: 15
A Scientific Paradox
1.3 Ovarian Carcinogenesis Under the Aura of Aurora A Kinase 16
1.3.1 Overview 16
1.3.2 A Plethora of Results Leading to Inconsistent Conclusions 18
1.4 Mitotic Checkpoint Regulation: A Masterful Process
Maintaining Genomic Stability (Most of the time….) 19
1.4.1 The Big Picture 19
1.4.2 The Role of the Spindle Assembly Checkpoint to
Ensure Chromosome Fidelity 21
1.4.3 Mitotic Regulation Gone Wrong: The Role of
Mitotic Errors on Tumor Development and Prognosis 22
1.5 Cell Models to Study the Development of Aneuploidy 26
1.5.1 Rationale 26

Chapter 2: Role of BRCA1 in Controlling Mitotic Arrest in Ovarian
Cystadenoma Cells 28
2.1 Abstract 28
2.2 Introduction 29
2.3 Materials and Methods 31
2.4 Results 34
2.5 Discussion 48

Chapter 3: Elucidating the Mechanism of Mitotic Arrest Induced by
Aurora A Down-Regulation 52
3.1 Abstract 52
3.2 Introduction 53
3.3 Materials and Methods 55
6
3.4 Results 58
3.5 Discussion 80

Chapter 4: The Road to Aneuploidy in High-grade Serous Ovarian Carcinoma 83
4.1 Abstract 83
4.2 Introduction 84
4.3 Materials and Methods 87
4.4 Results 88
4.5 Table 1 92
4.6 Discussion 97

Chapter 5: Summary and Future Directions 98
5.1 Summary 98
5.2 Future Directions 102

References 106

7
List of Figures

Figure 1: The Spindle Assembly Checkpoint 25
Figure 2: Cells with a similar genetic background to ovarian
carcinoma precursors are predisposed to aneuploidy 41
Figure 3: Differential expression of M phase regulators during ploidy driven crisis 42
Figure 4: Effect of BRCA1 down-regulation on M phase cell cycle regulators 43
Figure 5: BRCA1 down-regulation allows resumption of proliferation
in cells approaching crisis 44
Figure 6: Effect of BRCA1 down-regulation on ploidy 45
Figure 7: BRCA1 down-regulation promotes incomplete cytokinesis 46
Figure 8: A model for BRCA-induced cancer development 47
Figure 9: Influence of replicative age on BRCA1 and Aurora A expression 70
Figure 10: Effect of Aurora A down-regulation on ploidy status and growth rates 72
Figure 11: Effect of Aurora A down-regulation on cell cycle regulators 74
Figure 12: Effect of Aurora A down-regulation on cell cycle progression 75
Figure 13: Effect of Aurora A or BRCA1 down-regulation on cell fate
evaluated by time-lapse photography 76
Figure 14: Effect of Aurora A down-regulation on spindle assembly 77
Figure 15: Effect of Aurora A down-regulation on microtubule anchoring 78
Figure 16: Effect of Aurora A down-regulation on caspase activity 79
Figure 17: Molecular abnormalities associated with ovarian cancer 91
Figure 18: Effect of BRCA1 down-regulation on DNA ploidy in HOC7 cells 93
Figure 19: Effect of BRCA1 down-regulation on DNA ploidy in SKOV3 cells 94
Figure 20: Multinucleation counts following down-regulation of Auora A
and BRCA1 in SKOV3 and CAOV3 cells 95
Figure 21: Effect of Aurora A down-regulation on mitotic regulators 96

8
Chapter 1: Introduction

Ovarian cancer as a disease

1. Background

Ovarian cancer accounts for approximately 3% of cancers among women, causing the
most deaths of all of the reproductive cancers. Survival is hindered due to inadequate
screening methods and because in the early stages of disease there are minimal to no
symptoms. Current screening methods include transvaginal ultrasound (TVU) paired
with CA-125 biomarker sampling, however, these methods have not reduced mortality
rates associated with the disease, demonstrating a need for more sensitive screening
methods
1
. The National Institute of Health estimates that more than 60% of women are
diagnosed with advanced disease, which corresponds to a 5-year survival rate of 27.6%.
Only 15% of women are diagnosed with localized disease corresponding to a much more
favorable 5-year survival rate of 90%, highlighting the disparity in survival between
those diagnosed at earlier versus later stages
2
.
Epithelial ovarian carcinomas (EOCs) account for 85-90% of all cancers of the
ovaries. While literature suggests that the cell of origin is from the coelomic epithelium
on the ovarian surface, there is evidence that these tumors originate from tissues that are
embryologically derived from the Mullerian ducts. By this convention, ovarian epithelial
tumors, fallopian tube carcinomas, and primary peritoneal carcinomas are all of
Mullerian origin and could thus be classified as one disease entity
3
. Ovarian carcinomas
are divided into a variety of histologic subtypes, the most common of which is high-grade
serous carcinomas (40%), followed by endometroid (20%), clear cell (6%) and mucinoid
9
(1%). As with many tumors, histologic subtype dictates many aspects of clinical
treatment, management and prognosis
4
.
In many cases the cause of ovarian cancer is unknown, however there are some
risk factors found to be associated with the disease. Some of these are associated with
risk reduction whereas others are associated with increased risk. Pregnancy is known to
have a protective effect with studies showing up to a 50% risk reduction in women with a
history of one pregnancy. This protective effect increases with increasing parity. Oral
contraception is also associated with a decreased risk. On the other hand, age is
associated with an increased risk, specifically, 55 years of age is the threshold above
which an elevated risk is observed
5
. Finally, genetic predisposition is known to play a
role in approximately 10% of cases and mutations in DNA mismatch repair genes,
MLH1, MSH2, and MSH6 in addition to BRCA1/2 are the most prominent culprits.
Specifically, deleterious mutations in BRCA1 or BRCA2 confer a 20-70% risk of
developing ovarian cancer compared to a 1.4% lifetime risk in the general population.
Although the BRCA1/2 genes are rarely mutated in the non-familial form, they can be
silenced via DNA methylation of the promoter region
6
.
In terms of clinical course, debulking surgery, followed by adjuvant
chemotherapy is the standard of care for women with advanced stage disease. First-line
treatment typically includes Carboplatin, a platinum based agent, in combination with
Paclitaxel (Taxol), a microtubule stabilizer, which inhibits mitosis. For earlier stage
patients, treatment can be curative, however, for advanced stage patients, despite a
favorable response to front-line treatment, more than 80% relapse, usually with disease
that is resistant to chemotherapy. Depending on the duration of response to initial
10
therapy, patients may be treated with the same chemotherapeutic drugs or with other
compounds. Unfortunately, most women with advanced disease will develop many
episodes of recurrent disease with progressively shorter disease-free intervals. One of the
biggest hurdles for ovarian cancer patients and practitioners is the development of
treatment resistant disease, which is the primary reason why ovarian cancer is the fifth
leading cause of death from cancer in women in the United States
7
.

2. Molecular Abnormalities Associated with High-Grade Serous Ovarian
Carcinomas
High-grade serous ovarian carcinoma is a disease of chromosomes. Aneuploidy
is characterized by a change in chromosome number that is not a multiple of the haploid
number. While aneuploidy is a hallmark of many tumors, ovarian carcinoma is unique in
a sense that the only common genetic abnormality associated with these tumors, other
than P53 mutations, is their near tetraploid chromosomal content. These karyotypic
abnormalities are facilitated by the presence of P53 mutations, which are not only present
in nearly all high-grade serous ovarian carcinomas, but in their precursor lesions as well,
a phenomenon known as p53 signature. Pathological studies of samples from BRCA1/2
mutated patients following risk reducing salpingo-oophorectomy identified the distal
fallopian tube as a common site of tumor origin and are aptly known as serous tubal in
situ carcinoma (STIC). STIC-precursor lesions displayed alterations in p53
immunostaining, indicating that mutations in P53 are an early event in oncogenesis
8
.
Consistently, ploidy changes also occur early in tumor development and their magnitude
is a predictor of poor clinical outcome
9
. These ploidy changes are most likely triggered
11
by post-G1 cell cycle errors given their characteristic near polyploidy nature. Potential
mechanisms leading to such changes in DNA ploidy include, but are not limited to
endoreduplication, centrosome amplification, and mitotic slippage
10-12
.
It is generally accepted that it is precisely the karyotypic abnormalities inherent
in these tumors that drives the development of drug resistance in patients. While cancer
is clonal initially, it evolves over time due to different selective pressures from the
microenvironment. The near polyploid state of ovarian carcinomas is genetically
unstable, thus over time, cells will gain or lose chromosomes and acquire mutations,
which confer a more heterogeneous population of cells. Some of these mutations are
irrelevant, however, others confer a growth advantage. Treating tumor cells with
chemotherapy naturally selects for the drug resistant clone to expand, ultimately giving
rise to a population of proliferating cells that are no longer sensitive to the current
treatment modality. The development of drug resistance to conventional therapy has
prompted exploration for new therapies targeting dysregulated molecular pathways
associated with ovarian carcinomas. These include anti-angiogenic agents, such as
Bevacizumab, which was recently approved for the treatment of high-grade serous
ovarian carcinoma in combination with conventional chemotherapy. Poly ADP-ribose
polymerase (PARP) inhibitors represent another promising class of agents aimed at
causing catastrophic DNA double strand breaks, ultimately inducing apoptosis. Multiple
pharmaceutical companies are also allocating resources to the development of tyrosine
kinase inhibitors (TKIs) and many of these agents are currently undergoing clinical trials.
While these agents are better tolerated than standard cytotoxic chemotherapy, more
12
research needs to be conducted to establish firm conclusions with regards to the benefit of
these small molecular inhibitors in the epithelial ovarian cancer population
13
.

The Role of BRCA1 in Ovarian Carcinogenesis
1. Background
Familial ovarian and breast cancer cases typically have mutations in either the BRCA1
or BRCA2 genes. Hereditary ovarian cancer encompasses approximately 10% of all
cases and women who inherit a deleterious mutation in BRCA1 have a 40-60% chance of
developing the disease within their lifetime. While germline mutations do not contribute
to oncogenesis in sporadic ovarian carcinoma cases, studies have demonstrated that the
BRCA1 protein is either absent or present at reduced levels, most likely due to promoter
hypermethylation. These studies suggest a wider role for BRCA1 in controlling not only
hereditary cases, but sporadic cases as well. This notion of reduced BRCA1 levels
leading to earlier onset high-grade serous ovarian carcinoma supports the idea that
BRCAness is not only due to germline mutations but also to epigenetic events, which
alter protein levels
6
.
The first evidence of BRCA1 was provided by the King lab at UC Berkeley in 1990.
BRCA1 lies on chromosome 17q21 and encodes for 24 exons, the largest of which is exon
11, which corresponds to 61% of the coding region and contains the nuclear localization
signal
14
. There are three known products of BRCA1, full-length BRCA1, which encodes
the majority of the gene, BRCA1 delta-11b, which lacks the majority of exon 11, and
BRCA1-IRIS, which terminates within intron 11
15,16
. The full-length transcript is the
most widely studied, and is expressed ubiquitously in all tissues. This isoform encodes
13
for a 220 kD nuclear protein made up of 1836 amino acids. Since its discovery, it has
been widely studied, demonstrating that the gene encodes for a very large, complex
protein known to be involved in DNA damage repair,
17
cell cycle regulation,
18

transcriptional regulation,
19
and it is also known to exhibit ubiquitin ligase activity when
bound to its partner, BARD1
20,21
.
BRCA1’s role in DNA damage repair has been extensively studied and has led to the
development of PARP inhibitors which, based on the mechanism of action, are more
effective in a patient population that has a dysfunctional DNA repair pathway. BRCA1 is
part of a complex that repairs double-strand breaks in DNA, which is a physiological
event that naturally occurs as cells encounter outside factors leading to DNA damage.
BRCA1 employs homologous recombination, where repair proteins utilize a
complimentary sequence from either a sister chromatid, a homologous chromosome, or
from the same chromosome, depending on the phase of the cell cycle, as a template
22
.
BRCA1’s known interaction with RAD51 influences DNA damage repair and these two
proteins, along with BRCA2 play a significant role in maintaining genomic stability
23
.

2. BRCA1 as a Ubiquitin Ligase
In addition to its role in DNA damage repair, BRCA1 is known to act as a
ubiquitin ligase when bound to its partner, BARD1. In this context, BRCA1 exists as a
RING heterodimer with BARD1 and together, they exhibit ubiquitin E3 ligase activity.
The typical ubiquitination pathway involves collaboration between E3 ubiquitin ligase
with two other enzymes, E1, a ubiquitin activating enzyme, and E2, a ubiquitin
conjugating enzyme, to tag substrates for mono or poly-ubiquitination. Polyubiquitin
14
chains target proteins for degradation via the proteosomal pathway, while
monoubiquitination is known to play a role in membrane trafficking, endocytosis, and
viral budding
24
. Prior studies have demonstrated that BRCA1 catalyzes distinct
ubiquitination on different substrates, likely resulting in different fates for different
substrates
25
. In addition, it has been reported that polyubiquitin chains recruit BRCA1 to
damaged DNA
26-29
.
BRCA1’s role as a ubiquitin ligase were elucidated in studies demonstrating that
the protein plays a crucial role in controlling events within the centrosome. Centrosome
duplication is known to take place during the G1 and S phases of the cell cycle, resulting
in two centrosomes that form the poles of the bipolar mitotic spindle. Dysregulation of
this duplication process leads to multipolar spindle formation, chromosome mis-
segregation, and aneuploidy development. BRCA1 localizes to the centrosome and when
bound to BARD1, regulates centrosome duplication during S-phase. In pre-cancerous
breast cell lines, BRCA1 inhibition led to an accumulation of supernumerary
centrosomes
30
. The BRCA1/BARD1 complex is also known to ubiquitinate γ-tubulin, a
centrosomal protein at specific lysine residues. Forced expression of mutant γ-tubulin at
these specific lysine residues resulted in centrosome amplification, thereby revealing a
novel role for BRCA1 in cell cycle regulation
31
.
Another study completed by Parvin’s group established that BRCA1 ubiquitin
ligase activity directly inhibits centrosome-dependent nucleation potential, however, this
is somewhat paradoxical considering that centrosome nucleation potential is at its highest
during mitosis when BRCA1 localization to the centrosome is most prevalent. In an
effort to resolve this conundrum, his lab observed that BRCA1-dependent inhibition of
15
centrosome microtubule nucleation potential was at its highest during S-phase and then
dropped during M-phase, ultimately demonstrating that Aurora A kinase was responsible
for modulating BRCA1’s ubiquitin ligase activity via a phosphorylation event
32
. This
was the first evidence that a post-translational event on BRCA1 was crucial to regulate its
E3 ligase activity, to ensure accurate centrosome duplication and spindle formation,
thereby preventing mitotic errors and ensuring genomic stability.

3. BRCA1 as a Classical Tumor Suppressor Gene: A Scientific Paradox
Upon examination of breast and ovarian tumors from patients harboring BRCA1
germline mutations, who also experienced loss of heterozygosity at the BRCA1 locus, it
was demonstrated that these tumors maintain the mutant allele. This observation is
consistent with Knudson’s two hit hypothesis and led to the speculation that BRCA1
functioned as a classical tumor suppressor gene. Interestingly, loss of BRCA1 is
associated with suppressed growth and reduced cellular proliferation in the early stages of
embryogenesis, which is opposite of what one would expect of a classical tumor
suppressor gene
33
. Furthermore, homozygous BRCA1 mutant mice demonstrate
developmental defects, characterized by smaller embryos
34
. In addition, homozygous
deletion of BRCA1 resulted in embryonic lethality in nearly half of the mice
35
. It has
been suggested that the mechanism of cell death in these embryos is secondary to the
accumulation of chromosomal abnormalities, ultimately leading to activation of a p53 or
p21 dependent cell cycle checkpoint, and hence, early embryonic death
36
. It was also
considered that abolishing p53 activity may allow for survival, despite a null BRCA1
genotype, however, although embryonic survival time increases in the absence of p53,
16
these embryos eventually undergo embryonic death, seemingly due to continued
accumulation of chromosomal abnormalities.
Further characterization of BRCA1 demonstrated a role in the G2/M checkpoint
by activating Chk1 kinase upon DNA damage
37
. A dysfunctional BRCA1 pathway is
also associated with centrosome duplication, ultimately leading to aneuploidy
38
. Taken
together, BRCA1 mutant or null embryos are characterized by premature embryonic
death and reduced growth, which current literature suggests is a result of the
accumulation of chromosomal abnormalities. These chromosomal changes are the result
of a dysfunctional G2/M checkpoint as well as centrosome amplification, which suggests
that BRCA1 plays a crucial role in maintaining chromosomal fidelity by mediating
multiple pathways, including, the G2/M mitotic checkpoint.

Ovarian Carcinogenesis Under the Aura of Aurora A Kinase
1. Overview
Another protein that has been implicated more recently in cancer development is
Aurora A kinase. Aurora A is a member of the serine/threonine kinase family and is a
known mitotic regulator. Activity of Aurora A increases from late G2 onwards and
peaks during prometaphase. The protein localizes to the centrosome during mitosis and
ensures proper chromosome alignment, segregation, and completion of cytokinesis.
Aurora A polymorphisms are associated with an 18-25% increased risk of ovarian
cancer and over-expression is observed in a variety of human tumors originating from
the breast, colon, bladder, pancreas, and ovaries. Overexpression of Aurora A in ovarian
carcinoma cell lines is associated with mitotic dysregulation leading to genomic
17
instability
39
. Based on Aurora A’s implication in cancer, small molecule inhibitors of
the kinase have been developed and are currently undergoing clinical trials.
Potential roles of Aurora A in cell transformation were elucidated in studies
demonstrating that Aurora A kinase is involved in various post-translational
modifications of BRCA1. As stated earlier, one such modification includes a
phosphorylation on BRCA1, which, mediates BRCA1’s ubiquitin ligase activity and
prevents centrosome amplification. In a normal setting, BRCA1 exhibits ubiquitin ligase
activity leading to inhibition of centrosome nucleation potential. Such inhibition prevents
the formation of supernumerary centrosomes, thereby preventing aneuploidy in cells. At
mitosis, when nucleation potential is required, BRCA1’s ubiquitin ligase activity is
blocked when it is phosphorylated by Aurora A kinase. Overexpression of Aurora A
results in dysregulation BRCA1’s inhibitory effects, leading to centrosome amplification
and aneuploidy development
32
.
Another post-translational modification includes phosphorylation of BRCA1 by
Aurora A kinase at Serine 308, which is crucial for entry into mitosis. Previous studies
have demonstrated that, in the face of DNA damage, the unphosphorylated form of
BRCA1 is necessary to maintain the G2/M checkpoint, allowing cells to repair their DNA
prior to committing to mitosis. In an oncogenic setting, in the presence of increased
Aurora A expression, there is dysregulation of this process resulting in aberrant BRCA1
phosphorylation even in the face of DNA damage, leading to a loss of this G2/M
checkpoint and eventual genomic instability
40
.

18
2. A Plethora of Results Leading to Inconsistent Conclusions
Due to Aurora A’s known role in mitotic regulation and its implication in cancer,
multiple studies have been completed in a variety of species and cell types, all leading to
slightly different results. While multiple groups have demonstrated a growth arrest
following inhibition of Aurora A, species, cell type, and genetic background, most
importantly, P53 status, dictate the precise phase at which cells enter an arrest.
Synchronized HeLA cells treated with siRNA targeting Aurora A revealed a requirement
of Aurora A for initial activation of Cyclin B1-Cdk1 at the centrosome, and hence, entry
into mitosis. Cells arrest at G2 and do not commit to mitosis as evidenced by a
substantially reduced mitotic index upon Aurora A inhibition
41
. In contrast, a more
recent publication investigated the effect of a small molecule inhibitor MK8745 felt to be
specific to Aurora A, on multiple cell lines originating from colon, sarcoma, melanoma,
and pancreatic carcinomas. This particular group demonstrated that P53 status dictated
whether cells accumulate in mitosis as evidenced by an increase of p-histone-H3 and in
Aurora A expression levels (P53 wild-type), followed by apoptosis versus a prolonged
delay in mitosis with no apoptosis, resulting in ploidy changes (P53 null)
42
. Adding still
to the confusion is another paper that sought to elucidate the molecular mechanism of
Aurora A regulation in ovarian cancer. In all three ovarian carcinoma cell lines, which
are known to be P53 null, inhibition of Aurora A utilizing shRNA led to a decrease in
centrosome amplification and hence, a reduction in multipolar spindles, suggesting that
Aurora A over-expression can disrupt the formation of bipolar mitotic spindles, hence
affecting normal sister chromatid segregation, ultimately leading to genomic instability.
Furthermore, this group found that inhibition of Aurora A reduces cell cycle progression
19
through an attenuated G1 to S transition versus the standard G2 to M arrest described by
other groups
43
.
Another consideration complicating scientific investigation is the fact that many
of the small molecule inhibitors lack specificity to a single kinase. Studies completed in
HeLa and HCT116 cells have shown a dosage effect with regards to specificity, where a
different phenotype is observed at higher concentrations due to concurrent down-
regulation of Aurora B
44
. Taken together, it is clear that species, cell type, genetic
background, and method of silencing all play a significant role in dictating the function of
Aurora A within disparate experimental models. Despite these varying conclusions, one
consistent observation remains that Aurora A over-expression is associated with poor
clinical outcomes pointing to the validity of the use of these small molecule inhibitors in
clinic.

Mitotic Checkpoint Regulation: A Masterful Process Maintaining Genomic
Stability (Most of the time…)

1. The Big Picture
Both prokaryote and eukaryote survival is dependent on the replication of genetic
material prior to cellular division, resulting in two genetically identical daughter cells. In
eukaryotes, the sexual reproduction strategy typically results in sets of linear
chromosomes, with humans, for example, having two sets of 23. Genome replication and
and division are uncoupled, defining separate S and M phases for DNA synthesis and
segregation respectively. These phases are also separated by gaps or growth phases in
between, known as G1 and G2. Collectively, G1, S, and G2 are referred to as interphase.
Mitosis is divided into prophase, pro-metaphase, metaphase, anaphase, telophase and
20
finally, cytokinesis. During prophase, chromosome condensation begins along with the
formation of the mitotic spindle. Proteins on the nuclear envelope such as Histone H3 are
phosphorylated, which triggers nuclear envelope breakdown. At completion of nuclear
envelope breakdown, metaphase begins and chromosomes align on the metaphase plate,
which is the mid-line equator of the mitotic spindle. Kinetochore attachment to
microtubules is also crucial during metaphase. Anaphase is divided into anaphase A,
where kinetochore fibers shorten to pull sister chromatids to opposing spindle poles and
anaphase B, where spindle elongation push the poles apart causing further migration.
Finally, during telophase, two sets of daughter chromosomes arrive at their respective
spindle poles and a new nuclear envelope initiates prior to cyokinesis which involves the
cytoplasmic division by a contractile ring made up of actin/myosin filaments that
physically pinches the cell, resulting in two daughter cells, each with their own identical
nucleus.
Cell cycle checkpoints exist to ensure accurate division and segregation at G1/S,
S, and G2/M. These checkpoints are regulated by cyclin-cyclin dependent kinase (Cdk)
complexes, where cyclin expression fluctuates throughout the cell cycle. Cdks are only
active when bound to their respective cyclin. In the absence of cyclin B1, for example,
Cdk1 is maintained in an inactive form due to an inhibitory phosphorylations by Wee1
and Myt1 kinases. Activation of Cdk1 is mediated by Cdc25 phosphatase, which is
responsible for dephosphorylation of these inhibitory phospho-sites. These checkpoints
exist to ensure cellular repair in response to damage induced by internal or external
factors. If the DNA damage is too excessive, the cell will initiate apoptotic pathways,
resulting in cell death. These cell cycle checkpoints are controlled by inhibition of
21
cyclin/cdk activity as well as tumor suppressor genes, such as P53. P53 activates p21,
which binds to and inhibits activation of all cyclin/cdk complexes. Dysregulation of the
checkpoint pathways lead to accumulation of chromosomal errors, genomic instability
and potential oncogenic transformation.

2. The Role of the Spindle Assembly Checkpiont to ensure chromosome fidelity
In an effort to ensure accurate segregation of genetic material to two daughter
cells, cohesin, a multimeric protein ring structure, encircles and maintains bound sister
chromatids. Cohesin remains bound until the onset at anaphase at which point, it is
cleaved, chromatid cohesion is lost, and sister chromatids separate, allowing spindle
forces to pull them to opposite sides of the cell. Prior to this event, sister chromatid
cohesion must be maintained until all of the chromosomes are properly aligned on the
metaphase plate. The spindle assembly checkpoint (SAC), also known as the ‘mitotic
checkpoint’ or ‘M-phase checkpoint’, is responsible for monitoring proper alignment of
chromosomes and most importantly, kinetochore-microtubule attachments. Interestingly,
despite the implication in the name, the spindle assembly checkpoint does not monitor
spindle formation. Rather, in the presence of unattached kinetochores, the SAC is
activated and cells are not permitted to progress to anaphase. Progression from
metaphase to anaphase can only occur in the face of an inactive SAC when all
kinetochores are stably bound to microtubules.
The anaphase promoting complex (APC) is the downstream target of the SAC and
is responsible for executing the checkpoint with two cofactors, Cdc20 and Cdh1. The
APC is an E3 ubiquitin ligase that targets several proteins for proteolytic degradation,
22
including the cyclins. The mitotic checkpoint complex (MCC) is made up of Mad2,
Bub3, and BubR1. In the presence of unattached kinetochores, the MCC binds to and
sequesters Cdc20, which is the activating component of the APC responsible for
degradation of securin. When Cdc20 is bound by the MCC, securin levels are
maintained. Securin forms a complex with separase which prevents separase from
cleaving cohesin, thereby preventing sister chromatid separation (Figure 1)
45
. When all
chromosomes are aligned on the metaphase plate with proper kinetochore attachment, the
MCC dissociates from Cdc20 and the APC is considered activated. At this point, Cdc20
is free to bind to and ubiquitinate securin, targeting it for degradation, freeing separase to
cleave cohesin, resulting in progression to anaphase. Anaphase A is regulated by the
APC/Cdc20’s degradation of cyclin B1, which causes Cdk1 activity to fall. Cdh1 is a
target of Cdk1 phosphorylation, thus it is only when Cdk1 activity has dropped below a
certain threshold, that the unphosphorylated form of Cdh1 can bind to the APC, thereby
initiating anaphase B. Anaphase B is regulated by the APC/Cdh1, which targets
polypeptides whose destruction by the proteosome is required for the cell to exit from
mitosis and return to interphase
45
.

3. Mitotic Regulation Gone Wrong: The Role of Mitotic Errors on Tumor
Development and Prognosis

Aberrant mitosis events have long been accepted as the general cause of
chromosomal instability (CIN) in tumors. The majority of solid tumors are not only
aneuploid, but also have acquired mutations in either oncogenes or tumor suppressor
genes such as KRAS, P53, RB1, PTEN, APC, and BRCA1. These observations have led
to a debate over which event is the essential contributor to oncogenesis: aneuploidy or
23
mutations? While mutations in oncogenes and tumor suppressor genes certainly have
effects on proliferation and survival, aneuploidy is the most commonly observed genetic
observation in solid tumors. One possible resolution of these two commonly observed
events is the fact that chromosomal instability increases the chance that loss of
heterozygosity will occur on a tumor suppressor gene locus or perhaps lead to
amplification of oncogene by duplicating the chromosome. CIN is therefore thought to
contribute to tumor formation and is also predicted to contribute to cellular resistance in
oncology patients treated with standard therapeutic agents. The dual roles of CIN in both
tumorigenesis and treatment resistance, leading to poor outcomes in patients, has
prompted the investigation of agents that may inhibit the development of these mitotic
errors.
While normal cells have a robust mitotic checkpoint, where one or more
unattached kinetochores produces a signal that is sufficient to inhibit anaphase
progression, many tumor cells have diminished mitotic checkpoint response.
Interestingly, mitotic checkpoint genes are not commonly mutated in human tumors. It is
believed that mitotic checkpoint impairment and ensuing aneuploidy is a result of
alterations in levels of mitotic checkpoint proteins. These changes could be secondary to
altered transcriptional regulation or due to epigenetic events leading to gene silencing.
Studies aiming to determine the degree to which errors in the mitotic checkpoint
contribute to tumor formation is complicated by the fact that the components of the
mitotic checkpoint complex have roles outside of mitosis. For example, mice with
reduced BUBR1 levels prematurely age, ultimately expiring by 6 months, most likely
before a tumor had a chance to develop
46
. Another group’s investigation of Apc
Min/+

24
mice, who carry one mutant allele encoding for a truncated protein known to cause colon
neoplasms, develop approximately 0.4 colonic tumors per mouse. Crossing this mouse
with a mouse heterozygous for BubR1, thereby resulting in reduced BUBR1 levels,
resulted in a 10-fold increase in the number of tumors per mouse. These tumors were
also of higher histological grade than the Apc
Min/+
mice alone
47
. This group’s data
suggests that a weakened checkpoint might not necessarily drive tumorigenesis, but it
certainly facilitates tumor development, especially when paired with a mutated tumor
suppressor gene.
Taken together, a multitude of studies have been completed in which checkpoint
proteins were experimentally reduced. Tumor development within these models vary,
suggesting that it is dependent upon which gene is disrupted, most likely due to their
roles outside of checkpoint maintenance. It is certainly clear, however, that spindle
assembly checkpoint genes are essential for cell viability. Homozygous knockout mice
are embryonic lethal, however, knocking out one copy of Mad2, Bub3, or BubR1
predisposes mice to aneuploidy and a higher incidence of tumor development
46,47
.

25

Current Biology Vol 22 No 22

Figure 1: The Spindle Assembly Checkpoint
During the early stages of mitosis (prometaphase), unattached kinetochores catalyze the
formation of the mitotic checkpoint complex (MCC) composed of BubR1, Bub3, Mad2,
and Cdc20, leading to the inhibition of the APC/C. Once all of the chromosomes are
aligned on the metaphase place and kinetochore attachment is successful, generation of
the MCC ceases, allowing Cdc20 to activate the APC/C, leading to ubiquitylation and
degradation of securin and cyclin B1. Degradation of securin liberates sepaase which in
turn clearves the Scc1 kleisin subunit of the cohesin ring structure; this opens the ring,
allowing sister chromatids to separate (anaphase). Meanwhle, degredation of cyclin B1
inactivates Cdk1, leading to mitotic exit.

26
Cell Models to Study the Development of Aneuploidy
Our lab uses an in vitro epithelial cell model to study the development of
tetraploidy, which is a pre-cursor to aneuploidy. In this model, benign ovarian
cystadenomas are transfected with a SV40 Large T Antigen expression vector, which
binds to and inhibits Rb and p53 proteins, thus allowing cells to bypass senescence, the
first mortality checkpoint (M1). These cells continue to proliferate until they reach a
second mortality checkpoint known as crisis (M2). A previous student in our lab
demonstrated that this crisis event precedes and is independent of the classical crisis
event thought to be associated with telomere attrition. Rather, the crisis event observed
in our cells is driven by DNA ploidy changes, ultimately culminating in a mitotic arrest
63
.
In order for a cell to reach replicative immortality, it would have to overcome both of
these crisis events.

Rationale
Aneuploidy is a hallmark of many tumors, the development of which has been
linked to oncogenesis and treatment resistance. BRCA1 is known to play a role in
preventing aneuploidy development by exterting DNA damage control and demonstrating
ubiquitin ligase activity to prevent supernumerary centrosomes. I sought to further
elucidate BRCA1’s role in maintaining the physiological mitotic arrest that we observe in
our ovarian cystadenoma cell model. To our knowledge this is the first highly
characterized model of a physiological arrest versus an arrest induced by drug treatment.
My work extended to look for upstream regulators of BRCA1, hence my exploration of
Aurora A kinase. Many small molecule inhibitors of this kinase are undergoing clinical
27
trials, however, resultant phenotypes following inhibition have varied. I sought to
elucidate the mechanism of mitotic arrest mediated by Aurora A in our ovarian
cystadenoma cell model which allowed me to compare and contrast changes in relation to
a well characterized physiological mitotic arrest at the spindle assembly checkpoint.
Finally, I tested if the changes that we observe in our ovarian cystadenoma cell model,
which we regard as a precursor model to ovarian carcinomas, continue beyond oncogenic
transformation in ovarian carcinoma cell lines.

28
Chapter 2: Role of BRCA1 in Controlling Mitotic Arrest in Ovarian
Cystadenoma Cells

ABSTRACT
Cancers that develop in BRCA1 mutation carriers are usually near tetraploid/polyploid.
This led us to hypothesize that BRCA1 controls the mitotic checkpoint complex, as loss
of such control could lead to mitotic errors resulting in tetraploidy/polyploidy with
subsequent aneuploidy. We used an in vitro system mimicking pre-malignant conditions,
consisting of cell strains derived from the benign counterparts of serous ovarian
carcinomas (cystadenomas) and expressing SV40 large T antigen, conferring the
equivalent of a P53 mutation. We previously showed that such cells undergo one or
several doublings of their DNA content as they age in culture and approach the
phenomenon of in vitro crisis. Here we show that such increase in DNA content reflects a
cell cycle arrest possibly at the anaphase promoting complex, as evidenced by increased
expression of the mitotic checkpoint complex. Down-regulation of BRCA1 in cells
undergoing crisis leads to activation of the anaphase promoting complex and resumption
of growth kinetics similar to those seen in cells before they reach crisis. Cells recovering
from crisis after BRCA1 down-regulation become multinucleated, suggesting that
reduced BRCA1 expression may lead to initiation of a new cell cycle without completion
of cytokinesis. This is the first demonstration that BRCA1 controls a physiological arrest
at the M phase apart from its established role in DNA damage response, a role that could
represent an important mechanism for acquisition of aneuploidy during tumor
development. This may be particularly relevant to cancers that have a near
tetraploid/polyploid number of chromosomes.

29
INTRODUCTION

BRCA1, a protein associated with familial ovarian and breast cancer
predisposition,
49,50
acts as an E3 ubiquitin ligase when bound to its binding partner,
BARD1.
51
Investigation of its role in cell cycle regulation has focused primarily on its
control of progression from the G2 to the M phase in response to DNA damage. Specific
serine residues within the BRCA1 protein (Ser1387, Ser1457, and Ser988) are
phosphorylated in response to DNA damage. Such phosphorylation is controlled by ATM
(ataxia telangiectasia mutated), ATR, and Chk2 kinases, which are regulators of the G2/M
cell cycle checkpoint
52-54
. Additionally, BRCA1 seems to be involved in mitosis entry, as
phosphorylation of BRCA1 by the Aurora A Ser/Thr kinase is necessary for cellular G2
to M transition
40
.
There is much less data elucidating the role of BRCA1 on regulation of the M
phase during the cell cycle, although such a role is suggested by the fact that it regulates
centrosome duplication and microtubule nucleation activity
31,32
. Such a role may be an
important underlying mechanism for cancer predisposition in BRCA1 mutation carriers
because depletion of BRCA1 results in the formation of supernumerary centrosomes or
centrosome amplification, a hallmark of genomic instability which may lead to
aneuploidy
32
. Not only are cancers arising from BRCA1 mutation carriers typically
aneuploid, but also the number of chromosomes present in highly aneuploid cancer cells
is often near tetraploid/polyploid. This suggests that the aneuploid state is preceded by
tetraploidy/polyploidy, perhaps due to cytokinesis failure induced by defects in proteins
that comprise the mitotic spindle checkpoint, such as Mad2, BubR1, or Bub3
55-57
.
30
We hypothesized that BRCA1 controls the mitotic checkpoint complex and that
loss of BRCA1 control over this complex may lead to mitotic errors, resulting in
tetraploidy/polyploidy and subsequently to aneuploidy. We sought to test this hypothesis
using an in vitro system that mimics pre-malignant conditions as opposed to established
immortalized cell lines. We reasoned that such a system may provide better insights into
the role of BRCA1 in cancer development. We therefore used cell strains derived from
benign ovarian epithelial tumors
58
. These tumors, called cystadenomas, are the benign
counterpart of the same ovarian cancers that develop in BRCA1 mutation carriers. The
cells were transfected with SV40 large T antigen to increase their in vitro longevity
58
.
The resulting strains have the equivalent of a P53 mutation and the G1 phase of their cell
cycle is constantly activated because SV40 large T antigen binds to and inactivates p53
and Rb. Thus, this cell culture model parallels the situation preceding ovarian carcinoma
development because clonal P53 mutations (p53 signature) are regarded as hallmarks of
precursor lesions for such cancers
59,60
, especially in lesions that are mitotically active and
show dysplastic morphological changes
61
. In addition, most, if not all, high-grade serous
ovarian carcinomas harbor a p53 mutation
62
. We showed earlier that as our cultured
cystadenoma cells age in culture and approach the phenomenon of in vitro crisis, they
become tetraploid/polyploid. Although most of the cells eventually undergo apoptosis, an
occasional cell overcomes crisis and acquires replicative immortality
63
. This may be
reflective of in vivo events that occur during cellular aging and early tumorigenesis.
Here we show that the doubling in DNA content that is typically observed at the
time of crisis in our cell culture model is due to a cell cycle arrest at the M phase that can
be overcome by down-regulation of BRCA1. Our results also suggest that cells that
31
overcome such arrest fail to complete cytokinesis before re-entering a new cell cycle,
resulting in tetraploidy/polyploidy, which in turn may lead to aneuploidy.
MATERIALS AND METHODS
Cell Strains and Culture Conditions. The isolation and characterization of epithelial
strains derived from primary cultures of benign ovarian tumors (cystadenomas) and
expressing SV40 large T antigen was described
58
.

Antibodies. Antibodies specific to p-histone H3 (cat #: 8656-R), BRCA1 (cat #: 645),
cyclin A (cat #: 239), cyclin B1 (cat #: 245), Cdk1 (cat #: 8395), and Cdk2 (cat #: 6248)
(Santa Cruz Biotechnology Inc., Santa Cruz, CA) were used at a 1:200 dilution.
Antibodies against securin (MBL International, Woburn, MA, cat #: DCS-280), Bub3
(cat #: 611730), BubR1 (cat #: 612502), and Cdc27 (cat #: 61054) (BD Transduction
Labs, San Jose, CA) were used at a 1:1000 dilution.

Immunoprecipitation. Nuclear extracts were pre-cleared using 50 L of protein G or L
agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA), followed by centrifugation at
3000 rpm for 5 minutes. Reactions were set up using 200 g of nuclear extract, 2 g of
antibody, protease inhibitors, and triton X-100 to a final volume of 1 mL. Reactions with
antibody only (2 g) or nuclear protein only (200 g) were used as controls. Preparations
were placed on a rocker at 4°C for 2 hours, followed by incubation with 50 L of protein
L or G agarose beads at 4°C overnight. The preparations were washed 5 times by
centrifugation at 2500 rpm for five minutes, followed by re-suspension of the agarose
beads in 20 L of 2x Laemmli sample buffer. The samples were heated to near boiling
32
temperatures for 3 minutes and were electrophoresed on a SDS/polyacrylamide gel
followed by Western blotting.

BRCA1 down-regulation with siRNA. Two different siRNA sequences, both targeting
exon 11 of BRCA1, were used. While 5’-CGAUUUGACGGAAACAUCU-3’ was used
for the experiments shown in all illustrations, critical studies were repeated using 5'-
UGAUAAAGCUCCAGCAGGA-3’.

Quantitative Real-Time PCR. RNA extraction with Trizol Reagent (Invitrogen,
Carlsbad, CA) and cDNA synthesis with oligo dT (Invitrogen, Carlsbad, CA) were
carried out according to manufacturer’s protocols. We added 8 L of cDNA to a total
reaction volume of 20 L also containing Taqman Universal PCR master mix (PE
Applied Biosystems, Foster City, CA), probe and primers. The sequence of the BRCA1
probe was 5’-AAAATAATCAAGAAGAGCAAAGCATGGATTCAAACT- 3’ while
sequences of the forward and reverse primers were 5’-AAGAGGAACGGGCTTGGAA-
3’ and 5’-CACACCCAGATCCTGCTTCA-3’ respectively. Signal intensities were
determined using the ABI Prism 7700 Sequence Detection System (PE Applied
Biosystems, Foster City, CA).

DNA Profiling Studies. Cells were trypsinized, re-suspended in 0.2 mL of PBS, and
fixed in 2 mL of 70% ethanol.

After centrifugation, the cell pellets were resuspended in 1

mL of PBS, 10 µg/mL propidium iodide, and 100 g/mL

RNase. Fluorescence was
measured on a Coulter Profile II flow

cytometer (Beckman Coulter Inc., Fullerton, CA).
33

Metaphase Spreads. Cells were trypsinized, fixed in methanol, dropped onto glass
sides, dried at room temperature, stained with hematoxylin and eosin following published
protocols
87
, and examined under a light microscope.

Confocal Microscopy. Multi/bi-nucleated cells in siBRCA1-treated dishes were
observed using the PerkinElmer Spinning Disk Confocal Microscope with 10X
magnification and captured using the Ultra VIEW ERS system. DNA was visualized
with DAPI stain using a 405nm excitation wavelength. Tubulin was visualized using
rabbit anti-α tubulin-Alexa Fluor 647 (Cell Signaling Technology, Inc., Danvers, MA)
with excitation wavelength of 640nm.

Time-Lapse Photography Using Confocal Microscopy. ML10 cells were grown on
chambered glass cover slides and treated with siRNA against GFP (control) or Aurora A
for 48 hours then transferred to a confocal microscope fitted with an incubator under
controlled temperature and CO
2
conditions. Cells were filmed for 20 hours using a
PerkinElmer Spinning Disk Confocal Microscope with 10X magnification and captured
using the Ultra VIEW ERS system. Photographs were taken every minute for 20 hours.

34

RESULTS

Cells with a similar genetic background to ovarian carcinoma precursors are
predisposed to aneuploidy

Our lab utilizes epithelial cell strains derived from benign ovarian cystadenomas
to study the development of tetraploidy, which is a precursor to aneuploidy. These cells
are transfected with SV40 LTA, which allows them to bypass senescence (M1), the first
mortality checkpoint due in part to P53 inhibition. Considering that >90% of high-grade
serous ovarian carcinomas are P53 mutant and the fact that P53 mutations are thought to
occur early in oncogenesis, we feel that this model is an accurate representation of the
genetic background of ovarian carcinoma precursors. In this in vitro model, cells
continue to proliferate beyond senescence until they reach a second mortality checkpoint
known as crisis (M2) (Fig 2, upper panel). We previously demonstrated that this crisis
event is secondary to the development of a specific type of genomic instability,
characterized by numerical chromosomal alterations, that is independent of the classical
crisis event associated with telomere attrition
63
. An example of the ploidy changes that
we observe can be seen in the lower panel of Figure 2, where at low passage (<35
population doublings), the DNA profiles show no appreciable abnormalities, however at
high passage (>50 population doublings), the emergence of a proliferating tetraploid
population is apparent. An accumulation of cells in the 4N peak is also apparent at
higher passages, suggesting that cells experiencing these ploidy changes eventually
undergo a growth arrest. We also observed decreased BRCA1 levels as cells age in
culture as evidenced by the western blot shown (Fig 2, lower panel), and set out to
35
investigate whether this reduction in BRCA1 levels was permissive to allow cells to exit
the growth arrest and continue proliferating.

Ovarian cystadenomas undergo a mitotic arrest at the end of their in vitro lifespan
Although replicative senescence is overcome due to SV40 large T antigen
expression in our cultured ovarian cystadenomas, this is not sufficient for the acquisition
of replicative immortality as the cells eventually reach crisis. We previously showed that
crisis is preceded by one or several doublings in genomic DNA content
63
. We also
observed an accumulation of cells in the 4N peak as they age and suspected that aging
cells undergo a mitotic arrest a G2/M.
A previous student in lab observed a global decrease in the proportion of cells
undergoing DNA synthesis as measured by BrdU incorporation in cells at high passage,
supporting the notion that high passage cells undergo a cell cycle arrest post-G1
64
. We
sought to further support the presence of a cell cycle arrest by comparing the expression
and activity of various cell cycle regulators associated with G2 and M phases in cells
cultured in vitro at early passage or as they approached crisis. We focused on regulators
of the post-G1 phase because we reasoned that an arrest at an earlier phase would not
affect DNA content. Cyclin B1 and cyclin A, which respectively control progression
through the G2 and S phases by binding to and activating Cdk1 and Cdk2, were
immunoprecipitated from nuclear extracts obtained at early and late passages, followed
by Western blotting and hybridization to antibodies against Cdk1 and Cdk2. The results
showed a significant increase in Cdk1 activity in older cells when compared to younger
cells. Although the results for Cdk2 activity in the same cell populations were not
statistically significant, the trend showed a decrease as opposed to an increase in
36
enzymatic activity
64
. An increase in Cdk1 activity without a concomitant increase in
Cdk2 is further supportive of cells undergoing a G2-M arrest. The anaphase promoting
complex (APC) controls anaphase initiation. Its activation is inhibited by mitotic
checkpoint complex (MCC) proteins such as Mad2, BubR1, and Bub3 until all
chromosomes are properly aligned
65,66
. Given that BubR1 and Bub3 form a complex to
inhibit APC activation
67
, we sought to determine if this complex was differentially
expressed in cells cultured at an early passage compared to cells undergoing crisis. Bub3
was immunopreciptated in nuclear extracts collected at early and late passages, followed
by Western blotting using antibody probes against BubR1 and Cdc27. The results (Fig
3A) showed that Bub3-BubR1 interaction was up-regulated during crisis, indicating a
role for the MCC in the maintenance of the cell cycle arrest. Furthermore, interaction
with Cdc27, a component of the APC which catalyzes the ubiquitination of cycle B1
68

and signals exit from mitosis, was decreased during crisis (Fig 3A). Securin, another
product of APC ubiquitination
69,70
whose degradation acts as a signal to exit from
mitosis, was expressed at higher levels in nuclear extracts harvested from cultures
approaching crisis compared to extracts from younger cells (Fig. 3B). Levels of p-
histone H3, a marker for prophase activation, were also increased in cells approaching
crisis (Fig 3B). Taken together, these results strongly support the presence of an arrest at
the M phase in cells undergoing crisis.

Role of BRCA1 in controlling the M phase arrest associated with crisis
We hypothesized that BRCA1, a multi-functional protein involved in many
aspects of cell cycle regulation
71,72
, DNA repair
73
and chromatin remodeling
74
, may play
37
a role in the mitotic arrest associated with crisis because cancers arising in BRCA1
mutation carriers are often near polyploid. The mitotic arrest in cells approaching crisis is
characterized by elevated levels of p-histone H3 and Bub3-BubR1 and decreased levels
of Cdc27 (Fig. 3). We examined the expression of these mitotic regulators in nuclear
extracts harvested from cells in which siRNA had been used to down-regulate either
BRCA1 or GFP as a control. BRCA1 down-regulation led to a decrease in P-histone H3
and BubR1 levels and to an increase in levels of Cdc27 in cells approaching crisis while
expression of these proteins was unchanged in younger cells (Fig. 4). These results are
opposite to those observed in cells approaching crisis. We conclude that these proteins
act downstream of BRCA1 and play a role in the escape from mitotic arrest observed
following BRCA1 down-regulation. Similar studies using cells younger than 35
population doublings showed essentially no effect of BRCA1 down-regulation on the
expression of any of those mitotic regulators (Fig. 4).
We investigated whether down-regulation of BRCA1 in cells approaching crisis
would return their proliferation rate to the level seen in younger cells. Full length BRCA1
expression was down-regulated using siRNA targeting the C-terminus of this protein in
cystadenoma cell cultures either at the onset of crisis (Fig. 5B) or at early passages (Fig
5A) to verify this prediction. Growth curve analyses showed marked differences in the
growth rate of cultures at early passages compared to cells approaching crisis in controls
treated with siGFP (solid lines in Fig. 5), with doubling times of 1.3 and 2.7 days
respectively. Although down-regulation of BRCA1 (dotted lines in Fig. 5) did not
influence cell growth at early passages (Fig. 5A), it allowed cells undergoing crisis to
38
resume logarithmic growth and to sustain a doubling time of 1.2 days, which is similar to
the rates observed at early passages (Fig. 5B).

Reduced BRCA1 expression leads to cellular multinucleation, due to resumption of
the cell cycle without completion of cytokinesis

Cultures of cystadenoma cells approaching crisis include a subpopulation of
actively dividing tetraploid cells, as evidenced by the 8N peak present in their DNA
profiles (Fig. 1). We sought to test the hypothesis that BRCA1 down-regulation would
lead to an increase in tetraploid/polyploidy cells. Ovarian cystadenomas cultured either at
an early passage or while approaching crisis were treated with siRNA against BRCA1 or
GFP as a control for 72 hours. The cells were stained with propidium iodine and
analyzed for their DNA content by flow cytometry. There was no change in the DNA
profile of cells at early passages following treatment with siRNA against BRCA1 (Fig.
6A, top panel). In contrast, cells approaching crisis showed an increase in their
tetraploid/polyploid population following such treatment (Fig. 6A, bottom panel).
Additionally, 7 of 21 metaphase spreads of cells approaching crisis and treated with
siRNA against BRCA1 were near tetraploid (92 ± 5 chromosomes) compared to one of
21 metaphase spreads from cells treated with siRNA against GFP (Fisher’s exact t test:
0.03).
A plausible explanation for the increase in DNA content in cells treated with
siRNA against BRCA1 is that while most cells subjected to the M phase arrest eventually
undergo apoptosis, BRCA1 down-regulation allows some cells to overcome this arrest
and enter a new cell cycle without completion of the earlier cycle and, thus, without
cytoplasmic separation (cytokinesis). A prediction of this hypothesis would be that cells
39
that overcome the M phase arrest due to BRCA1 down-regulation should become bi- or
multi-nucleated. Cells at approximately 35 and 50 population doublings were treated with
siRNA against either BRCA1 or negative control scrambled sequence. The proportion of
cells in each population that were bi-/multi-nucleated was calculated from 9 random
microscopic fields by 3 independent observers who each evaluated 3 fields from each cell
population. BRCA1 down-regulation in younger cells led to an increase of borderline
statistical significance (P = 0.04) in bi/multi-nucleation from 5% to 15% (Fig. 6B, left
panel). In contrast, BRCA1 down-regulation in cells approaching crisis led to a much
larger and more significant (P = 0.002) increase in bi/multinucleation (8% to 34%, Fig.
6B, right panel). The difference in the proportion of multinucleated cells at high versus
low passage after treatment with siBRCA1 was also significant (2-sided P = 0.004,
unpaired t-test). A confocal image of a binucleated cell is shown in Fig. 6C.
These results suggest that cells treated with siRNA against BRCA1 exit the M
phase arrest and resume proliferation without completing cytokinesis. We corroborated
these results utilizing live-cell imaging to follow the fate of each cell following BRCA1
down-regulation with siRNA. Cells cultured at >50 population doublings were treated
with siRNA against a negative control scramble sequence or BRCA1 for 48 hours, then
transferred to a temperature and CO
2
controlled incubation chamber, mounted on a
confocal microscope. Time-lapse photographs were taken every minute over 20 hours.
In the cells treated with siRNA against BRCA1, there were many mono-nucleated cells
that eventually overcame the M phase arrest and exited mitosis without completing
cytokinesis, resulting in binucleated cells. Screenshots from such an example can be seen
in Figure 7A. Specifically, we observed a 4-fold increase in cytokinesis failure in cells
40
treated with siBRCA1 compared to cells treated with negative control siRNA (p=0.026, 2
sided t-test, Fig. 7B). As expected, cells treated with siRNA against BRCA1 were less
susceptible to cell death, compared to control cells (p-0.0083, 2 sided t-test, Fig. 7B).
Taken together, these findings confirm that aging cystadenoma cells eventually undergo a
physiological M phase arrest, which can be overcome by forced reduction of BRCA1
levels, leading to a proliferating tetraploid population, due to failure of cytokinesis.

41

Figure 2: Cells with a similar genetic background to ovarian carcinoma precursors
are predisposed to aneuploidy

Our lab utilizes epithelial cell strains derived from benign ovarian cystadenomas to study
the development of tetraploidy, which is a precursor to aneuploidy. These cells are
transfected with SV40 LTA, which allows them to bypass senescence, the first mortality
checkpoint due in part to P53 and Rb inhibition. These cells continue to proliferate until
they reach a second mortality checkpoint known as crisis. We previously demonstrated
that this crisis event is secondary to DNA ploidy changes and is independent of the
classical crisis event associated with telomere attrition. An example of the ploidy
changes that we observe can be seen in the lower panel, where at low passage (<35
population doublings), the DNA profiles look quite normal, however at high passage
(>50 population doublings), the emergence of a proliferating tetraploid population is
apparent. We also observed decreased BRCA1 levels as cells age in culture as evidenced
by the western blot shown. Ku70 was used as a loading control.

*The BRCA1 western blot for BRCA1 was completed by Vanessa Yu, PhD

Senescence
M2
Crisis (Apoptosis)
M1
Rb; p53

ML3, ML10
Mortal Cell
Strains
Logarithmic
Growth
SV40 LTA
2N 4N 8N 2N 4N 8N
DNA profiles
Velicescu M. et al., 2003
<35 PD >50 PD >50 PD
BRCA1
Ku-70
<35 PD
Telomerase
Activation
Inactive Cell Cycle
Replicative
Immortality
42

Figure 3: Differential expression of M phase regulators during ploidy driven crisis
Nuclear extracts harvested from cells at early passage (<35 population doublings) and at
the onset of crisis (>50 population doublins) were immunoprecipitated with Bub3
antibody followed by Western blotting against either BubR1 or Cdc27. Western blots of
Bub3 and Ku70 (loading control) using the same nuclear extracts are shown (a). Western
blotting was performed on nuclear extracts collected from early and late passage cells
using antibodies against securin and phosphorylated histone H3. Ku-70 served as a
loading control (b).

*The western blot in this figure was completed by Vanessa Yu, PhD

43

Figure 4: Effect of BRCA1 down-regulation on M phase cell cycle regulators
Western blotting was performed on nuclear extracts collected from cells either young
than 35 population doublings or approaching crisis and treated with siRNA against either
GFP or BRCA1 for 72 hours using the indicated antibody probes.

44

Figure 5: BRCA1 down-regulation allows resumption of proliferation in cells
approaching crisis

Cells either approaching crisis (B) or cultured at early passage (A) were treated with
siRNA against either GFP (control) or BRCA1. Growth curves were initiated in parallel
in each cell population starting 24 hours after the siRNA treatment and stopped at 120
hours. Cell numbers were determined from a minimum of 3 dishes using a Z2 Coulter
Particle Count and Size Analyzer (Beckman Coulter, Inc., Brea, CA).

45
Figure 6: Effect of BRCA1 down-regulation on DNA ploidy
(a) Cells with either less than 35 or more than 50 population doublings were treated with
siRNA against GFP or BRCA1 for 72 hr, fixed with 70% ethanol, and stained with
propidium iodide for DNA profiling analyses using flow cytometry. (b) Three
independent observers, two of which were blinded regarding siRNA treatment, each
calculated the proportion of bi/multinucleated cells in three separate microscopic fields
using the 32 objective of a Zeiss IM35 phase contrast microscope for each cell population
of interest. The results are expressed as averages of each of the nine data points for each
cell population 6 standard error. c: Confocal images of two binucleated cells stained for
a-tubulin and counterstained with DAPI
*The DNA profiles were completed by Vanessa Yu, Ph
46
Figure 7: BRCA1 down-regulation promotes incomplete cytokinesis
Cells cultured at >50 population doublings were treated with siRNA against a negative
control sequence or BRCA1 for 48 hours and then transferred to an incubation chamber
under controlled temperature and CO
2
mounted on a confocal microscope. Time-lapse
photographs were taken every minute over 20 hours. Many cells that were arrested in
mitosis eventually exited mitosis without undergoing cytokinesis, resulting in binucleated
cells (A). The end fate of each cell entering mitosis was recorded and followed by two
independent observers. Percent of cells experiencing incomplete cytokinesis and cell
death were quantified (B).
47

Figure 8: A model for BRCA-induced cancer development
The first step in this tumorigenesis model is based on the notion that cancer, in general,
arises in a background of increased cellular proliferation. We suggest that in BRCA1
mutation carriers, such increase in proliferative activity is induced by a cell
nonautonomous mechanism characterized by increased and prolonged estradiol secretion
by ovarian granulosa cells.
42,43
Increased proliferation eventually leads to senescence,
which can be overcome in the presence of mutations in cell cycle regulators such as RB
and p53. Indeed, p53 mutations are present in nearly 100% of high-grade serous ovarian
carcinomas. Eventually, the cells reach the second mortality checkpoint, crisis,
characterized by an M phase arrest as described in this manuscript. Under normal
circumstances, cells subjected to such arrest eventually die from apoptosis. Reduced or
absent BRCA1 expression facilitates overcoming the M phase arrest. Such recovery,
however, occurs without completion of cytokinesis. This leads to tetraploidy/polyploidy,
a genetically unstable state that culminates in aneuploidy and cancer development
48
DISCUSSION

Our results clearly show that the doubling in DNA content previously observed in
cultures of ovarian cystadenomas approaching crisis is due to a cell cycle arrest at the M
phase. Our results also show that decreased expression of BRCA1, a protein controlling
familial ovarian cancer predisposition, allows cells to overcome this M phase arrest,
leading to resumption of proliferation rates similar to those seen in cells cultured at early
in vitro passages, before they reach crisis. Cells that recover from crisis based on this
mechanism become tetraploid/polyploid and bi/multinucleated, suggesting that such
recovery may be characterized by cell cycle re-entry without cytokinesis.
Reduced BRCA1 expression allows cells to overcome a physiological M phase
arrest associated with changes in expression of mitotic regulators that comprise the
anaphase promoting complex. Although a definitive proof that BRCA1 directly controls
this complex is still lacking, the results nevertheless suggest that BRCA1 influences
mitotic progression upstream of the complex. Previous studies have shown that BRCA1
plays an important role in DNA damage repair by activating the G2/M checkpoint in
response to ionizing radiation, and loss of BRCA1 function causes abrogation of this
checkpoint and accumulation of cell in mitosis
75,76
. To our knowledge, our results are the
first to show that decreased BRCA1 expression leads to overcoming a physiological
arrest at the M phase, apart from its established role in DNA damage response. We have
not yet characterized this novel function of BRCA1. The mitotic regulators Bub3,
BubR1, Cdc27 are prime candidates to play a critical role in M phase arrest backed by
BRCA1 as suggested in Fig. 3. Another possibility would be known interactions between
49
BRCA1 and cell cycle-regulatory proteins
37
, as such mechanisms may not only control
the G2/M checkpoint, but could also be involved in M phase exit.
The idea that BRCA1 may function as a classical tumor suppressor comes
primarily from the observation that breast and ovarian tumors from patients harboring
BRCA1 mutations and showing loss of heterozygosity affecting the BRCA1 locus
typically show retention of the mutant allele, implying a survival advantage for tumors
lacking a functional BRCA1
77-79
. However, contrary to what would be expected of a
classical tumor suppressor, cancer cell lines derived from such tumors are typically
extremely difficult to culture in vitro, with doubling times measured in weeks rather than
days
80,81
. Additionally, loss of BRCA1 activity often results in growth retardation and
reduced cellular proliferation during early embryogenesis
33
. The notion that BRCA1 acts
as a classical tumor suppressor also fails to account for the fact that cancers arising in
BRCA1 mutation carriers, following chemotherapy, are under selective pressure to
undergo small deletions within the mutant BRCA1 allele, resulting in re-establishment of
the original reading frame and recovery of BRCA1 activity
82
. In addition, a classical
tumor suppressor mechanism, by itself, would not account for the fact that while BRCA1
is expressed in most cell types, germline mutations in this gene are associated with
predisposition to breast and female reproductive cancers only. To our knowledge, our
studies are the first to show a direct growth advantage associated with decreased BRCA1
activity. Although no such advantage could be seen with cells cultured at early passages,
decreased BRCA1 expression became clearly advantageous as the cells aged in culture
and approached crisis.

50
The diagram shown in Figure 8 summarizes our working hypothesis regarding
how a tumor suppressor mechanism and could cooperate with the cell non-autonomous
scenario previously reported by our group based on studies with experimental animals
83

to increase cancer predisposition in individuals with BRCA1 abnormalities. A
proliferation signal is the first step, as it is well known that cancer typically develops in a
background of increased proliferative activity. In the context of familial breast and
ovarian cancer development, we suggest, based on our earlier work, that such
proliferation signal is driven by a cell non-autonomous mechanism triggered by reduced
BRCA1 expression that leads to increased estrogen stimulation unopposed by
progesterone
84
. Increased proliferation eventually leads to the first mortality checkpoint,
the equivalent of in vitro senescence, which can be overcome in the presence of certain
genetic abnormalities such as P53 mutations, which are invariably present in tumors that
develop in BRCA1 mutation carriers
85
. The next mortality checkpoint, crisis, is
associated with a mitotic arrest, the existence of which is demonstrated in this chapter.
Although this usually leads to apoptosis, reduced or absent BRCA1 expression facilitates
escape from this arrest and cell cycle re-entry without completion of cytokinesis. This
leads to tetraploidy/polyploidy and ensuing aneuploidy, a hallmark of cancer.
The notion that the tumor suppressor function of BRCA1 comes, at least in part,
from its role in maintaining a mitotic arrest triggered by normal physiological phenomena
may explain why breast cancer cell lines that lack a functional BRCA1, such as
HCC1937, have highly abnormal karyotypes
17
. The idea that an escape from mitotic
arrest leading to tetraploidy/polyploidy and subsequent aneuploidy plays a role in
mediating BRCA1 induced tumorigenesis is also consistent with the fact that breast and
51
ovarian carcinomas that develop in individuals who carry germline BRCA1 mutations are
typically high grade and near tetraploid/polyploid
86.
Thus, our findings may be
particularly relevant to the development of cancers with near tetraploid/polyploid
karyotypes.

52
Chapter 3: Elucidating the Mechanism of the Mitotic Arrest Induced
by Aurora A Down-regulation

ABSTRACT
Aurora A belongs to a family of serine/threonine kinases known to play a role in
mitotic regulation. More recently, the kinase has been implicated in tumorigenesis.
Small molecule inhibitors of Aurora A are currently undergoing clinical trials, though
their effects on cellular phenotype have varied based on species, cell type, and P53 status.
While some studies have demonstrated a prolonged G2/M arrest post-treatment, others
have observed catastrophic polyploidy, leading to cell death. Investigations based on
these inhibitors are further complicated by their lack of specificity for a single kinase.
We sought to elucidate the mechanism of mitotic arrest mediated by Aurora A
and utilized siRNA technology to down-regulate Aurora A levels in order to avoid
potential complications due to lack of specificity of current small molecule inhibitors.
We also used a pre-cancerous cell model that we had previously extensively
characterized in order to avoid genetic defects inherent to carcinoma cell lines, which
would either enhance or limit the mitotic changes that we set out to observe. We
previously showed that such cells undergo a physiological mitotic arrest at the spindle
assembly checkpoint as they approach in vitro crisis, which can be overcome by reduced
BRCA1 levels. Multinucleation ensues as a result of incomplete cytokinesis. We
reasoned that Aurora A might exert some control over these phenomena given its
previously described role in post-translational modifications of BRCA1. Here we show
that inhibition of Aurora A leads to a mitotic arrest downstream of the spindle assembly
checkpoint between anaphase A and anaphase B due to cyclin B1 stabilization. This
53
arrest is not mediated by any effect of Aurora A on BRCA1. We suggest that this
represents a previously undescribed step at the M phase checkpoint protecting against
mitotic errors leading to polyploidy.

INTRODUCTION
Aurora A belongs to a family of serine/threonine kinases that have been widely
studied, all literature suggesting that these kinases play varying and crucial roles in
mitotic regulation. More recently, Aurora A has been implicated in tumorigenesis in a
variety of histologic cell types and is thought to be an oncogene when over-expressed.
Over-expression of the kinase is associated with aneuploidy and poor clinical outcomes
in patients
39
. Based on Aurora A’s implication in cancer, small molecule inhibitors of the
kinase have been developed and are currently undergoing clinical trials. Resultant
phenotypes following Aurora A inhibition with these small molecule inhibitors have
varied, indicating that species, cell type, and P53 status dictate Aurora A’s function
within these disparate experimental models
88-90
. While some studies demonstrate a
prolonged G2/M arrest
91
, others groups have observed catastrophic polyploidy, ultimately
leading to cell death
92
. Another consideration, which complicates scientific investigation,
is the fact that many of these small molecule inhibitors are not specific to one kinase.
Alisertib, also known as MLN8237, is generally thought to target Aurora A specifically,
thereby inducing a mitotic arrest at G2/M, however, HeLa and HCT116 cells treated with
higher concentrations experienced mitotic slippage or failure of cytokinesis and hence
polyploidy, secondary to concurrent down-regulation of Aurora B
44
.
54
We sought to elucidate the mechanism of mitotic arrest mediated by Aurora A
and used siRNA technologies to down-regulate Aurora A levels in order to avoid
potential complications due to lack of specificity of current small molecule inhibitors.
We also utilized a pre-cancerous cell model that we had previously extensively
characterized in order to avoid potential complications due to unknown genetic defects
associated with carcinoma cell lines and to investigate the role of Aurora A in the context
of a well-characterized mitotic checkpoint associated with mortal (not fully transformed)
cell cultures. This cell model is derived from ovarian serous cystadenomas, the benign
counterpart of ovarian serous carcinomas, which harbor the equivalent of a P53 mutation
due to forced expression of SV40 Large T Antigen
64
. We previously showed that such
cells undergo a physiological mitotic arrest at the spindle assembly checkpoint as they
approach in vitro crisis, which is associated with ploidy changes versus the classical crisis
event associated with telomere attrition. This crisis event can be overcome by reduced
levels of BRCA1, leading to multinucleation, due to incomplete cytokinesis
64
. We
reasoned that Aurora A might exert some control over these phenomena given its role in
post-translational modifications of BRCA1, including phosphorylation of BRCA1 at
Ser308, which regulates BRCA1’s role in G2 to M transition in the phase of DNA
damage
40
.
Our results demonstrate that Aurora A acts upstream of BRCA1 to mediate
BRCA1 expression levels, however, inhibiting Aurora A does not simply undo the
cytokinesis failure phenotype that we observe following BRCA1 inhibition. Rather,
inhibition of Aurora A leads to an independent mitotic arrest that is downstream of the
spindle assembly checkpoint, an interesting finding considering that the established
55
dogma suggests that this checkpoint marks the point of no return with regards to mitotic
regulation. Specifically, Aurora A inhibition induces a mitotic arrest in between
anaphase A and anaphase B due to cyclin B1 stabilization, thereby protecting against
mitotic events leading to polyploidy.

MATERIALS AND METHODS

Cell Strains and Culture Conditions. The isolation and characterization of epithelial
strains derived from primary cultures of benign ovarian tumors (cystadenomas) and
expressing SV40 large T antigen was previously described
58
.

Antibodies. Antibodies against the phosphorylated (cat #: 2914S) and unphosphorylated
(cat #: 4718S) forms of Aurora A (Cell Signaling Technology Inc., Danvers, MA) were
used at a 1:1000 dilution. Antibodies against Ku70 (cat #: 12729), BRCA1 (cat #: 642),
Mad2 (cat #: 47747) (Santa Cruz Biotechnology Inc., Santa Cruz, CA) were used at a
1:200 dilution. Antibodies against Bub3 (cat #: 611730), BubR1 (cat #: 612502), and
Cdc27 (cat #: 61054) (BD Transduction Labs, San Jose, CA) were used at a 1:1000
dilution.

Immunoprecipitation. Nuclear extracts were pre-cleared using 50 uL of protein G
beads (Santa Cruz Biotechnology, Santa Cruz, CA), followed by centrifugation at 3000
rpm for 5 minutes. Immunoprecipitations were set up using 200 ug of nuclear extract, 2
56
ug of antibody, protease inhibitors, and Triton X-100 in a final volume of 1 mL.
Reactions with antibody only (2 ug) or nuclear protein only (200 ug) were used as
controls. Preparations were placed on a rocker at 4°C for 2 hours, followed by incubation
with 50 uL of protein G agarose beads at 4°C overnight. The preparations were washed 5
times by centrifugation at 2500 rpm for five minutes, followed by re-suspension of the
agarose beads in 25 uL of 2x Laemmli sample buffer. The samples were heated to 85
degrees Farenheight for 3 minutes and were electrophoresed on a 10%
SDS/polyacrylamide gel followed by Western blotting.

Down-regulation with siRNA. Ovarian cystadeoma cell strains (ML-10 & ML-3) were
transfected with two different siRNA sequences targeting Aurora A for 72 hours. While
5’-AUGCCCUGUCUUACUGUCA-3’ was used for the experiments shown in all
illustrations, critical studies were repeated using 5'-AUUCUUCCCAGCGCAUUCC-3’
(Supplementary Fig. 1 and 2). Conditions and siRNA sequences for BRCA1 down-
regulation were published earlier
64
.

Quantitative Real-Time PCR. RNA extraction with Trizol Reagent (Invitrogen,
Carlsbad, CA) and cDNA synthesis with Oligo(dT) (Invitrogen, Carlsbad, CA) were
carried out according to manufacturer’s protocols. We added 8 uL of cDNA to a total
reaction volume of 20 uL also containing Taqman Universal PCR master mix (PE
Applied Biosystems, Foster City, CA), probe and primers. The sequence of the BRCA1
probe was 5’-AAAATAATCAAGAAGAGCAAAGCATGGATTCAAACT- 3’ while
sequences of the forward and reverse primers were 5’-AAGAGGAACGGGCTTGGAA-
57
3’ and 5’-CACACCCAGATCCTGCTTCA-3’ respectively. Signal intensities were
determined using the Fast Real-Time 7900HT PCR System (PE Applied Biosystems,
Foster City, CA). The Aurora A primer/probe set was acquired from Applied Biosystems
(Catelogue No. 4448489).

DNA Profiling Studies. Cells cultured on plastic dishes were trypsinized, re-suspended
in 0.2 mL of PBS, and fixed in 2 mL of 70% ethanol.

After centrifugation, the cell pellets
were resuspended in 1

mL of PBS, 10 µg/ ug/mL

RNase.
Fluorescence was measured on a EPICS XL-MCL flow cytometer (Beckman Coulter
Inc., Fullerton, CA).

Quantification of Multinucleated Cells. ML10 cells were treated with siRNA against
GFP (control) or Aurora A for 72 hours followed by fixation in 2% formaldehyde. The
number of multi-nucleated cells per high magnification (35x) field was determined in
triplicate dishes by 2 independent observers using a Zeiss IM 35 microscope. Three
fields per dish were examined by each observer. The results are expressed as an average
of the total of 18 observations for each condition.

Time-Lapse Photography Using Confocal Microscopy. ML10 cells were grown on
chambered glass cover slides and treated with siRNA against GFP (control) or Aurora A
for 48 hours then transferred to a confocal microscope fitted with an incubator under
controlled temperature and CO
2
conditions. Cells were filmed for 20 hours using a
58
PerkinElmer Spinning Disk Confocal Microscope with 10X magnification and captured
using the Ultra VIEW ERS system. Photographs were taken every minute for 20 hours.

RESULTS
Inverse correlation between Aurora A and BRCA1 expression levels
We sought to determine whether intracellular levels of Aurora A, either directly
or indirectly, influence BRCA1 expression given that BRCA1 is a known substrate of
Aurora A. We examined and compared levels of these two proteins in cultured ovarian
cystadenomas at early (<35 population doublings) and late (>50 population doublings) in
vitro passages in order to evaluate changes in their relative abundance in response to
physiological stimuli related to replicative aging. We found a significant decrease in the
levels of BRCA1 in cells at high passage, both at the mRNA and protein level (Fig. 9a).
In contrast, the phosphorylated and unphosphorylated forms of Aurora A kinase were
both increased in cells at high passage (Fig. 9b). Based on the inverse relationship in
expression levels that we observed in our cells as they age in culture, we hypothesized
that Aurora A over-expression during this crisis event may play a role in inhibiting
BRCA1 expression. We sought to further examine the relationship between Aurora A
kinase and BRCA1 expression by investigating the effects of down-regulating expression
of Aurora A on BRCA1 mRNA expression and protein levels. Treatment of two
independent strains of ovarian cystadenomas approaching crisis with two different siRNA
sequences against Aurora A led to an increase in the levels of phosphorylated and
unphosphorylated BRCA1 protein as well as an increase in BRCA1 mRNA expression
59
(Fig. 9c-9d), although Aurora A had no measureable consequences on BRCA1 in cells
cultured at earlier passages (Data not shown).
Given our previous finding that BRCA1 down-regulation leads to polyploidy due
to failure of cytokinesis in aging cystadenoma cells as they approach in vitro crisis due to
escape from a mitotic arrest at the spindle assembly checkpoint which typically occurs in
such cells, we reasoned that the decrease in BRCA1 expression observed in aging cells
may lead to polyploidy, another phenomenon typically associated with aging ovarian
cystadenoma cells. Separation of diploid cells based on DNA content by fluorescence
activated cell sorting showed lower BRCA1 mRNA expression in the polyploidy
population compared to diploid cells isolated from the same culture dish in support of this
hypothesis (Fig. 9e).

Aurora A inhibition decreases proliferation while stabilizing ploidy status
After establishing an inverse relationship between Aurora A and BRCA1
expression levels, we next investigated whether inhibition of Aurora A will have the
opposite effect of inhibiting BRCA1 with regards to ploidy changes and growth rates.
Specifically, we suspected that inhibition of Aurora A prevents cells from overcoming
the physiological mitotic arrest at the metaphase checkpoint complex that we observe as
they age in culture
64
. Aging cystadenoma cells in our cell culture model acquire an
actively dividing tetraploid sub-population as evidence by an 8N peak in their DNA
profiles
63,64
. Given our earlier findings that polyploid cell populations are the result of
recovery from an M-phase arrest in this experimental model
64
and due to Aurora A’s
known influence on BRCA1 and hence, maintenance of the G2/M checkpoint
40
, we
60
reasoned that blockage of such recovery by Aurora A inhibition might prevent acquisition
of such ploidy changes and increase genomic stability. Ovarian cystadenomas cultured at
either early or late passages were treated with siRNA against GFP (control) or Aurora A
for 72 hours and their DNA profiles were analyzed by flow cytometry. There were no
appreciable changes in the profiles of cells younger than 35 population doublings treated
with siRNA against Aurora A compared to control, ruling out the notion that Aurora A
inhibition can induce a mitotic arrest in such cells (Fig. 10a left panel). In contrast,
treatment of cells older than 50 population doublings resulted in an 18.1% decrease in the
post-4N cell population, which represents a pure population of proliferating polyploid
cells (p=0.03, Tukey’s HSD test) (Fig. 10a, right panel, Fig. 10b).
We further tested our hypothesis that down-regulation of Aurora A prevents
recovery from the physiological mitotic arrest by comparing the consequence of down-
regulation of either Aurora A or BRCA1 on the rate of multinucleation, which is a direct
consequence of the cytokinesis failure typically associated with such recovery.
Microscopic examination of the treated cells revealed that the average number of
multinucleated cells per 10X objective field was substantially lower in earlier passage
cells compared to higher passage, as expected (Fig. 10c). The rates in such cells, which
only rarely undergo a mitotic arrest at the metaphase checkpoint, were not substantially
altered by down-regulation of Aurora A, supporting the idea that multinucleation is the
result of recovery from a mitotic arrest triggered by decreased BRCA1 expression. In
contrast, high passage cells treated with siRNA against Aurora A showed a 39.3%
decrease in multinucleation compared to control cells (p=0.0014, 2-sided t-test),
supporting the notion that Aurora A inhibition results in a phenotype that opposes that
61
seen in cells with reduced BRCA1 expression. Cells cultured at low and high passages
showed a 37.5% and 57% respective increase in rates of multinucleation after treatment
with siRNA targeting BRCA1 compared to the control groups, (p=0.0051 & 0.0008
respectively, 2-sided t-test), indicating that the small proportion of younger cells
undergoing the mitotic arrest can also significantly contribute to changes in DNA ploidy
in the presence of reduced BRCA1 expression (Fig. 10c).
Considering that a mitotic arrest is the main determinant of crisis and crisis is
characterized by a decrease in proliferation, we sought to determine whether changes in
Aurora A expression influences crisis onset by examining the consequence of Aurora A
down-regulation on proliferation rates in cultured ovarian cystadenomas at early and late
in vitro passages. Although inhibition of Aurora A in cells cultured at early passages
showed no appreciable changes in ploidy status, we suspected that Aurora A inhibition
may result in growth rate changes that are independent of the physiological M-phase
arrest associated with BRCA1. Ovarian cystadenomas cultured at either early or late
passages were treated with siRNA against GFP (control), Aurora A, and a combination of
Aurora A/BRCA1 for 5 days. Number of cells per dish was measured every 24 hours. In
both low and high passages, reduced Aurora A expression resulted in a net decrease in
proliferation. Furthermore, the combination knockdown of Aurora A and BRCA1
abrogated the decrease in proliferation seen in the Aurora A knockdown alone, further
supporting the notion that BRCA1 is a key downstream target of Aurora A (Fig. 10d).
These findings suggest that the resultant phenotype following Aurora A inhibition
opposes that observed following BRCA1 inhibition with regards to ploidy changes,
however, the growth rate results suggest consequences are more complex than this. We
62
next wanted to evaluate the possibility that Aurora A inhibition induces a mitotic arrest
that is distinct from the physiological mitotic arrest that we observe in our aging ovarian
cystadenomas.

Mitotic arrest induced by inhibition of Aurora A is maintained despite securin
degradation and hence, APC activation
Progression from metaphase to anaphase is controlled by the anaphase promoting
complex (APC). The mitotic checkpoint complex (MCC), composed of the proteins
Mad2, Bub3 and BubR1, binds to the Cdc20 sub-unit on the APC and effectively inhibits
this complex until all chromosomes are properly attached and aligned at the equatorial
plate, at which time the MCC dissociates resulting in APC activation and progression to
anaphase
60-62
. Aging ovarian cystadenomas show an increase in components leading to
activation of the spindle assembly checkpoint due to an arrest in cell cycle progression at
this checkpoint
64
. In order to examine the role of Aurora A in this physiological arrest,
protein levels of mitotic regulators were examined 72 hours after transfecting cells
approaching crisis with siRNA targeting Aurora A. Increased BubR1 and Mad2
expression was seen compared to controls (Fig. 11a), suggesting that the metaphase
checkpoint had been stabilized, thereby preventing cellular progression to anaphase.
Protein immunoprecipitation using an antibody against Bub3, followed by western
blotting with an antibody targeting BubR1 likewise showed increased binding of these
two proteins to each other in aging cells following Aurora A down-regulation (Fig. 11b).
Down-regulation of Aurora A did not lead to any change in levels of MCC components at
earlier passages, when the arrest rarely takes place (Fig. 11a), indicating that such
63
inhibition does not induce the arrest that occurs spontaneously at high passages. These
results support the notion that Aurora A inhibition prolongs the mitotic arrest occurring
physiologically in aging cells, however, considering that Aurora A inhibition leads to a
decrease in proliferation in cells cultured at earlier passage, where the mitotic arrest
rarely takes place, we felt it plausible that Aurora A inhibition triggers a mitotic arrest
that is distinct from the physiological arrest that we observe in our cells as they age in
culture.
Considering that the known downstream target of APC activation is securin
degradation, we corroborated our mitotic regulator results by measuring protein
expression levels of securin following Aurora A inhibition using siRNA and a small
molecule inhibitor of the kinase, MLN8237 in aging cultured ovarian cystadenomas.
Upon dissociation of Mad2, Bub3, and BubR1 (components of the MCC), the Cdc20
subunit of the APC is free to bind to and ubiquitinate securin, thereby targeting it for
degradation. Securin degradation leads to activation of separase and cleavage of
cohesin., resulting in sister chromatid separation. We measured protein levels of securin
following Aurora A inhibition in aging cultured ovarian cystadenomas. Securin
expression was virtually absent in all cells in which this kinase was inhibited, indicating
that the arrest induced by Aurora A is maintained despite APC activation (Fig. 11c).
Thus, the components of the mitotic checkpoint, although increased following Aurora A
inhibition, are not bound to and therefore do not inhibit the ubiquitin ligase activity of the
APC/Cdc20 complex. This contrasts with our previous observations of increased
expression of mitotic checkpoint complex components and securin in aging ovarian
cystadenomas
64
, supporting the notion that the mitotic arrest induced by Aurora A
64
inhibition occurs downstream of the physiological arrest associated with the mitotic
checkpoint complex that we observe in aging ovarian cystadenomas. To confirm that
securin degradation was occurring via the APC/Cdc20 pathway, we performed protein
immunoprecipitation using an antibody against Securin, followed by western blotting
with an antibody targeting Cdc20 in cells treated with siRNA against Aurora A. When
comparing expression levels of Cdc20 in the DMSO control lane to the DMSO +
siAurora A lane, it is apparent not only that expression of Cdc20/securin increases, but
that the amount of native securin at 25 kD decreases, supporting the notion that more
securin is sequestered by Cdc20 and hence, degraded upon Aurora A inhibition (Fig.
11d). These results confirm our suspicions that the mitotic arrest induced by Aurora A
inhibition is distinct and downstream from the physiological arrest associated with
mitotic checkpoint complex that we observe as cells age in culture.

Aurora A inhibition prevents cells from progressing to anaphase B
We next sought to determine if Aurora A inhibition truly induced a mitotic arrest
despite APC activation or if the decrease in proliferation that we saw following down-
regulation of Aurora A was simply due to cell death. In order to explore this, cells at
both early and late in vitro passages were treated with siRNA against a negative control
scramble and Aurora A for 72 hours and examined by confocal microscropy. The
prevalence of metaphase versus anaphase cells was measured in a total of 50 mitotic
cells. As expected, we found a statistically significant difference (21.2% versus 53.4%,
P=0.003, two-sided t-test) between the percent of negative control cells in metaphase
from low to high passage (Fig. 12). Fewer cells progressed from metaphase to anaphase
65
when treated with siRNA against Aurora A at both passages (35.1% versus 12.9% at low
passage and 18% versus 7.3% at high passage, P=0.0014 and 0.003 respectively, 2-sided
t-test), despite the fact that securin had been degraded (Fig. 12). We also observed a
buildup of cells in metaphase post Aurora A inhibition at higher passages compared to
lower passages (72.2% versus 32.1%, P=0.002, 2-sided t-test), suggesting that the
population of cells arrested in aging cells is heterogeneous, including cells undergoing
the physiological arrest associated with crisis (which would not be sensitive to Aurora A
inhibition) and cells undergoing a mitotic arrest secondary to Aurora A inhibition (Fig.
12).

Aurora A inhibition prevents cells from undergoing cytokinesis failure due to
induction of a prolonged mitotic arrest and eventual cell death
We proposed earlier that physiologically arrested cells abruptly exit the mitotic
arrest without completing cytokinesis based on the increase in bi/multinucleation
observed following treatment with siRNA against BRCA1. Similarly, based on the
reduction in ploidy status and bi/multinucleation observed in cells treated with siRNA
against Aurora A, we suspected that Aurora A inhibition prevents cytokinesis failure
either by inducing an independent mitotic arrest or inducing cell death. Furthermore,
although lack of securin expression following Aurora A down-regulation is suggestive of
an active anaphase promoting complex, we did not observe any increase in progression to
anaphase following Aurora A inhibition. In order to reconcile these results, aging
cystadenoma cell strains treated with siRNA against GFP, BRCA1 or Aurora A for 48
hours were transferred to an incubation chamber under controlled temperature and CO
2
66
mounted on a confocal microscope. Time-lapse photographs were taken every minute
over 20 hours. The end fate of each cell entering mitosis was recorded and followed by
two independent observers. Several mononucleated cells that underwent a mitotic arrest
in the group treated with siBRCA1 eventually overcame this arrest and exited mitosis
without undergoing cytokinesis, resulting in binucleated cells as demonstrated by the
screenshots shown in Figure 13a. Specifically, we observed a 4-fold increase in
cytokinesis failure in cells treated with siBRCA1 compared to cells treated with siGFP
(p=0.026, 2-sided t-test). As anticipated, cells treated with siRNA against BRCA1 were
less susceptible to cell death compared to control cells (p=0.0083, 2-sided t-test), while
those treated with siRNA against Aurora A underwent a prolonged mitotic arrest leading
to a 1.8-fold increase in the rate of cell death compared to controls (p=0.0143, 2-sided t-
test) (Fig. 13b). Mononucleated cells undergoing an M-phase arrest only overcame such
arrest after forced reduction of BRCA1 expression and led to a proliferating tetraploid
population, due to failure of cytokinesis. In contrast, none of the cells treated with
siRNA against Aurora A overcame the mitotic arrest during the 20-hour observation
period, suggesting that the mechanism ultimately leading to a reduction in proliferating
polyploid cells is induction of a mitotic arrest. The lack of securin expression and the
limited number of cells progressing to anaphase suggest that the mitotic arrest associated
with Aurora A inhibition is downstream and independent of the M-phase arrest overcome
by BRCA1 down-regulation in our aging cystadenoma cells.

67
Mitotic arrest induced by inhibition of Aurora A is independent of spindle assembly
We next investigated whether all currently known criteria necessary for anaphase
progression were met in these cells in which Aurora A had been down-regulated. One
such requirement is a properly formed bipolar spindle apparatus, made up of
microtubules with asymmetric extremities, one end termed the ‘minus’ (-) end, relatively
stable and close to the centrosome and another end termed the ‘plus’ (+) end, with
alternating phases of growing-retraction
45,93
. Cultured ovarian cystadnomas were treated
with siRNA against Aurora A for 72 hours. Cells treated with Taxol for 24 hours were
used as a positive control for disrupted spindle formation. Cells were viewed using
confocal microscopy after staining with 4’-6’-diamino-2-phenylindole (DAPI) to
visualize DNA. Spindle assembly was visualized using an anti-tubulin antibody. Aurora
A inhibition did not affect spindle formation either at low or high passage (Fig. 14a). Out
of 26 metaphase cells observed at low passage and 20 metaphase cells observed at high
passage, there were no apparent defects in spindle assembly (Fig. 14b) (p=0.0002, 2-
sided, t-test), indicating that the arrest induced by Aurora A inhibition is independent of
spindle assembly formation. This experiment was repeated in another independent
ovarian cystadenoma cell line and similar results were observed (data not shown).

Aurora A inhibition leads to a mitotic arrest downstream of the mitotic checkpoint
complex and independent of kinetochore attachment
Another requirement for anaphase progression is kinetochore attachment at the
microtubule. The plus end tracking protein EB1 is a critical regulator of microtubule
dynamics, thereby maintaining chromosome fidelity. Studies suggest that EB1 modulates
68
kinetochore microtubule polymerization and/or attachment. This protein localizes to the
kinetochore at pro-metaphase/metaphase and is subsequently removed from the
kinetochore by the start of anaphase
94
. In order to determine whether the mitotic arrest
induced by Aurora A inhibition was related to incomplete microtubule attachment at the
kinetochore, ovarian cystadenomas cultured at either early or late passages were treated
with with siRNA against GFP or Aurora A for 72 hours followed by collection of nuclear
extracts. Immunoblotting for EB1 revealed an increase of nuclear protein expression at
both passages following Aurora A down-regulation, indicating stabilization of EB1
expression. (Fig. 15a) Cellular localization of EB1 was also examined via confocal
microscopy after treating cells at both low and high passage with siRNA against Aurora
A. Control cells at earlier passages showed EB1 localization to the DNA at metaphase
(Fig. 15b), while cells treated with siRNA against Aurora A, showed a predominant
intermediate phenotype post-metaphase between anaphase A and anaphase B,
characterized by sister chromatids approaching opposite poles, in the absence of spindle
pole migration. In these cells, EB1 was also localized with the DNA, seemingly in
greater abundance (Fig. 15b). We quantified the number of cells in true metaphase and
also cells with the intermediate phenotype and found that while only 20% of mitotic cells
at low passage were arrested at metaphase, 40% demonstrated the ‘intermediate’
phenotype, excluding the possibility that the phenotype is simply due to a normal mitotic
cell transitioning from metaphase to anaphase. In addition, the number of cells in true
anaphase B was only 12% of all mitotic cells at low passage and 9% of cells at high
passage supporting the notion that the arrest induced by Aurora A down-regulation
prevents cells from progressing through anaphase (check these numbers and maybe
69
describe this in ratios). Approximately 35% of mitotic cells were arrested with a true
metaphase plate in our aging cells, indicative of cells undergoing the physiological arrest
as previously described. Another 40% of mitotic cells demonstrated the intermediate
phenotype, indicating that those cells that were not subjected to the physiological arrest,
were sensitive to the arrest induced by Aurora A inhibition following siRNA treatment
(Fig. 15c, 15d).
Anaphase A is characterized by shortening of kinetochore fibers which pulls sister
chromatids to opposing spindle poles. Progression through anaphase A is regulated by
APC/Cdc20’s degradation of cyclin B1, which thereby causes a decrease in Cdk1
activity. Once Cdk1 activity has fallen below a specific threshold, cells progress to
anaphase B and the APC/Cdh1 complex is activated. During anaphase B, the spindle
elongates, pushing spindle poles apart causing further migration of sister chromatids.
The APC/Cdh1 complex then targets polypeptides whose destruction by the proteosome
is required to exit from mitosis and return to interphase
45
. Considering that cyclin B1
degradation is a crucial step to initiate the transition from anaphase A to anaphase B, we
next evaluated if Aurora A down-regulation led to stabilization of cyclin B1 expression
levels. Indeed, in both low and high passage nuclear extract fractions, cyclin B1
expression increased following Aurora A inhibition (Fig. 15e), suggesting lack of cyclin
B1 degradation in spite of lack of securin and thus, the presence of active APC/Cdc20
ubiquitin ligase activity.

70

Figure 9: Influence of replicative age on BRCA1 and Aurora A expression.

A: ML10 cystadenoma cells were harvested either at early (<35 population doublings) or
late >50 population doublings) passage. Relative levels of BRCA1 mRNA were
Ku-70
BRCA1
Aurora A
siGFP siAA
ML-10
siGFP siAA
ML-3 A
0
0.2
0.4
0.6
0.8
1
1.2
Ku-70
BRCA1
<35 >50
Population Doublings
Relative BRCA1
mRNA Expression
0
0.5
1
1.5
2
2.5
siGFP siAurora-A
Relative BRCA1
mRNA Expression
>50
Population
Doublings
Relative BRCA1
mRNA Expression
0
0.2
0.4
0.6
0.8
1
1.2
1.4
diploid
(2N)
tetraploid
(8N)
2N 4N 8N

Cell Number
DNA Content
B
D
E
ML-3
Aurora A
P-Aurora A
Ku-70
Population
Doublings:
ML-10
<35 <50 <35 <50
C
71
measured using quantitative real-time PCR and relative BRCA1 protein levels were
evaluated by western blotting for each passage. Ku70 served as a loading control for
protein analysis. B: Two different strains of ovarian cystadenomas, ML3 and ML10,
were harvested at either (<35 population doublings) or late (>50 population doublings)
passage and their relative abundance of either total Aurora A protein or the
phosphorylated form, representing the activated kinase were determined by western
blotting analysis. The blots were also probed using an antibody against Ku-70 as loading
control. C: ML10 cells at the onset of crisis (>50 population doublings) were treated
with siRNA directed against either GFP (control) or Aurora A and the relative level of
BRCA1 mRNA in each group was measured by quantitative real-time PCR. D: ML3 and
ML10 cystadenoma cells approaching crisis (>50 population doublings) were treated with
siRNA directed against either GFP (control) or Aurora A and BRCA1 protein levels in
each group was analyzed by western blotting. The blots were also probed with an
antibody against Aurora A to evaluate the extent of down-regulation achieved by the
siRNA treatment and against Ku-70 to evaluate protein loading. E: ML10 cells at the
onset of crisis (>50 population doublings) were stained with propidium iodide and
analyzed for DNA content by flow cytometry, followed by fluorescence activated cell
sorting to isolate 2N and 8N cell populations using the gating boundaries indicated by the
rectangles in the flow cytometry profile. The 2N and 8N sorted cell populations were
then analyzed for BRCA1 mRNA expression BRCA1 expression using quantitative real-
time PCR.

*Please note that the results presented in figure 9A, C, and E were completed by Vanessa
Yu, PhD

72

Figure 10: Effect of Aurora A down-regulation on ploidy status and growth rates

A: Cells at low (<35 population doublings) and high passage (>50 population doublings)
were treated with siRNA targeting either GFP (control) or Aurora A for 72 hours, stained
with propidium iodide, and their DNA profiles were obtained by flow cytometry. B: The
-2.0
2.0
6.0
10.0
14.0
18.0
22.0
26.0
30.0
34.0
38.0
42.0
46.0
<35 Population Doublings
Percent Multinucleation
siControl siAA siBRCA1
6.5 ±0.36
5.0 ±0.93
10.4 ±0.56
P=0.0051
-2.0
2.0
6.0
10.0
14.0
18.0
22.0
26.0
30.0
34.0
38.0
42.0
46.0
>50 Population Doublings
Percent Multinucleation
siControl siAA siBRCA1
15.5 ±0.40
9.4 ±0.58
36.1 ±2.2
P=0.0014
P=0.0008
Log of Cell Number
Days Post Transfection
<35 Population Doublings >50 Population Doublings
5
5.2
5.4
5.6
5.8
6
6.2
6.4
0 2 4 6
siGFP
siAA
siAA/BRCA1
5
5.2
5.4
5.6
5.8
6
6.2
6.4
0 2 4 6
siGFP
siAA
siAA/BRCA1
A B
D
>50 Population
Doublings
siGFP
siAurora A

<35 Population
Doublings
DNA content
2N 4N 8N 2N 4N 8N
2N 4N 8N 2N 4N 8N
Number of cells
-4
1
6
11
16
21
26
siGFP siAA1
Percent of Cells Post-4N
siGFP
siAurora A
P=0.03
C
73
number of cells within the rectangles shown in the profiles cells at high passage were
quantified and plotted. C: ML10 cystadenoma cells at low and high passage were treated
with siRNA against either GFP (control), Aurora-A, or BRCA1 for 72 hours and fixed in
2% paraformaldehyde. The average number of bi/multi-nucleated cells under the 10X
objective of an inverted phase contrast microscope was calculated from 3 randomly
selected fields in triplicate dishes (total of 9 observations per condition) by 2 independent
observers blinded to the treatment condition. D: ML10 cystadenoma cells at low and
high passage were treated with siRNA against GFP (control), Aurora A, or a combination
of Aurora A & BRCA1. Number of cells per dish were counted every 24 hours for 5
days following transfection.

*Please note that the results presented in figure 10A were completed by Vanessa Yu,
PhD

74

Figure 11: Effect of Aurora A down-regulation on cell cycle regulations

ML10 cystadenoma cells cultured at either low (<35 population doublings) or high (>50
population doublings) passage were treated with siRNA directed against either GFP
(control) or Aurora A kinase. A: Protein levels of selected components of the mitotic
checkpoint complex and anaphase promoting complex were measured by western blot in
each experimental group. The western blots were also probed with an antibody against
Ku-70 to evaluate protein loading. B: Each protein lysate was immunoprecipitated using
an antibody directed against BUB3 and the immunoprecipitates were analyzed by
western blotting using an antibody against BubR1. C: ML10 cystadenoma cells at high
passage were treated with DMSO (negative control), Taxol (positive control), siRNA
against Aurora A, and MLN8237, a small molecule inhibitor of Aurora A. Protein levels
of securin were measured in each experimental group. D: Each protein lysate was also
immunoprecipitated using an antibody directed against securin and immunoprecipitates
were analyzed by western blotting using an antibody against Cdc20.

>50 <35
siGFP siAurora-A
siGFP siAurora-A
P-Aurora A
Ku70
BubR1
Mad2
Population
Doublings
A
>50 Population Doublings
siGFP siAurora-A
_
_
_
150 kD
100 kD
50 kD
_
20%BubR1 Input
BubR1
B
Ab
Only
Protein
Only
Ku70
pAA
Securin
DMSO
Taxol
DMSO+siAA
Taxol+siAA
DMSO+MLN8237
Taxol+MLN8237
25 kD
50 kD
70 kD
_
_
_
Ab Only
Protein Only
DMSO IP
Taxol IP
DMSO+siAA IP
Taxol+siAA IP
_
_
_
25 kD
37 kD
50 kD
Cdc20
_
C
D
75

Figure 12: Effect of Aurora A down-regulation on cell cycle progression

Cells at low passage (<35 population doublings) and high passage (>50 population
doublings) were treated with siRNA against a negative control scramble sequence or
Aurora A for 48 hours. Cells were fixed and stained with DAPI prior to viewing on a
confocal microscope. The number of metaphase versus anaphase cells were counted.
Results plotted are the average of a set of 3 independent experiements.

0
10
20
30
40
50
60
70
80
Metaphase Anaphase
siNegative
Control
siAurora A
0
10
20
30
40
50
60
70
80
Metaphase Anaphase
siNegative
Control
siAurora A
Percent Cells
<35 Population Doublings >50 Population Doublings
76

Figure 13: Effect of Aurora A or BRCA1 down-regulation on cell fate evaluated by
time-lapse photography
Cells approaching crisis were treated with siRNA Aurora A or BRCA1 for 48 hours and
mounted on a confocal microscope under controlled temperature and CO
2
. Photographs
were taken every minute over 20 hours and assembled into a movie. A: Selected
screenshots demonstrating progression of a mononucleated cell treated with siBRCA1
undergoing a mitotic arrest and eventually recovering from this arrest and becoming
binucleated due to failure of cytokinesis is shown in the top panel while the lower panel
shows progression of a cell treated with siAurora-A undergoing a mitotic arrest
culminating in cell death. B: Cells undergoing a prolonged mitotic arrest in the time-
lapse studies shown in Fig. 5a were followed to determine compare the number times
such arrests led to cell death (left panel) versus the number of times they led to
multinucleation due to recovery followed by cytokinesis failure (right panel) in each
experimental group. The results are expressed as ratios of observation in cells treated
with either siAurora-A or siBRCA1 over observations in cells treated with siGFP, which
were normalized to one.

*Please note that the real-time movies were completed in collaboration with Theresa
Austria and Vanessa Yu, PhD
Mono-Nucleated Cell Mitotic Arrest
Incomplete Cytokinesis Multi-Nucleated
>50 Population Doublings
Multi-Nucleated Mitotic Arrest Cell Death
siBRCA1 siAurora A
Fold Change
Fate of Cells Entering Mitosis
-1
-0.5
0
0.5
1
1.5
2
2.5
-2
-1
0
1
2
3
4
5
6
B
**
**
**
Cell Death Cytokinesis Failure
siAurora A siBRCA1 siAurora A siBRCA1
A
77

Figure 14: Effect of Aurora A down-regulation on spindle assembly

Ovarian cystadenoma cells at low (<35 popluation doublings) and high (>50 population
doublings) passages were treated with Taxol (positive control), siRNA against a negative
scramble control or Aurora A for 48 hours prior to fixation and staining. A:
Immunofluorescence microscopy showing localization of alpha tubulin (green) and DNA
(blue). B: Quantitation of normal versus abnormal spindle assembly out of all
metaphase cells counted.

78
A
B
siNegative Control siAurora A
True M etaphase Interm ediate Phenotype
C
D
siNegative Control siAurora A
# cells/T otal Mitotic % # cells/T otal Mitotic %
Prophase 21/74 28.4 12/67 17.9
Metaphase 30/74 40.5 21/67 31.3
Intermediate 5/74 6.8 28/67 41.8
Anaphase 18/74 24.3 6/67 9.0
>50 PD <35 Population Doublings
siNC
siAA
Tax ol +siAA
DMSO+siAA
Tax ol
DMSO
Cyclin B1
Ku70
<35 PD >50 PD
siNC
siAA
siAA
siNC

Figure 16: Effect of Aurora A down-regulation on microtubule anchoring

A: Cells at low (<35 population doublings) and high (>50 population doublings) passage
were treated with siRNA targeting either a negative scramble control or Aurora A in the
presence of DMSO (negative control) or Taxol (positive control). Protein levels of EB1
were measured by western blot in each experimental group. B: Cells at high passage
were treated with siRNA targeting either a negative scramble control or Aurora A for 48
hours prior to fixation and staining. Immunofluorescence microscopy showing
localization of EB1 (green) and DNA (blue). C: Quantitation of cells in each phase of
the cell cycle was completed under immunofluorescence microscopy. D: Cells at low
and high passage were treated with siRNA targeting either a negative scramble control or
Aurora A for 72 hours. Protein levels of cyclin B1 were measured by western blot in
each experimental group.
79

Figure 16: Effect of Aurora A down-regulation on caspase activity

ML10 cystadenoma cells cultured at either low (<35 population doublings) or high (>50
population doublings passage were treated with siRNA directed against either GFP
(control) or Aurora A. Protein levels of select caspases were measured by western blot in
each experimental group.

>50 <35 Population Doublings:
siGFP siAurora-A
siGFP siAurora-A
PARP
Caspase 7
Caspase 9
Aurora A
P-Aurora A
Ku70
80

DISCUSSION

Our results demonstrate that Aurora A down-regulation induces a mitotic arrest
that is downstream of the spindle assembly checkpoint, in between anaphase A and
anaphase B, ultimately leading to reduced ploidy changes in aging ovarian cystadenomas.
Our in vitro cell culture model shares several characteristics with high-grade serous
ovarian carcionomas and their precursor lesions. These shared characteristics include a
common origin from serous mullerian epithelium and lack a functional P53. Repeated
cell division invariably leads to a physiological cell cycle arrest at the metaphase
checkpoint complex in this cell culture model, presumably due to genomic defects that
have accumulated because of a non-functioning P53. Aurora A and BRCA1 protein
levels fluctuate in opposite directions with increasing replicative age and an increase in
BRCA1 levels can be induced by forced down-regulation of Aurora A using siRNA. The
effect of Aurora A down-regulation on BRCA1 levels is only seen in aging cells
approaching in vitro crisis, suggesting that the interplay between these two proteins with
regards to protein expression is specifically associated with replicative aging. We
initially suspected that interplay between these two proteins was crucial to maintain the
M phase arrest in aging ovarian cystadenomas, however, here, we demonstrate that the
mitotic arrest induced by Aurora A down-regulation is not mediated by any effect of
Aurora A on BRCA1. Rather, Aurora A inhibition leads to a distinct mitotic arrest, that
is downstream of the physiological arrest that we observe in aging cystadenomas, which
we suggest is a undescribed step in mitotic checkpoint control.
81
We also show that securin is degraded by the APC/Cdc20 complex (Fig. 11c),
therefore, the ubiquitin ligase activity of Cdc20 is seemingly intact. It is possible that
Aurora A kinase is responsible for a phosphorylation event on either APC/Cdc20 which
alters substrate specificity from securin to Cyclin B1. Another plausible explanation is
that Aurora A kinase phosphorylates cyclin B1 itself, thereby making it a more attractive
substrate for ubiquitin mediated degradation. Taken together, our results demonstrate
that there is a secondary checkpoint in between anaphase A and B, downstream of the
spindle assembly checkpoint, which is regulated by Aurora A kinase, most likely due to
an activating phosphorylation event on cyclin B1 or Cdc20.
Aurora A is over-expressed in a variety of tumors including but not limited to
breast, ovarian, colorectal and esophageal carcinoma
95-104
. Its inhibition induces
apoptosis in many cell types although the precise mechanism of cell death is unknown
105
.
It has been suggested that inhibition of Aurora A leads to cell death due to catastrophic
polyploidy induced by Aurora A inhibition
106-108
. Recently, inhibition of Aurora A
activity using a small molecular inhibitor has been associated with severe polyploidy due
to chromosome endoreduplication in P53-deficient human colon carcinoma cells
42
.
However, our results show that Aurora A inhibition using siRNA reduces polyploidy in
P53-null ovarian serous cystadenomas. This is supported by observations Aurora A over-
expression leads to polyploidy in many cancers
102, 105, 109, 110)
. A potential explanation for
these apparent discrepancies is that studies published from other laboratories have used
small molecular inhibitors of Aurora A kinase while we used siRNA technologies, which
may target the kinase more specifically. A potentially more important explanation is that
our studies were performed in cells that were not immortal and focused on a phenomenon
82
associated with replicative aging that is over-shadowed in rapidly growing immortal cell
lines used in other studies. In fact, we also observed, in line with findings from other
laboratories, that Aurora A inhibition leads to activation of apoptotic pathways as
evidenced by an increase in caspase activity (Figure 16). These effects on apoptosis were
not limited to aging cells, but were also observed in cells cultured at early passages where
Aurora A inhibition does not lead to accumulation of cells arrested in mitosis. Thus, the
underlying mechanisms responsible for apoptosis following Aurora A down-regulation
are fundamentally different than those responsible for the effects on the cell cycle
reported here, which are associated with a mitotic arrest downstream of the physiological
mitotic arrest that is most prevalent in ovarian cancer precursor cells that have not
undergone complete malignant transformation.

83
Chapter 4: The Road to Aneuploidy in High-grade Serous Ovarian
Carcinoma

ABSTRACT

Aside from mutations in P53, severe aneuploidy, associated with a near polyploid
number of chromosomes, is the only genetic abnormality consistently associated with
non-familial high-grade serous ovarian carcinoma. It is also associated with poor clinical
outcome. Our laboratory has established an in vitro model derived from ovarian
cystadenomas, the benign counterparts of ovarian carcinomas, which further mimics the
genetic background of high-grade serous ovarian carcinomas and their precursors due to a
non-functional P53. These cells, which are not fully transformed and not immortal,
undergo a mitotic arrest as they age and approach in vitro crisis. We showed earlier that
recovery from this arrest can be induced by treatment with siRNA against BRCA1, a
protein associated with familial breast and ovarian cancer predisposition, leading to
polyploidy and multinucleated cells due to cytokinesis failure. We feel that this
mechanism provides insight to the underlying mechanism surrounding polyploidy
development of high-grade serous ovarian carcinomas. Such carcinoma cells frequently
show reduced BRCA1 expression either due to germline BRCA1 mutations in individuals
with familial predisposition or by BRCA1 promoter methylation in individuals without
such predisposition. These tumors often demonstrate increased Aurora A expression as
well, which is associated with poor clinical outcomes. We hypothesize that this
mechanism continues beyond oncogenic transformation in high-grade serous ovarian
carcinomas, resulting in a subset of cells that undergo a physiological mitotic arrest
which can be overcome due to reduced BRCA1 expression, thereby perpetuating the
chromosomal instability inherent in these tumors. We also propose that inhibition of
84
Aurora A will induce a mitotic arrest and prevent expansion of proliferating polyploidy
cells. Here we show that indeed, Aurora A knockdown reinstates mitotic checkpoint
control, resulting in increased genomic stability. We conclude that targeting Aurora A
kinase can reduce chromosomal instability in a subset of high-grade serous ovarian
carcinomas.

INTRODUCTION

Ovarian cancer accounts for 5% of all female cancer related deaths, more than any
other reproductive gynecological cancer. Most patients respond well to chemotherapy
following initial debulking surgery, but eventually undergo recurrence at which time the
cancer cells have become resistant to conventional chemotherapeutic agents. It is likely
that the genetic instability inherent to ovarian cancer cells, examples of which can be
seen in figure 17, facilitates the acquisition of such drug resistance
111
. The development
of strategies to reduce genetic instability early in the development of ovarian carcinomas
may therefore decrease the rate of drug resistance and diminish the morbidity and
mortality currently associated with these cancers.
Among the various histological subtypes of these tumors, those classified as high-
grade serous carcinomas are the most common. The Cancer Genome Atlas (TCGA)
recently completed an integrated genomic analysis on 489 high-grade serous ovarian
carcinomas and found that P53 was the only gene commonly mutated in the majority
(96%) of these carcinomas. Microscropic foci showing clonal P53 mutations have been
described in surgical specimens obtained from individuals harboring germline BRCA1 or
BRCA2 mutations. Such foci, referred to as P53 signature, are regarded as precursor
85
lesions for high-grade serous ovarian carcinomas
60
. Mutations in BRCA1 and BRCA2,
which are common in tumors from individuals with familial predisposition, were noted in
22.2% of all cases
112
. Although these two genes are rarely mutated in the non-familial
form, they can be silenced via DNA methylation of their promoter region
113
.

In addition to loss of a functional P53, a near polyploid DNA content is a
consistent genomic abnormality associated with high-grade serous ovarian carcinomas.
The mechanisms leading to such changes in DNA ploidy, the maintenance of which is
likely facilitated by the coexistence of P53 mutations, may provide important clues
surrounding the initial genetic events during the development of ovarian tumors. These
ploidy changes occur early in their development and their magnitude is a predictor of
poor clinical outcome
114
. They are most likely triggered by post-G1 cell cycle errors
given their characteristic near polyploid nature.

We recently showed that cultured ovarian serous cystadenomas, the benign
counterpart of ovarian serous carcinomas, undergo a mitotic arrest at the metaphase
checkpoint complex as they approach in vitro crisis if they harbor the equivalent of a P53
mutation, which we achieved by forced expression of SV40 Large T Antigen
64
. We
further showed that this arrest can be overcome by reduced expression of BRCA1. Cells
overcoming such arrest become multinucleated, suggesting a mechanism of polyploidy
development mediated by failure of cytokinesis
64
.
We hypothesize that high-grade serous ovarian carcinomas likewise undergo a
mitotic arrest possibly triggered by genetic abnormalities otherwise kept unchecked due
to the absence of a functional P53, and tested this in ovarian carcinoma cell lines with
both a function and dysfunctional P53 (Table 1). We further hypothesize that reduced
86
BRCA1 expression, which may be present in high-grade serous ovarian carcinomas, due
either to germ line mutations or promoter methylation, leads to recovery from this arrest
and cytokinesis failure, resulting in polyploidy and ensuing aneuploidy. Interfering with
this process may reduce the ability of cells to overcome such mitotic arrest and stabilize
their genome.
Aurora A, a member of the serine/threonine kinase family, is upregulated in many
cancers
39,115
. It also phosphorylates BRCA1 at Ser308, which regulates BRCA1’s role in
G2 to M transition in the face of DNA damage
40
. Another phophorylation event on
BRCA1, also catalyzed by Aurora A kinase, prevents multi-polar spindle formation by
inhibiting the ubiquitin ligase activity of BRCA1 at the centrosome during the M phase
24
.
Based on our previous findings (Chapter 3), we hypothesized that Aurora A inhibition
may induce a mitotic arrest downstream of the physiological arrest that we observe in our
ovarian cystadenoma cells. Such induction of an arrest would limit perpetuation of
proliferating polyploidy cells, thereby promoting genetic stability. We sought to test this
hypothesis by investigating the consequences of Aurora A down-regulation on
maintenance of this mitotic arrest and on overall ploidy status. Our results show that
targeting Aurora A kinase can indeed prevent the development of polyploidy by reducing
the rate of cytokinesis failure, underscoring the potential of this approach to stabilize the
genome of ovarian carcinoma cells. Considering that the scenario observed in our in
vitro cell culture model leading to polyploidy development is relevant not only to ovarian
cancer precursor cells, but also occurs after malignant transformation, raising the
possibility that targeting Aurora A kinase may also prevent genetic heterogeneity and
acquisition of drug resistance in ovarian carcinomas.
87

MATERIALS AND METHODS

Cell Strains and Culture Conditions. CAOV3, HEY, HOC7, and SKOV3 ovarian
carcinoma cell lines were cultured in Dulbecco’s Modified Eagle Medium (DMEM)
supplemented with 10% fetal-bovine serum, and 1% Penicillin-Streptomycin in a 37˚
Celsius, 5% CO
2
, humidified incubator.

Antibodies. Antibodies against the phosphorylated (cat #: 2914S) and unphosphorylated
(cat #: 4718S) forms of Aurora A (Cell Signaling Technology Inc., Danvers, MA) were
used at a 1:1000 dilution. Antibodies against Ku70 (cat #: 12729), Mad2 (cat #: 47747)
(Santa Cruz Biotechnology Inc., Santa Cruz, CA) were used at a 1:200 dilution.
Antibodies against Bub3 (cat #: 611730), BubR1 (cat #: 612502), and Cdc27 (cat #:
61054) (BD Transduction Labs, San Jose, CA) were used at a 1:1000 dilution.

DNA Profiling Studies. Cells cultured on plastic dishes were trypsinized, re-suspended
in 0.2 mL of PBS, and fixed in 2 mL of 70% ethanol.

After centrifugation, the cell pellets
were resuspended in 1

mL of PBS, 10 µg/mL propidium iodide, and 100 ug/mL

RNase.
Fluorescence was measured on a EPICS XL-MCL flow cytometer (Beckman Coulter
Inc., Fullerton, CA).

Quantification of Multinucleated Cells. CAOV3 and SKOV3 cells were treated with
siRNA against a negative scramble control, BRCA1 or Aurora A for 72 hours followed
by fixation in 2% formaldehyde. The number of multi-nucleated cells per high
88
magnification (35x) field was determined in triplicate dishes by 2 independent observers
using a Zeiss IM 35 microscope. Three fields per dish were examined by each observer.
The results are expressed as an average of the total of 18 observations for each condition.

RESULTS

BRCA1 down-regulation results in a proliferating polyploid population in a P53
deficient high-grade serous ovarian carcinoma cell line

We previously showed that benign ovarian cystadenomas undergo a physiological
mitotic arrest at the spindle assembly checkpoint as they age in culture which can be
overcome by forced reduction of BRCA1 levels. We suspected that this scenario
continued beyond oncogenic transformation and tested this by treating ovarian carcinoma
cell lines with siRNA targeting BRCA1 and evaluating DNA ploidy. We also felt that
treatment with siRNA targeting Aurora A may limit the ploidy changes in this same cell
population. We chose two ovarian carcinoma cell lines, HOC7 and SKOV3, to treat
based P53 status since our ovarian cystadenoma model exists in a genetic background
that is P53 null due to SV40 Large T antigen. Both cell lines originate from malignant
ascites from a serous high-grade ovarian carcinoma samples. While HOC7 has a wild-
type P53, SKOV3 cells harbor a mutant P53 (Table 1). While there were no appreciable
differences in DNA ploidy in the HOC7 cells following treatment with siRNA targeting
BRCA1 and Aurora A (Fig. 18), SKOV3 cells show a statistically significant difference
in proliferating polyploidy cells compared to control cells (14.2% versus 6.0%, p=0.0001,
two-sided t-test, Fig. 19). There was also a 12% increase in cells in the G2/M peak in the
cells treated with siRNA targeting BRCA1, which is most likely representative of
89
polyploidy cells in G1. Aurora A down-regulation also led to an statistically significant
increase in cells in cells post-G2/M which may be indicative of a mitotic arrest (6.0%
versus 15.7% , P=0.0001, two-sided t-test, Fig. 19). Taken together, these results suggest
that a small population of ovarian carcinoma cells undergo a mitotic arrest in culture
which can be overcome by forced reduction of BRCA1. Aurora A inhibition leads to an
increase in the number of cells in the G2/M peak which may be indicative of a mitotic
arrest. Further studies must be completed to confirm these findings.

BRCA1 down-regulation leads to an increase in bi/multinucleated cells in SKOV3
and CAOV3 cells

In an effort to evaluate the post-G2/M peak that we observed following BRCA1
down-regulation, we compared the consequence of down-regulation of either Aurora A or
BRCA1 on the rate of multinucleation, which is a direct consequence of the cytokinesis
failure typically associated with such recovery. Microscopic examination of the treated
cells revealed that the average number of multinucleated cells per 10X objective field
was substantially higher in both SKOV3 and CAOV3 cells treated with siRNA targeting
BRCA1 when compared to controls (4.3% versus 15.2% in SKOV3 cells and 8.0%
versus 15.1% in CAOV3 cells, p=0.0005 & 0.0034, respectively, two-sided t-test,
Fig.20). In contrast, both SKOV3 and CAOV3 cells treated with siRNA against Aurora
A showed a 46.8% and 48.1% decrease in multinucleation compared to control cells
(p=0.0349 & 0.0011 respectively, 2-sided t-test, Fig.20), supporting the notion that
Aurora A inhibition prevents genomic instability mediated by cytokinesis failure.

90
Aurora A down-regulation leads to an increase in mitotic regulators associated with
the spindle assembly checkpoint in HEY and SKOV3 cells

We next wanted to explore the mechanism by which Aurora A down-reguation
reduces bi/multinucleation in ovarian carcinoma cell lines and tested this by evaluating
protein levels of key mitotic regulators following Aurora A down-regulation using
siRNA. Progression from metaphase to anaphase is controlled by the anaphase
promoting complex (APC). The mitotic checkpoint complex (MCC), composed of the
proteins Mad2, Bub3 and BubR1, binds to the Cdc20 sub-unit on the APC and effectively
inhibits this complex until all chromosomes are properly attached and aligned at the
equatorial plate, at which time the MCC dissociates resulting in APC activation and
progression to anaphase
116-118
. Cdc27 catalyzes the transfer of ubiquitin to cyclin B1,
which is necessary for progression to anaphase
68
. In order to examine the role of Aurora
A in a mitotic arrest at the spindle assembly checkpoint, protein levels of mitotic
regulators were examined 72 hours after transfecting both SKOV3 (P53 mutant) and
HEY (P53 wild-type) cells with siRNA targeting Aurora A. Increased BubR1 and Mad2
expression and a concomitant decrease in Cdc27 were seen in both cell lines (Fig. 21),
suggesting that the metaphase checkpoint had been stabilized, thereby preventing cellular
progression to anaphase, and that stabilization of the spindle assembly checkpoint is
independent of P53 status. Considering our previous results in ovarian cystadenoma cells
where components of the mitotic checkpoint complex were up-regulated despite securing
degradation, more work needs to be completed to ensure that the increase in MCC
components that we observe in ovarian carcinoma cell lines is truly indicative of an
activated spindle assembly checkpoint.
91

Figure 17: Molecular abnormalities associated with ovarian cancer

A: Mononucleated and binucleated SKOV3 cells after BRCA1 downregulation. Cells
were fixed in 2% paraformaldehyde, stained against alpha-tubulin (red) and
counterstained with DAPI (blue). Pictures were taken at 100X magnification.
B: Example of an undifferentiated ovarian carcinoma showing abundant multinucleated
cells. C and D: More typical high-grade serous ovarian carcinoma; although it is difficult
to distinguish multinucleation from tangential cuts through multi-lobed nuclei, the 2 cells
indicated by thin arrows in C are very suggestive of bi-nucleated cells; more definite
evidence for the presence of multinucleated cells capable of further cell division comes
from the example of a multi-polar mitotic figure shown by the thick arrow in D. Bar: 50
microns

*H&E sections completed by Louis Dubeau, MD PhD

B A
C D
92

Table 1: Cell line characteristics

Cell lines selected for carcinoma experiments were CAOV3, HEY, HOC7, and SKOV3.
Origin, histologic sub-type, P53 status, doubling time, tumorigenicity, and karyotype
profiles can be seen in table 1.

Cell
Line Origin
Histo-
logy
P53
Status DT
Tumori-
genicity Karyotype
CAOV3
Malignant
ascites Serous MT 78 h Negative Near triploid
HEY
Peritoneal
deposit Serous WT 30 h Positive Highly aneuploid
HOC7
Malignant
ascites Serous WT 16.4 h Negative Pseudo-diploid
SKOV3
Malignant
ascites Serous MT 28.8 h Positive Near triploid
93

Figure 18: Effect of BRCA1 and Aurora A down-regulation on DNA ploidy in
HOC7 cells

HOC7 cells were treated with siRNA targetin a negative scramble control, BRCA1, or
Aurora A for 72 hours, stained with propidium iodide, and their DNA profiles were
obtained by flow cytometry. Percent of cells within each phase of the cell cycle were
quantified and plotted.

-10
0
10
20
30
40
50
60
70
80
90
100
G1 S G2/M S2 G2/M2
siNC
siBRCA1
siAA
94

Figure 19: Effect of BRCA1 and Aurora A down-regulation on DNA ploidy in
SKOV3 cells

SKOV3 cells were treated with siRNA targetin a negative scramble control, BRCA1, or
Aurora A for 72 hours, stained with propidium iodide, and their DNA profiles were
obtained by flow cytometry. Percent of cells within each phase of the cell cycle were
quantified and plotted.
-10
0
10
20
30
40
50
60
G1 S G2/M S2 G2/M2
siNC
siBRCA1
siAA
95

Figure 20: Multinucleation counts following down-regulation of Aurora A and
BRCA1 in SKOV3 and CAOV3 cells

CAOV3 and SKOV3 cells were treated with siRNA targeting either a negative scramble
control, Aurora A, or full-length BRCA1 for 72 hours, then fixed in 2%
paraformaldehyde. The average number of bi/multi-nucleated cells under the 10X
objective of an inverted phase contrast microscope was calculated from 3 randomly
selected fields in triplicate dishes (total of 9 observations per condition) by 2 independent
observers blinded to the treatment condition.

0
2
4
6
8
10
12
14
16
18
siNC siAA siFL2
CAOV3 Cells SKOV3 Cells
0
2
4
6
8
10
12
14
16
18
siNC siAA1 siFL2
P=0.0349
P=0.0005
P=0.0002
P=0.0011 P=0.0007
P=0.0034
96

Figure 21: Effect of Aurora A down-regulation on mitotic regulators

HEY and SKOV3 cells were treated with siRNA targeting a negative scramble control or
Aurora A for 72 hours. Nuclear extracts were harvested and prtein levels of selected
components of the mitotic checkpoint complex and anaphase promoting complex were
measured by western blot in each experimental group.

Aurora A

Ku70

BubR1

Mad2

Cdc27
siControl siAA siControl siAA
HEY Cells SKOV3 Cells
97

DISCUSSION

We previously demonstrated that benign ovarian cystadenomas, which we regard
as a precursor model for ovarian carcinomas undergo an M phase arrest at the spindle
assembly checkpoint as they age in culture. In such model, forced reduction of BRCA1
levels allows cells to overcome this M phase arrest, leading to a population of cells that
abruptly exit the mitotic arrest and resume proliferation without completing cytokinesis.
It is not inconceivable that the scenario observed in our in vitro cell culture model leading
to polyploidy development occurs after malignant transformation, raising the possibility
that reduced BRCA1 levels maintains genomic instability in ovarian carcinomas beyond
oncogenic transformation. Furthermore, targeting Aurora A kinase may also prevent
genetic heterogeneity by inducing a similar mitotic arrest to that observed in our ovarian
cystadenoma cell model (Chp. 3).
Recent studies demonstrate a growing role for Aurora A in chromosome
instability. Specifically, a study conducted in colorectal tumors revealed that a majority
of patient samples displayed elevated AURKA, which in turn enhanced Aurora A kinase
at the centrosome leading to dysregulated microtubule dynamics and genomic instability.
Mediating AURKA levels rescued microtubule dynamic, resulting in enhanced genomic
stability
119
. We also observe enhanced genomic stability in both benign ovarian
cystadenomas as well as ovarian carcinoma cells treated with siRNA targeting Aurora A.
Our results in benign ovarian cystadenomas reveal a potentially novel role for Aurora A
in mitotic checkpoint control that is downstream of the spindle assembly checkpoint.
Inhibition of Aurora A in ovarian carcinoma cells, may reduce chromosomal instability
98
by a similar mitotic arrest scenario. While the mechanism described by our group differs
from the mechanism previously described in colorectal cancer cells, the resultant
phenotype is the same, suggesting that Aurora A may play varying roles depending on
the tumor cell origin and, hence, variable genetic backgrounds. Regardless of the
mechanism, from a clinical perspective, inhibition of Aurora A resulting in enhanced
genomic stability may ultimately reduce or delay acquisition of drug resistance in ovarian
carcinomas.
Although our results are preliminary, they suggest that the mechanism of
polyploidy development observed in our precancerous cell model continues beyond
oncogenic transformation and that polyploidy development can be enhanced by forced
down-regulation of BRCA1. Aurora A inhibition can be utilized as a means to increase
genomic stability in ovarian carcinoma cell lines that harbor the equivalent of a P53
mutation, as evidenced by a decrease in multinucleation in both SKOV3 and CAOV3 cell
lines. While components of the mitotic checkpoint complex increase both HEY and
SKOV3 cell lines, with concurrent decrease in Cdc27, an activating component of the
anaphase promoting complex, it remains to be seen whether the mitotic arrest induced by
Aurora A inhibition is secondary to the classical spindle assembly checkpoint or to an
arrest in between anaphase A and anaphase B due to cyclin B1 stabilization, which we
observed in our benign ovarian cystadenomas. Further studies need to be conducted in
order to confirm these findings.

99

Chapter 5: Summary and Future Directions
Summary
Our laboratory has established an in vitro epithelial cell model to study the
development of tetraploidy, which is a precursor to aneuploidy. In this model, epithelial
strains derived from benign ovarian tumors (cystadenomas) are transfected with a SV40
Large T Antigen vector, which binds to and inhibits p53 and Rb in these cells. These
cells bypass senescence, the first mortality checkpoint (M1), which is characterized by a
growth arrest at G1, due in part to p53 and Rb inhibition. These cells continue to
proliferate until they reach a second mortality crisis (M2), which is characterized by
widespread apoptosis. The classical crisis event is associated with telomere attrition,
however, our lab demonstrated that the crisis event observed in our benign tumor model
precedes the telomere-associated crisis and instead, is associated with ploidy changes,
ultimately culminating in a mitotic arrest. Cells that overcome this ploidy dependent
crisis eventually lose chromosomes, resulting in aneuploidy. These cells continue to
proliferate until they reach a secondary crisis event associated with telomere shortening.
In order for a cell to reach replicative immortality, it needs to overcome both of these
crisis events
63
. Previous studies conducted to evaluate crisis were completed utilizing
fibroblasts. In these cell models, telomere attrition was the main determinant of crisis in
this cell type
120
. Considering that most human cancers are epithelial in nature, our
epithelial cell model may be more reflective of events that take place during tumor
development, especially with regards to DNA ploidy changes.
100
Germline mutations in BRCA1 account for approximately 10% of all ovarian
cancer cases, however, epigenetic events leading to reduced BRCA1 levels suggest a
wider role for the gene in sporadic cases as well
6
. We previously observed reduced levels
of BRCA1 as ovarian cystadenoma cells age in culture and suspected that it was precisely
this reduction that allowed cells to overcome the ploidy driven M phase arrest, giving rise
to a proliferating polyploid population. Indeed, forced reduction of BRCA1 utilizing
siRNA, allowed cells to overcome this mitotic arrest, giving rise to a more robust
proliferating polyploid population. Based on these results, we suspected that BRCA1
played a role in mediating this mitotic arrest. My work further supported data of a
previous student, Vanessa Yu, PhD, which ultimately concluded that BRCA1 played a
role in mediating the mitotic arrest that we observe in benign ovarian tumors undergoing
ploidy dependent crisis. We feel that our data supports a model for BRCA-induced
cancer development, which may be particularly relevant for the development of cancers
with near tetraploid/polyploidy karyotypes.
We next sought to identify upstream regulators of BRCA1 expression levels and
felt that Aurora A was a strong candidate based on its influence on post-translational
modifications of BRCA1. Although it was apparent that Aurora A mediated BRCA1
expression levels, my work concluded that the mitotic arrest induced by Aurora A
inhibition was not mediated via its influence on BRCA1. Rather, Aurora A inhibition
induces a mitotic arrest that is downstream of the physiological arrest at the spindle
assembly checkpoint, which is observed in aging ovarian cystadenomas. Specifically,
down-regulation of Aurora A utilizing siRNA leads to a mitotic arrest in between
101
anaphase A and B, due to cyclin B1 stabilization. We propose that this is a previously
undescribed step in the M phase checkpoint.
While Aurora A over-expression has been described in a variety of tumors, the
mechanism by which it contributes to chromosomal instability is unclear. Furthermore,
while many groups have described a G2/M phase arrest following Aurora A inhibition,
the mechanism by which this occurs is poorly described. We feel that our cell model
provides a useful tool to compare and contrast changes in mitotic regulators following
Aurora A inhibition to a well characterized model that is based on a physiological mitotic
arrest versus a mitotic arrest induced by drug treatment. Furthermore, our cell model
provides a purer genetic background as opposed to carcinoma cell lines which may have
genetic alterations that either enhance or limit the mitotic changes we set out to observe.
Such observations regarding the precise mechanism of the mitotic arrest induced by
Aurora A inhibition may have clinical implications to help identify which
chemotherapeutic agents would be most effective when paired with Aurora A kinase
inhibitors.
The majority of my work has been completed in benign ovarian cystadenomas
which we regard as a precursor model for ovarian carcinomas. We suspected that the
observations that we have made in our ovarian cystadenoma model, continue beyond
oncogenic transformation. In an effort to demonstrate this, I evaluated the effect of
BRCA1 and Aurora A down-regulation on DNA ploidy status and multi-nucleation and
considered P53 status when choosing candidate cell lines, since our ovarian cystadenoma
cell model as well as nearly all high-grade serious ovarian carcinomas lack a functional
p53. Although preliminary, my results suggest that a small population of P53 mutant
102
ovarian carcinoma cell lines undergo a mitotic arrest which can be overcome due to
forced reduction of BRCA1 levels. This mechanism may perpetuate the ploidy changes
that we observe in ovarian carcinoma cells and lend to tumor heterogeneity that is
responsible for development of disease resistance. Understanding the mechanism
surrounding such ploidy changes, may help develop therapeutic agents that target such
genetic instability, ultimately leading to reduced treatment resistance.

Future Directions
1. Identify the activating phosphorylation event mediated by Aurora A that
contributes to cyclin B1 degradation

My findings demonstrate that cyclin B1 stabilization results in a mitotic arrest in
between anaphase A and anaphase B following Aurora A inhibition. Cyclin B1 is a
target of APC/Cdc20’s ubiquitin ligase activity and its degradation is essential for mitotic
exit. Securin, which is also a target of APC/Cdc20, is degraded following Aurora A
inhibition, which implies that the APC/Cdc20’s ubiquitin ligase activity is seemingly
intact. Thus, it is possible that an activating phosphorylation event on Cdc20 is necessary
to dictate specificity of the APC/Cdc20 to cyclin B1, as opposed to securin. It is also
plausible that the phosphorylation event takes place on cyclin B1, thereby making it a
better substrate for ubiquitin mediated degradation. In an effort to identify novel
phosphorylation sites on either Cdc20 or cyclin B1 by Aurora A kinase, an in vitro kinase
assay will be completed first to confirm phosphorylation mediated by Aurora A kinase
versus another kinase. In these experiments, ovarian epithelial cells approaching ploidy
dependent crisis (ML10) will be cultured and harvested. A pull-down assay can be
completed using protein A agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA)
103
and an anti-Aurora A kinase antibody (Cell Signaling, Danvers, MA). Incubation of the
Aurora A/protein A agarose complex with purified Cdc20 and cyclin B1 will be
completed prior to performing a kinase reaction using a kinase reaction mixture, which
containing
32
P-ATP. After the kinase reaction has been completed samples can be run on
a polyacrylamide gel, transferred to a PVDF membrane, and exposed the gel to x-ray
film. If, upon exposure,
32
P band is visualized, one will assume that this represents
phosphorylated Cdc20 or cyclin B1. In an effort to confirm that differences in
phosphorylation status is not secondary to different amounts of antibody or substrate in
each well, gels will be stained with Coomassie blue. This stain can be utilized to
visualize the phosphorylated Cdc20 or cyclin B1 and the band(s) can be excised and sent
to the mass spectroscopy core in order to identify novel phosphorylation sites. The mass
spectroscopy core will perform trypsin digestion, followed by phospho-peptide
enrichment since many times, phosphorylated proteins only make up 2-3% of the total
protein. Upon phospho-peptide enrichment, the samples will be spotted onto a plate for
mass spectroscopy. Once the mass spectrum is obtained, virtual digest can be performed
using the accession number for Cdc20 or cyclin B1 and a program developed by UCSF
(http://prospector.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msdigest). This virtual
digest can identify the putative phosphorylation sites and will also show accompanying
mass peaks for these sites. These peaks can then be matched to the peaks in acquired MS
spectrum and manually input the phospho-peaks for tanden MS/MS. A true phospho-site
by Aurora A kinase on Cdc20 or cyclin B1 will yield a high ion score, which is obtained
by MS/MS.
104
If the experiment above yields exciting results, the next step would entail site-directed
mutagenesis of that specific phosphorylation site of Cdc20 or cyclin B1 to confirm the mass
spectroscopy findings in vivo. In these experiments, in vitro kinase assays will be performed using
wild-type Cdc20 or cyclin B1 and Cdc20 and cyclin B1 phospho-site specific mutant as substrates. If
this is a true phosphorylation site of Aurora A kinase, then one would expect to see a reduced
32
P band
upon film exposure.
2. Further elucidate the phenotype following Aurora A down-regulation in
ovarian carcinoma cell lines

My preliminary results in ovarian carcinoma cell lines suggest that a small subset
undergo a mitotic arrest which can be overcome with forced down-regulation of BRCA1
levels using siRNA, based on DNA ploidy changes and multinucleation counts.
Furthermore, Aurora A inhibition seems to induce a mitotic arrest in these cells, resuling
in a decrease in multinucleation. While I see an increase in mitotic regulators associated
with the mitotic checkpoint complex and a concurrent decrease in Cdc27, a marker for
anaphase promoting complex activation in both HEY and SKOV3 cell lines, more work
needs to be conducted to determine the status of securin protein levels upon BRCA1 and
Aurora A down-regulation. It is possible that Aurora A inhibition induces a mitotic
arrest that is downstream of the spindle assembly checkpoint, similar to what we
observed in our benign ovarian cystadenomas. If this is the case, we may see reduced
securin levels following Aurora A down-regulation which would suggest a similar
mechanism of mitotic arrest. If this is the case, levels of EB1 can be evaluated by
western blot and localization of EB1 can be investigated utilizing confocal microscopy by
similar mechanisms described in this chapter 3. Such elucidation of the mitotic arrest in
105
ovarian carcinoma cells will help identify which therapeutic agents should be paired with
Aurora A kinase inhibitors when treating ovarian carcinoma patients.

106
References

1
Jacobs IJ, Menon U. Progress and challenges in screening for early detection of ovarian
cancer. Mol Cell Proteomics. 2004 Apr;3(4):355-66.

2
Kurta ML, Edwards RP, Moysich KB, McDonough K, Bertolet M, Weissfeld JL, Catov
JM, Modugno F, Bunker CH, Ness RB, Diergaarde B. Prognosis and conditional
disease-free survival among patients with ovarian cancer. J Clin Oncol. 2014 Nov 17.

3
Dubeau L. The cell of origin of ovarian epithelial tumours. Lancet Oncol. 2008
Dec;9(12):1191-7.

4
Jayson GC, Kohn EC, Kitchener HC, Ledermann JA. Ovarian cancer. Lancet. 2014
Oct;384(9951):1376–88.

5
Vargas AN. Natural history of ovarian cancer.
Ecancermedicalscience. 2014 Sep 25;8:465.

6
Ruscito I, Dimitrova D, Vasconcelos I, Gellhaus K, Schwachula T, et al. BRCA1 gene
promoter methylation status in high-grade serous ovarian cancer patients – A study of the
tumour Bank ovarian cancer (TOC) and ovarian cancer diagnosis consortium (OVCAD).
Eur J Cancer. 2014 Aug;50(12):2090-8.

7
Ehrlichova M, Mohelnikova-Duchonova B, Hrdy J, Brynychova V, Mrhalova M, Kodet
R, Rob L, Pluta M, Gut I, Soucek P, Vaclavikova R. The association of taxane resistance
genes with the clinical course of ovarian carcinoma. Genomics. 2013 Aug;102(2):96-101.

8
Horn LC1, Kafkova S, Leonhardt K, Kellner C, Einenkel J. Serous tubal in situ
carcinoma (STIC) in primary peritoneal serous carcinomas. Int J Gynecol Pathol. 2013
Jul;32(4):339-44.

9
Voeghtly LM, Mamula K, Campbell JL, Shriver CD, Ellsworth RE. Molecular
alterations associated with breast cancer mortality. PLos One. 2012;7(10):e46814.

10
Davaadelger B, Shen H, Maki CG. Novel roles for p53 in the genesis and targeting of
tetraploid cancer cells. PLos One 2014 Nov 7;9(11):e110844.

11
Ohshima S. Centrosome aberrations associated with cellular senescence and p53
localization at supernumerary centrosomes. Oxid Med Cell Longev. 2012;2012:217594.

12
Zhang Z, Arora S, Zhou Y, Cherry A, Wang TS. Replication-compromised cells
require the mitotic checkpoint to prevent tetraploidization. Cell Cycle. 2013
May1;12(9):1424-32.

107
13
Morotti M, Becker CM, Menada MV, Ferrero S. Targeting tyrosine-kinases in ovarian
cancer. Expert Opin Investig Drugs. 2013 Oc;22(10):1265-79.

14
Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, harshman K, Tavtigian S, Liu Q,
Chchran C, Bennett LM, Ding W, et al. A strong candidate for the breast and ovarian
cancer susceptibility gene BRCA1. Science 1994;266:66-71.

15
Wilson CA, Payton MN, Elliott GS, Buaas FW, Cajulis EE, Grosshans D, Ramos L,
Reese DM, Slamon DJ, Calzone FJ. Differential subcellular localization, expression and
biological toxicity of BRCA1 and the splice variant BRCA1-delta11b. Oncogene
1997;14:1-16

16
El Shamy WM, Livingston DM. Identification of BRCA1-IRIS, a BRCA1 locus
product. Nat Cell Biol 2004;6:954-67.

17
Scully R, Ganesan S, Vlasakova K, Chen J, Socolovsky M, Livingston DM. Genetic
analysis of BRCA1 function in a defined tumor cell line. Mol Cell 1999;4:1093-1099.

18
Xu X, Weaver Z, Linke SP, Li C, Gotay J, Wang XW, Harris CC, Ried T, Deng CX.
Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic
instability in BRCA1 exon 11 isofomr deficient cells. Mol Cell 1999;3:389-95.

19
Monteiro AN, August A, Hanafusa H. Evidence for a transcriptional activation
function of BRCA1 C-terminal region. Proc Natl Acad Sci USA 1996;93:3595-13599.

20
Brzovic PS, Rajagopal P, Hoyt DW, King MC, Klevit RE. Structure of a BRCA1-
BARD1 heterodimer RING-RING complex. Nat Struct Biol 2001;8:833-837.

21
Mallery DL, Vandenberg CJ, Hiom K. Activation of the E3 ligase function of the
BRCA1/BARD1 complex by polyubiquitin chains". EMBO J. 21 (24): 6755–62.

22
Warmoes M, Jaspers JE, Pham TV, Piersma SR, Oudgenoeg G, Massink MP, Waisfisz
Q, Rottenberg S, Boven E, Jonkers J, Jimenez CR (July 2012). Proteomics of mouse
BRCA1-deficient mammary tumors identifies DNA repair proteins with potential
diagnostic and prognostic value in human breast cancer". Mol. Cell Proteomics 11 (7):
M111.013334.

23
Boulton SJ (November 2006). Cellular functions of the BRCA tumour-suppressor
proteins". Biochem. Soc. Trans. 34 (Pt 5): 633–45.

24
Starita LM, Parvin JD. Substrates of the BRCA1 dependent ubiquitin ligase. Cancer
Biol Ther 2006;5:137-141.

25
Wang RH, Yu H, Deng CX. A requirement for breast-cancer-associated gene 1
(BRCA1) in the spindle checkpoint. Proc Natl Acad Sci USA 2004;101:7108-7113.

108

26
Wang B, Matsuoka S, Ballif BA, Zhang D, Smogorzewska A, Gygi SP, Elledge SJ.
Abraxas and RAP80 form a BRCA1 protein complex required for the DNA damage
response. Science. 2007;316:1194–1198.

27
Sobhian B, Shao G, Lilli DR, Culhane AC, Moreau LA, Xia B, Livingston DM,
Greenberg RA. RAP80 targets BRCA1 to specific ubiquitin structures at DNA damage
sites. Science. 2007;316:1198–1202.

28
Kim H, Chen J, Yu X. Ubiquitin-binding protein RAP80 mediates BRCA1-dependent
DNA damage response. Science. 2007;316:1202–1205.

29
Yan J, Kim YS, Yang XP, Li LP, Liao G, Xia F, Jetten AM. The ubiquitin-interacting
motif containing protein RAP80 interacts with BRCA1 and functions in DNA damage
repair response. Cancer Res. 2007;67:6647–6656.

30
Parvin JD, Sankaran S. The BRCA1 E3 ubiquitin ligase controls centrosome
dynamics. Cell Cycle. 2006 Oct 15;15(20):3012-23.

31
Starita LM, Machida Y, Sankaran S, Elias JE, Griffin K, Schlegel BP, Gygi SP, Parvin
JD. BRCA1-dependent ubiquitination of γ-tubulin regulates centrosome number. Mol
Cell Biol 2004;24:8457-8466.

32
Sankaran S, Crone DE, Palazzo RE, Parvin JD. Aurora-A kinase regulates breast
cancer-associated gene 1 inhibition of centrosome-dependent microtubule nucleation.
Cancer Res 2007;67:11186-11194.

33
Hakem R, de la Pompa JL, Sirard C, Mo R, Woo M, Hakem A, Wakeham A, Potter J,
Reitmair A, Bilia F, Firpo E, Hui CC, Roberts J, Rossant J, Mak TW. The tumor
suppressor gene Brca1 is required for embryonic cellular proliferation in the mouse. Cell
1996;85:1009-1023.

34
Ludwig T, Chapman DL, Papaioannou VE, Efstratiadis A. Targeted mutations of
breast cancer susceptibility gene homologs in mice: lethal phenotypes of Brca1, Brca2,
Brac1/Brca2, Brca1/p53, and Brca2/p53 nullizygous embryos. Genes Dev 1997;11:1226-
1241.

35
Shen SC, Weaver Z, Xu X, Li C, Weintstein W, Guan XY, Ried T, Deng CX. A
targeted disruption of the murine Brca1 gene causes gamma-irradiation hypersensitivity
and genetic instability. Oncogene 1998;17: 3115-24.

36
Deng CX. Tumorigenesis as a consequence of genetic instability in Brca1 mutant mice.
Mutat Res 2001;477:183-189.

109
37
Yarden RI1, Pardo-Reoyo S, Sgagias M, Cowan KH, Brody LC. BRCA1 regulates the
G2/M checkpoint by activating Chk1 kinase upon DNA damage. Genet. 2002
Mar;30(3):285-9.

38
Tarapore P, Hanashiro K, Fukasawa K. Analysis of centrosome localization of BRCA1
and its activity in suppressing centrosomal aster formation. Cell Cycle. 2012 Aug
1;11(15):2931-46.

39
Landen CN Jr, Lin YG, Immaneni A, Deavers MT, Merritt WM, et al. Overexpression
of the centrosomal protein Aurora-A kinase is associated with poor prognosis in epithelial
ovarian cancer patients. Clin Cancer Res. 2007 Jul 15;13(14);4098-104.

40
Ouchi M, Fujiuchi N, Sasai K, Katayama H, Minamishima YA, Ongusaha PP, Deng
C, Sen S, Lee SW, Ouchi T. BRCA1 phosphorylation by Aurora-A in the regulation of
G2 to M transition. J Biol Chem 2004;279:19643-19648.

41
Hirota T1, Kunitoku N, Sasayama T, Marumoto T, Zhang D, Nitta M, Hatakeyama K,
Saya H. Aurora-A and an interacting activator, the LIM protein Ajuba, are required for
mitotic commitment in human cells. Cell. 2003 Sep 5;114(5):585-98.

42
Nair JS, Ho AL, Schwartz GK. The induction of polyploidy or apoptosis by the
Aurora A kinase inhibitor MK8745 is p53-dependent. Cell Cycle. 2012 Feb
15;11(4):807-17.

43
Yang G, Chang B, Yang F, Guo X, Cai KQ, et al. Aurora kinase A promotes ovarian
tumorigenesis through dysregulation of the cell cycle and suppression of BRCA2. Clin
Cancer Res. 2010 Jun 15;16(12):3171-81.

44
Marxer M, Ma HT, Man WY, Poon RY. p53 deficiency enhances mitotic arrest and
slippage induced by pharmacological inhibition of Aurora kinases. Oncogene. 2014 Jul
3;33(27):3550-60.

45
Lara-Gonzalez P, Westhorpe FG, Taylor SS. The spindle assembly checkpoint. Curr
Biol. 2012 Nov 20;22(22):R966-80.
46
Baker DJ, Jeganathan KB, Cameron JD, Thompson M, Juneja S,. BubR1 insufficiency
causes early onset of aging-associated phenotypes and infertility in mice. Nature Genet.
2004;36:744–749.
47
Rao CV, Yang YM, Swamy MV, Liu T, Fang Y. Colonic tumorigenesis in BubR1
+/-
Apc
Min/+
compound mutant mice is linked to premature separation of sister chromatids
and enhanced genomic instability. Proc. Natl Acad. Sci. USA. 2005;102, 4365–4370.
48
Rieder CL, Maiato H. Stuck in division or passing through: what happens when cells
cannot satisfy the spindle assembly checkpoint. Dev Cell. 2004;7: 637–651.
110
49
Futreal PA, Liu Q, Shattuck-Eidens D, Cochran C, Harshman K, Tavtigian S, et al.
BRCA1 mutations in primary breast and ovarian carcinomas. Science. 1994; 266(3):
120–2.
50
Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Harshman K, Tavtigian, et al. A
strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science.
1994; 266: 66–71.
51
Wu LC, Wang ZW, Tsan JT, Spillman MA, Phung A, Xu XL, Yang MC, Hwang LY,
Bowcock AM, Baer R. Identification of a RING protein that can interact in vivowith the
BRCA1 gene product. Nat Genet. 1996;14: 430–40.
52
Bell DW, Varley JM, Szydlo TE. Heterozygous germ line hCHK2 mutations in Li-
Fraumeni syndrome. Science. 1999; 286: 2528–31.
53
Cortez D, Wang Y, Qin J, Elledge SJ. Requirement of ATM-dependent
phosphorylation of brca1 in the DNA damage response to double-strand breaks. Science.
1999; 286: 1162–6.
54
Gatei M, Zhou BB, Hobson K, Scott S, Young D, Khanna KK. Ataxia telangiectasia
mutated (ATM) kinase and ATM and Rad3 related kinase mediate phosphorylation of
Brca1 at distinct and overlapping sites. In vivo assessment using phospho-specific
antibodies. J Biol Chem. 2001; 276: 17276–80.
55
Babu JR, Jeganathan KB, Baker DJ, Wu X, Kang-Decker N, van Deursen JM. Rae1 is
an essential mitotic checkpoint regulator that cooperates with Bub3 to prevent
chromosome missegregation. J Cell Biol. 2003; 160: 341–53.
56
Dai W, Want Q, Liu T, Swamy M, Fang Y, Xie S, Mahmood R, Yang YM, Xu M, Rao
CV. Slippage of mitotic arrest and enahanced tumor development in mice with BubR1
haploinsufficiency. Cancer Res 2004;62:440–5.
57
Michael LS, Liberal V, Chatterjee A, Kirchwegger R, Pasche B, Gerald W, Dobles M,
Sorger PK, Murty VV, Benezra R. Mad2 haplo-insufficiency causes premature anaphase
and chromosome instability in mammalian cells. Nature 2001;409:355–9.
58
Luo MP, Gomperts B, Imren S, DeClerck YA, Ito M, Velicescu M, Felix JC, Dubeau
L. Establishment of long-term in vitro cultures of human ovarian cystadenomas and LMP
tumors and examination of their spectrum of expression of matrix- degrading proteinases.
Gynecol Oncol 1997; 67:277–84.
59
Folkins AK, Jarboe EA, Saleemuddin A, Lee Y, Callahan MJ, Drapkin R, Garber JE,
Muto MG, Tworoger S, Crum CP. A candidate precursor to pelvic serous cancer (p53
signature) and its prevalence in ovaries and fallopian tubes from women with BRCA
mutations. Gynecol Oncol 2008; 109:168–73.
60
Lee Y, Miron A, Drapkin R, Nucci MR, Medeiros F, Saleemuddin A, Garber J, Birch
C, Mou H, Gordon RW, Cramer DW, McKeon FD, et al. A candidate precursor to serous
carcinoma that originates in the distal fallopian tube. J Pathol 2007;211:26–35.
111
61
Norquist BM, Garcia RL, Allison KH, Jokinen CH, Kernochan LE, Pizzi CC, Barrow
BJ, Goff BA, Swisher EM. The molecular pathogenesis of hereditary ovarian carcinoma.
Cancer 2010;116: 5261–71.
62
Ahmed AA, Etemadmoghadam D, Temple J, Lynch AG, Riad M, Sharma R, Stewart
C, Fereday S, Caldas C, Defazio A, Bowtell D, Brenton JD. Driver mutations in TP53 are
ubiquitous in high grade serous carcinoma of the ovary. J Pathol 2010;221: 49–56.
63
Velicescu M, Yu J, Herbert BS, Shay JW, Granada E, Dubeau L. Aneuploidy and
telomere attrition are independent determinants of crisis in SV40-transformed epithelial
cells. Cancer Res 2003;63: 5813–20.
64
Yu VM, Marion CM, Austria TM, Yeh J, Schonthal AH, Dubeau L. Role of BRCA1
in controlling mitotic arrest in ovarian cystadenoma cells. Int J Cancer 2012;130:2495-
2504.
65
Hoyt MA, Totis L, Roberts BT. S. cerevisiae genes required for cell cycle arrest in
response to loss of microtubule function. Cell 1991;66:507–17.
66
Li R, Murray AW. Feedback control of mitosis in budding yeast. Cell 1991;66: 519–
31.
67
Taylor SS, Ha E, McKeon F. The human homologue of Bub3 is required for
kinetochore localization of Bub1 and a Mad3/Bub1-related protein kinase. J Cell Biol
1998;142:1–11.
68
King RW, Peter JM, Tugendreich S, Rolfe M, Hieter P, Kirschner MW. A 20S
complex containing Cdc27 and Cdc16 catalyzes the mitosis-specific conjugation of
ubiquitin to cyclin B. Cell 1995 Apr 21;81:279-288.
69
Hagting A, Den Elzen N, Vodermaier HC, Waizenegger IC, Peters JM, Pines J. Human
securin proteolysis is controlled by the spindle checkpoint and reveals when the APC/C
switches from activation by Cdc20 to Cdh1. J Cell Biol 2002;157: 1125–37.
70
Thornton BR, Toczyski DP. Securin and B- cyclin/CDK are the only essential targets
of the APC. Nat Cell Biol 2003;5: 1090–4.
71
Chen Y, Farmer AA, Chen CF, Jones DC, Chen PL, Lee WH. BRCA1 is a 220-kDa
nuclear phosphoprotein that is expressed and phosphorylated in a cell cycle- dependent
manner. Cancer Res 1996;56: 3168–72.
72
Vaughn JP, Cirisano FD, Huper G, Berchuck A, Futreal PA, Marks JR, Iglehart JD.
Cell cycle control of BRCA2. Cancer Res 1996;56:4590–4.
73
Scully R, Chen J, Plug A, Xiao Y, Weaver D, Feunteun J, Ashley T, Livingston DM.
Association of BRCA1 with Rad51 in mitotic and meiotic cells. Cell 1997;88: 265–75.

112
74
Bochar DA, Wang L, Beniya H, Kinev A, Xue Y, Lane WS, Wang W, Kashanchi F,
Shiekhattar R. BRCA1 is associated with a human SWI/SNF-related complex: linking
chromatin remodeling to breast cancer. Cell 2000;102:257–65.
75
Wu W, Sato K, Koike A, Nishikawa H, Koizumi H, Venkitaraman AR, Ohta T.
HERC2 is an E3 ligase that targets BRCA1 for degradation. Cancer Res 2010;70: 6384–
92.
76
Xu B, Kim ST, Kastan MB. Involvement of Brca1 in S-phase and G(2)-phase
checkpoints after ionizing radiation. Mol Cell Biol 2001;21:3445–50.
77
Cornelis RS, Neuhausen SL, Johansson O, Arason A, Kelsell D, Ponder BA, Tonin P,
Hamann U, Lindblom A, Lalle P. High allele loss rates at 17q12-q21 in breast and
ovarian tumors from BRCAl-linked families. The Breast Cancer Linkage Consortium.
Genes Chromosomes Cancer 1995;13:203–10.
78
Neuhausen SL, Marshall CJ. Loss of heterozygosity in familial tumors from three
BRCA1-linked kindreds. Cancer Res 1994;54:6069–72.
79
Smith SA, Easton DF, Evans DG, Ponder BA. Allele losses in the region 17q12–21 in
familial breast and ovarian cancer involve the wild-type chromosome. Nat Genet
1992;2:128–31.
80
Rice JC, Massey-Brown KS, Futscher BW. Aberrant methylation of the BRCA1 CpG
island promoter is associated with decreased BRCA1 mRNA in sporadic breast cancer
cells. Oncogene 1998;17: 1807–12.
81
Tomlinson GE, Chen TT, Stastny VA, Virmani AK, Spillman MA, Tonk V, Blum JL,
Schneider NR, Wistuba II, Shay JW, Minna JD, Gazdar AF. Characterization of a breast
cancer cell line derived from a germ-line BRCA1 mutation carrier. Cancer Res
1998;58:3237–42.
82
Swisher EM, Sakai W, Karlan BY, Wurz K, Urban N, Taniguchi T. Secondary
BRCA1 mutations in BRCA1-mutated ovarian carcinomas with platinum resistance.
Cancer Res 2008;68:2581–6.
83
Chodankar R, Kwang S, Sangiorgi F, Hong H, Yen HY, Deng C, Pike MC, Shuler CF,
Maxson R, Dubeau L. Cell-nonautonomous induction of ovarian and uterine serous
cystadenomas in mice lacking a functional Brca1 in ovarian granulosa cells. Curr Biol
2005;15:561–5.
84
Hong H, Yen HY, Brockmeyer A, Liu Y, Chodankar R, Pike MC, Stanczyk FZ,
Maxson R, Dubeau L. Changes in the mouse estrus cycle in response to Brca1
inactivation suggest a potential link between risk factors for familial and sporadic ovarian
cancer. Cancer Res 2010; 70:221–8.

113
85
Eerola H, Heikkila P, Tamminen A, Aittomaki K, Blomqvist C, Nevanlinna H.
Histopathological features of breast tumours in BRCA1, BRCA2 and mutation- negative
breast cancer families. Breast Cancer Res 2005;7:R93–R100.
86
Turner NC, Reis-Filho JS. Basal-like breastcancer and the BRCA1 phenotype.
Oncogene 2006;25:5846–53.
87
Hsieh CL. Basic cytogenetic techniques: culture, slidemaking and bandinged, 2nd edn.
New York: Academic Press, 1997.
88
Crane, R. Gadea B, Littlepage L, Wu H, Ruderman JV. Aurora A, Meiosis and Mitosis.
Biology of the Cell 2003;96(3): 215–229.
89
Hannak E, Kirkham M, Hyman AA, Oegema K. Aurora-A kinase is required for
centrosome maturation in Caenorhabditis elegans. J. Cell Biol. 2001;155 (7):1109–16.
90
Marumoto T, Honda S, Hara T, Nitta M, Hirota T, Kohmura E, Saya H. Aurora-A
Kinase Maintains the Fidelity of Early and Late Mitotic events in HeLa Cells. J. Biol.
Chem. 2003; 278:51786-95.

91
Keen N, Taylor S. Aurora-kinase inhibitors as anti-cancer agents. Nat Rev Cancer 2004
Dec;4(12):927-936.
92
Zhou N1, Singh K, Mir MC, Parker Y, Lindner D, Dreicer R, Ecsedy JA, Zhang Z, Teh
BT, Almasan A, Hansel DE. The investigational Aurora kinase A inhibitor MLN8237
induces defects in cell viability and cell-cycle progression in malignant bladder cancer
cells in vitro and in vivo. Clin Cancer Res. 2013 Apr 1;19(7):1717-28.
93
Sun L, Gao J, Dong X, Liu M, Li D, et al. EB1 promotes Aurora-B kinase activity
through blocking its inactivation by protein phosphatase 2A. PNAS 2008 May
20;105(20):7153-7158.
94
Tirnauer JS, Canman JC, Salmon ED, Mitchison TJ. EB1 targets to kinetochores with
attached, polymerizing microtubules. Mol Biol Cell 2002 Oct;13(12):4308-16.
95
Bischoff JR, Anderson L, Zhu Y, et al. A homologue of Drosophila aurora kinase is
oncogenic and amplified in human colorectal cancers. EMBO J 1998; 17: 3052-65.

96
Gritsko TM, Coppola D, Paciga JE, et al. Activation and overexpression of centrosome
kinase BTAK/Aurora-A in human ovarian cancer. Clin Cancer Res 2003; 9: 1420-6.

97
Lingle WL, Lutz WH, Ingle JN, Maihle NJ, Salisbury JL. Centrosome hypertrophy in
human breast tumors: implications for genomic stability and cell polarity. Proc Natl Acad
Sci U S A 1998; 95: 2950-5.

98
Takahashi T, Futamura M, Yoshimi N, et al. Centrosomal kinases, HsAIRK1 and
HsAIRK3, are overexpressed in primary colorectal cancers. Jpn J Cancer Res 2000; 91:
1007-14.
114

99
Tanaka E, Hashimoto Y, Ito T, et al. The clinical significance of Aurora-
A/STK15/BTAK expression in human esophageal squamous cell carcinoma. Clin Cancer
Res 2005; 11: 1827-34.
100
Tanaka T, Kimura M, Matsunaga K, Fukada D, Mori H, Okano Y. Centrosomal kinase
AIK1 is overexpressed in invasive ductal carcinoma of the breast. Cancer Res 1999; 59:
2041-4.

101
Tanner MM, Grenman S, Koul A, et al. Frequent amplification of chromosomal region
20q12-q13 in ovarian cancer. Clin Cancer Res 2000; 6: 1833-9.

102
Tong T, Zhong Y, Kong J, et al. Overexpression of Aurora-A contributes to malignant
development of human esophageal squamous cell carcinoma. Clin Cancer Res 2004; 10:
7304-10.

103
Watanabe T, Imoto I, Katahira T, et al. Differentially regulated genes as putative
targets of amplifications at 20q in ovarian cancers. Jpn J Cancer Res 2002; 93: 1114-22.

104
Zhou H, Kuang J, Zhong L, et al. Tumour amplified kinase STK15/BTAK induces
centrosome amplification, aneuploidy and transformation. Nat Genet 1998; 20: 189-93.

105
Jeng YM, Peng SY, Lin CY, Hsu HC. Overexpression and amplification of Aurora-A
in hepatocellular carcinoma. Clin Cancer Res 2004; 10: 2065-71.
106
Cross SM, Sanchez CA, Morgan CA, et al. A p53-dependent mouse spindle
checkpoint. Science 1995; 267: 1353-6.

107
Elhajouji A, Cunha M, Kirsch-Volders M. Spindle poisons can induce polyploidy by
mitotic slippage and micronucleate mononucleates in the cytokinesis-block assay.
Mutagenesis 1998; 13: 193-8.

108
Harrington EA, Bebbington D, Moore J, et al. VX-680, a potent and selective small-
molecule inhibitor of the Aurora kinases, suppresses tumor growth in vivo. Nat Med
2004; 10: 262-7.

109
Kamada K, Yamada Y, Hirao T, et al. Amplification/overexpression of Aurora-A in
human gastric carcinoma: potential role in differentiated type gastric carcinogenesis.
Oncol Rep 2004; 12: 593-9.

110
Xu J, Wu X, Zhou WH, et al. Aurora-A identifies early recurrence and poor prognosis
and promises a potential therapeutic target in triple negative breast cancer. PLoS One
2013; 8: e56919.

111
Struski S, Doco-Fenzy M, Koehler M, et al. Cytogenetic evolution of human ovarian
cell lines associated with chemoresistance and loss of tumorigenicity. Anal Cell Pathol
2003; 25: 115-22.
115

112
TCGA Research Network: http://cancergenome.nih.gov/

113
Wang YQ, Yan Q, Zhang JR, Li SD, Yang YX, Wan XP. Epigenetic inactivation of
BRCA1 through promoter hypermethylation in ovarian cancer progression. J Obstet
Gynaecol Res 2013; 39: 549-54.

114
Pothuri B, Leitao MM, Levine DA, et al. Genetic analysis of the early natural history
of epithelial ovarian carcinoma. PLoS One 2010; 5: e10358.

115
Naruganahalli KS, Lakshmanan M, Dastidar SG, Ray A. Therapeutic potential of
Aurora kinase inhibitors in cancer. Curr Opin Investig Drugs 2006; 7: 1044-51.

116
Caydasi AK, Lohel M, Grunert G, Dittrich P, Pereira G, Ibrahim B. A dynamical
model of the spindle position checkpoint. Mol Syst Biol 2012; 8: 582.

117
Lopes CS, Sampaio P, Williams B, Goldberg M, Sunkel CE. The Drosophila Bub3
protein is required for the mitotic checkpoint and for normal accumulation of cyclins
during G2 and early stages of mitosis. J Cell Sci 2005; 118: 187-98.

118
Musacchio A, Ciliberto A. The spindle-assembly checkpoint and the beauty of self-
destruction. Nat Struct Mol Biol 2012; 19: 1059-61.

119
Ertych N, Stolz A, Stenzinger A, Weichert W, Kaulfub S, Burfeind P, et al. Increased
microtubule assembly rate increases chromosomal instability in colorectal cancer cells.
Nat Cell Biol 2014: 16: 779-791.

120
Bodnar AG, Ouellette M, Frolkis M, Holt SE, Chiu CP, et al. Extension of life span
by introduction of telomerase into normal human cells. Science 1998; 279: 349-352.

Abstract (if available)

Abstract Aurora A belongs to a family of serine/threonine kinases known to play a role in mitotic regulation. More recently, the kinase has been implicated in tumorigenesis. Small molecule inhibitors of Aurora A are currently undergoing clinical trials, though their effects on cellular phenotype have varied based on species, cell type, and P53 status. While some studies have demonstrated a prolonged G2/M arrest post-treatment, others have observed catastrophic polyploidy, leading to cell death. Investigations based on these inhibitors are further complicated by their lack of specificity for a single kinase. ❧ We sought to elucidate the mechanism of mitotic arrest mediated by Aurora A and utilized siRNA technology to down-regulate Aurora A levels in order to avoid potential complications due to lack of specificity of current small molecule inhibitors. We also used a pre-cancerous cell model that we had previously extensively characterized in order to avoid genetic defects inherent to carcinoma cell lines, which would either enhance or limit the mitotic changes that we set out to observe. We previously showed that such cells undergo a physiological mitotic arrest at the spindle assembly checkpoint as they approach in vitro crisis, which can be overcome by reduced BRCA1 levels. Multinucleation ensues as a result of incomplete cytokinesis. We reasoned that Aurora A might exert some control over these phenomena given its previously described role in post-translational modifications of BRCA1. Here we show that inhibition of Aurora A leads to a mitotic arrest downstream of the spindle assembly checkpoint between anaphase A and anaphase B due to cyclin B1 stabilization. This arrest is not mediated by any effect of Aurora A on BRCA1. We suggest that this represents a previously undescribed step at the M phase checkpoint protecting against mitotic errors leading to polyploidy.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Understanding and controlling mitotic errors leading to aneuploidy in early ovarian cancer development

PDF

Mitotic regulation in ovarian epithelial tumors approaching in vitro crisis

PDF

Cell non-autonomous control of proliferation in tissues of elevated cancer risk in BRCA1 mutation carriers

PDF

POLR2B and its contributions to cancer cell growth

PDF

The discovery of a novel 105kDa complex comprised of BUB3, C-JUN and SUMO-1 and its relationship to ovarian cancer development

PDF

Integrative genomic and epigenomic analysis of human cancer

PDF

The role of PAX8 in epithelial ovarian carcinoma

PDF

Construction of plasmids for constitutive and induced expression of secreted alkaline phosphatase in mammalian ovarian carcinoma cells

PDF

Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme

PDF

Induction of hypersignaling as a therapeutic approach for treatment of BCR-ABL1 positive Acute Lymphoblastic Leukemia (ALL) cells

PDF

The role of endoplasmic reticulum chaperones in adipogenesis, liver cancer and mammary gland development

PDF

Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis

PDF

The role of Prkci in stem cell maintenance and cell polarity using a 3-D culture system

PDF

Investigating the role of STAT3 in mouse and rat embryonic stem cell self-renewal and differentiation

PDF

Investigating the role of SASH1 gene located on chromosome 6 in ovarian cancer

PDF

The calcium-sensing receptor in the specification of normal and malignant hematopoietic cell localization in the bone marrow microenvironment

PDF

The role of miRNA and its regulation in pulmonary hypertension in sickle cell disease

PDF

The essential role of histone H2A deubiquitinase MYSM1 in natural killer cell maturation and HSC homeostasis

PDF

Role of the bone marrow niche components in B cell malignancies

PDF

The mechanism by which extracellular Hsp90α promotes cell migration: implications in wound healing and cancer

Asset Metadata

Creator Marion, Christine M. (author)

Core Title Understanding the role of BRCA1 and Aurora A kinase in polyploidy development in ovarian carcinoma precursor cells

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 08/17/2015

Defense Date 12/04/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Aurora A,BRCA1,cell cycle,OAI-PMH Harvest,ovarian carcinoma,polyploidy

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Hofman, Florence M. (committee chair), Dubeau, Louis (committee member), Press, Michael F. (committee member), Schönthal, Axel H. (committee member)

Creator Email christine.marion@astellas.com,christine.marion@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-636634

Unique identifier UC11306658

Identifier etd-MarionChri-3830.pdf (filename),usctheses-c3-636634 (legacy record id)

Legacy Identifier etd-MarionChri-3830.pdf

Dmrecord 636634

Document Type Dissertation

Format application/pdf (imt)

Rights Marion, Christine M.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA