University of Southern California Dissertations and Theses

Investigating novel avenues for effective treatment of ovarian cancer

(USC Thesis Other)

Investigating novel avenues for effective treatment of ovarian cancer

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

by
Francesca Ferri

A Thesis Presented to the
FACULTY OF THE USC KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
MOLECULAR MICROBIOLOGY AND IMMUNOLOGY

August 2020

Copyright 2020 Francesca Ferri
Investigating Novel Avenues for
Effective Treatment of Ovarian Cancer
ii

Acknowledgements
First, off, I would like to acknowledge my PI, Dr. Axel Schönthal. Thank you for allowing
me to be a part of your lab, but more importantly for giving me the responsibility and dignity of a
colleague. Throughout my time at USC, you were the first to treat me with such respect, which has
been both a terrifying and gratifying experience. What a journey!
Next, I would like to thank my other committee members: Dr. Louis Dubeau, Dr. Weiming
Yuan, and Dr. Radu Minea; whose guidance have greatly impacted this thesis. In addition, all of
the Chen Glioma lab group has had a valuably influence throughout my master’s program. I would
like to give a special thank you to Dr. Stephen Swenson, Dr. Diganta Das, Dr. Camelia Danilov,
Dr. Catalina Silva-Hirschberg, Dr. Hee-Yeon Cho, Samantha Stack, Nazleen Mohseni, Robert
Herrera, (almost Dr.) Krutika Deshpande, and (almost Dr.) Vahan Maritirosian. All these
wonderful people have given me such academic and emotional support, which I will forever be
grateful. In addition, this thesis could not have taken form without the generous advice by fellow
colleagues: Libéré Ndayishimiye, David Lee, and Stephen Docherty.
This experience has been so rewarding and challenging in every regard, so I could not have
accomplished all this without the love and support from my friends and family. I would like to
give an immense heart-felt thank you to my wonderful father who shows endless pride and
encouragement; my dedicated mother who has been my ally and advocate through every endless
draft, stress frenzy, and academic feat; and my incredible brother who believed in me when I didn’t
believe in myself. I would not be who I am today without each of you, so thank you. I would also
like to acknowledge Imani Mitchell and Melissa Venegas for their constant reassurance and
comedic relief.
Finally, I would like to acknowledge past teachers and professors who were instrumental
in my development. An extraordinary thank you to Ms. Theresa Peters (who originally cultivated
my love of science) and Dr. Frank Jones (who initiated my journey through cancer biology); they
are the unsung heroes of academia. Their compassion and fervor for teaching has helped me
become the academic I am today. Thank you for teaching me to stay curious, to keep reading, to
value integrity over prestige, and most importantly showing me how to be a mature and
compassionate adult through your example.
I am so fortunate that God put all of these exceptional people in my life and for giving me
strength and perseverance through this thesis project.
iii

Contents
Acknowledgements………………………………………………………………………………..ii
List of Figures...…………………………………………………………………………………...v
List of Abbreviations...……………………………………………………………………………vi
Abstract…...……………………………………………………………………………………...vii
Chapter 1 – Introduction…………………………………………………………………………...1
1.1 Ovarian Cancer…………………………………………………………………………….1
1.1.1 Definition and Classification…………………………………………………….1
1.1.2 Epidemiology and Etiology………………………………………………….......2
1.1.3 Treatment……………………………………………………………………......3
1.1.4 Understanding Recurrence……………………………………………………....4
1.2 Integrins……………………………………………………………………………………5
1.2.1 Structure and Function…………………………………………………………...5
1.2.2 Uses in Cancer…………………………………………………………………...6
1.3 NEO212……………………………………………………………………………………9
1.4 Hypothesis…………………………………………………………………………………9
Chapter 2 – Materials and Methods………………………………………………………………11
2.1 Cell Lines and Maintenance………………………………………………………………11
2.2 Pharmacological Agents………………………………………………………………….11
2.3 Development of Clinically-Relevant Models…………………………………………….12
2.4 MTT Assay……………………………………………………………………………….13
2.5 Fluorescence-activated cell sorting (FACS) Analysis……………………………………14
2.6 Immunoblotting…………………………………………………………………………..15
2.7 Statistical Analysis……………………………………………………………………….15

iv

Chapter 3 – Results……………………………………………………………………………….16
3.1 Cytotoxicity effects of pharmacological agents used in clinical treatment………………..16
3.1.1 Cytotoxicity of Carboplatin…………………………………………………….16
3.1.2 Cytotoxicity of Paclitaxel………………………………………………………18
3.2 Delineation of differences in clinically-relevant models…………………………………20
3.2.1 Cytotoxicity differences………………………………………………………..20
3.2.2 ABC transporter profile………………………………………………………...21
3.2.3 Differential expression of cell proliferation and cell survival markers………….22
3.3 Integrin profile after chemotherapeutic pressure…………………………………………24
3.3.1 Western blot analysis of Integrin β1 expression………………………………...24
3.3.2 Levels of active integrin αv and α5 species……………………………………25
3.4 Determining chemotherapeutic sensitization using NEO212…………………………….27
3.4.1 Expression levels of cell survival markers in MGMT-positive ovarian
cancer cells…………………………………………………………………………...28
3.4.2 Expression levels of cell survival markers in ovarian cancer cells
minimally-expressing MGMT……………………………………………………….29
Chapter 4 – Discussion…………………………………………………………………………..31
References………………………………………………………………………………………..37

v

List of Figures
Figure 1.1: Active integrin conformers contribute to various outside-in signaling pathways that
promote cell survival and tumor progression……………………………………………………...8
Figure 2.1: Workflow for selection of chemotherapy-exposed subpopulations…………………13
Figure 3.1: Effect of carboplatin on cell viability in ovarian cancer cell lines used……………..16
Figure 3.2: Comparison of half maximal inhibitory concentrations of cell lines exposed to
carboplatin. ………………………………………………………………………………………17
Figure 3.3: Effect of paclitaxel on cell viability in ovarian cancer cell lines used………………18
Figure 3.4: Comparison of half maximal inhibitory concentrations of cell lines exposed to
paclitaxel…………………………………………………………………………………………19
Figure 3.5: Comparison of cytotoxicity curves of wild-type and recovered models…………….20
Figure 3.6: ABC transporter levels of the three clinically-relevant models used………………..22
Figure 3.7: Expression levels of cell viability and cell proliferation markers in all wild-type
and recovered cell populations…………………………………………………………………..23
Figure 3.8: Expression levels of integrin β1 in all wild-type and recovered models……………25
Figure 3.9: Levels of active αv and α5 integrin species in all clinically-relevant models of
HEYA8…………………………………………………………………………………………..26
Figure 3.10: Changes in expression levels of cell survival markers and integrin β1 in A2780
from NEO212 treatment. ………………………………………………………………………..28
Figure 3.11: Changes in expression levels of cell survival markers and integrin β1 in ES2
from NEO212 treatment…………………………………………………………………………29

vi

List of Abbreviations

ABC: ATP-binding cassette
AIC:5-aminoadazole-4-carboxamide
ATCC: American Type Culture Collection
BD: Becton Dickinson
CA125: cancer antigen 125
Cp: carboplatin
DMEM: Dulbecco’s Modification of Eagle’s
Medium
DMSO: dimethyl sulfoxide
ECD: extracellular domain
EOC: epithelial ovarian cancer
ECM: extracellular matrix
FACS: fluorescence-activated cell sorting
FAK: focal adhesion kinase
FBS: fetal bovine serum
GBM: glioblastoma multiform
HRP: horseradish peroxidase
HRT: hormone replacement therapy
IC50: half maximal inhibitory concentration
ICD: intracellular domain
ITG: integrins
MGMT: O-6-methylguanine
methyltransferase
MTIC: 5-(3-methyltriazen-1-yl) imidazole-
4-carboxamide
MTT: methylthiazoletetrazolium
ns: not significant
OD: optical densities
PARP: poly ADP-ribose polymerase
PBS: phosphate buffered saline
Ptx: paclitaxel
PVDF: polyvinylidene fluoride
R: Recovered
RIPA: radioimmunoprecipitation assay
RGD: Arginine-Glycine-Aspartate
RPM: rotations per minute
SDS: sodium dodecyl sulfate
T: Treated
TBST: 1x tris-buffered saline plus 0.1%
tween 20
TMZ: temozolomide
UT: untreated
VEGF: vascular endothelial growth factors
vh.: vehicle
WT: wild-type
vii

Abstract
Ovarian cancer is the most lethal gynecologic cancer bolstered by insufficient treatment
options and substantial recurrence rates. A well-studied proponent to this detriment is a protein
family of Arginine-Glycine-Aspartate (RGD)-recognizing integrins, which are commonly
exploited in ovarian cancer for metastasis, drug resistance, and tumor progression. In this study,
clinically-relevant models of different histotypes were developed in order to determine the
relationship between RGD receptors and chemotherapy; meanwhile, wild-types were subjected to
treatment by a temozolomide (TMZ) affiliate, NEO212, for determination of increased
sensitization to the standard of care. The veracity of the clinically-relevant models is dubious;
however, they clearly convey that histotype classification is consequential for patient outcome. All
models exhibited a persistent and robust integrin αv, α5, and β1 expression which poses a
promising inference of RGD receptor behavior. Despite vast heterogeneity between histotypes,
RGD receptors remain a constancy. Furthermore, analysis of cell cycle and anti-apoptotic proteins
suggest NEO212 perpetuates a diminishing native chemoprotective effect. Hence, NEO212
supplementation as well as integrin-targeted therapies prove as encouraging new developments for
more efficacious treatment for ovarian cancer.

1

Chapter 1: Introduction
1.1 Ovarian Cancer
1.1.1 Definition and Classification
Ovarian cancer is an anomaly among the pernicious field of malignancies due to its vast
heterogeneity (1). Since not all cases emanate from the ovary nor do all have the same tissue of
origin, the overarching title of “ovarian cancer” is quite misleading. There are some reported cases
distinguishing ovarian cancer as small-cell carcinoma or carcinosarcoma; however, the vast
majority spans a class of adenocarcinoma, otherwise known as a malignancy of a glandular
epithelium (2). Modern research in this field resolves that ovarian cancer is in fact a collection of
several distinct histotypes ranging in fundamental parameters including location of origin, clinical
manifestations, and molecular makeup (3). The subset of epithelial ovarian cancers (EOCs) –
which includes serous, endometrioid, clear-cell, and mucinous – account for over 90% of all cases
worldwide; thus, they are the main concern for ovarian cancer researchers in the arduous task of
advancing medical practices for treatment of this disease.
The serous histotype, which is further subdivided into high-grade or low-grade, originates
from the fallopian tubes (4) while all others hail from the peritoneum. Despite the common tubal
origin, only low-grade serous arises from a benign serous cystadenoma which progresses to a low-
malignant, borderline serous tumor and finally culminating into a low-grade carcinoma (5, 6).
Clear cell and endometrioid histotypes arise from endometrial tissue, while mucinous is from
transitional cell nests at the tubal-mesothelial junction or the gastrointestinal tract (7). Thus, in all
of these instances, the ovary is a secondary location for tumor metastasis. Low-grade serous,
mucinous, endometrioid, and clear cell all progress in a step-wise manner, yielding a better
understanding of disease progression; hence, they are classified as type I ovarian cancers. These
2

are also associated with specific mutations including aberrations in BRCA, KRAS, ERBB2, and
TP53 in serous; PI3KCA and PTEN in clear cell and endometrioid; and KRAS in mucinous (3).
Since its progression has yet to be fully elucidated and has a propensity for genetic
instability, high-grade serous remains an outlier among histotypes. Therefore, it is distinguished
as type II and remains more aggressive than those in type I (8).
Histotypes vastly differ in overall risk of development with high-grade serous occurring
most prominently at 70% of all EOC cases while endometrioid and clear cell each manifest at 10%,
mucinous at 3%, and low-grade serous at less than 5%.
1.1.2 Epidemiology and Etiology
With over 239,000 new cases and over 152,000 cancer-specific deaths every year, ovarian
cancer is the leading gynecologic cancer for women worldwide. Regrettably, women have a 1 in
75 risk of developing this nefarious disease during their lifetime and a 1 in 100 chance of related
death. Overall survival for all cases is roughly 45.6%; yet, approximately three-quarters have
already reached advanced stages at the time of diagnosis and face a notably grim 25% survival rate
(9). An unfortunate proponent to these devastating statistics is the absence of effective screening
tests and the stagnation in medical advancement, which has persisted for over 40 years.
Ovarian cancer is more common in predominantly Caucasian regions with North America,
Central Europe, and Eastern Europe experiencing the highest rate of incidence at over 8 in every
100,000. Conversely, Africa and Asia maintain the lowest occurrences at less than 3 in 100,000.
Even in more heterogenous societies, women of White descent are more at-risk than those with
Asian, Hispanic, or Black genealogies. Thus, the disparity between ethnicities corroborates the
geographical variation (9).
3

The etiology of ovarian cancer has yet to be fully expounded (10); however, its prevalence is
affected by several factors. Women positive for certain genetic markers, such as BRCA1 or BRCA2
mutations, have an increased propensity for development of ovarian cancer; meanwhile, parity,
breastfeeding, and gynecological surgeries – including, but not limited to, tubal ligation,
salpingectomy, and oophorectomy – can all decrease their likelihood. Research has also implicated
exogenous hormones in the development of this gynecological disease; however, the results are
variable. Oral contraceptives diminish a woman’s risk while hormone replacement therapy (HRT)
can cause an increase. Not unlike other diseases of cellular misfunction, ovarian cancer is also
detrimentally influenced by an unhealthy lifestyle. Substance abuse, obesity, smoking, and
persistent depression can all propel a woman into the vicissitudes of poor physical health and
probable cancer development (11).
1.1.3 Treatment
Despite distinctive molecular signatures among histotypes, all women receive the same
standard of care for primary tumors which include surgical debulking followed by intravenous
administration of a conglomerate of drugs as adjuvant chemotherapy (12). Carboplatin, which has
a platinum-conjugated structure, works by diffusing into the nuclear membrane and crosslinking
with DNA to prevent proper genomic replication and thus stimulating the apoptotic pathway.
Carboplatin typically causes both intrastrand and interstrand cross-linkages with preference to
neighboring purine bases (13). Meanwhile, paclitaxel functions by stabilizing microtubule
dynamics and preventing normal cytoskeletal trafficking and spindle fiber formation necessary for
division (14). If the cancer has disseminated drastically or if the tumor burden is deemed too high,
then the patient will undergo neoadjuvant chemotherapy whereby drug treatments are administered
before and after the cytoreductive surgery.
4

Even with prompt and belligerent treatment, there are some women who will not respond to
chemotherapy; yet, for the majority of them who found initial success, progression-free survival
is short-lived. Over 80% of these women will experience an aggressive recurrence for which there
is no standard of care (3). Women experiencing a relapse are currently treated with poly ADP-
ribose polymerase (PARP) inhibitors – for individuals with BRCA mutations – or angiogenesis
inhibitors with limited success. These forms of immunotherapy directly inhibit DNA damage
repair or vessel growth downstream of vascular endothelial growth factor (VEGF), respectively.
Recurrent cases are often divided into platinum-sensitive and platinum-resistant groups whereby
the former experience a 50% response to treatment and the latter experience a mere 10-15%
response rate with a limited progression-free survival of 3 to 4 months (15). These statistics have
not altered significantly in the past several decades. Hence, there is a desperate need for more
sensitive screening tests and additional effective treatment options.

1.1.4 Understanding Recurrence
Unless the root causes of the high recurrence rate are considered, current endeavors to
improve patient outcome is futile. Diagnostic tests – which analyze levels of peripheral cancer
antigen 125 (CA 125) – are minimally sensitive, conveying a severe lack of detectability (16). In
addition, these heterogenous masses include inherently drug resistant cells (17), which poses an
added detriment to cancer therapy. Research has shown that these cells, questionably pluripotent
in nature (18), reside in the mass even prior to chemotherapy exposure; so, this form of treatment
merely provides an evolutionary sieve with which the sensitive progeny will dissipate while those
implementing mechanisms for survival and drug evasion will endure.
5

Efforts for divulging these mechanisms involved are underway; however, the true nature of
their resistance – whether innate or acquired – is still elusive (19). Many findings suggest cancer
cells can acquire a transient resistance by entering dormancy (20, 21). If cells can evade the effects
of chemotherapeutics by limiting proliferation, then these drugs not only isolate innately resistant
progeny, but may also stimulate quiescence in cancer as a response to environmental stress.
This model mimics clinical outcome whereby patients enter remission despite retaining
dormant cancer cells. Quiescence and low metabolic activity allows for negative screening and
restoration of normal bodily functioning. Unfortunately, as other organs rehabilitate, so will the
microscopic tumor niche, which can once more become hospitable for accelerated growth and
erupt into a truculent recurrence.

1.2 Integrins
1.2.1 Structure and Function
Integrins (ITG) make up a class of heterodimeric transmembrane proteins made from alpha
and beta noncovalently-linked glycoproteins (22). Eighteen alpha subunits and eight beta subunits,
translated from the human genome, permutate to make 24 distinct integrins (23). Both the alpha
and beta subunits have their own sizeable extracellular domain (ECD), transmembrane domain,
and short intracellular domain (ICD). The ECDs of each subunit are bent when idle, but upon
ligation and subsequent activation, they become upright which enables the formation of the ligand-
binding domain (25). Similarly, the ICDs change conformation in response to activation and allow
for interaction with intracellular partners (22). Balancing these active and inactive conformations
allows for mediation of extracellular matrix (ECM) and cytoskeletal components and the
coordination of necessary cell functions like cell migration, proliferation, and polarity. In addition,
6

anchorage-dependent growth, which characterizes noncancerous cells, is heavily mediated by
integrins (25). Cells will not replicate unless their integrins are in contact with ECM components,
which suggests integrin influence on the cell cycle.
Though some integrins are involved in structural integrity via cell-cell junctions, the
majority participate in cell-matrix interactions (26). As chief detectors of matrix tension and
composition, they play critical roles in a cell’s communication with its environment. In the case of
hemidesmosomes, integrins mediate the interaction between the basal lamina and intermediate
filaments, which allows for structural stability (27); however, in most other scenarios, integrins
associate with a plethora of partners and stimulate pathways for cell motility or transformation
(28). The ECDs canonically recognize various ECM components – including collagen I,
fibronectin, laminin, and vitronectin – and transmit messages across the plasma membrane. Hence,
they are often deemed receptors to these matrix proteins. Complementing the variety of ECD
binding partners, the ICD can associate with an array of cytosolic proteins including actin, adaptor
proteins including talin and vinculin, and a multitude of kinases (28). The proteins involved, as
well as their order of association, dictate the cellular changes that will ensue downstream of
integrin signaling (29). One of the key features of integrins is their ability to participate in both
outside-in and inside-out communication, meaning both the cell’s microenvironment and its
internal dynamics have the capabilities to modify the other’s behavior (30).
1.2.2 Uses in Cancer
Integrins are subcategorized based on ligand recognition (24); however, the most significant
in cancer are the RGD receptors, which are a subset of eight integrins containing a receptor domain
for a tripeptide motif made of arginine, glycine, and aspartate (31). This Arg-Gly-Asp (RGD) motif
resides on matrix components like fibronectin, thrombospondin, vitronectin, laminin, and the von
7

Willebrand, which are all expressed on different tissue types to varying degrees. Hence, these
integrins – α5β1, αvβ1, αvβ3, αvβ5, αvβ6, αvβ8, and α8β1 – are used for recognition of and
adaptation to these different environments, a tactic which is particularly advantageous in metastatic
cancers. However, the eighth RGD-recognizing integrin – αIIbβ3 – is only expressed on platelets
for regulation of blood coagulation (32), and is therefore commonly disregarded in the context of
cancer. Current literature has proven that these RGD-recognizing integrins partake in critical
functions during wound healing, tissue remodeling, epithelial-mesenchymal transition, and
metastasis (33, 34). Naturally then, these RGD receptors are overexpressed by many forms of
cancer, particularly aggressive conditions like glioblastoma multiform (GBM), triple negative
breast cancer, and ovarian cancer (23).
Changes in extracellular matrix targets can stimulate pro-survival pathways. Mechanical
signals generated through tension in matrix fibers can stimulate outside-in signaling using integrins
(35). When ligated to these ECM components, the active heterodimer binds directly to focal
adhesion kinase (FAK) intracellularly, which then recruits another kinase named Src, and
stimulates a phosphorylation cascade leading to anchorage-dependent invasion and proliferation
(Figure 1.1, left panel) (36). This same pathway also perpetuates migration for both cancerous and
noncancerous cells whereby integrin dynamics dictate undulating cells along a surface. The
leading edges maintain active integrins for binding to ECM – or glycoproteins on endothelial cells
in the case of intravascularized cells – while the trailing end necessitates inactive integrins to
prevent extraneous binding. Thus, coordination from active to inactive conformations, and vice
versa, is vital for coordinated movement (22).
Normal integrin function is often exploited in cancer, however, it juxtaposes that of normal
tissue particularly in the case of unligated receptors. Though undiscovered in native tissue, active
8

integrin heterodimers can exist in cancer cells without an environmental stimulus. These unligated
active integrin species can act in a Src-dependent manner to induce signal transductions pathways
promoting anchorage-independent survival and anoikis evasion (37, 38), otherwise known as
apoptosis stimulated by a lack of tissue interaction. (Figure 1.1, middle panel) Since these
pathways help circulating spheroids stimulate pro-survival pathways while suppressing
anchorage-independent apoptosis, respectively, they are both critical for metastatic survival.
Furthermore, the β1 conformer is an especially detrimental integrin species since it can act in
Kras-dependent manner and stimulate pathways promoting stemness, angiogenesis, drug
resistance, and metastasis (Figure 1.1, right panel) (39, 40). Therefore, the beta species has been
well correlated with failed chemotherapy and poor prognosis (41, 23).
Figure 1.1: Active integrin conformers contribute to various outside-in signaling pathways that
promote cancer cell survival and tumor progression. The active alpha species is represented in red
while the active beta species is represented in blue.

9

1.3 NEO212
NEO212 is a novel experimental chemotherapeutic formed from the conjugation of
temozolomide (TMZ) and perillyl alcohol (POH), a cyclic monoterpene (42). POH is a naturally-
occurring constituent of plants canonically used in herbal apothecary and essential oils, including
– but not limited to – mint, lavender, citrus, lemongrass, cranberries, and sage (43). POH has
previously failed in clinical trials for use in ovarian cancer patients (44); however, due to the
presence of TMZ, NEO212 has a greater potential as an anti-cancer agent.
Current research proposes NEO212 utilizes the same mechanism of action as TMZ
whereby it intracellularly degrades into 5-(3-methyltriazen-1-yl) imidazole-4-carboxamide
(MTIC), then 5-aminoadazole-4-carboxamide (AIC) and a methyldiazonium cation. The latter
functions as the active component inflicting DNA damage through N7- and O6-methylation of
purines (45). Without a key DNA repair protein, O-6-methylguanine methyltransferase (MGMT),
the cells will experience several irreparable double stranded DNA breaks followed by intrinsic
apoptosis. Recent studies show NEO212 has superior sensitivity to TMZ, especially in cells
utilizing MGMT machinery, including GBM and primary cutaneous T-cell lymphoma (46).
1.4 Hypothesis
Integrins maintain paramount functions in tumor survival and growth; therefore, they have
a high potential for persistent expression throughout disease progression: in a chemonaïve tumor
prior to diagnosis, concurrent with chemotherapy treatment, and at time of recurrence. If all these
instances of ovarian cancer express this heterodimeric protein, then they can all be subjected to
integrin-targeting agents and engender a more favorable patient outcome. However, these women
still face another complication of limited response to standard chemotherapeutics. Since NEO212
10

has shown promising results in previous cancer studies, it could potentially be used as a
supplemental therapy prior to standard of care in order to increase the tumor’s susceptibility to
chemotherapy.
Therefore, it is hypothesized that RGD-recognizing integrins are expressed at stable levels
throughout the course of ovarian cancer; meanwhile, NEO212 administration is hypothesized to
diminish the native chemoprotective effect of cancer cells. Both postulations, if found true, could
reform current treatment and prognosis for ovarian cancer.

11

Chapter 2: Materials and Methods
2.1 Cell Lines and Maintenance
Four different ovarian cancer cell lines ranging in histotypes were provided by the
American Type Culture Collection (ATCC) (Manassas, VA). The wild types of A2780, ES2, and
HEYA8 were extracted from women during debulking surgery – prior to any exposure to
chemotherapy – and cultured in vitro to generate immortalized cell lines. However, A2780 WT
was also further manipulated to develop a progeny line that is resistant to carboplatin; this is
deemed A2780CP. All cells were propagated in Dulbecco’s Modification of Eagle’s Medium
(DMEM) with 4.5 g/L glucose, L-glutamine, and sodium pyruvate, and supplemented with 10%
fetal bovine serum (FBS), 100 ug/mL penicillin, and 100 ug/mL streptomycin. DMEM, FBS, and
antibiotics were provided by the Cell Culture Core at Norris Comprehensive Cancer Center (Los
Angeles, CA). All lines were maintained in a humidified incubator at 37°C with 5% CO2 atm.
2.2 Pharmacological Agents
Carboplatin (Cp) was obtained from Cayman Chemical Company (Ann Arbor, MI) or
Tokyo Chemical Industry (Tokyo, Japan) and dissolved in phosphate buffered saline (PBS),
bought from the Cell Culture Core at Norris Cancer Center, to a concentration of 10 mM or 13.5
mM, respectively. Stock solutions were stored at -80°C while dry carboplatin was stored at -20°C,
per company recommendations. Paclitaxel (Ptx), obtained from Alfa Aesar (Haverhill, MA), was
dissolved in PBS to a concentration of 5 mM. Stock solutions were also stored at -80°C.

12

2.3 Development of Clinically-Relevant Models
All three ovarian cancer cell lines – A2780, ES2, HEYA8 – were kept in culture while a
collection of each was seeded into four T75 flasks for combination chemotherapy treatment and
optimal retention of cell number. Once each flask became confluent, they were treated with 1.35
µM of carboplatin and 50 nM of paclitaxel for four consecutive weeks. During this time, the
combination chemotherapy was not removed except during the time of medium changes in which
both drug concentrations were restored along with fresh supplemented DMEM. At the end of the
four weeks, these populations were harvested and seeded into two groups. The first continued to
receive treatment in order to maintain the characteristic phenotype of the chemotherapy-tolerant
population, while the other flask was allowed to recover in chemotherapy-free medium for four
consecutive weeks.
Thus, three models are established with different relationships to chemotherapy as depicted
in Figure 2.1. The first, which remained chemonaïve throughout model progression best mimics
the disease at time of diagnosis (red panel in Figure 2.1). The population of cells which received
persistent chemotherapy of both carboplatin and paclitaxel throughout the duration is deemed
microscopic residual disease (green panel in Figure 2.1), since they are the subset which survived
throughout therapy. Finally, the population which received chemotherapy and then was allowed
to recover in growth medium (blue panel in Figure 2.1) models the disease at time of recurrence
since these cells are chemo-experienced, yet it is to be determined if they are chemorefractory or
if they have regained sensitivity to the chemotherapeutics used in the clinic.
13

Figure 2.1: Workflow for selection of chemotherapy-exposed subpopulations. Time progression
diagram illustrating the development of the three different models used: disease at diagnosis (WT),
microscopic residual disease (T), and recurrent disease (R).
2.4 MTT Assay
For the cytotoxicity analysis, 96-well plates were seeded with 2.0 – 5.0×10
3
cells/well in a
50 µL volume and allowed to adhere overnight. Cells were seeded at these low volumes to prevent
total exhaustion of medium nutrients over the course of the experiment. The following day, each
well was treated with 50 µL of various concentrations of chemotherapeutics dissolved in
supplemented DMEM or 50 µL of chemotherapy-free medium in the case of the untreated wells
used as controls. Drug treatment lasted 72 hours at which point, all solution was removed from the
wells and replaced with 100 µL of 10% thiazolyl blue tetrazolium bromide
(Methylthiazoletetrazolium, MTT) in DMEM. Plates were then incubated for 3.5 hours for
reduction of MTT by mitochondrial NADH-dependent oxidoreductases. Reaction termination then
14

occurred via cell lysis using 100 µL of solubilization solution (.01 M hydrochloric acid in 10%
sodium dodecyl sulfate, SDS). In order for even distribution of MTT crystals, plates were
visualized the following day using Varioskan Lux Reader by Thermo Fisher Scientific (Waltham,
MA). Optical densities (OD) were noted using absorbance at 560 nm. Background values,
provided by the cell-free control wells containing 10% MTT and solubilization solution, were
averaged and subsequently subtracted from all measured values. Finally, percent cell viability (%)
was calculated by proportioning each condition OD with respect to the control OD. Each condition
was set up in duplicate or triplicate while each individual experiment was repeated at three
different time points for statistical analysis.
2.5 Fluorescence-activated Cell Sorting (FACS) Analysis
Cytometric analysis was performed using Aria flow cytometers from Becton Dickinson
(BD) Biosciences Ltd. (Franklin Lakes, NJ). in the Flow Cytometry Core of USC. Cells were
harvested in 1.5 mL Eppendorf tubes and incubated with 2 –5 µL of fluorophore-conjugated
antibodies in supplemented DMEM. Each Eppendorf tube was allowed to incubate on a rotator
inside a humidified incubator at 37°C for 45 to 60 minutes. After incubation, cells were pelleted
in order to aspirate out the medium and resuspend the cells in 1 mL PBS. Negative controls were
established using a sample incubated sans fluorophores. Antibodies synthesized by Novus
Biologicals (Littleton, CO) are as follows: PE/Cy-7 anti-ABCA1 (HJ1, #2068), PE anti-MRP1
(IU2H10, #156). Antibodies from BD are as follows: APC anti-CD338 (5D3, #561451), BV421
anti-CD243 (UIC2, #566015).

15

2.6 Immunoblotting
Cells were grown on 10-cm dishes and harvested using 1 mL ice cold PBS. All contents
were then centrifuged at 10,000 rotations per minute (RPM) for 30 seconds. Supernatant was
removed so cell pellets could be lysed using radioimmunoprecipitation assay (RIPA) buffer from
Thermo Fisher Scientific. For determination of protein concentration, Pierce BCA protein assay
reagent from Thermo Fisher Scientific was utilized. Fifty µg of protein of each sample added to
wells in 10-15% SDS polyacrylamide gels. Polyvinylidene fluoride (PVDF) membranes from
BioRad Laboratories (Hercules, CA) were utilized for semi-dry transfer. 5% milk in 1x tris-
buffered saline plus 0.1% Tween 20 (TBST) was used as a blocking buffer.
Polyclonal primary antibodies from Cell Signaling Technology (Danvers, MA) was used
for detection of integrin B1 (#4706S), c-myc (#13987), Mcl-1 (#5453), cyclin D (#55506), MGMT
(#2739), and survivin (#2808). Santa Cruz Biotechnology (Dallas, TX) provided the primary
antibody for detection of actin (sc-47778). Horseradish peroxidase (HRP)-conjugated anti-IgG
antibodies, used for detection of the primary antibodies, were obtained from Jackson
ImmunoResearch Laboratories Inc. (West Grove, PA). Dilutions and storage were in accordance
with the suppliers’ recommendations. Detection of protein levels was carried out using ProSignal
Femto Enhanced Chemiluminescence Reagent by Genesee Scientific (San Diego, CA). Finally,
blots were imaged using ImageQuant LAS-4000 Chemiluminescence and Fluorescence Imaging
System from Fujitsu Life Sciences (San Jose, CA).
2.7 Statistical Analysis
Unpaired t-testing with two tails was used for determination of all statistical significance.
All values were compared to standard p-values.

16

Chapter Three: Results
3.1 Cytotoxicity effects of pharmacological agents used in clinical treatment
The purpose of this study is to determine inhibitory concentrations of chemotherapeutics
currently used in the clinic for the treatment of ovarian cancer. In order to characterize differences
between chemonaïve and chemoresistant cell lines, it is imperative to first quantify the cytotoxic
effects of the drugs of interest in vitro. The half maximal inhibitory concentration (IC50) is a
commonly used quantitation of the effectiveness of a substance in inhibiting a biological function.
In this experiment, inhibition by these drugs will cause cell death and thus each condition can be
compared to an untreated condition, yielding a percentage of viable cells by the end of treatment.
Overall cytotoxicity was determined using an MTT assay with a 72-hour treatment period. Each
condition was carried out in duplicate, while each experiment was conducted at a minimum of
three separate times for statistical analysis.
3.1.1 Cytotoxicity of Carboplatin

Figure 3.1: Effect of carboplatin on cell viability in ovarian cancer cell lines used. Cell viability
was determined relative to the untreated samples. *p-value ≤ 0.05; ns = not significant.
17

Carboplatin differs tremendously between cell lines as shown by the variation in
concentration range. A2780 showed the most sensitivity to Cp where almost all cells were no
longer viable at 150 µM Cp. The same cannot be said for ES2 or HEYA8, however, since they did
not experience full cell death until much higher concentrations of 400 µM and 450 µM,
respectively. When whole sets of each density were compared, there was no significant difference
which conveys that trends in cytotoxic effect is not affected by total tumor volume in vitro. It is
interesting to note that the largest divergences between cytotoxicity curves occurs around the IC50,
thus when comparing these values individually, there is significant difference between cell
densities. IC50 values shown in Figure 3.2 were interpolated from each individual cytotoxicity
curve that was averaged to yield the quantitation above in Figure 3.1.
The 5k and 2k IC50s of A2780 are 69.2 µM and 17.4 µM, respectively. The ES2 cell
densities reached 50% cell viability at 90.6 µM and 60.5 µM, respectively. For HEYA8 wild-types,
Cp concentrations need to reach 280.7 µM and 117 µM to diminish a 5k cell mass and a 2k cell
Figure 3.2: Comparison of half maximal inhibitory concentrations of cell lines exposed to
carboplatin. **p-value ≤ 0.001; ***p-value ≤ 0.0001; **** p-value ≤ 0.00001.
18

mass, respectively, by half. An unpaired t-test showed these disparities are highly significant.
Though there is no significant difference in overall cytotoxicity of a chemonaïve cell line despite
different seeding densities, drug concentration necessary for ablation of half of the entire mass in
vitro is contingent upon the total tumor volume, which confirms that a larger tumor volume would
limit accessibility of the anti-cancer drugs. These results clearly convey the vast heterogeneity
between cell lines, which suggests differential patient outcome with respect to histotypes.
3.1.2 Cytotoxicity of Paclitaxel
All cell lines used were very susceptible to paclitaxel with cell death occurring even before
1 nM Ptx. Overall range of concentrations used was much less severe than that of carboplatin,
suggesting similar cytotoxic effects of Ptx on all cell lines used. Though A2780 again showed the
most sensitivity to Ptx, they all maintained a similar trend of immediate loss of cell viability
Figure 3.3: Effect of paclitaxel on cell viability in ovarian cancer cell lines used. Cell viability
was determined relative to the untreated samples. *p-value ≤ 0.05; ns = not significant.

19

followed by a plateau in cytotoxicity without ever reaching total cell death. Differences in overall
cytotoxicity between the propagation of 5k cells and 2k cells were not significant, except in the
case of ES2, which demonstrated a dose-dependent response.
Yet, by comparing individual values at the IC50, there is a clear difference between cell
densities as shown in Figure 3.4.
Again, interpolated IC50 values prove that tumor volume influences the concentration
necessary for diminished tumor size by half in the cases that exhibit a dose-dependent response.
For ES2, these highly significant values are 19.5 nM and 7.4 nM. A2780 IC50 values for 5k and
2k seeding are .62 nM and .46 nM Ptx, respectively; while HEYA8 showed half cell death at .24
nM and .08 nM Ptx, respectively. These discrepancies in A2780 and HEYA8 are not significant.
This finding could be substantiated by the immediate depletion of cell viability at the lowest
concentration.

Figure 3.4: Comparison of half maximal inhibitory concentrations of cell lines exposed to
paclitaxel. **p-value ≤ 0.001; ns = not significant.
20

3.2 Delineation of differences in clinically-relevant models
The purpose of this study is to better characterize the nature of the three models based on
differences in cytotoxicity, established markers of drug resistance, and markers of cell cycle and
survival. These parameters help juxtapose the models – and thus the stages of ovarian cancer
development – in order to determine their validity and speculate on potential areas of targeting for
more efficacious treatment.
3.2.1 Cytotoxicity differences
Cells were treated with either carboplatin or paclitaxel for 72 hours and analyzed using
MTT assay. Seeding of 5k cells are only shown for simplicity. Each condition was carried out in
duplicate or triplicate.
Figure 3.5: Comparison of cytotoxicity curves of wild-type and recovered models. Cell viability
was determined relative to the untreated samples. ns was determined for all experiments. *p-
value ≤ 0.05; ns = not significant.

21

3.2.2 ABC transporter profile
Efflux of xenochemicals by ATP-binding cassette (ABC) transporters is well correlated
with drug resistance in ovarian cancer. There are several mechanisms by which cancer cells can
abrogate chemotherapy; however, current research suggests ovarian cancer commonly employs
ABC transporters. Whether this is the only mechanism or one of many with which this cancer
survives is not well elucidated. Despite this uncertainty, ABC transporter levels can suggest
differences in phenotypes with respect to drug sensitivity. This study simply uses ABC transporter
levels as a marker for drug tolerance and not an irrefutable determination of drug resistance.
In this experiment, all 9 models – three from each cell line – were incubated with four
fluorophore-conjugated antibodies specific for ABCA1, ABCB1, ABCC1, or ABCG2. Using flow
cytometry, populations were analyzed for two ABC transporters at a time. ABCA1 and ABCC1
levels were analyzed first, then the double positive population (ABCA1
+
/ABCC1
+
) was further
selected for analysis of the remaining two ABC transporters. Therefore, the population density of
quadruple positive cells (ABCA1
+
/ABCC1
+
/ABCB1
+
/ABCG2
+
) was determined. This
quantification was used as a parameter for comparison of the different models since a larger
presence of quadruple positive subpopulation suggests a higher drug tolerance

22

3.2.3 Differential expression of cell proliferation and cell survival markers
After development of clinically-relevant models, all cells were grown in drug-free medium
until confluency for harvesting and whole cell lysate analysis by immunoblotting. Fifty µg of
protein from each sample was loaded into their corresponding wells. WT and R models,
representing disease at time of diagnosis and recurrence, respectively, were compared for
differences in levels of cell proliferation and cell survival markers. Since recurrent cases are
generally more aggressive than their corresponding primary conditions, the R models are expected
to have higher levels of both cell cycle and cell survival markers. A2780CP, a carboplatin-resistant
Figure 3.6: ABC transporter levels of the three clinically-relevant models used. Cells in gate
P3, representing the double positive ABCA1
+
/ABCC1
+
, were reanalyzed in the right panel for
other ABC transporter levels. Thus, the P5 gate represents the quadruple positive
ABCA1
+
/ABCC1
+
/ABCG2
+
/ABCB1
+
.
23

derivative of A2780, was also used as a control for a truly drug resistant protein profile. This
analysis was repeated for validation.
The cell proliferation markers, c-myc and cyclin D, are expressed in all populations;
however, there is a clear difference only between the WT and R models of the A2780 and ES2 cell
lines. C-myc is upregulated slightly in the chemoexposed models of A2780 while it is
downregulated in the chemoexposed condition of ES2. This same behavior is also seen in cyclin
D, a downstream target of c-myc whereby the A2780CP and A2780-R maintain higher levels than
the A2780-WT; yet, the ES2-WT has a more robust cyclin D signal than ES2-R. In analyzing the
HEYA8-WT and HEYA8-R cell cycle protein profiles, there is a slight downregulation of cyclin
Figure 3.7: Expression levels of cell proliferation and cell survival markers in all wild-type and
recovered cell populations. Actin was used as a loading control. WT = wild-type; CP=
carboplatin resistant cell line; R = recovered.

24

D in the chemoexposed model; however, this difference is quite discreet and could be considered
nominal since this method does not allow for quantification.
Mcl-1 levels only change between the A780 models where A2780-WT has a very low
expression compared to both A2780CP and A2780-R. Their corresponding survivin profiles mirror
this behavior as well, which concludes both of the chemoexposed populations produce more anti-
apoptotic machinery. Yet, this development cannot be generalized to all ovarian cancer cell lines
since ES2 and HEYA8 are not in accordance with this finding. Instead, these two cell lines sustain
Mcl-1 levels between the WT and R models with minimal discrepancies. In contrast, survivin
expression is altered to a greater extent: after chemotherapy exposure, it is bolstered in ES2 and
wanes in HEYA8.
Overall, all cell lines have responded differently to the amalgam of carboplatin and
paclitaxel with pronounced differences seen mainly in the A2780 cell line.
3.3 Integrin profile after chemotherapeutic pressure
The purpose of this study is to better characterize the relationship between RGD-recognizing
integrins and chemotherapy. Individual integrin species αv, α5, and β1 were analyzed in this study
which could yield insight to the nature of RGD receptors under chemotherapeutic pressure;
however, generalizations about all these integrins cannot be concluded.
3.3.1 Western blot analysis of Integrin β1 expression
All cells were grown in growth medium until confluency for whole cell lysate analysis of
integrin β1 expression by immunoblots. This experiment was repeated for validation.
25

Similar to the findings of cell proliferation and cell survival markers above in section 3.2.3,
the most striking difference between WT and R models is seen in the A2780 cell line. There is no
salient difference between models of ES2 or HEYA8 cell lines despite the addition of
chemotherapy in the Recovered model. These results deduce the functional value of the β1 species;
however, this is not fully indicative of RGD-recognizing integrins since these groups are not
mutually inclusive. Since there is an upregulation in β1 species – which is also characteristic of
integrin receptors for laminin, collagen, and leukocytes, it can only be hypothesized that A2780-
R upregulates its RGD receptors. Further studies need to ensue before validating this syllogism.
3.3.2 Levels of active integrin αv and α5 species
Populations of each model from the HEYA8 cell line were harvested at the same time with
two fluorophore-conjugated antibodies, each specific for the active αv or α5 integrin species, then
analyzed using flow cytometry. A negative control was established by incubating a sample without
fluorophores. Thus, this iso control, which is negative for both fluorophores, allowed for gating of
the P3 double positive αv
+
/α5
+
subpopulation.
Figure 3.8: Expression levels of integrin β1 in all wild-type and recovered models. Actin was
used as a loading control. WT = wild-type; CP= carboplatin resistant cell line; R = recovered.

26

These results between the HEYA8-WT and HEYA8-T show a prominent upregulation
from 7.9% to 49.2% in population double positive for the αv and α5 active species under the
pressure of chemotherapy. More interestingly, however, is the maintenance of this αv
+
/α5
+
double
positive population even after a month of recovery in chemotherapy-free growth medium. The
HEYA8-R subpopulation has a decreased double positive subpopulation at 39.6%, but this
difference is minimal when compared to the miniscule subpopulation of the HEYA8-WT.
This FACS data suggests a robust upregulation of integrin αv and α5 species in HEYA8;
yet, this is not fully corroborated in section 3.3.1 with respect to integrin β1 expression, which
shows no drastic change after chemotherapy exposure. This could be explained by differing
behaviors of the alpha and beta species; however, it suggests the necessity for quantification of
integrin β1expression for a more accurate comparison.
Though these results display at most half of the cancer population maintaining αv
+
/α5
+
populations, this is not inclusive of cells which are positive for one or the other. Therefore, the
proportion of cells which express RGD receptors may be much higher than this projection.
Figure 3.9: Levels of active αv and α5 integrin species in all clinically-relevant models of
HEYA8. Gating for the double positive population of αv
+
/α5
+
was based on the signal from a
negative control, which was incubated without antibodies.

27

3.4 Determining chemotherapeutic sensitization using NEO212
The purpose of this study is to determine if NEO212 can sensitize ovarian cancer cells to
current chemotherapeutics used in the clinic. In this regard, NEO212 will not supplant this mode
of therapy, but enhance the effects of it. Previous studies have shown that some cells which
regularly express c-myc are more responsive to a new kind of chemotherapeutic: NEO212.
Therefore, the two cell lines which maintained higher levels as shown in Figure 3.7 were employed
for this study. In order to determine the therapeutic potential, drug treatment was followed by
western blot analysis of cell survival markers. Furthermore, expression of MGMT and integrin B1
were analyzed for true determination of NEO212 sensitivity and response to chemotherapy,
respectively.
Six plates were seeded with the same concentration of A2780 or ES2 cells and allowed to
adhere overnight. The following morning, A2780 plates received a single treatment of 60 µM
NEO212 while ES2 plates received 50 µM NEO212. Untreated (UT) plates were not disturbed
while A2780 and ES2 vehicle (vh.) conditions were given .06% or .05% dimethyl sulfoxide
(DMSO), respectively. Cells were then harvested on days 2, 3, 4, and 5. Medium was not changed
throughout treatment, except for the untreated conditions which were split on day 4 and harvested
the following day along with the vh. plates. Expression levels of actin were used as a loading
control for all samples.
28

3.4.1 Expression levels of cell survival markers in MGMT-positive ovarian cancer cells
NEO212 showed no stark change in expression of MGMT except on day 3 where there is
a slight upregulation. When visualizing the antiapoptotic markers, Mcl-1 showed an immediate
downregulation after exposure to NEO212. This diminished signal is maintained throughout
treatment with a slight variation at day 3; therefore, duration of treatment does not have a clear
effect on this downregulation. Meanwhile, survivin expression differs from that of Mcl-1 despite
their functional commonality. A2780 natively expresses a higher level of survivin and only
experiences a downregulation in this pro-survival protein at day 5 of treatment. The vh. -treated
condition also exhibited a diminished survivin signal, but this could be the result of confluency at
Figure 3.10: Changes in expression levels of cell survival markers and integrin β1 in A2780
from NEO212 treatment. Cells were treated with 60 µM NEO212 or .06% DMSO, as the
vehicle, and harvested on four consecutive days. Actin was used as the loading control. UT =
untreated; vh. = vehicle-treated.

29

time of harvesting. In addition, the medium was not deprived of essential nutrients, however, it
was beginning to be spent.
Integrin β1 expression was downregulated initially at day 2 of treatment, which suggests
initial cytotoxic effect since this integrin is commonly linked to cell viability. However, this effect
is only short-lived which implies a quick response to NEO212 after initial exposure. This finding
is possibly corroborated by the minor augmentation in expression of MGMT, survivin, and Mcl-1
at day 3 which suggest a coordinated and compensatory response to evade further
chemotherapeutic damage. Yet, this postulation is negligible without quantification.
3.4.2 Expression levels of cell survival markers in ovarian cancer cells minimally expressing
MGMT
Figure 3.11: Changes in expression levels of cell survival markers and integrin β1 in ES2 from
NEO212 treatment. Cells were treated with 50 µM NEO212 or .05% DMSO, as the vehicle,
and harvested on four consecutive days. Actin was used as the loading control. UT = untreated;
vh. = vehicle-treated.

30

In chemonaïve ES2 cells, MGMT is not strongly expressed; nevertheless, this signal is
expunged after NEO212 treatment. Mcl-1 depicts a similar trend as MGMT whereby the untreated
and vh.-treated conditions maintain very low levels with total loss of expression with the presence
of NEO212. Survivin expression is more robust for the negative controls; however, there is still a
clear incremental downregulation with increasing incubation time. Therefore, NEO212 shows an
immediate decrease of MGMT, Mcl-1, and survivin.
Expression levels of integrin β1 enervate after several days of treatment most prominently
on day 5. This downregulation validates the change in morphology whereby cells appeared
apoptotic from day 3 to day 5 of treatment. The lysates collected on the last days were from dying
cells; thus, it is clear why there is hardly any sustaining survivin or integrin β1.

31

Discussion
With respect to Cp treatment, a larger tumor volume most likely has a chemoprotective
effect since IC50 significantly differs between cell densities. A direct comparison of these values
clearly illustrates the vast difference in cytotoxicity between cell lines. Since all ovarian cancer
cases are treated in the same manner – with doses only contingent on body surface area – then it
can be deduced that there will be drastically different effects depending on histotype. Naturally,
some histotypes are more likely to persevere and increase the propensity for recurrence. Relative
to those used in this study, clinical concentrations administered intravenously – 175 mg/m
2
Ptx
over 3 hours and 300 mg/m
2
Cp every 3-4 weeks – are much higher to compensate for
pharmacodynamics and drug elimination. Meaning, before reaching the intraperitoneal site of
cancer growth, the drug may pass through the liver and kidneys which – along with natural
degradation – reduce the molecules available to reach the tumor. Regardless, these phenomena of
human physiology do not entirely negate the likely difference in outcome, observed in vitro, which
is dictated by histotype.
Intraperitoneal levels of paclitaxel are maintained longer than that in plasma (48),
suggesting the true concentration at the tumor site may be higher than projected; however, this
may not exhibit the most beneficial consequences. Liebmann et al. demonstrated an increase in
surviving cell fraction when administering Ptx concentrations above 250 nM. Therefore, the
efficacy of this taxane-derived drug is questionable as damage inflicted on the tumor may or may
not outweigh adverse side effects. Furthermore, these conclusions corroborate this study’s findings
on paclitaxel toxicity whereby there was an initial dose-dependent response followed by a plateau
in cytotoxicity without accomplishing total cell death. They observed optimal cell damage with a
72-hour incubation and lower nM concentrations – which have been demonstrated as clinically-
32

relevant doses (14). However, Ptx IC50 values were smaller in comparison which could be
accounted for by the differences in cell type or the lack of unidirectionality in drug response. These
cells exhibited an overall initial loss of cell viability; however, between 1 and 10 nM, a slight
increase in surviving fraction of cells was regularly observed.
Overall, variability in IC50 values between cell lines further conveys the heterogeneity
between histotypes and gives a possible explanation why many therapies fail to fully ablate the
cancer cells. It is important to note that these drugs were only tested individually and not in
conjunction with each other; thus, a more accurate quantitation would include the cooperative use
of these drugs. Regardless, the vast heterogeneity between histotypes should no longer be ignored.
Hence, if histological profile has no implications on treatment dosing, then surely some
chemotherapeutics will work better for some women while others will simply fail chemotherapy
all together or experience recurrence after initial success.
Creating a model for each stage of disease progression would yield substantial significance
in determining the efficacy of chemotherapeutics; however, many other parameters should be
included for their development. For the models used here, cytotoxic analysis via MTT assays
showed no significant discrepancies between the WT and R models. The latter does not necessarily
need to be chemoresistant; yet, it should maintain a lower sensitivity to the chemotherapeutics it
once evaded as observed by McGrail, who demonstrated that a truly paclitaxel-resistant cell line
exhibited an increase in order of magnitude of its IC50 (36). This expectation was not observed,
thus suggesting future modification of the model for better clinical imitation.
FACS analysis of ABC transporter profile is more promising with respect to characterizing
each model. Current research acknowledges ABC transporters as a mode of chemoresistance in
ovarian cancer (41); therefore, the larger the quadruple positive
33

ABCA1
+
/ABCC1
+
/ABCG2
+
/ABCB1
+
subpopulation, the higher the likelihood of drug evasion.
Higher quadruple positive subpopulation suggests the Treated model may maintain better
resistance to chemotherapy; however, this was not permeated into the R models, despite common
origin. This observation is somewhat substantiated in the clinic whereby microscopic residual
disease lies growth arrested and resistant until recurrence, which is – in some cases – accompanied
by drug sensitivity. Yet, this does not explain other recurrent cases which are characterized by
platinum resistance. Since recurrence is not fully understood in the clinic, no conclusions can be
made regarding the validity of this models.
The fleeting ABC transporter profile could also be explained by another argument where
the quadruple positive population became overwhelmed by the non-side population majority that
is ABCA1
-
/ABCC1
-
/ABCG2
-
/ABCB1
-
due to differences in proliferative capacities. However, He
et al proves this is not in fact the case (49). Instead, they show ABC transporter expression is a
consequence of phenotypic plasticity, which commonly characterizes cancer stem cells or tumor
infiltrating cells. Therefore, these findings support the increased aggressiveness of the Treated
model relative to the WT or Recovered. However, phenotypic plasticity is bilateral. The T models
appear more truculent by this analysis; yet, the WT or R models are just as likely to establish ABC
+

progeny as the T models are to produce ABC
-
progeny. Furthermore, ABC transporter expression
also varies based on passaging, which was not held constant between models. Therefore, ABC
transporter profile may yield trivial implications on the distinctions between models.
When analyzing the cell proliferation markers via western blot, c-myc seems transitory
since it is not well expressed in HEYA8 models despite the high level of one of its downstream
targets, cyclin D, which suggests previous activation. Conversely, c-myc could have be expressed
at a lower level and stay constitutively active and thus perpetuate a high downstream cyclin D
34

signal. However, the former is the more likely hypothesis since this relationship between c-myc
and cyclin D is only seen in HEYA8. Further analysis needs to be conducted before any
conclusions can be made with regard to the proliferative profile of each model. It is suggested that
cancer cells can enter a period of dormancy in order to evade the effects of cell cycle-targeting
chemotherapy; yet, this does not appear to define the WT or R models, which are all positive for
c-myc and cyclin D.
Mcl-1 expression is held fairly constant before and after chemotherapy treatment except in
the case of A2780 where there is a stark upregulation in both the A2780CP and A2780-R. This
finding implies a persistent pro-survival response after exposure to chemotherapy. Also in A2780
models, a strong Mcl-1 and MGMT expression suggest a tendency towards drug tolerance,
especially since the former is an essential oncogene in drug resistant ovarian cancer (50). Survivin
also exhibits the same upwards trend in chemoexposed lines for the A2780 and ES2 models;
however, the reverse is true for the HEYA8 model. This anomaly could be accounted for by
considering its IC50 relative to the other cell lines; HEYA8 can withstand much higher
concentrations of carboplatin, which suggests it subjectively experienced a less draconian
chemotherapy regimen. HEYA8 models are readily proliferating with little regard for anti-
apoptotic precautions, as evidence by low Mcl-1 and survivin expression. Meanwhile, the A2780
models subjected to chemotherapy demonstrate a stronger anti-apoptotic response while
proliferating. This juxtaposition merely highlights the drastic differences between histotypes,
which should be implicated in the standard of care. In comparing the anti-apoptotic markers,
survivin levels vacillated more so in all cell lines after drug exposure suggesting a more prominent
role in abrogating apoptosis in ovarian cancer cells.
35

Western blot analysis confirms that these cell lines all maintain levels of integrin β1,
however to varying degrees. ES2 and HEYA8 sustained this expression without notable change;
however, A2780 did show a stark increase in ITG β1 levels after chemotherapeutic exposure. The
A2780-R, which is not truly drug resistant, expressed a much higher level of ITG β1 than the
bonafide resistant population of A2780CP. Thus, the relationship between ITG β1 and
chemotolerance is not as clear cut as that seen with Mcl-1. This suggests ITG β1 expression could
elevate subsequent to chemotherapy exposure and independently of drug resistance. In this
experiment, A2780-R was recently subjected to drug treatment while A2780CP was only
propagated in growth medium, which could account for the differences in ITG β1 expression.
Furthermore, A2780-R was exposed to paclitaxel; but, A2780CP was only exposed to carboplatin.
Hence, this discrepancy could also justify the observed ITG levels; presence of paclitaxel,
independent of carboplatin, could potentiate pathways resulting in an upregulation of ITG β1.
FACS and western blot analysis demonstrate a maintenance of both of the active integrin
species αv and α5 as well as integrin β1 expression perpetuated by chemotherapy. Although cell
cycle and cell proliferation markers alter by stage of disease, integrin expression persists.
However, further probing should be done at a later time point in recovery. In addition, co-
immunoprecipitations could be conducted to pull down whole integrin heterodimers, which could
conclude more accurate results about RGD receptors than when evaluating individual alpha or beta
subunits. These results merely imply the behavior of RGD-recognizing integrins but may contain
outliers. Despite speculations, it is clear that integrins are linchpins throughout disease progression
and thus are viable targets for newer therapies aimed at targeted administration of chemotherapy.
This includes integrin antagonism, brachytherapy via radioactive isotope-conjugated ligands of
RGD-recognizing integrins, and RGD-integrin-targeting immunotherapy (31, 51).
36

Western blots and current literature confirm that MGMT is heterogeneously expressed in
ovarian cancer. Expression of this protein canonically implies resistance to TMZ; however, some
studies have shown an increased sensitivity to NEO212 despite MGMT presence. Though
promising, these results are not inclusive of ovarian cancer and thus, further determination by this
study was necessary to deem NEO212 an effective treatment supplement for this affliction.
Cells that are less responsive to carboplatin have shown a sensitivity to NEO212. For
example, the ES2 cell line, which has an IC50 of 90.6 µM Cp, experienced a loss of anti-apoptotic
proteins after only one treatment of 50 µM NEO212. Mcl-1 levels decreased during treatment
which suggests an interference with the cell’s ability to impede apoptosis; however, this is truly
only corroborated at later time points when survivin levels diminish as well. When comparing
current standard of care and NEO212 in terms of anti-apoptotic marker response, it is clear that
the former induces a chemoprotective effect while the later invokes a decrease in chemoprotection,
which is more optimal for better patient outcome. A reduction in both survivin and Mcl-1 after
NEO212 treatment – seen in MGMT-positive and minimally-expressing MGMT cells in vitro –
implies a universal heightened vulnerability to further treatment. With diminished machinery to
fight off apoptotic signals, cancer cells could be much more responsive to clinical standard
chemotherapies.
NEO212 addition could yield better efficacy at lower concentrations and even a better
outcome with fewer side effects. This hypothesis should be analyzed in vivo after administering
intraperitoneal ovarian cancer cells and treating with NEO212 before intravenous administration
of paclitaxel and carboplatin. Therefore, it is still questionable if NEO212 would actually be an
effective treatment for all cases; however, NEO212 shows promising results as a supplemental
medication for the use of sensitization to other chemotherapeutics.
37

References
(1) Mallen AR, Townsend MK, Tworoger SS. Risk Factors for Ovarian Carcinoma. Hematol
Oncol Clin North Am. 2018 Dec 1; 32(6):891-902.
(2) Cho KR, Shih I. Ovarian Cancer. Annu Rev Pathol. 2010 Jan 1; 4:287-313.
(3) Matulonis UA, Sood AK, Fallowfield L, et al. Ovarian Cancer. Nature Reviews Disease
Primers. 2016 Aug 25; 2:16061
(4) Lisio MA, Fu L, Goyeneche A, Gao Z, Telleria C. High-Grade Serous Ovarian Cancer:
Basic Sciences, Clinical and Therapeutic Standpoints. Int J Mol Sci. 2019 Feb; 20(4):952.
(5) Kurman RJ, Shih IM. Molecular pathogenesis and extraovarian origin of epithelial
ovarian cancer – Shifting the paradigm. Hum Pathol. 2011 March 23; 42:918-931.
(6) Feeley KM, Wells M. Precursor lesions of ovarian epithelial malignancy. Histopathol.
2001; 38:87-95.
(7) Vaughan S, Coward JI, Bast jr RC, et al. Rethinking ovarian cancer: recommendations
for improving outcomes. Nat Rev Cancer. 2011 Sep 23; 11(10): 719-25.
(8) Bowtell DB, Bohm S, Ahmed AA, et al. Rethinking ovarian cancer II: reducing mortality
from high-grade serous ovarian cancer. Nat Rev Cancer. 2015 Nov 15; 15(11): 668-79.
(9) Reid BM, Permuth JB, Sellers TA. Epidemiology of ovarian cancer: a review. Cancer
Biol Med. 2017 Feb; 14(1): 9-32.
(10) Hunn J, Rodriguez GC. Ovarian Cancer: Etiology, Risk Factors, and Epidemiology. Clin
Obstet Gynecol. 2012 Mar; 55(1): 3-23.
(11) Mayo Clinic. [Internet]. Ovarian Cancer; 2020; Available from:
https://www.mayoclinic.org/diseases-conditions/ovarian-cancer/symptoms-causes/syc-
20375941
38

(12) Hasegawa K, Shimada M, Takeuchi S, et al. phase 2 study of intraperitoneal carboplatin
plus intravenous dose-dense paclitaxel in front-line treatment of suboptimal residual
ovarian cancer. Br J Cancer. 2020 Jan 31.
(13) Dasari S, Tchounwou B. Cisplatin in cancer therapy: molecular mechanisms of action.
Eur J Pharmacol. 2014 Oct 5; 0:364-378.
(14) Zasadil LM, Andersen KA, Yeum D, et al. Cytotoxicity of Paclitaxel in Breast Cancer Is
due to Chromosome Missegregation on Multipolar Spindles. Sci Transl Med. 2014 Mar
26; 6(229):229ra43.
(15) Viscarra T, Buchegger K, Jofre I, et al. Functional and transcriptomic characterization of
carboplatin-resistant A2780 ovarian cancer cell line. Biol Res. 2019 Mar 21; 52(1): 13.
(16) Trinidad CV, Tetlow AL, Bantis LE, Godwin AK. Reducing Ovarian Cancer Mortality
Through Early Detection: Approaches Using Circulating Biomarkers. Cancer Prev Res
(Phila). 2020 Mar; 13(3): 241-52.
(17) Johnson SW, Ozols RF, Hamilton TC. Mechanisms of Drug Resistance in Ovarian
Cancer. Cancer. 1993 Jan 15; 71(2 Suppl):644-9.
(18) Wiechert A, Saygin C, Thiagarajan PS, et al. Cisplatin induces stemness in ovarian
cancer. Oncotarget. 2016 May 24; 7(21): 30511-22.
(19) Christie EL, Bowtell DDL. Acquired chemotherapy resistance in ovarian cancer. Ann
Oncol. 2017 Nov 1; 28(Suppl 8):viii13-viii15.
(20) Dey-Guha I, Wolfer A, Yeh AC, et al. Asymmetric cancer cell division regulated by
AKT. Proc Natl Acad Sci USA. 2011 Aug 2; 108(31): 12845-50.
39

(21) Ranganathan AC, Adam AP, Zhang L, Aguirre-Ghiso JA. Tumor cell dormancy induced
by p38SAPK and ER-stress signaling: an adaptive advantage for metastatic cells? Cancer
Biol THer. 2006 Jul; 5(7): 729-35.
(22) Alberts B, Johnson A, Lewis J, Morgan D, Raff M, Roberts K, Walters P. Molecular
Biology of THE CELL. Sixth Edition. New York: Garland Science, Taylor & Francis
Group, LLC; 2015.
(23) Seguin L, JS, Weis SM, Cheresh DA. Integrins and cancer: regulators of cancer
stemness, metastasis, and drug resistance. Trends Cell Biol. 2015 Apr; 25(4): 234-40
(24) Hynes RO. Integrins: Bidirectional, Allosteric Signaling Machines. Cell. 2002 Sep 20;
110(6): 673-87.
(25) Chiarugi P, Giannoni E. Anchorage-Dependent Cell Growth: Tyrosine Kinases and
Phosphatases Meet Redox Regulation. Antioxid Redox Signal. 2005 May-Jun; 7(5-6):
578-92.
(26) Cukierman E, Pankov R, Yamada KM. Cell interactions with three-dimensional
matrices. Curr Opin Cell Biol. 2002; 14:633-39.
(27) Walko G, Castanon MJ, Wiche G. Molecular architecture and function of the
hemidesmosome. Cell Tissue Res. 2015; 360:529-44.
(28) Miyamoto S, Teramoto H, Coso OA, et al. Integrin function: molecular hierarchies of
cytoskeletal and signaling molecules. J Cell Biol. 1995 Nov 1; 131(3): 791-805.
(29) Damsky CH, Ilic D. Integrin signaling: it’s where the action is. Curr Opin Cell Biol.
2002; 14(5):594-602.
40

(30) Bill HM, Knudsen B, Miranti C, et al. Epidermal Growth Factor Receptor-Dependent
Regulation of Integrin-Mediated Signaling and Cell Cycle Entry in Epithelial Cells. Mol
Cell Biol. 2004 Oct; 24(19): 8586-8599.
(31) Kwan BH, Zhu EF, Tzeng A, et al. Integrin-targeted cancer immunotherapy elicits
protective adaptive immune responses. J Exp Med. 2017; 214(6): 1679-90.
(32) Stouffer GA, Smyth SS. Effects of thrombin on interactions between beta3-integrins and
extracellular matrix in platelets and vascular cells. Arterioscler Thromb Vasc Biol. 2003
Nov 1; 23(11): 1971-8.
(33) Hamidi H, Ivaska J. Every step of the way: integrins in cancer progression and
metastasis. Nat Rev Cancer. 2018 Sep; 18(9): 533-48.
(34) Oudin MJ, Jonas O, Kosciuk T, et al. Tumor Cell-Driven Extracellular Matrix
Remodeling Drives Haptotaxis during Metastatic Progression. Cancer Discov. 2016 May;
6(5): 516-31.
(35) Levental KR, Yu H, Kass L, et al. Matrix Crosslinking Forces Tumor Progression by
Enhancing Integrin Signaling. Cell. 2009 Nov 25; 139: 891-906.
(36) McGrail DJ, Khambhati NN, Qi MX, et al. Alterations in ovarian cancer cell adhesion
drive taxol resistance by increasing microtubule dynamics in a FAK-dependent manner.
Sci Rep. 2015 Apr 17; 5: 9529.
(37) Desgrosellier JS, Barnes LA, Cheresh DA, et al. Integrin avB3/c-src “Oncogenic Unit”
Promotes Anchorage-independence and Tumor Progression. Nat Med. 2009 Oct; 15(10):
1163-9.
(38) Desgrosellier JS, Cheresh DA. Integrins in cancer: biological implications and
therapeutic opportunities. Nat Rev Cancer. 2010 Jan; 10(1): 9-22.
41

(39) Seguin L, Desgrosellier JS, Weis SM, Cheresh DA. Integrins and cancer: regulators of
cancer stemness, metastasis, and drug resistance. Trends Cell Biol. 2015 April; 25(4):
234-40.
(40) Seguin L, Kato S, Cheresh DA. An integrin B3-KRAS-RaIB complex drives tumor
stemness and resistance to EGFR inhibition. Nat Cell Biol. 2014 May; 16(5): 457-68.
(41) Zheng W, Ge D, Meng G. Reversing microtubule-directed chemotherapeutic drug
resistance by co-delivering a2B1 inhibitor and paclitaxel with nanoparticles in ovarian
cancer. Cell Biol Int. 2020 Feb; 44(2): 610-20.
(42) Chen TC, da Fonseca CO, Schönthal AH. Intranasal Perillyl Alcohol for Glioma
Therapy: Molecular Mechanisms and Clinical Development. Int. J. Mol. Sci. 2018 Dec
6;19(12): 3905.
(43) Memorial Sloan Kettering Cancer Center. Perillyl Alcohol [Internet]. updated 24 Feb
2020. Available from: https://www.mskcc.org/cancer-care/integrative-
medicine/herbs/perillyl-alcohol#references-7
(44) Bailey HH, Levy D, Harris LS, et al. A Phase II Trial of Daily Perillyl Alcohol in
Patients with Advanced Ovarian Cancer: Easter Cooperative Oncology Group Study
E2E96. Gynecol Oncol. 2002 Apr 24;85(3):464-8.
(45) Zhang J, Stevens MFG, Bradshaw TD. Temozolomide: Mechanisms of Action, Repair
and Resistance. Curr Mol Pharmacol. 2012;5(1):102-114.
(46) Cho HY, Wang W, Jhaveri N, et al. NEO212, Temozolomide Conjugated to Perillyl
Alcohol, Is a Novel Drug for Effective Treatment of a Broad Range of Temozolomide-
Resistant Gliomas. Mol Cancer Ther. 2014 Aug;13(8)2004-17.
42

(47) Silva-Hirschberg C, Hartman H, Stack S, et al. Cytotoxic impact of a perillyl alcohol-
temozolomide conjugate, NEO212, on cutaneous T-cell lymphoma in vitro. Ther Adv
Med Oncol. 2019 Dec 6;11:1-15.
(48) Liebmann JE, Cook JA, Lipschultz C, et al. Cytotoxic studies of paclitaxel (Taxol®) in
human tumor cell lines. Br J Cancer. 1993: 68:1104-9.
(49) He K, Xu T, Goldkorn A. Cancer Cells Cyclically Lose and Regain Drug-Resistant
Highly Tumorigenic Features Characteristic of a Cancer Stem-like Phenotype. Mol
Cancer Ther. 2011 Jun; 10(6): 938-48.
(50) Wu X, Luo Q, Zhao P, et al. MGMT-activated DUB3 stabilizes MCL1 and drives
chemoresistance in ovarian cancer. PNAS. 2019 Feb 19; 116(8): 2961-66.
(51) Zheng Y, Lv L, Wang J, et al. rLj-RGD3 Suppresses the Growth of HeyA8 Cells in
Nude Mice. Molecules. 2017 Dec; 22(12): 2234.

Abstract (if available)

Abstract Ovarian cancer is the most lethal gynecologic cancer bolstered by insufficient treatment options and substantial recurrence rates. A well-studied proponent to this detriment is a protein family of Arginine-Glycine-Aspartate (RGD)-recognizing integrins, which are commonly exploited in ovarian cancer for metastasis, drug resistance, and tumor progression. In this study, clinically-relevant models of different histotypes were developed in order to determine the relationship between RGD receptors and chemotherapy

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Radiosensitization of nasopharyngal carcinoma cells by NEO212, a perillyl alcohol-temozolomide conjugate drug

PDF

Insights from mouse models on the origin of ovarian cancer and its predisposing mechanisms

PDF

Mitotic regulation in ovarian epithelial tumors approaching in vitro crisis

PDF

Effect of NEO212, a novel perillyl alcohol-temozolomide conjugate, on mycosis fungoides and Sézary syndrome

PDF

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

PDF

Reactivation of the Epstein Barr virus lytic cycle in nasopharyngeal carcinoma by NEO212, a novel temozolomide-perillyl alcohol conjugate

PDF

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

PDF

Novel approaches of mobilizing human iNKT cells for cancer immunotherapies

PDF

Telomerase expression as a marker for minimal residual disease detection in patients undergoing second-look laparotomy for ovarian cancer: a follow-up study

PDF

The effects of hormonal exposures on ovarian and breast cancer risk

PDF

Let there be LIGHT: TNFSF14 immunomodulating properties for cancer immunotherapy

PDF

Modeling SynGAP1 truncating mutations in neurodevelopmental disease using iPSC-derived neurons

PDF

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

PDF

Role of a novel transmembrane protein, MTTS1 in mitochondrial regulation and tumor suppression

PDF

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

PDF

Anti-cancer effects of novel glidobactin type proteasome inhibitors

PDF

Design, synthesis and validation of Axl-targeted monoclonal antibody probe for microPET imaging of human lung cancer

PDF

Molecular mechanism for the immune evasion of CD1d antigen presentation by herpes simplex virus-1 UL56 protein

PDF

Characterization of invariant natural killer T cells in a novel humanized HBV-transgenic model

PDF

The opening of the blood brain barrier by homogenized perillyl alcohol

Asset Metadata

Creator Ferri, Francesca (author)

Core Title Investigating novel avenues for effective treatment of ovarian cancer

School Keck School of Medicine

Degree Master of Science

Degree Program Molecular Microbiology and Immunology

Publication Date 06/09/2020

Defense Date 03/20/2020

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

Tag A2780,ABC transporters,carboplatin cytotoxicity,ES2,HEYA8,integrins,NEO212,OAI-PMH Harvest,ovarian cancer,paclitaxel cytotoxicity

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Schonthal, Axel H. (committee chair), Dubeau, Louis (committee member), Minea, Radu (committee member), Yuan, Weiming (committee member)

Creator Email fferri@usc.edu,francescakferri@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-317779

Unique identifier UC11664323

Identifier etd-FerriFranc-8582.pdf (filename),usctheses-c89-317779 (legacy record id)

Legacy Identifier etd-FerriFranc-8582.pdf

Dmrecord 317779

Document Type Thesis

Rights Ferri, Francesca

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