INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the

text directly from the original or copy submitted. Thus, some thesis and

dissertation copies are in typewriter face, while others may be from any type of

computer printer.

The quality of this reproduction is dependent upon the quality of the copy

submitted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleedthrough, substandard margins, and improper alignment

can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and

there are missing pages, these will be noted. Also, if unauthorized copyright

material had to be removed, a note will indicate the deletion.

Oversize materials (e-g., maps, drawings, charts) are reproduced by sectioning

the original, beginning at the upper left-hand comer and continuing from left to

right in equal sections with small overlaps. Each original is also photographed in

one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6" x 9" black and white photographic

prints are available for any photographs or illustrations appearing in this copy for

an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning 300 North Zeeb Road, Ann Arbor, MI 48106-1346 USA

800-521 -0600

Hypoxia-lnduced Upregulation of VEGF mRNA

in Cervical Cancer Cell Lines

James Anthony Chiarotto

A thesis submitted in conformity with the requirements for the degree of Master of Science

Graduate Department of Medical Biophysics University of Toronto

@Copyright by James Anthony Chiarotto 1998

National Library I * m of Canada Bibliotheque nationale du Canada

Acquisitions and Acquisitions et Bibliographic Services services bibliographiques

395 Wellington Street 395, rue Wellington OttawaON K 1 A O M Ottawa ON K1 A O N 4 Canada Canada

The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of thls thesis in microform, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fkom it may be printed or otherwise reproduced without the author' s permission.

L'auteur a accorde me licence non exclusive permettant a la Bibliotheque nationale du Canada de reproduire, przter, distribuer ou vendre des copies de cette these sous la forme de microfiche/@ de reproduction sur papier ou sur format electronique.

L'auteur conserve la propriete du droit d'auteur qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent &tre imprimes ou autrement reproduits sans son autorisation.

Hypoxia-Induced Upregulation of VEGF rnRNA in Cervical Cancer Cell Lines Master of Science, 1998 James Anthony Chiarotto

Department of Medical Biophysics University of Toronto

ABSTRACT

Malignant cells produce an angiogenesis factor, vascular endothelial growth factor (VEGF), which

is upregulated by hypoxia and may increase metastatic ability. Clinical data for cervical cancer

relate oxygen levels in tumours to treatment outcome. This thesis addresses whether the presence

of VEGF mRNA could be used as a marker for relevant levels of hypoxia. VEGF mRNA levels

in three cervical cancer cell lines were found to be increased at oxygen concentrations welL above

that causing radiobiologic hypoxia, indicating that VEGF mRNA may not be useful as a direct

marker for radioresistant cells. Differences in the oxygen concentration associated with an

increased VEGF mRNA level, in the patterns of hypoxic upregulation of the VEGF gene, and

in oxygen consumption at low oxygen concentrations were noted between the different cell lines.

The results suggest a separation in the level of hypoxia which affects radiosensitivity and that

which may affect increased metastatic potential.

ACKNOWLEDGEMENTS

This thesis would not have existed without the help and encouragement of numerous people.

A large debt is owed to all my teachers, especially to Dr. Michael Thirlwell, who unselfishly

gave of their time to impart to me the skills, wisdom, and altruism gained over the length of their

careers. Similarly, I wish to thank Dr. Michael Rauth, Dr. Anthony Fyles, and Dr. Solomon

Minkin.

Once in the Iab, my every move needed to be guided (and carefully monitored to ensure that

I caused no disasters). This function was performed by Anne Jang with unfailing persistence and

patience. I have since forgiven her for leaving me all alone while taking six months off for

maternity leave. (I'm certain that I was more helpless than Cameron.)

My other lab mates were always ready to offer encouragement and guidance. Les (allowing

me to listen to the CBC on the lab radio), Marie-Claire (always making sure my name was on

lab greeting cards), Bob (updates on how far from the bottom of the standings the Leafs or Jays

were), Haihua (for pointing out how messy my desk was), Katrien (so that's were Belgium is),

and Wilson (my fingers now glow in the dark) became part of my family and offered every type

of support possible.

The father of this family was my supervisor, Dick Rill. To me, he will always be a model of

- the selfless, patient, and astute scientist and teacher. I can do little more than offer my profound

thanks and acknowledge his significant contribution to this work.

Finally, for their emotional support, my thanks to my mother, sisters, and Denise whose

constant presence made the long hours endurable.

DEDICATION

Per le donne Calabrese

specialmen te la nonna.

For the Etruscans among us.

TABLE OF CONTENTS Page

. . Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 ... Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Chapter I : Introduction

1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Hypoxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.2.1 The tumor microenvironment . . . . . . . . . . . . . . . . . . . . . . . 5 1.2.2 Chronic and acute hypoxia . . . . . . . . . . . . . . . . . . . . . . . . 5 1 .2.3 Radiobiologic hypoxia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.2.4 Chemotherapy and hypoxia . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.2.5 Apoptosis and anticancer agents . . . . . . . . . . . . . . . . . . . . . 7 1.2.6 Hypoxia and cellular behaviour . . . . . . . . . . . . . . . . . . . . . . 8

. . . . . . . . . . . . . . . . . . . . . . 1.2.7 Hypoxia and metastatic ability 8 1.2.8 Hypoxia and genetic instability . . . . . . . . . . . . . . . . . . . . . 10 1.2.9 Techniques of measurement of oxygen

status in turnours . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2.10 Determination of oxygen status in

human tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.3 Vascular Endothelial Growth Factor . . . . . . . . . . . . . . . . . . . . . . . 12 1.3.1 The historical perspective . . . . . . . . . . . . . . . . . . . . . . . . . 12

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 .3.2 Characterization 13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.3 Actions 13

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.4 Receptors 15 . . . . . . . . . . . . . . . . . . . . . . . 1.3.5 Signal transduction pathway 16

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.6 Oncogenes 17 1 -3.7 The VEGF message: HIF- 1 -mediated

. . . . . . . . . . . . . . . . . . . . . . . . . . . . transcription 17 1.3.8 The VEGF message: stabilization . . . . . . . . . . . . . . . . . . . 18

. . . . . . . . . . . . . . . . . . . . . . . . . 1.3.9 Localization-- physiologic 18 . . . . . . . . . . . . . . . . . . . . . 1 .3.10 Localization- pathophysiologic 18

1.3.1 1 Stimuli causing upregulation . . . . . . . . . . . . . . . . . . . . . . 19 1.3.12 VEGF presence in human turnours and

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . correlations 19 . . . . . . . . . . . . . . . . f . 3.13 VEGF and intratumoural microvesseis 20

1 -3.14 Angiogenesis: stimulators and . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . inhibitors 21

1.3.15 Dependency of hypoxic upregulation of VEGF on . . . . . . . . . . . . . . . . . . . . . . . oxygen concentration 22

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Thesis Outline and Rationale 24

Page

Chapter 2 : Hypoxia-lnduced Upregulation of VEGF mRNA Levels in Cervical Cancer Cell Lines

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Abstract 26

2.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials and Methods 29 2.3.1 Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3.2 Oxygenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.3 -3 Oxygen measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.3.4 RNase protection assay . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2.3.5 Probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 -6 Statistical analysis 33

2.4 ResuIts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 2.4.1 Net increase in VEGF mRNA levels under long-term

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . hypoxia 34 2.4.2 Analysis of VEGF mRNA levels as a hnction of oxygen

concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.4.3 Determination of VEGF mRNA half-life in the presence of

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . oxygen 43 2.4.4 Pattern of VEGF gene upregulation . . . . . . . . . . . . . . . . . . 46

2.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Chapter 3: Future Work

3.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Future Work 55

Chapter 4: References 4.1 List of References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Page List of Tables

Table 2.1 Table 2.2

List of Figures

CHAPTER I Figure 1.1

CHAPTER 2 Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 Figure 2.6

Oxygen concentrations in cell-containing medium . . . . . . . . . . . . . 3 1 Oxygen concentrations associated with K, values and half-lives . . . 40

Structure of VEGF splice variants . . . . . . . . . . . . . . . . . . . . . . . . 14

Relative VEGF mRNA Ievels in cell lines exposed to anoxia or air . 35 Relative VEGF mRNA levels as a fbnction of oxygen concentration 39 Determination o f VEGF rnRNA half-life in air in ME-1 80 cell line . 41 Representative RNase protection assays . . . . . . . . . . . . . . . . . . . . 42 Determination of VEGF mRNA half-life in air in ME-180 cell line . 44 Patterns of VEGF gene upregulation . . . . . . . . . . . . . . . . . . . . . . 46

vii

Chapter One

Introduction

2

1.1 OVERVIEW

The metabolic processes responsible for the maintenance of life are dependent upon the

presence of molecular oxygen. Sensors for monitoring the levels o f ambient oxygen are present

in single-celled organisms and in the individual cells of multi-celled organisms. In addition,

these multi-celled organisms have developed elaborate circulatory systems to ensure that each cell

receives an adequate and uninterrupted supply of oxygen. When the concentration of oxygen

departs from a well-defined range, cellular genetic programs are activated to ensure survival.

Hypoxia is said to exist when the oxygen concentration falls below what is considered necessary

to carry on normal cellular finction. The most important factor is the production of ATP which

is more efficient in the presence of oxygen. The amount of ATP available to the cell influences

such things as rate of cell growth and the quantity and types of metabolic end products (1).

A group of physiologically relevant hypoxia response genes produce a range of proteins

including: i) hypoxia-inducible factor 1, a transcription factor shown to be upregulated in the

presence of hypoxia (2) and thought to be the regulator of the hypoxia response genes (3) ii)

erthyropoietin, the primary regulator of erythropoiesis (1) ; iii) vascular endothelial growth factor

(VEGF), the primary regulator of angiogenesis (1) ; iv) a group of glycolytic enzymes which

provide a metabolic pathway for ATP production when no oxygen is present (1) ; v) a group of

- enzymes, such as nitric oxide synthase, which produce vasoactive molecules which cause dilation

and increased blood flow (4).

Hypoxia is the main factor in several human diseases which cause significant morbidity and

mortality. Heart attack and stroke are caused by damage to the heart muscle and brain tissue

respectively due to lack of oxygen. Diabetes causes changes in small blood vessels which

3

compromise the supply of oxygen to the heart, nerves, skin, and other major organs (5). Cancer,

a major cause of death in our society, is considered to be caused by a disorder of gene regulation

(6). However, in solid turnours such as cancer of the uterine cervix (7-lo), and cancer of the

head and neck (1 1, 12), hypoxia is believed to play a major role in modifying the natural history

of the disease to a more aggressive phenotype and in rendering anticancer therapies less effective.

Modem medicine has had notable success with chemotherapy and radiation therapy in the

treatment and cure of the hematologic cancers Hodgkin's disease, lymphoma, and leukemia. The

same is not true for the solid turnours. If the cancer has left the original focus and spread to

other parts of the body in a process called metastasis, it is more difficult to use surgical

treatments to remove all existing cancer cells, and systemic treatment such as chemotherapy and

occasionally radiotherapy must be used (13). Treatments which can cure cancer when metastases

are clinically evident exist only for two rare tumour types: cancer of the testicle (14) and

gestational trophoblastic disease (1 5). This leaves patients with meiastatic cancer with a reduced

hope for cure.

We are learning more about why anticancer treatments have had such a difficulty in eradicating

metastases. Resistance to chemotherapy is thought to be innate in some cancer cells and acquired

in others. For example, some cancer cells possess a naturally occurring transmembrane protein

- called P-glycoprotein, which has the ability to pump certain classes of chemotherapeutic agents

out of the cell (16). Radiation is less effective at killing cancer cells at Iow oxygen

concentrations (17). We are now also learning about how hypoxia can select for cancer cells

with a diminished zbility to undergo programmed cell death, or apoptosis (18). The ability to

trigger apoptosis may be necessary for many anticancer therapies to be effective (19) .

4

Thus, the presence of an hypoxic environment in solid tumours not only makes anticancer

therapies less effective but also may select for cancer cells which behave in a more aggressive

fashion- Because these effects make cancer treatment more difficult, much effort has been

directed at assessing the oxygen status in tumours. Methods in current use include needle

electrodes which can quantitate oxygen concentration (20), drugs which are metabolized to

reactive intermediates only in the absence of oxygen and whose presence bound in the tissue

denotes hypoxia (20), and histomorphornetric analysis (21) which assesses the presence of

vascular density as a sign of hypoxia. It is believed that this new blood vessel formation is due

the production of the potent angiogenic protein VEGF. Each method has its own limitations,

however needle electrodes have been able to identify patients with poor prognosis on the basis

of intratumoural oxygen tension (20).

VEGF is produced by many cancer cells and its production is upregulated by hypoxia. The

relationship between oxygen level and VEGF upregulation is not well understood. The purpose

of the present work was to examine this relationship in detail and to determine the extent of its

variation between different tumor cells of similar histopathological type. Cancer lines derived

from human cancer of the uterine cervix were chosen based on the abundance of clinical data

with needle electrodes relating hypoxia to prognosis and the possibility that VEGF might be a

- usable marker for relevant levels of hypoxia in tumours.

1.2 HYPOXIA

1.2.1 The tumour microenvironment

In contrast to the orderly proliferation of normal tissue and the organized system of oxygen and

nutrient supply, turnours are much more chaotic. Firstly, demand for oxygen and nutrients in

turnours is greater due to relatively high proliferation rates and aberrant metabolism. This

aberrant metabolism results in high rates of anaerobic glycolysis and acid production (22).

Secondly, the vascular network which supplies the tumour has structural and architectural

abnormalities. Abnormalities in the vessel wall include defects in the endothelial lining,

basement membranes, contractile components, and physiologic receptors. Architectural

abnormalities inciude tortuosity, contour irregularities, and the presence of arteriovenous shunts

(4). The resulting plasma leakage into the interstitiurn, in addition to the absence of lymphatic

vessels in tumours (22), causes accumuIation of plasma, increased extravascular pressures and

collapse of vessels. The result is a blood and oxygen supply that is erratic.

1.2.2 Chronic and Acute Hypoxia

The above abnormalities in blood flow lead to the development of two different types of

- hypoxia. Chronic hypoxia occurs because oxygen is consumed as it diffuses fiom a blood vessel

into the tissue. At a maximum of about 200 p m from a vessel, there is expected to be almost

no oxygen available for metabolism, which is consistent with the presence of tumour cords

around blood vessels with a layer of viable cells of similar thickness (23). Beyond this distance

necrosis usually develops.

6

Transient interruptions in blood flow can create acute hypoxia. The vascular malformations and

high intratumoral pressures contribute to an unstable flow. Reversal of flow or stoppage of flow

occurs in tumour vessels and reduces the amount of oxygen available to tissues at any given time

(22)-

1.2.3 Radiobiologic Hypoxia

The damage sustained by a cell exposed to ionizing radiation is dependent on the presence of

oxygen. Oxygen is a radiation sensitizer and the mechanism believed responsible for this is

described by the oxygen-fixation hypothesis. Free radicals are produced when radiation is

absorbed in biological tissues. These free radicals are highly reactive, interacting with

macromolecules in the tissues causing breakage of chemical bonds and other damage, and are

thus capable of altering the structure of cellular molecules. Most of the radiation is absorbed by

water molecules and the free radicals which are formed interact with macromo1ecules such as

cellular proteins or nucleic acids causing the production of radical sites which can form stable

addition products with oxygen, such as peroxides. These addition products are difficult to repair

and hence oxygen is said to "fix" the damage. In the absence of oxygen, the free radical may

acquire a proton from hydrogen atom-donating groups such as sulphydryls to restore its original

- form. Maximal radiosensitization occurs at oxygen tensions above 14 mmHg (20 pM) (17). The

oxygen tension at which the sensitizing effect is half of maximum (the K, value) is about 3-4

mmHg (5 pM) (17, 24). In the absence of oxygen, the dose of radiation required to cause a

given level of damage is increased to about three times that when oxygen is plentifbl.

7

1.2.4 Chemotherapy and Hypoxia

The presence or absence of oxygen can modify the effectiveness of cytotoxic chemotherapy.

Drugs which are more effective under oxygenated conditions include bleomycin, procarbazine,

actinomycin D, and vincristine (25). Hypoxia can increase the cytotoxicity of mitomycin C and

doxorubicin (25) as well as melphalan (26).

1.2.5 Apoptosis and Anticancer Agents

Since ionizing radiation and many chemotherapeutic agents cause DNA damage, and DNA is

considered to be a key cellular molecule for cell killing, their tumour-specific cytotoxic action

has been attributed to the genotoxic effect on the rapidly proliferating cells. Some drugs,

however, do not produce sufficient DNA damage to explain their toxicity (27, 28) without

invoking another mechanism. The notion of programmed cell death, or apoptosis, has allowed

a better explanation for the observed cytotoxicity of these agents.

Apoptosis can be described as physiologic, or programmed, cell death. This process has been

known to be a fundamental feature of development and to occur in mature tissues (29). The

histologic hallmarks of cells undergoing apoptosis include shrinkage, loss of cell-cell contact, and

chromatin condensation. Internucleosornal degradation of DNA, also known as DNA laddering,

- can usually be seen on DNA eIectrophoresis (30).

An intact and hnctional p53 tumour suppressor gene is necessary for the activation of one of

the apoptotic pathways in cells (3 1). The production of cells deficient in p53 function showed

that activation of apoptosis following exposure to chemotherapeutic compounds and radiation is

dependent on p53. Cell lines grown in culture which had normal p53 function could be induced

8

to undergo apoptosis following exposure to anticancer agents. Inactivation of p53 rendered cells

less susceptible to the cytotoxic effects of radiation and a cross-section of chemotherapeutic

agents: 5-fluoruracil (an antimetabolite), etoposide (topoisomerase I1 inhibitor), and adriamycin

(DNA intercalator) (32).

A similar dependency on p53 was demonstrated in vivo. Oncogenically transformed cells both

possessing intact p53 or disrupted p53 gene function were able to form tumours. However, cells

lacking p53 were resistant to radiation and adriamycin; cells with intact p53 function undenuent

apoptosis, and tumour shrinkage (3 3).

These findings help make sense of clinical observations regarding the success of anticancer

therapies. A high percentage of cancers of the lung, colon, bladder, prostate, and malignant

melanoma possess p53 mutations (32), which may explain the difficulty experienced in

eradicating metastatic cancer cells originating from these cancers. Cancer of the testicle and

childhood acute lymphoblastic leukemia rarely exhibit p53 mutations (32); patients with these

cancers can be cured even in the metastatic setting.

1.2.6 Hypoxia and Cellular Behaviuur

Hypoxia is a potentially reversible stress with which a cell must deal if it is to survive. Along

- with the induction of hypoxia response genes, growth arrest is a possible response (34). Hypoxia

causes a GI-phase arrest and is one of the nongenotoxic stresses to do so. Ionizing radiation

causes a GI-phase arrest by inducing p53 and p21 (34). Hypoxia can also cause an accumulation

of p53 through HIF-la-dependent stabilization of p53 protein (35). However the cell-cycle arrest

is not dependent on p53 (34).

9

Hypoxia induces cell death by triggering apoptosis. In this situation, the apoptosis is p53-

dependent (18, 36), so that ceils containing mutant p53 may be more capable of surviving

exposure to hypoxia.

1.2.7 Hypoxia and Metastatic Ability

Exposure of cancer cells to hypoxia may enhance their metastatic ability. On exposure to severe

hypoxia followed by reoxygenation, cultured rodent tumour cells demonstrated a transient

increase in their ability to form metastases in mice (37, 38).

1.2.8 Hypoxia and Genetic Instability

The increased malignant behaviour exhibited by cancers as they develop has been associated

with genomic instability (39). Many mutations can be found in malignant cells. The change in

behaviour means less success in cancer treatments and shorter survival times for patients. DNA

replication occurring in cells which have been exposed to hypoxic environments is prone to error.

When cells were exposed to severely hypoxic conditions (4 mmHg 0,) for 4 hours, a 3-4 fold

increase in mutation frequency was noted with transversions and deletions being the mutations

most often seen (39).

- Mutations occurring in key genes will alter cellular behaviour dramatically. The overexpression

of c-Myc is associated with oncogenic traneformation (40). Paradoxically, this oncogene is also

involved in preparing the cell for apoptotic cell death (18). c-Myc-induced apoptosis in the

setting of hypoxia can be suppressed by increasing the expression of the antiapoptotic gene bcl-2

(41). This situation would give cells a survival advantage in tumours with hypoxic regions.

10

Hypoxia has, in fact, been found to select for cells which are less able to undergo apoptosis

(18). Thus, tumours with hypoxic regions might be expected to slowly become populated with

cells which possess a diminished ability to undergo apoptosis. Since many anticancer therapies

require an intact apoptotic pathway, such tumours would be more difficult to treat effectively.

1.2.9 Techniques of Measurement of Oxygen Status in Tumours

The most commonly used technique to measure tumour oxygenation is the Eppendorf pO,

Histograph which uses a polarographic needle electrode (22). This technique is in clinical use

and, because it has been able to identify patients with poor prognosis on the basis of oxygen

tension, it has become a gold standard (20). However, there are problems with this technique.

A low signal-to-noise ratio at low oxygen concentrations in the range of radiobioIogic hypoxia

limits its accuracy. It is sometimes difficult to determine if the tip of the electrode is in necrotic

areas or normal tissue thus causing over- or underestimation respectively of the hypoxic fraction.

In addition, the use of this technique is feasible only when a tumour is readily accessible, such

as cancer of the uterine cervix and soft-tissue sarcomas (20).

Another major clinically applicable technique is the selective binding of nitroimidazoles in

hypoxic cells. These compounds are metabolized by nitroreductases only at oxygen tensions

- below about 3 mmHg (20, 42), producing species which bind to cellular macromolecu~es.

Accumulation of the metabolites can be used to identify hypoxic regions within a turnour either

using radiolabelled drugs or by imrnunohistochemistry. The amount of nitroimidazole metaboIite

produced depends on the level o f nitroreductases within the target cells and this can produce

considerable variability between cell lines in vitro (20). In addition, large quantities of the parent

1 I

drug and metabolites accumulate in the liver and gastrointestinal tract which can make imaging

of cancer in these organs with radiolabelled drugs difficult (20).

1.2.10 Determination of Oxygen Status in Human Turnours

The presence of hypoxic cells in human tumours has been linked to poor radiotherapeutic

response and to poor survival in clinical studies. Needle electrodes have been used to measure

oxygen tension in lymph node metastases of squamous cell carcinoma of the head and neck and

it was found that radiation was less effective at shrinking hypoxic lymph nodes (11). Studies

using the same disease and methodoIogy showed that the presence of hypoxic lymph nodes could

identify patients with poor locoregional tumour control (12). Similar work using advanced cancer

of the uterine cervix (7-10) showed that oxygen measurements of the main tumor mass indicating

hypoxia correlated with poor patient survival. Extension of this work found that hypoxia in

cervical cancers correlated with a greater likelihood of nodal metastases (8). This propensity to

metastatis in hypoxic cervical tumours was also found in soft tissue sarcoma of the extremities

(43)-

1.3 VASCULAR ENDOTHELIAL GROWTH FACTOR WEGF)

1.3.1 The Historical Perspective

The notion that there were angiogenic growth factors important in the growth of a cancer was

pioneered by Judah Folkman (44) in his consideration of the problem of oxygen supply to a

tumour. Before a newly transformed clone of cells could begin to grow beyond a microscopic

focus, and later on metastasize, it must first acquire an independent blood supply. Malignant

cells were placed on such model systems as the chick chorioallantoic membrane and rabbit iris,

and it was noted that the cells did not initially undergo any growth. During the initial three to

five days, blood vessels were seen to grow in towards the "dormant" malignant cells. Once the

new blood vessels reached the malignant cells, the celIs underwent rapid growth. Folkman

postulated that the malignant cells released a humoral angiogenic factor upon which the newly

formed clone of malignant cells was vitally dependent for the development of a blood supply

(44). The process of a malignant cell acquiring the ability to produce angiogenic factors he

called the angiogenic switch. Once a cell had acquired this ability, its growth as a tumour mass

would not be limited by the diffusion of oxygen and nutrients from a nearby vessel, but it would

be able to induce its own independent supply.

- The major angiogenic growth factor was first identified through its ability to cause increased

permeability of cancer-associated blood vessels and extravasation of plasma into body cavities,

known as malignant ascites or a malignant pleural effusion. It was named vascular permeability

factor (VPF) (45, 46). When it was discovered that this protein had mitogenic activity for

endothelial cells, it became known as vascular endothelial growth factor. Many other growth

13

factors with angiogenic activity have been discovered and are described in section 1.3.14.

1.3.2 Characterization

The VEGF protein has a molecular weight of approximately 34-46-kDa and is an homodimeric

glycoprotein linked via disulphide bridges. It shares sequence homologies with platelet derived

growth factor and with placenta growth factor. The cDNA sequence encodes for a hydrophobic

secretory leader sequence that promotes active secretion (47).

The human VEGF gene is composed of 8 exons (figure 1.1). Multiple VEGF isoforms are

generated through alternative mRNA splicing (48). Three isoforms expressing 12 1 -, 165-, or

189-amino acid are commonly produced by malignant cells. The 165-amino acid isoform is the

most abundant. A 206-amino acid isoform was discovered through examination of cDNA

libraries and is uncommon (49). Cardiac myocytes were found to produce a 23-amino acid

isoform comprising exons one and eight (50).

Even though the 189-aa and 206-aa isoforms have secretion signals, they are poorly secreted;

the presence of exon 6 is responsible for these isoforms being cell-associated (49). The 189-aa

and 206-aa isofoms bind to heparin more readily than the 165-aa or 12 I-aa isoforms, and are

sequestered on cell surface heparan-sulphates, fiom which they are presumably Liberated when

required (51). The unique biological roles of each of these isofoms, if any, is yet to be

determined.

1.3.3 Actions

VEGF is a potent endothelial cell mitogen. Acting via high affinity receptors, it is regarded

Fimrre 1.1 The structure of VEGF isoforrns. The exons and corresponding peptides are shown

and not drawn to scale. The number of amino acids in each of the exon encoded peptides is

shown in the bottom box. (Taken fiom (51)).

15

as the most selective endothelial cell rnitogen known (52). VEGF's most potent activity, and one

not yet attributable to any other angiogenic growth factor, is its promotion of vascular

permeability. This permeability is made possible by a transcellular pathway known as vesiculo-

vascular organelles (VVO). W O ' s are clusters of interconnecting vesicles that extend across the

entire thickness of the endothelial cytoplasm (53). It has been suggested that this facilitates

extravasation of fibrinogen and other plasma proteins resulting in the formation of a fibrin

pIatform which supports the growth of new blood vessels (53). VEGF also elicits endothelial cell

chemotaxis and induces the expression of plasminogen activators and collagenase which promote

degradation of the supporting tissue to facilitate migration of endothelial cells (52).

VEGF acts as a survival factor for endothelial cells in the setting of retinopathy of prematurity

(54); it likely has the same role generally and can delay and reverse endothelial cell senescence

(55). VEGF is ubiquitous during vasculogenesis and inactivation of just one allele in the mouse

is embryonic lethal with impaired angiogenesis (56, 57). This represents the most severe

heterozygous phenotype reported to date (56).

1.3.4 Receptors

There are two known high affinity VEGF receptors in humans: FIt-l(in rodents)NEGFRI

- (£ins-like tyrosine kinaseNEGF receptor I ) and Flk-l(in rodents)KDWEGFR2 &nase insert

domain containing receptorNEGF receptor 2). They are receptor tyrosine kinases and contain

a single membrane spanning domain, seven extracelIular immunoglobulin-like domains and a

kinase domain (58). Heparan sulfate proteoglycans are required for binding of VEGF to its

receptors (58).

16

It was thought that VEGF receptors were expressed exclusively by endothelial cells (53) but

they are now being identified on cancer cells themselves as well as other cells. VEGFR2 (KDR)

was found on normal human hematopoietic stem cells, megakaryocytes, platelets, as well as

human leukemia cell lines (59). VEGF protein had no mitogenic effect on a human leukemia

cell line, however it was able to suppress radiation-induced apoptotic cell death in this line as

well as normal hematopoietic stem cells. Recently, both VEGF receptors were found on Capan-

1, a human pancreatic cell line. These receptors were functional since the presence of VEGF

protein stimulated the growth of these cells, and was associated with VEGFRZ phosphorylation

(60), indicating the initiation of signal transduction.

There is now evidence that the receptors themselves can be upregulated by hypoxia. This was

observed for VEGFRl (Flt-1) but not VEGFR2 in human umbilical vein endothelial cells. The

mouse Flt-1 gene promoter has been found to contain a functional hypoxia-inducible factor-1

(HIFl) binding site (6 1) (see section 1.3.7 for discussion of HIF1).

1.3.5 Signal Transduction Pathway

Little is known about the signal transduction pathway VEGF triggers when it binds with its

receptor. Receptor dimerization and either autophosphorylation or transphosphorlyation occurs,

- followed by binding and phosphorylation of specific downstream signal transduction proteins.

Tyrosine phosphorylations of four signaling molecules that contain SH2 domains have been

identified. By inhibiting these tyrosine phosphorylations induced by VEGF, the mitogenic effect

of VEGF on endothelial cells is diminished (62).

17

1.3.6 Oncogenes

Oncogenes which not only cause cellular transformation but also induce VEGF production have

been identified. Mutant K-RAS and H-RAS cause production of VEGF and disruption of ras

protein function by a protein farnesyltransferase inhibitor causes suppression (63, 64). A similar

situation was found with C-SRC (65) and V-SRC (66). Another investigator confirmed that V-

SRC induced VEGF expression but found that C-SRC did not (67). It was found that suppression

of VEGF production could be achieved in cells transiently transfected with wild-type p53 (66).

However, this suppressive effect was not seen in cells stably transfected with wild-type p53 (68).

This evidence indicates that one of the ways that oncogenes may contribute to growth of solid

tumours is by influencing the production of tumour angiogenesis factors (63) .

1.3.7 The VEGF Message: HIF-l -Mediated Transcription

The hypoxia-mediated increase in transcription of the VEGF gene is controlled by the

transcription factor hypoxia-inducible factor 1 (HIFI) (69). HE1 was first discovered through

its role as the hypoxia-inducible transcription factor for erythropoietin (70). It is now recognized

to be a universal regulator of the hypoxia response genes (3). HIF1 is a heterodimer of HIFla,

a unique protein, and HIFP, the aryl hydrocarbon receptor nuclear translocator, which is a

- common subunit for a family of heterodimeric transcription factors (71). HIFla is continuously

synthesized and degraded by the ubiquitin-proteosome system when oxygen is plentiful with a

half-life of 5-10 minutes (72). Hypoxic upregulation occurs with inhibition of this proteosome

system.

18

2.3.8 The VEGF Message: Stabilization

Afier transcription, the VEGF mRNA can be stabilized in the presence of hypoxia, thus

increasing its half-life (73). This hypoxic stabilization is mediated by the RNA-binding protein

HuR, which binds to sites in the 3'-untranslated region (74). Both increased transcription and

stabilization of VEGF mRNA have been demonstrated in rat glioma cells exposed to hypoxia

(75)-

1.3.9 Localization--physiologic

VEGF is ubiquitous in the embryo due to its vital role in vasculogenesis. VEGF is expressed

at low levels in a wide variety of normal tissues in the adults and at higher levels in cells of the

renal glomerulus, cardiac myocytes, prostatic epithelium and semen, and certain cells of the

adrenal cortex (53) and lung (76). Expression in these sites is not accompanied by angiogenesis.

VEGF is prominent in the female menstrual cycle; development of ovarian follicles, corpus

luteum, and endometrium is VEGF dependent (47).

I .3.10 Localization--pathophysiologic

VEGF is present in numerous disease states. There is a large amount of evidence which links

- malignant transformation with the ability to produce VEGF. Almost all in vivo cancers and all

transformed cell lines show the ability to produce VEGF constitutively (53). This finding

supports the Folkman notion of the angiogenic switch (44). VEGF can also be found in the

following disease states: rheumatoid arthritis, retinal neovascularization, psoriasis, and infantile

hemangiomas (77).

19

1.3.1 1 Stimuli Causing Upregulation

Hypoxia is a potent stimulus for VEGF production, especially for cancer cells (78, 79). This

hypoxic induction of VEGF is a general response since many types of cultured cancer cells have

been observed to increase VEGF mRNA levels by 10-50-fold when the percentage of oxygen was

dropped from 2 1% to roughly 3% or less (47).

Another stress able to upregulate VEGF is hypoglycemia (glucose deficiency) (80). Induction

of VEGF is not seen when anoxia and hypoglycemia are present at the same time (80). Thus, in

the absence of both oxygen and glucose, cells are unable to upregulate VEGF expression.

Recently, low intracellular pH has been found to be a stimulus for VEGF expression (81).

Induction was found to occur at a pH of 7.0 with dramatic induction at a pH of 6.8. Both

increased transcription of the VEGF gene and stabilization of the VEGF message were seen.

1.3.12 VEGF Presence in Human Tumours and Correlations

The presence of VEGF mRNA in tumours of various histologies has been correlated with poor

prognosis (82) and its protein to poor prognosis in breast cancer (83). VEGF mRNA has been

correlated to increased metastasis in breast cancer (84) and its protein to increased metastasis in

human colon cancer (85).

- The ability of human melanoma cells to upregulate vascular endothelial growth factor secretion

under hypoxia has been associated with an increased metastatic efficiency in immune-deprived

mice (86). When a human fibrosarcoma cell line was transfected with truncated Flt-1, causing

sequestering of VEGF and formation of heterodirners with VEGF receptors, a significant decrease

in the ability of its cells to implant and grow in the lungs of nude mice was seen (87). Thus,

VEGF may be a key regulator of metastatic ability of cancer cells.

1.3.13 VEGF and Intratumoural Microvessels

The measurement of the density of microvessels within a tumour has been attempted as an

estimate of the angiogenic activity within a tumour (88), the oxygenation status of a tumour (21),

and as an indicator of prognosis. However, controversy exists as to how best to stain, select, and

count tumour microvessels (89, 90). As a result, correlations between microvessel density,

hypoxia (21), and prognosis (91) in human tumours are unclear.

The best evidence that increased intratumoural microvessel density (iMVD) correlates with

prognosis exists for early stage breast cancer (90-92). Some of the studies which have not shown

this correlation for breast cancer have been criticized for their methodology (90). In cancer of

the cervix, both increased iMVD (93) and decreased iMVD (91, 94) have been correlated with

poor prognosis. Attempts at correiating iMVD with prognosis with other tumours have yielded

conflicting results as has been noted in recent reviews (90-92).

Since VEGF is considered the major angiogenesis factor in solid tumours (52), correlations

between its presence and iMVD have been sought. Initial resuIts in breast (83) and lung cancer

(95) show that increased iMVD is correlated with increased VEGF protein expression and that

- both of these measures are associated with poor prognosis. However, in cancer of the cervix no

correlation was found between iMVD and VEGF protein expression and neither was correlated

to prognosis (96).

Increased iMVD counts were significantly correlated with VEGF mRNA as measured by in situ

hybridization in cervical cancer (97) and by northern analysis in renal cell carcinoma (82, 98).

21

1.3.14 Angiogenesis: Stimulators and Inhibitors

Angiogenesis is a compiex process which invdves the endothelial cells breaking their contacts,

migrating through and breaking down the extracellular matrix, and organizing a new, patent

vessel. This process is controlled by proteins which can be divided into stimulators and

inhibitors and has been discussed in a number of recent reviews (99-101).

Basic fibroblast growth factor (bFGF) was the first angiogenic cytokine to be identified (100).

bFGF stimulates the growth of endothelial cells in vitro and promotes angiogenesis in viva,

although it stimulates the growth of other types of cells as well. bFGF may fbnction as an

angiogenesis factor by inducing or interacting with a direct-acting growth factor such as VEGF.

The same may be true for a number of other cytokines that are associated with angiogenesis such

as transforming growth factor a, acidic fibroblast growth factor, and transforming growth factor

P*

VEGF is just one member of the VEGF family. The proteins VEGF-B and -C are abundant

at sites of angiogenesis. VEGF-B binds to flt-4, a receptor found only in lymphatic endothelium

and could not compensate for the lack of VEGF in null or heterozygotic animals. VEGF-C

promotes endothelial cell growth. Little else is known about these family members (101).

The ligands for the endothelial cell-specific receptors tie-1 and tie-2 (or tek) are another group

- of angiogenic stimulators. Loss of tie-1 results in neonatal death and the loss of tie-2 results in

early embryonic lethality. One of the ligands for tie-2 has been identified and named

angiopoietin. It may be a survival factor for mature capillaries (100).

Endogenous inhibitors of angiogenesis can be divided into those present physiologically and

those present pathophysiologically. Thrombospondin- 1 (TSP- 1) has recently been identified as

22

an angiogenic inhibitor which functions as a physiologic regulator of vessel growth. It is present

in a wide variety of normal tissues. Vascular density of regressing mammary glands in TSP-1

knock-out mice is significantly higher than that of control mice, indicating that TSP-1 may not

be the only molecule with this inhibitory effect, otherwise the outcome of the TSP-1 null mice

would have been far more severe (100).

Two angiogenic inhibitors were first identified as cancer cell products. Angiostatin, a fragment

of plasminogen (102), and endostatin (103), a C-terminal hgment of collagen XVIII have been

found to suppress vessel growth by directly inhibiting endothelial cell proliferation. Angiostatin

was found in the urine or blood of mice carrying a large tumour burden. Endostatin was isolated

from cultures of malignant endothelial cells and was not produced by normal counterparts.

Another major inhibitor of angiogenesis is interferon a. It acts by antagonizing mitosis and

migration of endothelial cells, and by blocking bFGF production by parenchymal cells (101).

1-3-15 Dependency of Hypoxic Upregulation of VEGF on Oxygen Concentration

While it has been clearly established that hypoxia is a potent inducer of VEGF expression,

during studies of hypoxia-induced upregulation of VEGF the exact concentration of oxygen at

which this occurs is poorly defined. This is due to poor control of ambient oxygen concentration.

-

In reported studies cells have been grown in a culture flask with no-oxygen (65, 78) or 1%-

oxygen (79) environments above the medium to induce VEGF, and an air control. Due to the

formation of oxygen gradients caused by oxygen consumption by the respiring cells (I04), the

actual oxygen concentration to which the cell is exposed is unknown.

Other methods of inducing VEGF via hypoxia include using cancer-cell spheroids (80, 105) to

create hypoxic gradients, but the actual oxygen tension is again unknown. Similarly, the use of

sandwich cultures (106), where cells are grown between glass slides and oxygen is only allowed

to diffuse in through the exposed edges, creates oxygen gradients but, in this published

experiment, the oxygen concentration was not measured-

Only one study has attempted to correlate the production of VEGF in cancer cells to the

ambient oxygen concentration (107). VEGF protein production in two colon cancer cell lines

exposed to a range of oxygen concentrations was measured. An increase in VEGF protein

production was noted with decreasing ambient oxygen concentration, however, the cells were kept

in monolayer culture with a 2.3 mm layer of medium overlying them. This can allow the

formation of significant oxygen gradients (108) which results in uncertainty in the actual oxygen

concentration to which the cells were exposed.

24

Exposure of cancer cells to hypoxia is now thought to increase their resistance to anticancer

therapies and increase their ability to metastasize. Thus, the ability to determine the oxygen

status within a solid tumour will help in designing more effective cancer therapies and

determining prognosis. VEGF is an angiogenesis factor produced by cancer cells and whose

expression is upregulated by hypoxia. This thesis will address whether the presence of VEGF

mRNA in cancer cells can be used as a marker for relevant levels of hypoxia, such as

radiobiological hypoxia. Chapter 2 will present data correlating VEGF mRNA levels in human

cervical cancer cell lines in vitro to accurately measured ambient oxygen concentrations. Chapter

3 will summarize the experimental findings and discuss possible future experiments.

Chapter Two

Hypoxia-In duced Upregulation of VE GF mRNA Levels

in Cervical Cancer Cell ~ i n e s *

-

his chapter is the text of a paper entitled "A Quantitative Analysis of the Reduction in Oxygen

Levels Required to Induce Upregulation of VEGF mRNA Levels in Cervical Cancer Cell Lines"

authored by "James A. Chiarotto and Richard P. Fill" submitted for publication to British Journal

of Cancer.

2.1 ABSTRACT

The presence of hypoxia (low oxygen concentrations) in solid turnours correlates with poor

prognosis, increased metastasis, and resistance to radiotherapy and some forms of chemotherapy.

Malignant cells produce an angiogenesis factor, vascular endothelial growth factor (VEGF), which

may increase metastatic ability and is upregulated in the presence of hypoxia. Clinical data for

cancers of the cervix and head and neck relate oxygen IeveIs in the turnour to treatment outcome.

This suggests the possibility that the presence of VEGF mRNA might be used as a marker for

relevant Ievels of hypoxia. Suspension cultures of three human cervical cancer cell lines, SiHa,

ME-180, and HeLa, were used to investigate upregulation of VEGF mRNA levels following

exposure to precisely defined oxygen concentrations for 2 or 4 h. An oxygen sensor was used

to confirm the actual levels of dissolved oxygen present. Differences in the oxygen concentration

which caused half-maximal upregulation (the Y, value) of VEGF mRNA level were seen among

the cell lines (K,,, at 4 h: SiHa 27.0*5.7pM, ME-1 80 16.8*3.3pM, HeLa l3.O*l.8pM, SiHa and

HeLa p=0.01). The Y, values for the HeLa cell line as measured at 2 h (24.9h0.8pM) and 4 h

(13.0*1.8pM) were also significantly different (p<0.0001). VEGF mRNA half-lives measured

in air were consistent with values in the literature (SiHa 59.8k5.8 min, ME-1 80 44.4*7.2 min,

HeLa 44.5t6.3 rnin). Differences in the patterns of hypoxic upregulation of the VEGF gene and

in oxygen consumption at Low oxygen concentrations were noted between the different cell lines.

Stirring in suspension culture was found to induce VEGF mRNA in S B a cells. Since the

for oxygen sensitization is about 5 pM, the results suggest a separation in the level of hypoxia

which affects radiosensitivity and that which may affect increased metastatic potential.

2.1 INTRODUCTION

Regions of hypoxia, or low oxygen tension, are known to exist within tumours (109). Since

radiotherapy and some forms of chemotherapy are less effective at killing cancer cells in hypoxic

environments, much effort has been directed toward identifying tumours containing such regions.

Measurement of oxygen tension by needle electrodes in lymph node metastases of cancer of the

head and neck found that radiation was less effective at inducing regression when the lymph

nodes were hypoxic (1 1) and such measurements have been reported to identify patients with

poor locoregional tumour control (12). Similar work in advanced cancer of the uterine cervix

(7-10) showed that increased levels of hypoxia in the primary tumour mass correlated with poorer

treatment outcome. These results suggested that hypoxia in cervical cancers correlated with a

greater likelihood of both local failure and nodal metastases. Metastases were also found to be

more fiequent in patients with the most hypoxic soft tissue sarcoma of the extremities (43).

Exposure of cancer cells to hypoxia (37, 38) and hypoxia-induced increases in vascular

endothelial growth factor (VEGF) secretion have been associated with a n increased metastatic

ability (86).

VEGF is the most selective vascular endothelial cell mitogen known (53). In cancer cells, the

- four VEGF isofoms which have been most frequently described contain 1 2 1, 165, 189, and 206

amino acids (48). In some studies, increased intratumoural VEGF mRNA and protein has been

associated with poor prognosis (82, 83), increased metastasis (84, 85), and increased microvessel

density (83, 95, 97). Antibodies directed towards VEGF can inhibit angiogenesis and the

proliferation of cancer cells in vivo (1 10, 1 11). Cancer cells constitutively produce VEGF (53)

28

and can upregulate its expression under hypoxic stress (78) via the transcription factor hypoxia-

inducible factor 1 (69) and by stabilization of the mRNA (73, 74).

The qualitative relationship between oxygen level and VEGF upregulation has been examined

in a variety of different turnour systems but has not been quantitatively documented in most of

the studies performed (65, 78, 79, 107). The purpose of the present work was to examine the

relationship between VEGF mRNA level and oxygen concentration in detail and to determine the

extent of its variation between different tumour cells of similar histopathoIogica1 type. Cancer

lines derived fkom human cancer of the uterine cervix were chosen for the study based on the

abundance of clincal data relating tumour oxygenation as measured by the Eppendorf PO,

Histograph to treatment outcome and the possibility that VEGF mRNA might be usable as a

marker for relevant levels of hypoxia in turnours. VEGF mRNA was chosen for study because

it is localized to the cell which is under hypoxic stress and is thus able to localize this

environment.

2.3 MATERIALS AND METHODS

2.3.1 Cells

Cell lines used in the experiments were SiHa, ME-180, and HeLa which are derived from human

cancer of the uterine cervix. The SiHa and ME-180 cells were obtained from ATCC (American

Type Culture Collection, Manassas, VA), while the HeLa cells were obtained from the laboratory

of Dr. Michael Rauth at CCWMH where they had been grown for many years. These cells were

grown in plastic tissue culture flasks (Gibco BRL, Burlington, ON, Canada) in a-minimal

essential medium (a-MEM, Gibco BRL, Burlington, ON, Canada) plus 10% fetal bovine serum

(FBS, Wisent, Quebec, Canada) plus antibiotics for SiHa and HeLa and in McCoy's 5A medium

plus 10% FBS plus antibiotics for ME-180. The cells were grown to about 70% confluence, then

trypsinized, counted, and a volume containing 3x10~ cells was spun down at 130g at 4OC. The

supernatant was then poured off leaving a cell pellet which was resuspended in the remaining

medium, approximately 50 p1, and introduced into the vials within 10 min of centrifugation as

described below.

2.3.2 Oxygenation

- The apparatus used for accurate control of the level of oxygen exposure of the cells has been

Mly described (1 12). Briefly, the apparatus consisted of a 37°C water bath into which a set of

glass vials, each containing a small magnetic stir bar and 10 ml of medium, was placed. The

medium was stirred at 200 rpm and humidified gas was flowed through an inlet in the stopper

for 90 min to achieve an equilibrium between the 0, in the gas phase and the liquid medium.

30

The small volume of medium containing the cells was introduced into each of the vials (final

concentration: 3x10' cells/ml) by sterile Pasteur pipet after the 90-min gassing period and the

gassing and stimng was continued. At defined time intervals later, a vial was removed fkom the

37°C water bath and placed on ice prior to total RNA extraction. Gases containing O2

concentrations of 2l%(air), 6.25%, 4.85%, 3.46%, 2.1 1%, 1.57%, 1.00%, or 0% (c 10 ppm) each

with 5%CO, and balance N2 were used. The composition of the gases was analyzed to be within

2% or better of the value given by the supplier (Praxair, Toronto, ON, Canada) and was

confirmed in our laboratory.

Each cell line was checked twice for the occurrence of cellular aggregation after 4 h of stimng.

No cellular aggregates were seen. The effect of the stimng on the cells was checked after 4 h

of stirring in both N, and air environments by plating the cells in tissue culture dishes for colony

formation. Plating efficiencies ranged from 50-80% of that found for cells which were not stirred

(data not shown). The pH of the stirred cell suspension was checked for each cell line and was

found to be unchanged at about 7.5 throughout the length of a 4-h gassing period.

2.3.3 Oxygen Measurements

Since the cells can be expected to consume some of the oxygen and hence influence the level

- of oxygen to which they are exposed, a Clark-type polarographic electrode (1 13) was used to

measure 0, concentration in the cell-containing medium. A glass vial and stopper, into which

an extra hole was cut to accommodate the oxygen sensor, was prepared as above. The sensor

was first calibrated for 0, concentration using the above series of gases. A total of 3 x 1 0 ~ cells

was introduced into stirred medium, as described above, and the resulting 0, concentration was

Table 1.1. Oxygen concentrations in pM measured in medium containing 3x105 cel lshl of one of the cell lines compared to the oxygen tension in the overlying gas in mm Hg and ta the expected oxygen concentration in the medium in the absence of cells in pM.

Ambient Oxygen Tension (rnmHg) [% 0, in gassing

mixture]

Expected Oxygen

Concentration (PM)

10.6

' Oxygen concentration in medium containing S N a cells significantly different from medium containing ME-1 80 cells (p=0.0 1) or HeLa cells (pc0.00 1)

32

measured once a stable signal indicating equilibration had been achieved, a process requiring

approximately 5 min. Results obtained for the different mixtures are shown in Table 2.1.

2.3.4 RNase protection assay

After exposure to gassing, the cells were spun into a pellet at 200g at 4°C for 5 min and total

RNA extracted with Trizol (Gibco BRL, Burlington, ON, Canada) using the manufacture's

protocol and stored at -70°C until analysis. An RNase protection assay was used to quantitate

VEGF and 36B4 mRNA levels (1 14). The 36B4 mRNA codes for an acidic ribosomal protein

(115) and served as a loading control. Briefly, riboprobes were purified on a 6%

polyacrylamide/urea gel then eluted overnight at 37°C in elution buffer (0.5M EDTA, 0.1% SDS,

0.1M EDTA), precipitated, and their radioactivity quantified. The probes were hybridized in

excess to 10 pg of total RNA overnight at 52OC. Samples were digested with 40 pg/ml of

RNaseA and 2 &ml of RNaseTl for 30 min at 30°C. Ten microlitres of 20% SDS and 50 pg

of proteinase K were then added, followed by incubation for 15 min at 37"C, phenol-chloroform

extraction, and ethanol precipitation with glycogen. The protected probes were then resolved on

a 6% potyacrylarnidehrea gel and quantitated on a Phosphorimager (Molecular Dynamics).

- 2.3.5 Probes

A probe for detecting VEGF mRNA was generated using the StyI hgmen t (nucleotides 99-352)

(48) obtained from the cDNA for VEGF165 (kind gift of Dr. Keith Laderoute, Stanford Research

Institute). The hgment was blunted and subcloned into the Smol site of the pB1uescript II KS(-)

cloning vector, which was then linearized with XbaL Correct orientation of the insert was

33

confirmed with the production of a 12 1 -bp DraII hgment. The 32~-radiolabeled antisense

riboprobe, containing 253 nucleotides was capable of recognising all four cancer-related isofoms.

It was transcribed using a T7 RNA polymerase.

A cloning vector containing the cDNA for 36B4 (1 15) (kind gift of Dr. Linda Penn, Ontario

Cancer Institute) was linearized with EcoR V. The '*P-radiolabeled antisense riboprobe containing

63 nucleotides, was transcribed using an SP6 RNA polymerase. For a11 the results presented, the

amount of VEGF mRNA was normalized to the amount of 36B4 mRNA detected in the same

lane on the polyacrylamide gel. A comparison of the 36B4 levels in cells gassed with 95%

N2/5% CO, for Oh or 4h collected over 48 experiments for all three cell lines showed no evidence

of hypoxic exposure on the expression level (paired t-test, p=0.69).

2.3.6 Statistical Analysis

The curves for VEGF mRNA upregulation as a function of oxygen concentration were fitted

to a logistic function using the Levenburg-Marquardt algorithm in the Origin 5.1 software

package (Microcal, Northampton, MA, USA) as described in the text. The K, values for VEGF

mRNA upregulation and the VEGF mRNA half-lives were compared in a two-tailed test using

the Z statistic. Comparisons of oxygen concentrations in cell-containing medium were made

- using a two-tailed Student's t-test. In both cases significance was determined by a p value less

that 0.05.

2.4 RESULTS

2.4.1 Net Increase in VEGF mRNA Levels Under Long-Tern Anoxia.

Initially we examined the time course of the increase in VEGF mRNA levels during exposure

to anoxia ( 4 0 ppm 03. Cells were placed in a vial containing medium which had been

equilibrated either with N2/5%CO2 (anoxic) or air/S%CO,. The cells were then exposed to this

environment for up to 12 h with samples taken at 0, 2, 4, 8, and 12 h for analysis of mRNA

levels. Figure 2.1 shows the results for each cell line under anoxic and air conditions. The

degree of VEGF upregulation was normalized to time 0 h for each cell line. The points represent

the mean(&S.E.) of at least 3 independent experiments. By 4 h, VEGF mRNA in ME-180 cells

had reached its maximum level. The HeLa line showed a gradual upregulation of VEGF mRNA

over the 12 h, however the degree of upregulation is similar to that in ME-180 cells. In both

these cell lines only slight changes in VEGF mRNA levels occurred over 12 h under air

conditions. In the SiHa cells the VEGF mRNA level reached a plateau after 8 h of anoxia.

However, there was also an increase in the control air conditions particularly after 2 h,

presumably as a result of the stress of being in the stirred suspension. Thus, a plateau for SiHa

cells likely occurs earlier than 8 h.

Figure 2.1. Relative VEGF mRNA level in cells exposed to anoxia or air in stirred cell suspension as a function of time. Solid symbols with solid lines indicate exposure to anoxia. Open symbols with dashed lines indicate exposure to air.

TIME (h)

36

2.4.2 Analysis of VEGF mRNA Levels in Cells Exposed to Direrent Oxygen Concentration.

This set of experiments was performed to determine in detail the range of oxygen concentration

over which VEGF mRNA is upregulated. Cells were gassed with various oxygen concentrations

using groups of three vials, that included a vial which contained cells exposed to the oxygen

concentration in question, a via1 containing cells which were gassed with 95% NJ5% CO, (the

positive control), and a vial containing cells gassed with 95% air/5% CO, (the negative control).

Since the results in figure 2.1 suggest that much of the effect of the hypoxic exposure occurred

in the first 4 h of gassing, the cells were sampled at time 0, 2, and 4 h after the start of the

gassing and total RNA extracted. The ratio of VEGF mRNA to 36B4 mRNA at 2 and 4 h was

normalized to the ratio at 0 h. For each of the three cell lines, measurements at each of the

above oxygen concentrations was performed at least in triplicate.

Figure 2.2a shows the relative VEGF mRNA level as a fbnction of oxygen concentration for

ME-180 cells measured at 2 and 4 h. Figures 2.2b and 2 . 2 ~ show the 2- and 4-h data for SiHa

and HeLa celIs respectively. The d a b were fitted to a logistic function and the oxygen

concentration at which half-maximal upregulation occurs was determined. During curve fitting,

the maximum and minimum values were fixed based on observations of the data. Tine minimum

values were the lowest points at the highest oxygen concentration with the single exception of

- 4-h data for the HeLa cells where an average of the values at the two highest oxygen

concentrations were used. For the ME-180 cell line, the maximum level was chosen to be the

highest single point (at 5.5 pM) in the 2-h and 4-h data set. The maximum value was an average

of the two values (at 2.1 p M and 8.9 pM) which appeared to be on the upper plateau in the case

of the SiHa cells. For the SiHa and ME-180 cell lines, this calculation did not include the 95%

37

NJS% CO, point because it was generally below the maximum value observed at intermediate

oxygen levels and it is possible that a cell's ability to produce mRNA may be compromised at

very low oxygen concentrations. This effect was not seen in the HeLa cell line, consequently

for the HeLa cells, the maximum value was chosen to be the relative VEGF mRNA level at the

anoxic point. Figure 2.2 indicates that the lines generated are generally a good fit to the data.

Modifying these choices of maximum levels (or minimum levels for the HeLa 4-h data) did not

affect the conclusions drawn fiom the data.

The data shows that at high oxygen concentrations, i.e. above 60 pM, there is relatively little

upregulation of VEGF mRNA. In the region between 10 pM and 50 pM, there is a dramatic

increase in the amount of VEGF mRNA present in the cell. Below 10 pM, the upregulation of

VEGF mRNA appears to reach maximal levels. There may be a trend towards VEGF mRNA

levels which are lower than maximal at very low oxygen levels, but this does not reach statistical

significance.

Table 2.2 shows the oxygen concentrations which produce half-maximal upregulation at the 2-

and 4-h time points. There were no statistical differences between to 2- and 4-h values for the

SiHa and ME-180 cell lines. The difference between to 2- and 4-h vaIues for the HeLa cell line

was significant (p<0.0001) and remained so with alternate choices for maximum and minimum.

- No statistical differences were seen among the 2-h values for the three cell lines. The 4-h value

for the SiHa cell line and the HeLa cell line was statistically different (p=0.01) and remained so

with recalculation using alternate choices of maximum and minimum. Figure 2.3 illustrates this

difference by showing the percentage change in VEGF mRNA levels in SiHa and HeLa cells at

4 h as a function of oxygen concentration. The data for ME 180 cells at 4 h lies intermediate

between these two curves and has been omitted for clarity.

A representative set of RNase protection assays showing the changes in the intensity of the

VEGF mRNA bands at various concentrations of oxygen as compared to air and nitrogen controls

is presented in Figure 2.4.

Figure 2.2. Relative VEGF mRNA level as a function of oxygen concentration. Data for ME- 180 cell line (a), SiHa cell line (b), and HeLa cell line (c) showing VEGF mRNA levels as measured at 2 h, open squares, and 4 h, solid squares.

0.001 0.01 0.1 1 10 100

Oxygen Concentration (pM)

Table 2.2. Oxygen concentrations which cause half-maximal upregulation of VEGF mRNA at 2 and 4 h and half-lives in presence of air.

Cell Line

SiHa

ME- 180

HeLa

2 h(pMh1SE)

2 9 - 1 6 I

18.7*6.9

24.9k0.7

4 h(pM+ISE)

27.0S.7

16.8*3.3

13.0*1.8

Half-life(min* 1 SE)

59.8*5.8

44.4k7.2

445t6.3

Oxvaen Concentration (uM1

Figure 2.3. Percentage change in VEGF mRNA level in the SiHa and HeLa cell lines as a hnction of oxygen concentration.

AIR VARIABLE NITROGEN OXYGEN

Time(h) 0 2 4 0 2 4 0 2 4

VEGF

3684

Figure 2.4. Representative RNase protection assays. Each experiment comprises two sets of three timed samples (0, 2, and 4 h) at three different oxygen concentrations: air (negative control), a measured oxygen concentration shown along the left margin in pM, and nitrogen (positive control). The top band in each experiment is the protected VEGF probe and the bottom band, the loading control, is a doublet representing the protected 36B4 probe. The bands representing 36B4 in the experiment using 8.9 p M are quite dif%se.

2.4.3 Determination of VEGF rnRlVA Half-Zzye in the Presence of Oxygen.

In the final series of experiments we examined the stability of the VEGF mRNA when the cells

were returned from anoxic exposure ((10 ppm 02) to exposure to air. The cells were placed in

vials equilibrated with either 95% N J5% CO, or 95% aids% CO,, and stirred in their respective

environments for 3 h, then 1 vial of each was sampled and the remaining 95% NJ5% CO, vials

were switched to gassing with 95% air/5% C 0 2 . The cell-containing medium required 4 min to

reach an oxygen concentration of 100 pM at which point the effect of hypoxia on upregulation

of VEGF mRNA was considered to be negligible. The vials were subsequently sampled at 1,

2, and 3 h. The level of VEGF mRNA in the N2-gassed vials was divided by the values in the

air controls and this ratio plotted against time. Figure 2.5 shows the resu!ts for ME-180, fitted

to an exponential decay curve. Similar results were obtained for SiHa and HeLa cells (data not

shown). The half-life for reduction of the mRNA levels was determined for each cell line from

+L Lllb - fitted exponential decay curve. The values obtained are shown in table 2.2. These data

suggest a similar decay time for a11 three cell lines. The calculated half-lives of the VEGF

mRNA in ME-1 80 cells, HeLa cells, and SiHa cells are consistent with values reported in the

literature: approximately 40 min in rat glioblastoma (116) and 43h6 min in rat

p heochromocytorna cells (73).

2.4.4 Pattern of VEGF gene upregulation

The relative level of VEGF mRNA at a given oxygen concentration was graphed as a function

of time in figure 2.6. The ME180 and SiHa cells, figures 2.6a and 2.6b, show a scattered

increase in VEGF mRNA levels with decreasing oxygen tension. However, figure 2 . 6 ~ shows

that the 2-h VEGF levels in HeLa cells are either unchanged or 2.5-fold upregulated.

. 4- nitrogen

5- - 5.5

. -x- 53.7

f

: 4- nitrogen - - 2.1

b.' -

7- -16.5 - + 30.7 -

5- --x-59.6 - - - - -

t

0 1 2 3 4

. -4- nitrogen - 6.4 - C . 5- -

-6 10.8 - -16.8

4- + 29.5 - . - 40.3

-

-

1 7 - . 1 - - 7 - -1 T - -

0 1 ; 3 i TIME (h)

Figure 2.6. Relative VEGF mRNA level in ME-180 cell line (a), SiHa cell line @), and HeLa cell line (c) plotted for each individual oxygen concentration tested as a function of time.

47

2.5 DISCUSSION

In the present study we examined the effect of different oxygen concentrations on the

upregulation of VEGF mRNA in three cervical cancer cell lines. One aim of the experiments

was to examine whether the presence of high levels of VEGF mRNA could serve as a suxrogate

marker for radiobiological hypoxia. Three different cell lines were studied using the same

techniques to determine if differences existed. VEGF mRNA was chosen for study over VEGF

protein because the mRNA stays localized to the cell which is under hypoxic stress and is thus

able to localize this environment.

Figure 2.1 shows that the rate of the hypoxia-induced increase in the relative level of VEGF

mRNA differs amongst cell lines. To our knowledge, the increase in the relative level of VEGF

mRNA associated with the physical stress of stirring, as was the case especially for the SiHa cell

line, has not been described before. A metastatic cell presumably would be exposed to physical

stress in the circulation and this might have a similar effect. The increased vascular permeability

caused by VEGF (53) would presumably allow a metastatic celI producing it to penetrate a

capillary wall more easily.

The pattern of induction of VEGF mRNA at different oxygen leveIs in the three cell lines is

shown in figure 2.6. The relative VEGF mRNA levels in the HeLa cell line (Figure 2 . 6 ~ ) at 2

- h is what one would expect if there were a threshold oxygen concentration below which the

VEGF gene was activated. The other cell lines show a gradual increase in relative VEGF mRNA

levels with decreasin oxygen concentration. It is not known how cells sense different oxygen

concentrations (1). Furthermore, for most oxygen-regulated genes, it is not known whether their

expression responds linearly to oxygen concentrations or if expression responds to oxygen

48

thresholds (1). The results for the cell lines studied here suggest that the linear pattern is more

common, but both patterns of induction of VEGF mRNA may occur in different cells.

The oxygen concentrations at which the VEGF mRNA is half-maximally upregulated (Y,

value) appear to be cell line specific (figures 2.2 and 2.3, table 2.2), indicating differences in the

ability of cervical cancer cell lines to react to hypoxic stress. Differences in the sensitivity of

the oxygen sensor within the cell, thought to be a heme protein (I), could explain these

differences. One possible explanation, cell-size related differences in intraceIlular oxygen

gradients (104) seems unlikely, since SiHa cells are the largest and ME-180 cells are the smallest

of the three cell types (based on microscope observations). A shift in the Y, value to a lower

oxygen concentration at 4 h (vs. 2h) for the HeLa cell line may indicate a change in the cell

line's ability to react to hypoxic stress with time.

The oxygen dependence of HIF-1 protein production and DNA-binding activity has been studied

previously using the HeLa cell line (2). The half-maximal value for these two activities,

measured after 4 h of treatment, occurred between 1.5 and 2% oxygen (15-20 pM). This value

is close to that observed for HeLa cells in the present study (13.0*1.8 pM). In deriving their

value, Jiang et al. circumvented the problem of oxygen gradients created by cellular respiration

by inhibiting oxidative phosphorylation with KCN. The presence of KCN altered the HIF-1

- subunit levels, however the oxygen value associated with the half-maximal HIF-1 level in the

presence or absence of KCN were about the same.

A number of other studies have shown upregulation of VEGF mRNA with hypoxia, but in most

cases have not examined its dependence on oxygen concentration quantitatively (65,78, 79). To

our knowledge only one study has examined the oxygen concentration dependence of VEGF

49

production in cells (107). In this study VEGF protein production was examined in two colon

cancer cell lines exposed to a range of oxygen concentrations. Similar VEGF protein secretion

rates were observed at oxygen concentrations in the gas phase below 0.3%. The apparatus for

control of oxygen concentration consisted of cells in monolayer culture with an overlying 2.3-rnm

layer of medium over which flowed gas of accurately known oxygen concentration. With this

depth of medium, a significant oxygen gradient due to cellular respiration would exist (108) and

would result in uncertainty in the actual oxygen concentration to which the cells were exposed.

In fact, the cell density was similar to that in the present study (2.8~10' vs. 3.0~10' cells/ml)

where a small but significant decrease in oxygen concentration due to cellular respiration was

clearly demonstrated (Table 2.1). The present study was conducted in a well-controlled

environment with direct oxygen measurements and specifically measured the Y, value for

upregulation of the mRNA, not protein, and showed that differences in that value may exist in

cells of the same cancer type. Interestingly, the K,,, values that we have measured for VEGF

upregulation are similar to the oxygen concentrations which stimulate the ability of endothelial

cells to form capillary networks (1 17).

One possible use for a detailed understanding of the oxygen dependence of VEGF mRNA

upregulation would be as a marker for hypoxia, especially radiobiologic hypoxia. Oxygen acts

- as a radiation sensitizer with a & at oxygen tensions of about 2-4 rnrn Hg (3-6 pM) (4, 24).

The values for the SiHa and ME480 cell Iines used in this study are similar (1 18) although the

exact Q value for radiosensitization appears to be affected by sulfiydryl levels within the cells.

Our results suggest that the presence of VEGF mRNA will not be a useful marker for this level

of hypoxia since upregulation occurs at oxygen concentrations below about 40 pM. Thus, VEGF

50

mRNA levels would be increased at oxygen concentrations above that considered to be important

for radiobiological hypoxia. In addition, there may be differences in Y, values (for VEGF

upregu!ation) among cell lines.

Physiologic conditions other than hypoxia are known to cause upregulation of VEGF. Low

glucose levels in the presence of oxygen (80) and low pH (81) can both increase VEGF levels

in vitro. This complicates the interpretation of changes in the level of VEGF in vivo. The levels

of hypoxia required for its upregulation in vivo need to be established. This might be done with

the use of a marker for hypoxia such as the 2-nitroimidazole EF-5, which has a Y, value for

binding of about 1 mrn Hg (1.5 pM)(119). The amount of EF-5 bound in celIs can be used to

determine their oxygen level during exposure to EF-5 and seems to correlate with VEGF protein

expression in spheroids (1 05).

Increased VEGF production by a cell may enhance its ability to form metastases. This may

occur through the ability of VEGF of promote growth and permeability of blood vessels, thus

allowing access of the cancer cells to the venous circulation. Evidence exists linking increased

production of VEGF in rodent turnour cells (38), human melanoma cells (86, 120, 121), and

human fibrosarcoma cells (87) to increased ability to form metastases. Thus, tumours in which

the cells upregulated VEGF at higher oxygen concentrations may be more likely to fonn

- metastases. The results of this study indicate that these effects would be maximal in the range

of 10-30 pM, which suggests a separation in the level of hypoxia which affect radiation

sensitivity and that which may affect metastatic ability.

Chapter 3

Discussion

Summary and Future Work

52

3.1 SUMMARY

The experiments presented in this thesis were designed to investigate the suitability of using

the presence of VEGF mRNA as a marker for radiobiologic hypoxia. As shown in chapter 2,

the K, values for VEGF mRNA levels as a function of oxygen concentration occur at values

higher than the K, for the oxygen sensitization effect for radiation. Thus, VEGF mRNA will be

of limited use as a marker for radiobiologic hypoxia. However, there is evidence showing that

VEGF may be invoived in increasing the metastatic ability of cells (38, 86, 120, 121). If this

is the case, then there may be an important separation between the level of hypoxia in turnours

which causes poor treatment response and the level of hypoxia responsible for increased

metastases.

Another novel finding presented in this thesis deals with potential differences in K, values of

VEGF mRNA level as a function of oxygen concentration amongst the three cell lines tested.

These differences were most clearly seen for the data at 4 h where SiHa cells had the highest Y,

value and HeLa cells had the lowest. There could be numerous physical explanations for this.

Oxygen gradients exist within a cell due to oxygen consumption causing the oxygen

concentration at the center to be less than that at the cell membrane (104). If the oxygen sensor

were in or near an organelle at the center of the cell, such as the nucleus or perhaps the

- mitochondria, a larger cell would have a lower oxygen concentration at the sensor than a smaller

cell even if the oxygen concentration at the cell membrane were the same for both cells. Given

that the location of the oxygen sensor within a cell is unknown, it is uncertain whether this could

explain the differences in Y, values seen between cell lines. In any case, the potential

importance of this finding relates to the presence of VEGF protein at higher oxygen

53

concentrations for some cells than others. This presence may increase the cell's metastatic ability

(38, 86, 120, 121).

A difference in the Y, value with respect to time of exposure to hypoxia was seen in the HeLa

cell line. This could indicate that the ability of this cell line to produce VEGF mRNA may be

altered by factors other than oxygen concentration. Since the experimental system which

contained the cells held constant all known important stimuli for VEGF production, such as

oxygen concentration and pH, perhaps this cell line has a negative-feedback loop which senses

the production of VEGF and adjusts its sensitivity to hypoxia.

The next novel finding in this thesis is the observation of the pattern of the VEGF gene

upregulation. It is not known how cells sense different oxygen concentrations (1). For most

oxygen-regulated genes, it is not known if their expression responds linearly to oxygen

concentrations or if instead expression responds to oxygen thresholds (1). The oxygen

concentrations used in these experiments were chosen to allow observation of VEGF mRNA

levels as they increased most dramatically. The VEGF gene seems to upregulate in hypoxia in

a linear fashion. The HeLa cells may have a gradient response to hypoxia (figure 2.6).

The final novel finding presented in this thesis is the ability of stirring to induce an increase

in VEGF mRNA levels in SiHa cells, and to a much lesser extent in the other cell lines.

- PhysicaI stress has not been described as an inducer of VEGF. The stress of stirring may

simulate the same stresses a metastatic cancer cell experiences as it leaves its initial focus and

enters the blood stream. The production of VEGF, given its ability to increase vascular

permeability, may aid a potentially metastatic cell which is already primed for VEGF production

to penetrate another vessel and, through its ability to stimulate angiogenesis, quickly acquire its

54

own blood supply. Inhibition of a human fibrosarcoma cell's production of VEGF was associated

with significant decrease in their implantation and growth in the lungs of nude mice (87).

3.2 FUTURE WORK

The unique roles, if any, of each of the four cancer related VEGF isoforms is currently

unknown. The different isofoms are formed from differential splicing of the mRNA and the

exons which are used to form each of these isoforms has an impact on their ability to bind to the

cell surface and hence their diffusibility (51). Little is known about why a cancer cell should

produce more than one isoform. The ability of certain isoforms to bind to heparan-sulphates on

the surface of cells may allow them to form a reserve source of VEGF which could be released

quickly when needed (51). The question as to whether each isoform may be produced in

response to a specific environmental stimulus has yet to be answered. This is an attractive

hypothesis since each isoform may be particularly suited for a specific environment. For example

in acidic environments, one particular isoform may dominate because of its unique stability.

The question as to whether one of the isoforms is preferentially produced under hypoxia has

been addressed in a cardiac myocyte system (50). Using an RNase protection assay, and specific

probes for each of the currently known isoforms as well as for a 23-amino acid isoform, no

difference was seen in the relative expression of the different mRNAfs produced by cardiac

myocytes under hypoxia (1% 0,) for 24 h. However, this experiment has not been done with

- cancer cells. A probe for each of the isoforms could be constructed using RT-PCR and used on

the samples already collected. If there were a difference in relative expression within a cell line,

perhaps other isoforms are specifically tailored for other environments, and it would be of interest

to test for these. If it is true that each isoforrn is preferentially produced by a specific

environmental stimulus, then expression of this isoform could provide a probe specific for that

56

environment. The question as to whether there is a difference in the relative amounts of the

different isoforms produced between cell lines or cancer types should also be addressed.

The ability of hypoxia to alter the genome of a cell has been established. There is evidence

that hypoxia can increase a cancer cell's ability to establish metastases (38, 86, 120, 121).

However, the genetic changes induced by hypoxia which give rise to this increased metastatic

ability are unknown. Whether these changes are produced by hypoxia or by physiologic changes

associated with hypoxia are not known. If the VEGF isoforms are specific for certain

environments, a clinical biopsy sample of a solid tumour could be probed, using in situ

hybridization and areas of the cancer which have been exposed to a specific environment, might

be identified. Further analysis of the cells in such areas would then be possible using the

technique of laser-capture microdissection. This technique allows one or a few cells to be

isolated from a tissue section. The cell's genome can then be amplified by RT-PCR and genetic

changes which have taken place can be analyzed by a gene array or gene chip technology.

The expression of many genes may be increased in hypoxic conditions. However, the number

of genes which are involved in increasing metastatic efficiency is unknown. The correct choice

of the key genes for further study from among the numerous upregulated genes is an obvious

necessity for success. Of the known hypoxia-response genes, ones that are associated with the

- production of glycolytic enzymes or vasoactive molecules may be less useful in the enhancement

of a metastatic phenotype. Genes which show a high degree of upregulation and are found to

do so across cells lines for which hypoxia exposure is believed to modify metastatic ability (eg.

head and neck cancer, cervical cancer, soft-tissue sarcoma) may indicate genes which are

important for metastasis. Once these genes have been identified, they could be cloned into a

57

plasmid and stably transfected into cancer celIs. To determine if they can affect metastatic

ability, transfected cells would be implanted into mice and the number of resulting metastatic foci

quantified. In this way, the genetic changes responsible for the increased metastatic ability of

cells can be investigated.

In the current studies three cell lines were tested for hypoxic induction of VEGF rnRNA. In

the first set of experiments (Figure 2 3 , the ceIls were exposed to two environments, air and

anoxia. The air environment was designed to be the control and little upregulation of VEGF

mRNA was expected. This was true for the ME-180 and HeLa cell Iines but in the SiHa cell

line, there was clear upregdation of VEGF mRNA in air conditions. This was presumably due

to the physical stress of being in the stirred cell apparatus. Stress-inducible protein kinases such

as SAPWNK are known to respond to the physical stresses of heat, UV irradiation, and

chemotherapy. SAPK has been shown to be upregulated by hypoxia in the SiHa cell line as well

(122). Whether SAPK is involved in the upregulation under air conditions could easiIy be tested.

The first step would be to perfom a kinase assay to determine if there is any SAPK in an active,

phosphorylated state after 4h of stimng. If there is an increase in phosphorylated SAPK in the

SiHa cells a causative relationship would have to be confirmed. The SiHa cells could be stably

transfected with a plasmid containing SEK-AL, which is a mutant form of SEK, the kinase which

- phosphorylates SAPK. SEK-AL inhibits the phosphorylation and thus the activation of SAPK.

If the SEK-AL transformed SiHa cells did not upregulate VEGF under stimng, this would be

strong evidence for a causative link.

The Y, values for the VEGF mRNA levels as a function of oxygen concentration were

determined for each of the three cell lines. For two of the cell lines, SiHa and ME-180, the K,

58

values at two and four hours were the same (table 2.2, figure 2.2). However, the two and four

hour values for the HeLa cell line were significantly different. The four hour value was

associated with a much lower oxygen concentration. One possible explanation for this behaviour

is the presence of a feedback loop which makes the cell less responsive to hypoxic stimuli when

VEGF is being produced. This presumably would require the presence of VEGF receptors in

HeLa cells. VEGF receptors were originally thought to occur only on endothelial cells, but there

is now evidence that receptors can also be expressed on cancer cells (59, 60). These receptors

seem to be functional. The VEGF receptors found on leukemia cells (59) were not mitogenic

in the presence of VEGF protein but did protect the cells from radiation-induced apoptotic cell

death. The receptors found on a pancreatic cell line (60) were mitogenic in the presence of

VEGF protein. Therefore, the possibility exists that VEGF receptors are present on the surface

of HeLa cells and form part of a negative feedback loop which make HeLa cells more resistant

to VEGF induction by hypoxia in the presence of VEGF. The existence of VEGF receptors

could be determined by the binding of specific antibodies or by the identification of VEGF

receptor mRNA by RT-PCR. However, the presence of receptors does not mean they are

functional. To determine whether they are hnctional two experiments could be done. Anti-

VEGF antibodies have been used before as possible antiangiogenic therapy (1 10, 11 1) with an

- apparently effective inhibition of angiogenic activity. Antibodies directed towards the receptors,

VEGFRl and VEGFR2, could also be used. The lack of phosphorylations on the receptor would

prove effective blockade. The other strategy would be to stably transfect the HeLa cells with

plasmids coding for the antisense RNA of each of the two receptors and testing for down-

regulation of their expression. This strategy would be more definitive and would likely be

59

necessary if the antibody method indicated functional receptors. Once the receptors have been

neutralized, the determination of the Y, values at 2 and 4 h would be repeated and if a negative

feedback loop does exist, the K, values should not be significantly different.

In summary, these proposed experiments will extend the findings outlined in chapter 2. They

may lead to an understanding of the role of VEGF isoforms, one of the possible signal

transduction pathways involved in VEGF upregulation, the possible presence of a VEGF

negative-feedback loop in a cancer ceil line, and an understanding of the genes which may be

responsible for increased metastatic ability in cells exposed to hypoxia.

Chapter 4

References

4.1 LIST OF REFERENCES

Bunn, H. F. and Poyton, R. 0. Oxygen sensing and molecular adaptation to hypoxia,

Physiol. Rev. 76: 839-885, 1996.

Jiang, B.-H., Semenza, G. L., Bauer, C., and Marti, H. H. Hypoxia-inducible factor 1

levels vary exponentially over a physiologically relevant range of 0, tension, Am. J.

Physiol. 271: C1172-C 1 180, 1996.

Guillemin, K. and Krasnow, M. A. The hypoxic response: huffing and HIFing. weview]

[21 refs], Cell. 89: 9-12, 1997.

Vaupel, P., Kallinowski, F., and Okunieff, P. Blood Flow, Oxygen and Nutrient Supply,

and Metabolic Microenvironment of Human Tumors: A Review., Cancer Res. 49:

6449-6465, 1989.

Foster, D. W. Diabetes mellitus, lir: A S . Fauci, E. Braunwald, Kurt J. Isselbacher, Jean

D. Wilson, Joseph B. Martin, Dennis L. Kasper, Stephen L. Hauser, Dan L. Longo (eds),

Hamson's Principles and Practice of Internal Medicine. 14 ed. McGraw-Hill: 2060-2080,

1998.

Squire, J. A., Whitmore, G. F., and Phillips, R. A. Genetic basis of cancer, In: I.F.

Tannock, R.P. Hill (eds), The Basic Science of Oncology. 3 ed. McGraw-Hill: 48-78,

1998.

Hockel, M., Knoop, C., Schlenger, K., Vorndran, B., Baussman, E., Mitze, M., Knapstein,

P. G., and Vaupel, P. Intratumoral PO, predicts survival in advanced cancer of the uterine

cervix, Radiother. Oncol. 26: 45-50, 1993.

62

8. Hockel, M., Schlenger, K., Aral, B., Mitze, M., Schaffer, U., and Vaupel, P. Association

between tumor hypoxia and malignant progression in advanced cancer of the uterine

cervix, Cancer Res. 56: 4509-4515, 1996.

9. HockeI, M. and Vaupel, P. The prognostic significance of hypoxia in cervical cancer: a

radiobiological or tumor biological phenomenon?, In: M. Molls, P. Vaupel (eds.), Blood

Perfusion and Microenvironment of Human Tumors. New York: Springer 73-79, 1998.

10. Fyles, A. W., Milosevic, M., Wong, R., Kavanagh, M.-C., Pintilie, M., Chapman, W.,

Levin, W., Manchul, L., Keane, T. J., and Hill, R. P. Oxygenation predicts radiation

response and survival in patients with cervix cancer., Radiother. Oncol. 48: 149- 156,

1998.

11. Gatenby, R. A., Kessler, H. B., Rosenbloom, J. S., Coia, L. R., Moldofsky, P. J., Hartz,

W. H., and Broder, G. J. Oxygen distribution in squamous cell carcinoma metastases and

its relationship to outcome of radiation therapy; Int. J. Radiat. Oncol. Biol. Phys. 14:

831-838, 1988.

12. Nordsmark, M., Overgaard, M., and Overgaard, J. Pretreatment oxygenation predicts

radiation response in advanced squamous cell carcinoma of the head and neck, Radiother.

Oncol. 41: 3 1-39? 1996.

- 13. Kohn, E. C. and Liotta, L. A. Invasion and metastasis, In: A.S. Fauci, E. Braunwald, Kurt

J. Isselbacher, Jean D. Wilson, Joseph B. Martin, Dennis L. Kasper, Stephen L. Hauser,

Dan L. Longo (eds), Harrison's Principles and Practice of Internal Medicine. 14 ed.

McGraw-Hill: 520-522, 1998.

63

14. Motzer, R. J. and Bosl, G. J. Testicular cancer, in: A S - Fauci, E. Braunwald, Kurt J.

Isselbacher, Jean D. Wilson, Joseph B. Martin, Dennis L. Kasper, Stephen L. Rauser, Dan

L. Longo (eds), Harrison's Principles and Practice of Internal Medicine. 14 ed.

McGraw-Hill: 602-604, 1998.

15. Young, R. C. Gynecologic malignacies, In: A.S. Fauci, E. Braunwald, Kurt J. Isselbacher,

Jean D. Wilson, Joseph B. Martin, Dennis L. Kasper, Stephen L. Hauser, Dan L. Longo

(eds), Harrison's Principles and Practice of Internal Medicine. 14 ed. McGraw-Hill:

605-610, 1998.

16. Tannock, I. F. Drug resistance and experimental chemotherapy, In: I.F. Tannock, R.P. Hill

(eds), The Basic Science of Oncology. 3 ed. McGraw-Hill: 402-404, 1998.

17. Hill, R. P. Molecular and cellular basis of radiotherapy, In: I.F. Tannock, R.P. Hill (eds),

The Basic Science of Oncology. 3 ed. McGraw-Hill: 309, 1998.

18. Graeber, T. G., Osmanian, C., Jacks, T., Housman, D. E., Koch, C. J., Lowe, S. W., and

Giaccia, A. J. Hypoxia-mediated selection of cells with diminished apoptotic potential in

solid turnours, Nature. 379: 88-9 1, 1996.

19. Kerr, J. F. R., Winterford, C. M., and Harmon, B. V. Apoptosis: its significance in

cancer and cancer therapy, Cancer. 73: 20 13-2026, 1993.

20. Stone, H. B., Brown, J. M., Phillips, T. L., and Sutherland, R. M. Oxygen in human

tumors: correlations between methods of measurement and response to therapy, Radiat.

Res- 136: 422-434, 1993.

21. Raleigh, J. A., Dewhirst, M. W., and Thrall, D. E. Measuring tumor hypoxia, Semin.

Radiat. Oncol. 6: 37-45, 1996.

64

Gulledge, C. J. and Dewhirst, M. W. Tumor oxygenation: a matter of supply and demand,

Anticancer Research. 16: 74 1-750, 1996.

Thomlinson, R. H. and Gray, L. H. The histological structure of some human lung cancers

and the possible implications for radiotherapy, Br. J. Cancer. 9: 539-549, 1955.

Chapman, J. D., Dugie, D. L., Reuvers, A. P., Meeker, B. E., and Borsa, J. Studies on the

radiosensitizing effect of oxygen in Chinese hamster cells, Int. J. Radiat. Biol. 26:

383-389, 1974.

Teicher, B. A., Lazo, J. S., and Sartorelli, A. C. Classification of antineoplastic agents by

their selective toxicities toward oxygenated and hypoxic tumor cells., Cancer Res. 41:

73-81, 1981.

Skarsgard, L. D., Skwarchuk, M. W., Vinczan, A., Kristl, J., and Chaplin, D. J. The

cytotoxicity of rnelphalan and its relationship to pH, hypoxia and drug uptake., Anticancer

Res. 15: 219-224, 1995.

Aliegra, C. J. Antifolates, In: B. Chabner, J . Collins (eds.), Cancer Chemotherapy:

Principles and Practice. New York: J. B. Lippincott: 126, 1990.

Myers, C. and Chabner, B. Anthracyclines, In: B. Chabner, J . Collins (eds)., Cancer

Chemotherapy: Principles and Practice. New York: ' B. Lippincott: 359, 1990.

Raff, M. C. Social controls on cell survival and cell death, Nature. 356: 397-399, 1992.

WylIie, A. H. Glucocorticoid-induced thymocyte apoptosis is associated with endogenous

endonuclease activation, Nature. 284: 555-556, 1980.

65

3 1, CIarke, A. R., Purdie, C. A., Harrison, D. J., Moms, R. G., Bird, C. C., Hooper, M. L.,

and Whllie, A. H. Thyrnocyte apoptosis induced by p53-dependent and independent

pathways, Nature. 362: 849-852, 1993.

32. Lowe, S. W., Ruley, H. E., Jacks, T., and Housman, D. E. p53-dependent apoptosis

modulates the cytotoxicity of anticancer agents., Cell. 74: 957-967, 1993.

33. Lowe, S. W., Bodis, S., McClatchey, A., Rernington, L., Ruley, H. E., Fisher, D. E.,

Housman, D. E., and Jacks, T. p53 status and the efficiency of cancer therapy in vivo.,

Science. 266: 807-8 10, 1994.

34. Graeber, T. G., Peterson, J. F., Tsai, M., Monica, K., Fornace, A. J., Jr., and Giaccia, A.

J. Hypoxia induces accumulation of p53 protein, but activation of a GI-phase checkpoint

by low-oxygen conditions is independent of p53 status, MoI. Cell. Biol. 14: 6264-77,

1994.

35, An, W. G., Kanekal, M., Simon, M. C., Maltepe, E., Blagosklonny, M. V., and Neckers,

L. M. Stabilization of wild-type p53 by hypoxia-inducible factor l a , Nature. 392:

405-408, 1998.

36. Giaccia, A. J. Hypoxic stress proteins: survival of the fittest, Seminars in Radiation

Oncology. 6: 46-58, 1996.

- 37. Young, S. D., Marshall, R. S., and Hill, R. P. Hypoxia induces DNA overreplication and

enhances metastatic potential of murine tumor cells., Proc. NatI. Acad. Sci. USA. 85:

9533-9537, 1988.

66

Jang, A. and Hill, R. P. An examination of the effects of hypoxia, acidosis, and glucose

starvation on the expression of metastasis-associated genes in murine tumor cells., Clin.

Exp. Metastasis. 15: 469-483, 1997.

Reynolds, T. Y., Rockwell, S., and Glazer, P. M. Genetic instability induced by the tumor

microenvironment, Cancer Res. 56: 5754-5757, 1996.

Hendriksson, M. and Luscher, B. Proteins of the Myc network: essentiaI regulators of

cell growth and differentiation, Advances in Cancer Research. 68: 109- 182, 1996.

Alarcon, R. M., Rupnow, B. A., Graeber, T. G., b o x , S. J., and Giaccia, A. J.

Modulation of c-myc activity and apoptosis in vivo, Cancer Res. 56: 43 15-43 19, 1996.

Franko, A. J., Koch, C. J., Garrecht, B. M., Sharplin, J., and Hughes, D. Oxygen

dependence of binding of misonidazole to rodent and human tumors in vitro, Cancer Res.

47: 5367-5376, 1987.

Brizel, D. M., Scully, S. P., Harrelson, J. M., Layfield, L. J., Bean, J. M., Prosnitz, L. R.,

and Dewhirst, M. W. Tumor oxygenation predicts for the likelihood of distant metastases

in human soft tissue sarcoma, Cancer Res. 56: 941-943, 1996.

Folkman, J. Tumor angiogenesis: a possible control point in tumor growth. peview] [22

refs], Ann. Intern. Med. 82: 96-100, 1975.

Senger, D. R., Galli, S. J., Dvorak, A. M., Permzzi, C. A*, Harvey, V. S., and Dvorak,

H. F. Tumor cells secrete a vascular permeability factor that promotes accumulation of

ascites fluid, Science. 219: 983-5, 1983.

67

Connolly, D. T., Olander, J. V., Heuvelman, D., Nelson, R., Monsell, R., Siegel, N.,

Haymore, B. L., Leimgruber, R., and Feder, I. Human vascular permeability factor.

Isolation from U937 cells, J. Biol. Chem. 264: 20017-24, 1989.

Thomas, K. A. Vascular Endothelial Growth Factor, a Potent and Selective Angiogenic

Agent, J. Biol. Chem. 271: 603-606, 1996.

Tischer, E., MitcheII, R., Hartman, T., Silva, M., Gospodarowicz, D., Fiddes, J. C., and

Abraham, J. A. The human gene for vascular endothelial growth factor, J. Biol. Chem.

266: 11947-11954, 1991.

Houck, K. A., Ferrara, N., Winer, J., Cachianes, G., Li, EL, and Leung, D. W. The

vascular endothelial growth factor family: identification of a fourth molecular species and

characterization of alternative splicing of RNA, Mol. Endocrinol. 5: 1806- 18 t 4, 199 1.

Levy, A. P., Levy, N. S., Loscalzo, J., Calderone, A., Takahashi, N., Yeo, K.-T., Koren,

G., Colucci, W. S., and Goldberg, M. A. Regulation of Vascular Endothelial Growth

Factor in Cardiac Myocytes, Circulation Research. 76: 758-766, 1 995.

Neufeld, G., Cohen, T., Gitay-Goren, H., Poltorak, Z., TessIer, S., Sharon, R.,

Gengrinovitch, S., and Levi, B.-2. Similarities and differences between the vascular

endothelial growth factor (VEGF) splice variants, Cancer and Metastasis Reviews. 15:

153-158, 1996.

Ferrara, N. The role of vascular endothelial growth factor in pathological angiogenesis,

Breast Cancer Res. & Treat. 36: 127-137, 1995.

68

Dvorak, H. F., Brown, L. F., Detmar, M., and Dvorak, A. M. Vascular permeability

factor/vascular endothelial growth factor, microvascular hyperpermeability, and

angiogenesis, Am. J. Pathol. 146: 102% 1039, 1995.

Alon, T., Hemo, I., Itin, A., Pe'er, J., Stone, J., and Keshet, E. Vascular endothelial

growth factor acts as a survival factor for newly formed retinal vessels and has

implications for retinopathy of prematurity, Nat. Med. 1: 1024-1028, 1995.

Watanabe, Y., Lee, S. W., Detmar, M., and Dvorak, A. M. Vascular permeability

factor/vascular endothelial growth factor (VPFNEGF) delays and induces escape £?om

senescence in human dermal microvascular endotherial cells, Oncogene. 59: 100-1 15,

1997.

Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kleckens, L., Gertsenstein, M.,

Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhart, C., Declercq, C., Pawling, J., Moons,

L., Collen, D., Risau, W., and Nagy, A. Abnormal blood vessel development and lethality

in embryos lacking a single VEGF allele, Nature. 380: 435-439, 1996.

Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O'Shea, K. S.,

Powell-Braxton, L., HiIlan, K. J., and Moore, M. W. Heterozygous embryonic lethality

induced by targeted inactivation of the VEGF gene, Nature. 340: 439-442, 1996.

Terman, B. I. and Dougher-Vemazen, M. Biological properties of VEGFNPF receptors.

Review1 f42 refsl. Cancer & Metastasis Reviews. 15: 159-63, 1996.

69

59. Katoh, O., Tauchi, H., Kawaishi, K., Kimura, A., and Satow, Y. Expression of the

vascular endothelial growth factor (VEGF) receptor gene, KDR, in hematopoietic cells

and inhibitory effect of VEGF on apoptotic cell death caused by ionizing radiation,

Cancer Research. 55: 5687-92, 1995.

60. Itakura, J. and Korc, M. Activation of mitogenic signaling by VEGF in Capan-1 human

pancreatic cancer cells., Proceedings of the 89th Annual Meeting of the AACR. 39:, 1998.

6 1. Gerber, H.-P., Condorelli, F., Park, J., and Ferrara, N. Differential transcriptional

regulation of the &vo vascular endothelial growth factor receptor genes., J. Bio. Chem.

272: 23659-23667, 1997.

62. Guo, D., Jia, Q., Song, H. Y., Warren, R. S., and Donner, D. B. Vascular endothelial cell

growth factor promotes tyrosine phosphorylation of mediators of signal transduction that

contain SH2 domains. Association with endothelial cell proliferation, J. Biol. Chem. 270:

6729-33, 1995.

6 3 . Rak, J., Filmus, J., Finkenzeller, G., Grugel, S., Marme, D., and Kerbel, R. S. Oncogenes

as inducers of tumor angiogenesis, Cancer Metas. Rev. 14: 263-277, 1995.

64. Rak, J., Mitsuhashi, Y., Bayko, L., Filmus, J., Shirasawa, S., Sasazuki, T., and Kerbel,

R. S. Mutant ras Oncogenes Upregulate VEGFNPF Expression: Implications for

Induction and Inhibition of Tumor Angiogenesis, Cancer Res. 55: 4575-4580, 1995.

65. Mukhapadhyay, D., Tsiokas, L., Zhou, X.-M. Z., Foster, D., Brugge, J. S., and Sukhatme,

V. P. Hypoxic induction of human vascular endothelial growth factor expression through

c-Src activation, Nature. 3 75: 577-58 1, 1995.

70

66. Mukhopadhyay, D., Tsiokas, L., and Sukhatme, V. P. Wild-Type p53 and v-src Exert

Opposing Influences on Human Vascular Endothelial Growth Factor Gene Expression,

Cancer Research. 55: 6 16 1-6 165, 1995.

67. Jiang, B.-H., Agani, F., Passaniti, A., and Semenza, G. L. V-SRC induces expression of

hypoxia-inducible factor 1 (HE- 1 ) and transcription of genes encoding vascular

endothelial growth factor and enolase I : involvement of HIF-1 in tumor progression.,

Cancer Res. 57: 5328-5335, 1997.

68. Agani, F., Kirsh, D. G., Friedman, S. L., Kastan, M. B., and Semenza, G. L. p53 does not

repress hypoxia-induced transcription of the vascular endothelial growth factor gene.,

Cancer Res. 57: 4474-4477, 1997.

69. Forsythe, J. A., Jiang, B.-H., Iyer, N. V., Agani, F., Leung, S. W., Koos, R. D., and

Semenza, G. L. Activation of vascular endothelial growth factor gene transcription by

hypoxia-inducible factor 1, Mol. Cell. Biol. 16: 4604-46 13, 1996.

70. Goldberg, M. A. and Schneider, T. J. Similarities between the oxygen-sensing

mechanisms regulating the expression of vascular endothelial growth factor and

erythropoietin, J. Biol. Chem. 269:, 1994.

71. Jiang, B.-H., Zheng, J. Z., Leung, S. W., Roe, R., and Semenza, G. L. Transactivation and

inhibitory domains of hypoxia-inducible factora, J. BioI. Chem. 272: 19253-19260, 1997.

72. Salceda, S. and Caro, J. Hypoxia-inducible factor la (HE-la) protein is rapidly degraded

by the ubiquitin-proteasome system under normoxic conditions., J. Biol. Chem.

22642-22647, 1997.

71

Levy, A. P., Levy, N. S., and Goldberg, M. A. Post-transcriptional regulation of vascular

endothelial growth factor by hypoxia, J. Biol. Chem. 271: 2746-2753, 1996.

Levy, N. S., Chung, S., Furneawc, H., and Levy, A. P. Hypoxic stabilization of vascular

endothelial growth factor mRNA by the RNA-binding protein HuR, J. BioZ. Chem. 273:

6417-6423, 1998,

Damert, A., Machein, M., Breier, G., Fujita, M. Q., Hanahan, D., Risau, W., and Plate,

K. £3. Up-regulation of vascular endothelial growth factor expression in a rat glioma is

conferred by two distinct hypoxia-driven mechanisms., Cancer Res. 57: 3 860-3 864, 1997.

Tuder, R. M., Flook, B. E., and Voelkel, N. F. Increased Gene Expression for VEGF and

the VEGF Receptors KDWFZk and FZt in Lungs Exposed to Acute or to Chronic Hypoxia,

J. Clin. Invest. 95: 1798-1 807, 1995.

Folkman, J. Angiogenesis in cancer, vascular, rheumatoid and other disease, Nature Med.

1: 27-3 1, 1995.

Shweiki, D., Itin, A., Soffer, D., and Keshet, E. Vascular endothelial growth factor

induced by hypoxia may mediate hypoxia-ioitiated angiogenesis, Nature. 359: 843-848,

1992.

Minchenko, A., Bauer, T., Salceda, S., and Caro, J. Hypoxic Stimulation of Vascular

Endotheliai Growth Factor Expression In vitro and In vivo, Lab. Invest. 71: 374-379,

1994.

Shweiki, D., Neeman, M., Itin, A., and Keshet, E. Induction of vascular endothelial

growth factor expression by hypoxia and by glucose deficiency in multicell spheroids:

Implications for tumor andiogenesis, Proc. Natl. Acad. Sci. USA. 92: 768-772, 1995.

8 1. Xie, K., Huang, S., Xu, L., and Fidler, I. J. Molecular mechanisms for the regulation of

vascular endothelial growth factor expression by extracelluIar and intracellular pH.,

Proceedings of the 89th annual meeting of American Association for Cancer Research.

39: 378, 1998.

82. Berger, D. P., Herbstritt, L., Dengler, W. A., Marme, D., Mertelsmann, R., and Fiebig,

H. H. Vascular endothelial growth factor (VEGF) mRNA expression in human tumor

models of different histologies, Ann. Oncol. 6: 8 17-825, 1995.

83. Toi, M., Inada, K., Sumki, H., and Tominaga, T. Tumor angiogenesis in breast cancer:

its importance as a prognostic indicator and the association with vascular endothelial

growth factor expression, Breast Cancer Res Treat. 36: 193-204, 1995.

84. Brown, L. F-, Berse, B., Jackman, R. W., Tognazzi, K., Guidi, A. J., Dvorak, H. F.,

Senger, D. R., Connolly, J. L., and Schnitt, S. J. Expression of vascular permeability

factor (vascular endothelial growth factor) and its receptors in breast cancer, Human

Pathology. 26: 86-9 1, 1995.

85. Takahashi, Y., Kitadai, Y., Bucana, C. D., Cleary, K. R., and ElIis, L. M. Expression of

vascular endothelial gowth factor and its receptor, KDR, correlates with vascularity,

metastasis, and proliferation of human colon cancer, Cancer Res. 55: 3964-3968, 1995.

- 86. Danielsen, T. and Rofstad, E. K. Hypoxia-induced metastasis of human melanoma cells:

involvement of vascular endothelial growth factor-mediated angiogenesis., Forty-sixth

Annual Meeting of the Radiation Research Society 168, 1998.

73

87. Goldman, C. K., Kendall, R. L., Cabrera, G., Soroceanu, L., Heike, Y., GilIespie, G . Y. ,

Siegal, G. P., Mao, X., Ben, A. J., Huckle, W. R., Thomas, K. A., and Curiel, D. T.

Paracrine expression of a native soluble vascular endothelial growth factor receptor

inhibits tumor growth, metastasis, and mortality rate., Proc. Natl. Acad. Sci. USA. 95:

8795-8800, 1998.

88. Weidner, N., Semple, J. P., Welch, W., and Folkman, I. Tumor angiogenesis and

metastasis--correlation in invasive breast carcinoma, N Engl J Med. 324: 1-8, 199 1.

89. Barbareschi, M., Gasparini, G., Morelli, L., Forti, S., and Dalla Palma, P. Novel methods

for the determination of the angiogenic activity of human tumors, Breast Cancer Res

Treat. 36: 181-192, 1995.

90. Weidner, N. Current pathologic methods for measuring intratumoral microvessel density

within breast carcinoma and other solid tumors, Breast Cancer Res Treat. 36: 169-180,

1995.

9 1. Endrich, B. and Vaupel, P. The Role of the microcirculation in the treatment of malignant

tumors: facts and fiction, In: M. Molls, P. Vaupel (Eds), Blood Perfusion and

Microenvironment of Human Tumors. Springer- Verlag: 19-39, 1 998.

92. Craft, P. S. and Harris, A. L. Clinical prognostic significance of turnour angiogenesis,

Ann. Oncol. 5: 305-3 11, 1994.

93. Schlenger, K., Hockel, M., Mitze, M., Schaffer, U., Weikel, W., Knapstein, P. G., and

Lambert, A. Tumor vascularity-a novel prognostic factor in advanced cervicaI carcinoma,

Gynecol. Oncol. 59: 57-66, 1995.

74

94. Awwad, H. K., el Naggar, M., Mocktar, N., and Barsoum, M. Intercapillary distance

measurement as an indicator of hypoxia in carcinoma of the cervix uteri, Int. J. Radiat.

Oncol. Bio1. Phys. 12: 1329-1333, 1986.

95. Fmtini, G., Vignati, S., Lucchi, M., Mussi, A., Calcinai, A., Boldrini, L,, Chine, S.,

Silvestri, V., Angeletti, C. A., Basolo, F., and Bevilacqua, G. Neoangiogenesis and p53

protein in Iung cancer: their prognostic role and their relation with vascuIar endothelial

growth factor (VEGF) expression, Br J Cancer. 75: 1295-130 1, 1997.

96. Hawighorst, H., Knapstein, P. G., Knopp, M. B., Weikei, W., Brix, G., Zuna, I.,

Schonberg, S. O., Essig, M., VaupeI, P., and Kaick, G. v. Uterine cervical carcinoma:

comparison of standard and pharmacokinetic analysis of time-intensity curves for

assessment of tumor angiogenesis and patient survival, Cancer Res. 58: 3598-3602, 1998.

97. Guidi, A. J., Abu-Jawdeh, G., Berse, B., Jackman, R. W., Tognazzi, K., Dvorak, H. F.,

and Brown, L. F. Vascular permeability factor (vascular endothelial growth factor)

expression and angiogenesis in cervical neoplasia, J. NatI. Cancer Inst. 87: 1237-1244,

1995.

98. Takahashi, A., Sasaki, H., Kim, S. J., Tobisu, K.-i., Kakizoe, T., Tsukamoto, T.,

Kumamoto, Y., Sugirnura, T., and Terada, M. Markedly increased amounts of messenger

- RNAs for vascular endothelial growth factor and placenta growth factor in renal cell

carcinoma associated with angiogenesis, Cancer Res. 54: 423 3-423 7, 1994.

99. Kirkpatrick, C. J., Wagner, M., HErmanns, I., Klein, C. L., Kohler, H., Otto, M., van

Kotten, T. G., and Bittinger, F. Physiology and cell biology of the endothelium: a

dynamic interface for cell communication, Int J Microcirc. 17: 23 1-240, 1997.

75

100. hela-Arispe, M. L. and Dvorak, H. F. Angiogenesis: a dynamic balance of stimulators

and inhibitors, Thrombosis and Haemostasis. 78: 672-677, 1 997.

101. Dedhar, S., Gregory E. Hannigan, Rak, J., and Kerbel, R. S. The extracellular

environment and cancer, in: I.F. Tannock and R.P. Hill (eds), The Basic Science of

Oncology. 3 ed. McGraw-Hill: 208-2 15, 1998.

102. O'Reilly, M. S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R. A., Moses, M., Lane,

W. S., Cao, Y., Sage, E. H., and Folkman, J. Angiostatin: A Novel Angiogenesis Inhibitor

That Mediates the Suppression of Metastases by a Lewis Lung Carcinoma, Cell. 79:

315-328, 1994.

103. O'Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E.,

Birkhead, J. R., Olsen, B. R., and Folkman, J. Endostatin: an endogenous inhibitor of

angiogenesis and tumor growth, Cell. 88: 277-285, 1997.

104. Boag, J. W. Cell respiration as a function of oxygen tension, Int. J. Radiat. Biol. Relat.

Stud. Phys. Chem. Med. 18: 475-8, 1970.

105. Waleh, N. S., Brody, M. D-, Knapp, M. A*, Mendonca, H. L,, Lord, E. M., Koch, C. J.,

Laderoute, K. R., and Sutherland, R. M. Mapping of the vascular endothelial growth

factor-producing hypoxic cell in multicellular tumor spheroids using a hypoxia-specific

marker, Cancer Res. 55: 6222-6226, 1995.

106. Hlatky, L., Tsionou, C., Hahnfeldt, P., and Coleman, C. N. Mammary fibroblasts may

influence breast tumor angiogenesis via hypoxia-induced vascular endothelial growth

factor up-regulation and protein expression, Cancer Res. 54: 60 83 -608 6 , 1 994.

76

Leith, J. T. and Michelson, S. Secretion rates and levels of vascular endothelial growth

factor in clones of HCT-8 human colon tumour cells as a function of oxygen

concentration., Cell Prolif. 28: 415-430, 1995.

Koch, C. J. A thin-film culturing technique allowing rapid gas-liquid equilibration (6 sec)

with no toxicity to mammalian cells., Radiat. Res. 97: 434-442, 1984.

Brown, J. M. and Giaccia, A. J. The unique physiology of solid tumors: opportunities

(and problems) for cancer therapy, Cancer Res. 58: 1408-141 6, 1998.

Kim, K. J., Li, B., Winer, J., Annanini, M., Gillett, N., Phillips, H. S., and Ferrara, N.

Inhibition of vascuiar endothelial growth factor-induced angiogenesis suppresses tumour

growth in vivo, Nature. 362: 841-844, 1993.

Kondo, S., Asano, M., and Suzuki, H. Significance of vascular endothelial growth

factor/vascular permeability factor for solid tumor growth, and its inhibition by the

antibody, Biochem. Biophys. Res. Commun. 194: 1234- 124 1, 1993.

Whillans, D. W. and Rauth, A. M. An experimental and analytical study of oxygen

depletion in stirred cell suspensions, Radiat Res. 84: 97-1 14, 1980.

Marshall, R. S., Koch, C. J., and Rauth, A. M. Measurement of low levels of oxygen and

their effect on respiration in cell suspensions maintained in an open system, Radiat Res.

108: 91-101, 1986.

Zim, K., DiMaio, D., and Maniatis, T. Identification of two distinct regulatory regions

adjacent to the human B-interferon gene, Cell. 34: 865-879, 1983.

Laborda, J. 3684 cDNA used as an estradiol-independent mRNA control is the cDNA for

human acidic ribosomal phosphoprotein PO, NucI. Acids Res. 19: 3998, 199 1.

77

116. Stein, I., Neeman, M., Shweiki, D., Itin, A., and Keshet, E. Stabilization of vascular

endothelial growth factor mRNA by hypoxia and hypoglycemia and coregulation with

other ischemia-induced genes, Mol. Cell. Biol. IS: 5363-53 68, 1995.

117. Helmlinger, G. and Jain, R. Accepted for publication, Proc. Natl. Acad. Sci. USA, 1998.

1 18. Vukovic, V., Hill, R. P., Rauth, A. M., and HedIey, D. W. Effects of BSO on NPSH

profiles, DNA single strand breaks and cell survival in ME180 and SiHa cells irradiated

at intermediate oxygen concentrations, Radiat. Res. (submitted), 1998.

119. Koch, C. J., Evans, S. M., and Lord, E. m. Oxygen dependence of cellular uptake of EF5

[2-(nitro- 1 H-imidazo le- 1 -yl)-N-(2,2,3,3,3 -pentafluoropropyI)acetamide] : analysis of drug

adducts by fluorescent antibodies vs bound radioactivity, Br. J. Cancer. 72: 869-874,

1995.

120. Slaven, P., Heikkila, P., and loensuu, H. Enhanced expression of vascular endothelial

growth factor in metastatic melanoma, Br. 3. Cancer. 76: 930-934, 1997.

121. Claffey, K. P., Brown, L. F., del Aguila, L. F., Tognaui, K., Yeo, K.-T., Manseau, E. I.,

and Dvorak, H. F. Expression of vascular permeability factor/vascular endothelial growth

factor by melanoma cells increases tumor growth, angiogenesis, and experimental

metastasis, Cancer Res. 56: 172- 18 1, 1996.

- 122. Laderoute, K. R., Mendoca, H. L., Calaoagan, J. M., Knapp, A. M., Giaccia, A. J., and

Sutherland, R. M. Induction and regulation of SAP- pathways in hypoxic mouse and

human transformed cells., Proceedings of the 89th Annual Meeting of the AACR. 39:

408, 1998.