Modulation of tumor angiogenesis through the use of antisense oligodeoxynucleotodes targeted to VEGF and BFGF.

MISSING IMAGE

Material Information

Title:
Modulation of tumor angiogenesis through the use of antisense oligodeoxynucleotodes targeted to VEGF and BFGF.
Physical Description:
vii, 171 leaves : ill. ; 29 cm.
Language:
English
Creator:
Shi, Wenyin, 1974-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Endothelial Growth Factors -- genetics   ( mesh )
Endothelial Growth Factors -- physiology   ( mesh )
Fibroblast Growth Factors -- genetics   ( mesh )
Fibroblast Growth Factors -- physiology   ( mesh )
Angiogenesis Inducing Agents -- genetics   ( mesh )
Angiogenesis Inducing Agents -- physiology   ( mesh )
Carcinoma, Renal Cell -- genetics   ( mesh )
Carcinoma, Renal Cell -- therapy   ( mesh )
Oligodeoxyribonucleotides, Antisense -- genetics   ( mesh )
Oligodeoxyribonucleotides, Antisense -- therapeutic use   ( mesh )
Department of Pharmacology and Therapeutics thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Pharmacology and Therapeutics -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2002.
Bibliography:
Bibliography: leaves 135-170.
Statement of Responsibility:
by Wenyin Shi.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002829474
oclc - 50681577
notis - ANV9844
System ID:
AA00011835:00001


This item is only available as the following downloads:


Full Text











MODULATION OF TUMOR ANGIOGENESIS THROUGH THE USE OF
ANTISENSE OLIGODEOXYNUCLEOTIDES TARGETED TO VEGF AND BFGF












By

WENYIN SHI













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


2002





























Dedicated to
my parents, Xuehui and Ying
my wife, Weiwen
and
my daughter, Julia.















ACKNOWLEDGMENTS


I would like to express my sincere gratitude to Dr. Dietmar W. Siemann for

providing me the very precious opportunity to work in a wonderful laboratory, and for his

immeasurable support and encouragement.

I would also like to extend my gratitude to the members of my supervisory

committee, Dr. Ian Phillips, Dr. Steven Sugrue, and Dr. Edwin Meyer, for their valuable

advice and continuous encouragement in the completion of my studies. I also would like

to thank Dr. Clare Yuan Zhang for her kind help and assistance in establishing the

studies. I also would like to express my gratitude to Dr. Jeffrey Hughes, Dr. Fuxing Tang

for assistance in liposome preparation and to Neal Benson for his help with FACS

analysis.

In addition, I would like to express my appreciation to the past and present

members of Dr. Siemann's laboratory, including Dr. Lingyun Li, Sharon Lepler, Chris

Pampo, Dr. Gustavo Cabrera, Dr. Kenneth Warrington, Jr., Howard Salmon, Heather

Newlin, Emma Mercer, and Destry Taylor, for their help and providing a pleasant

working environment. I also greatly appreciate all the faculty, staff and student groups at

the Department of Pharmacology and Therapeutics for the countless help they have

rendered and the stimulating intellectual atmosphere they provided.















TABLE OF CONTENTS
page

ACKNOW LEDGM ENTS............................................................ ..........................iii

A B ST RA C T ............................................................................ ..................................... vi

CHAPTERS

1 IN TRO D U CTIO N .............................. ..................................... .......................... 1

Tum or A ngiogenesis.................................................................... ............................. 1
Antiangiogenesis Targets in the Treatment of Cancer................................................ 4
Antisense Oligodeoxynucleotides Technology............................................... 16
Renal Cell Carcinom a ............................ .................... .............. .......... 19
Significance .......................................................................... ...................................22

2 CELLULAR DELIVERY OF ANTISENSE OLIGODEOXYNUCLEOTIDES......... 28

Introduction ............................................................. 28
M materials and M ethods..................................... ................................................. 30
R results ........................................................................ ..................... 34
D discussion ............................................................................ .................................. 36

3 AS-ODNS DESIGN AND IN VITRO ASSESSMENT............................................. 49

Introduction .................. .................... ................................................ 49
M materials and M ethods................................... ....................................................... 52
Results ................................................. ........................ ...................... 57
Discussion ........................................ ...................... ........................ 60

4 ANTI-ANGIOGENIC EFFICACY STUDIES......................... ......... ............. 79

Introduction ........................................ ....................... ....................... 79
Materials and Methods.................. .......................................................... 81
Results ................................................. ....................... ...................... 85
Discussion ........................................... ........................ ...................... 87









5 EFFICACY OF AS-ODNS TREATMENT IN CAKI-I XENOGRAFTS ................. 97

Introduction ............................ .... ..................................................................... 97
M materials and M ethods......................................... ............................................. 99
Results ............................................... ....................... ....................... 102
Discussion ........................................ ...................... ........................ 104

6 COMBINATION STUDIES................................................. .. 114

Introduction ................................... ........................................................................ 114
M material and M ethods............................................................ .......................... 116
Results ............................................. ....................... ....................... 118
Discussion ................................................................ ......................... 120

7 SUMMARY AND PERSPECTIVE............................. ................ ......... 130

R EFER EN C ES........................................................ .................................................. 135

BIOGRAPHICAL SKETCH ......................................................................... 171
































v















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MODULATION OF TUMOR ANGIOGENESIS THROUGH THE USE OF
ANTISENSE OLIGODEOXYNUCLEOTIDES TARGETED TO VEGF AND BFGF
By

Wenyin Shi

August 2002

Chair: Dietmar W. Siemann
Department: Pharmacology and Therapeutics

Angiogenesis is critical for the growth and metastatic spread of solid tumors, It is

tightly controlled by specific regulatory factors. Vascular endothelial growth factor

(VEGF) and basic fibroblast growth factor (bFGF) have been implicated as the key

factors in tumor angiogenesis. The present studies were undertaken to evaluate the effects

of blocking VEGF/bFGF production by antisense phosphorothioate

oligodeoxynucleotides (AS-ODNs) on the angiogenic activity and growth of a preclinical

model of renal cell carcinoma (Caki-1).

Efficient deliveries of AS-ODNs were achieved using cationic liposome

(DOTAP:DOPE) based delivery systems both in vitro and in vivo.

AS-ODNs sequences against VEGF and bFGF have been designed and their

efficacies were tested in vitro. Effective AS-ODNs against VEGF (V515) and bFGF

(B460) were identified. Treatment of Caki-l cells with V515 or B460 led to a reduction

in VEGF or bFGF expression levels sufficient to impair the proliferation and migration









potential of co-cultured endothelial cells. The observed effects were AS-ODNs sequence

specific, dose dependent and were achieved at a low, non-toxic dose. The treatment of

Caki-1 cells with V515 or B460 was also sufficient to impairthe Caki-l tumor cell

induced angiogenesis in vivo. When V515 or B460 treated Caki-l cells were injected into

nude mice and evaluated for their angiogenic potential, the number of vessels initiated

were significantly reduced.

To test antitumor efficacy of VEGF/bFGF AS-ODNs treatment, V515 and B460

were administrated to Caki-l xenograft tumor bearing mice. The results showed that

systemic administration of VEGF/bFGF AS-ODNs significantly inhibited the growth of

Caki-1 tumors. More importantly, a better response was observed when these two AS-

ODNs treatments were combined. A combination of VEGF/bFGF AS-ODNs treatment

with VEGF/bFGF receptor inhibitor or single dose local radiation also showed enhanced

tumor responses when compared to single treatment alone.

These results indicate that AS-ODNs against pro-angiogenic factors VEGF and

bFGF may have great utilities in the treatment of renal cell carcinoma either alone or in

combination with other anti-cancer therapies.













CHAPTER 1
INTRODUCTION

Cancer is a group of diseases characterized by uncontrolled growth and spread of

abnormal cells. If the spread is not controlled, it can result in death. In spite of ever

increasing efforts to understand its process and improve treatment, its incidence in the

population is rising. In the US, about 1 in 2 men and 1 in 3 women will develop cancer

in their lifetime (American cancer society, 2002). As a cause of mortality overall in the

Western World, cancer is second only to cardiovascular disease. Cancer can be treated by

surgical removal or destroyed with toxic chemicals or radiation. However, these

approaches all have drawbacks. Surgery will work for many primary tumors, but

metastases are difficult to identify let alone remove at an early stage. Radiation and

chemotherapy are generally toxic to normal cells as well. If even a few cancerous cells

remain, they can proliferate to produce a resurgence of the disease; moreover, unlike the

normal cells, cancer cells are genetically dynamic and may evolve resistance to the

chemicals used against them. New therapeutic approaches providing more tumor-specific

targeting still need to be exploited. One promising new development of cancer therapy is

the anti-angiogenic strategies, following the recognition that tumor growth and metastasis

depend on establishment of new blood vasculature (Folkman, 1971; Folkman, 1972b).

Tumor Angiogenesis

Angiogenesis is the formation of new blood vessels out of pre-existing capillaries.

It is a sequence of events that is of key importance in a broad array of physiologic and

pathologic processes (Folkman and Shing, 1992). While it plays a key role in








development, in adults, it is a rare event under normal physiological circumstances,

occurring almost exclusively in the female reproductive system. Under normal conditions

such as wound healing, the angiogenic process switches on and then off at the appropriate

times indicating tight regulation of stimulatory and inhibitory factors (Hanahan and

Folkman, 1996). However, angiogenesis can be activated with a variety of pathological

conditions and occur in a less controlled manner (O'Reilly, 1997). These including

cardiovascular diseases (atherosclerosis), rheumatoid arthritis, diabetic retinopathy,

psoriasis, etc (Folkman, 2001). In addition, angiogenesis is critical in the growth and

metastatic dissemination of cancer (Folkman, 1995; Folkman, 1992; Folkman, 1972a;

Folkman and Shing, 1992).

It has been observed for more than one hundred years that tumors appear to be

more vascular than normal tissues. It was not until in the early 1970s that Drs Folkman

and Denekamp put forward the idea that tumors are highly vascularized and thereby

vulnerable at the level of their blood supply. This is the initial recognition of

angiogenesis being a therapeutically interesting process in the area of oncology. Folkman

proposed the hypothesis that angiogenesis was a requirement for the growth and

metastatic spread of solid tumors (Figure 1-1) (Folkman, 1971; Folkman, 1972a). He

further hypothesized that solid tumors could only grow to a size of-1-2 mm in diameter

without developing new blood supply, and if the development of vascular supply could

be prevented, tumor growth could be limited to a small size (Folkman, 1971; Folkman,

1972a). This hypothesis implied that by destroying the newly developing vessels of the

tumor, all the tumor cells supported by these vessels could also be killed (Denekamp J,

1972).









The process of angiogenesis consists of multiple, sequential, and interdependent

steps (Figure 1-2). It begins with local degradation of the basement membrane

surrounding capillaries, which is followed by the invasion of the surrounding stroma by

underlying endothelial cells in the direction of the angiogenic stimulus. Endothelial cell

migration is accompanied by the proliferation of endothelial cells and their organization

into three-dimensional structures that join with other similar structures to form a network

of new blood vessels (Figure 1-2) (Auerbach and Auerbach, 1994).

The process of angiogenesis is mediated by the balance between pro-angiogenic

and anti-angiogenic factors. Angiogenesis is rapidly initiated in response to hypoxic or

ischemic conditions. In all types of angiogenesis, under either physiologic or pathologic

conditions, endothelial cell activation seems to be the first process to take place. In

tumors, angiogenesis begins by mutual stimulation between tumor cells and endothelial

cells by paracrine mechanisms (Gasparini, 1999). Cytokines from various sources

including tumor and stromal cells are released in response to hypoxia or ischemia It is

suggested that vascular endothelial growth factor (VEGF) is a major player in

angiogenesis initiation (Ziche et al., 1997). Besides affecting vasodilation and vascular

permeability, VEGF can induce the expression ofproteases and receptors important in

cellular invasion and tissue remodeling and is able to prevent endothelial cell apoptosis

(Ferrara and Keyt, 1997; Gupta et al., 1999). Following the releasing of pro-angiogeneic

factors, endothelial cells can release proteolytic enzymes (matrix metalloproteinases,

MMPs) to degrade the extracellular matrix for migration, proliferation, and endothelial

penetration into new areas of the body (Stetler-Stevenson, 1999).









Endothelial cell proliferation and migration is stimulated by pro-angiogenic

growth factors, like VEGF and bFGF. VEGF and bFGF are direct acting pro-angiogenic

growth factors. BFGF exists in both low molecular weight form and high molecular

weight forms due to alternative translation (Florkiewicz et al., 1991). It is suggested that

during angiogenesis, low molecular weight bFGF binding to endothelial cell surface FGF

receptors leads to increased motility, proliferation and proteinase activity, whereas the

high molecular weight forms may act on endothelial cell proliferation after nuclear

translocation (Gleizes et al., 1995; Klein et al., 1997). VEGF, besides its effect on

angiogenesis initiation, also affects endothelial cell proliferation through high affinity

receptors (KDR/flk-1 and Fit-1) expressed on endothelial cells (Ferrara, 1999; Veikkola

and Alitalo, 1999). Finally, the neovasculature become mature and stable through the

interaction of endothelial cells with extracellular matrix and mesenchymal cells. After

endothelial cell proliferation and migration and maturation and formation of endothelial

tube structures, surrounding vessel layers composed of mural cells need to be recruited.

Endothelial cells may accomplish this via the synthesis and secretion of platelet-derived

growth factor (PDGF), a mitogen and chemo-attractant for a variety of mesenchymal

cells (George, 2001; LaRochelle et al., 2002). Subsequent differentiation of the mural

precursor cells into pericytes and smooth muscle cells is believed to be cell-cell contact

dependent process (Griffioen and Molema, 2000).

Anti-angiogenic Targets in the Treatment of Cancer

Understanding angiogenesis and its unique characteristics in tumor growth and

metastasis has provided insights to a variety of ways to interrupt the process. During the

past decade, research on anti-angiogenic agents has exploded and with ever increasing

interest in its potential (Kerbel, 2000). We now have a much clearer understanding of









tumor angiogenesis, including key regulatory factors, differences between normal and

tumor vasculature, along with endogenous inhibitors and methods to study and quantify

angiogenesis (Kerbel, 2000).

The complex process of tumor angiogenesis of tumor provides multiple potential

targets for anti-angiogenic strategies. The formation of new blood vessels involves

basement membrane degradation, endothelial cell migration, endothelial proliferation and

tube formation. Anti-angiogenic strategies under evaluation target at least one of the

several stages (Figure 1-2). These strategies vary from regulation of angiogenic factor

expression in tumors, to endogenous inhibitors of angiogenesis. Currently, there are over

80 clinical trials employing such strategies underway (http://cancertrials.nci.nih.gov/)

reflecting the high pace of development. Based on the biological activities, these

strategies can be categorized into several broad classes. The first class consists ofMMP

inhibitors, compounds that block the degradation of the basement membrane. The second

class of agents includes those designed to inhibit endothelial cell function, such as TNP-

470, thalidomide, endostatin, etc. The third class of agents specifically targets angiogenic

growth factors. It includes trysine kinase inhibitors of VEGF/bFGF, antibodies or AS-

ODNs against pro-angiogenic growth factors or their receptors. The last class of agents

target survival factors ofneovascular blood supply, such as intergrin antagonists, or anti-

VEGF therapy (Reinmuth et al., 2001; Fidler et al., 2000).

I will now review some of the important angiogenic factors that have potential as

therapeutic targets in anti-angiogenic therapies.









VEGF-A and Its Receptors

VEGF-A and its receptor system is among the most substantial mediators of

angiogenesis. Considerable evidence has accumulated indicating that VEGF is an

angiogenic cytokine of central importance. Its angiogenic activities have been

demonstrated in numerous experimental models (Takeshita et al., 1994; Wilting et al.,

1992; Wilting et al., 1993; Potgens et al., 1995; Kondo et al., 1995). The central role of

VEGF in tumor angiogenesis has also been suggested. Over-expression of VEGF has

been reported to occur in the vast majority of clinically important cancers (Zhu and

Witte, 1999; Hemmerlein et al., 2001; Ferrara and Keyt, 1997). High serum and urine

levels of VEGF have been associated with poor survival and treatment outcome to

patients of different cancers (Hemmerlein et al., 2001; Maeda et al., 1996; Gasparini et

al., 1997; Edgren et al., 1999; Sliutz et al., 1995). Tumor associated endothelial cells

frequently demonstrate increased expression of VEGF receptors (Zhu and Witte, 1999).

Moreover, high VEGF expression is found to associate with increased microvessel

density and increased metastatsis in cancers (Zhu and Witte, 1999; Tsuji et al., 2002;

Fontanini et al., 2002; Ng et al., 2001).

Disruption of VEGF signal transduction provides a potential effective target for

anti-angiogenic approaches. Different strategies have been designed and evaluated.

Specific VEGF antibodies are a way of stopping the angiogenic effects of this growth

factor. Systemic administration of monoclonal VEGF antibody into tumor bearing nude

mice significantly suppressed tumor growth in several tumor models (Kim et al., 1993;

Borgstrom et al., 1998). A humanized antibody, rhuMAb-VEGF also has been developed

and is undergoing Phase II clinical trials (Lin et al., 1999; Presta et al., 1997).









Alternatively, AS-ODNs or antisense RNA were also been used to directly disrupt VEGF

protein expression and lead to inhibition of tumor growth in different tumor systems

(Smyth et al., 1997; Ellis et al., 1996; Nguyen et al., 1998). Another approach is the

coupling of a toxin to VEGF itself When active parts of diphtheria toxin (DT390) are

linked to VEGF165 or VEGFI21, the chimeric molecule exerts highly selective toxic

effects on endothelial cells. It disrupts neovascularisation in the chicken chorioallantoic

membrane assay and slows down the growth of tumors in preclinical tumor models

(Arora et al., 1999).

Blocking the interaction of VEGF with it receptor and receptor signal

transduction pathway provides another option for anti-angiogenic treatment. VEGF

receptors (flt-1, flk-l and fit-4) are almost exclusively expressed on endothelial cells,

with VEGFR-2 (flk-1) believed to play a central role in VEGF signal transduction

(Ortega et al., 1999; Bematchez et al., 1999). Antibodies against VEGF receptors also

showed efficacy in inhibiting tumor growth in preclinical tumor models (Brekken et al.,

2000; Klement et al., 2000; Witte et al., 1998). Targeting VEGF receptor expression

through the use of antisense also has been shown to be effective at inhibiting tumor

angiogenesis and growth (Kamiyama et al., 2002; Marchand et al., 2002). Purified

soluble VEGFR-1 binds VEGF with high affinity and blocks VEGF induced endothelial

cell proliferation (Kendall et al., 1996). Over-expression of VEGF soluble receptor

through gene therapy significantly inhibits tumor growth and metastasis and leads to

higher survival rates (Goldman et al., 1998; Hasumi et al., 2002; Shiose et al., 2000;

Takayama et al., 2000). Recently, small molecular compounds that can inhibit VEGF

receptor tyrosine kinase activities have been developed and initial preclinical studies









showed promising anti-angiogenic and antitumor effects (Solorzano et al., 2001; Ning et

al., 2002; Hess et al., 2001). To interfere in the binding of VEGF with its receptors, novel

peptides have been developed (Fairbrother et al., 1998). This approach provides another

potentially effective means to disrupt the VEGF signal transduction pathway and a

provide treatment for cancer.

FGFs and Their Receptors

Fibroblast growth factor was originally identified as an activity in pituitary

extracts that stimulates the proliferation of Balb/c 3T3 cells (Armelin, 1973;

Gospodarowicz, 1974; Gospodarowicz, 1975). Currently, FGFs consist of a family of

over 20 structurally related proteins (Basilico and Moscatelli, 1992;Omitz and Itoh,

2001). They bind and activate high-affinity tyrosine kinases, FGFR 1-4 (Lee et al., 1989;

Dionne et al., 1990; Ruta et al., 1989; Reid et al., 1990). Among all the FGFs, FGF-2 also

called basic fibroblast growth factor (bFGF) is one most extensively investigated and

important in angiogenesis.

BFGF was originally purified from bovine pituitary gland (Esch et al., 1985). It

acts in a paracrine and autocrine manner and is released by tumor cells, macrophages or

the extracellular matrix. BFGF can stimulate endothelial proliferation and also is

chemotactic for endothelial cell migration (Gospodarowicz et al., 1987; Moscatelli et al.,

1986; Moscatelli et al., 1986). It can also up-regulate other important pro-angiogenic

factors like VEGF or plasminogen activator (Seghezzi et al., 1998; Montesano et al.,

1986). Blocking bFGF expression and function can be achieved by using vaccine,

antisense against bFGF or its receptors (Plum et al., 2000; Maret et al., 1995; Wang and









Becker, 1997; Redekop and Naus, 1995; Ensoli et al., 1994; Murphy et al., 1992; Becker

et al., 1989).

Blocking the intrinsic tyrosine kinase activity of FGF receptors is a promising

new approach in anti-angiogenic strategies targeting the bFGF signal transduction

pathway. Some experimental compounds have been found to specifically block signaling

of FGFR- I and inhibit angiogenesis (Mohammadi et al., 1998; Perollet et al., 1998).

The Tie-angiopoietin System

The tie-receptor family consists of two know endothelial tyrosine kinases: TIE

and TIE2/Tek. They are identified in vascular endothelium and hematopoietic cells

(Dumont et al., 1992; Iwama et al., 1993; Partanen et al., 1992; Schnurch and Risau,

1993). Mice lacking TIE 1 or TIE 2 are lethal (Puri et al., 1995). Ties may represent the

earliest endothelial cell lineage marker and may regulate the endothelial cell proliferation,

differentiation, and proper patterning during vasculogenesis.

The first Tie-2 ligand, Aniopoietin-1 (Ang-1) was identified from human

neuroepithelioma and mouse myoblast cell lines (Davis et al., 1996). Ang-1 is a novel

endothelial regulatory factor that has been found to promote angiogenic remodeling by

vascular supporting as well as vessel maturation and stabilization. Another related ligand,

Ang-2 also has been found. However, binding of Ang-2 to Tie-2 did not induce

phosphorylation of Tie-2 in endothelial cells. Moreover, it seems to block Ang-l activity

and suggesting Ang-2 may be antagonize the activation of Tie-2 (Maisonpierre et al.,

1997). These findings together underline the feasibility to use the Tie-angiopoietin

system to control angiogenesis (Lin et al., 1998). But before these molecules can be used

for therapy, their effects on adult human vasculature and their interaction with other









angiogenic molecules during physiological and pathological angiogenesis need to be

further investigated.

Angiogenin

Angiogenin is a 14 kD single chain basic protein found in human adenocarcinoma

cells. It is a potent inducer of angiogenesis in vivo, which functions in the picomolar

range (Vallee and Riordan, 1997; Fett et al., 1985). It stimulates the proliferation of

endothelial cells and promotes the adhesion of endothelial and tumor cells (Soncin et al.,

1994; Hu et al., 1997). Blocking angiogenin with a monoclonal antibody can impair

subcutaneous tumor growth of a colon adenocarcinoma in a dose dependent manner

(Olson et al., 1994). In 40-50% of the cases, growth of human breast carcinoma

xenografts in athymic mice could be completely inhibited by a humanized version of the

monoclonal antibody (Piccoli et al., 1998). In addition to its efficacy of inhibit tumor

growth, it may also inhibited the establishment and metastatic growth of tumor cells

(Olson et al., 2002). Besides the use of antibodies, other strategies to abolish angiogenin-

induced angiogenesis include the use of DNA aptamers or antisense (Olson et al., 2001;

Nobile et al., 1998). Anti-angiogenic therapy using angiogenin as a target may become an

important tool because angiogenin mediates angiogenesis by mechanisms distinct from

VEGF and bFGF (Lixin et al., 2001; Moroianu and Riordan, 1994)

Endogenous Inhibitors (Endostatin, Angiostatin)

Angiostatin and endostatin are two endogenous peptides that have been found to

have potent anti-angiogenic effect (O'Reilly et al., 1997; O'Reilly et al., 1994b; O'Reilly

et al., 1994a). Angiostatin is a prolytic fragment of plasminogen and endostatin is a 20

kD fragment of collagen XVIII. These factors make endothelial cells resistant to









angiogenic stimuli and induce "dormancy" of metastases. Administration of the

recombinant protein, or expressing angiostatin or endostatin by means of gene therapy,

illicits potent anti-tumor effects in the preclinical studies (O'Reilly et al., 1997; O'Reilly

et al., 1994b; O'Reilly et al., 1994a; Bertolini et al., 2000; Jin et al., 2001; Yamanaka et

al., 2001; Feldman et al., 2001; Sacco et al., 2001; Szary and Szala, 2001; Wen et al.,

2001; Shi et al., 2002). The anti-tumor activities of angiostatin and endostatin are

currently undergoing clinical evaluation.

In addition to angiostatin and endostatin, there are other endogenous angiogenic

inhibitors including restin, vasostatin, etc (Pike et al., 1998; Ramchandran et al., 1999).

Also, human prolactin, growth hormone, placental lactogen and growth hormone variant

are angiogenic factors, whereas their 16 kD N-terminal fragments are anti-angiogenic

(Struman et al., 1999).

Integrins

Integrins are heterodimeric transmembrane proteins consisting of a and p

subunits with large ectodomains and short cytoplasmic tails. They control cell motility,

differentiation and proliferation via interactions with extracellular matrix molecules.

Integrins avP3 and avPi are up-regulated on proliferating endothelial cells in angiogenic

blood vessels during wound healing as well as in tumor vasculature (Brooks et al., 1994;

Friedlander et al., 1996). The avyf integrin, an adhesion receptor for extracellular matrix

components with an exposed RGD sequence, is an attractive target for anti-angiogenic

therapy. This integrin is almost exclusively present on the cell surface of activated

endothelial cells, but absent on quiescent endothelium or other cell types (Eliceir and

Cheresh, 1999). Antibodies against avP3 were found to inhibit adhesion-dependent signal









transduction by angiogenic factors, leading to apoptosis of activated endothelial cells.

Consequently, these compounds could block endothelial tube formation and angiogenesis

in tumors (Brooks et al., 1994; Brooks et al., 1995). Currently, integrin antagonists are

being evaluated in phase I and phase II clinical trials (Brower, 1999).

Matrix Metalloproteinases (MMPs) and Tissue Inhibitor of Metalloproteinases

(TIMPs)

To form new blood vessels, endothelial cells of existing blood vessels must

degrade the underlying basement membrane and invade into the stroma of the

neighboring tissue (Mignatti and Rifkin, 1993; Mignatti and Rifkin, 1996). These

processes of endothelial cell invasion and migration require the cooperative activity of

the plasminogen activator and the MMPs.

The MMPs are a family of structurally related zinc-dependent endopeptideases

collectively capable of degrading extracellular matrix (ECM). MMPs play an important

role in the degradation of ECM, both in physiological conditions, such as morphogenesis

and tissue repair and in pathologic conditions, such as tumor invasion and metastasis. The

activity of MMPs is controlled at different levels (Liekens et al., 2001) First, the

expression of MMPs is up-regulated by angiogenic growth factors (Giuliani et al., 1999;

Bond et al., 1998; Wang and Keiser, 1998). Secondly, MMPs need to be activated

proteolytically (Murphy et al., 1999). Lastly, the MMPs activities are also regulated by

their inhibitors TIMPs (Blavier et al., 1999; Henriet et al., 1999). However, a large body

of evidence suggests that this regulation is lost during tumor growth and metastasis

(Rasmussen and McCann, 1997). An imbalance between MMPs and their TIMPs is

responsible for the invasive phenotype of breast, colon and lung tumors and a low









survival rate in urothelial cancers (Kossakowska et al., 1996; Gohji et al., 1996b; Gohji et

al., 1996a).

Inhibition of MMPs activities thus has been extensively studied as an approach to

inhibit growth and invasion of neoplastic cells. Important anigiogenesis inhibitors in

clinical trials based on MMP blocking are Metastat, Neovastat, BMS-2752291,

Mariamstat, AG3340, Bay 12-9556 and CGS 27023A (Vihinen and Kahari, 2002).

MMPs inhibitors currently in clinical trials are synthetic peptides or non-peptidic

molecules, chemically modified tetracyclines, bisphosphonates or natural MMP inhibitors

(neovastat). Further trials using MMPs in combination with classical chemotherapy are

also underway.

Plasminogen Activator (uPA) and its Inhibitors (PAl-1)

Proteases of the fibrinolytic cascade also contribute to the regulation of

angiogenesis. Expression of urokinase-type plasminogen activator (uPA) by malignant

cells results in an aggressive phenotype with increased tumor angiogenesis and metastatic

invasion. PAl-I, the natural uPA inhibitor, is paradoxically also up-regulated in human

tumor samples (Landau et al., 1994). Clinically, expression of both, uPA and PAI-I

correlate with a poor prognosis of several cancers (Rosenquist et al., 1993; Heiss et al.,

2002; Osmak et al., 2001). Taken together, inhibition of uPA rather than PAI-1 activity

might be a possible therapeutic target to treat cancer and other angiogenesis dependent

diseases.

Thrombospondin (TSP)

Extracellular matrix molecules play an important role in maintaining tissue

integrity and endothelial cell viability. However, thrombospondin, one such extracellular









matrix molecules, first identified in 1979 from platelets is also a very powerful inhibitor

of endothelial cell adhesion, migration, motility and proliferation and angiogenesis in

vivo (Lawler et al., 1978; Taraboletti et al., 1990; Good et al., 1990). TSP-1 is a member

of a family of structurally related proteins encoded by different genes, which includes 4

recent identified members, TSP 2-5 (Bomstein and Sage, 1994). TSP-I inhibits

endothelial cell proliferation, migration and can induce endothelial apoptosis (Vogel et

al., 1993; Tolsma et al., 1993). In addition, tumor cells transfected with TSPI developed

smaller tumors than the parental cell lines (Volpert et al., 1998; Streit et al., 1999). In

clinical studies, expression of TSP1 has been inversely correlated with malignant

progression of breast cancer, melanoma, and lung carcinomas (Zabrenetzky et al., 1994).

These data indicate that TSP-I can be utilized to inhibit tumor growth by an anti-

angiogenic mechanism.

Platelet Factor 4 (PF-4)

Platelet factor 4 belongs to the CXC cytokine superfamily (Strieter et al., 1995). It

is a 7.8 kD protein of 70-amino acid in length that shares homologies with p-

thromboglobulin and interleukin-8 (Deuel et al., 1977). It has been known for a while that

PF-4 inhibits angiogenesis (Maione et al., 1990). First, PF-4 inhibits endothelial cell

proliferation, migration and angiogenesis in vivo (Gupta and Singh, 1994; Maione et al.,

1990). Second, PF-4 is targeted to endothelial cells that undergo angiogenesis in vivo

(Hansell et al., 1995), Moreover, it has been shown the tumor angiogenesis could be

inhibited by PF-4 (Sharpe et al., 1990; Kolber et al., 1995). In addition, human glioma

cells infected with a secretable PF-4 cDNA grew slowly in vivo and were hypovascular









(Tanaka et al., 1997). Finally, data exist suggesting the PF-4 may counteract angiogenic

factor activity at the sties of platelet activation (Watson et al., 1994).

Administration of recombinant PF-4 protein or delivery of the PF-4 gene showed

efficacy against tumor growth in preclinical models (Maione et al., 1990; Tanaka et al.,

1997; Kolber et al., 1995; Maione et al., 1991). These findings suggest that it may have

great potential as an effective anti-angiogenic factor in the treatment of cancer. Currently,

recombinant PF-4 is being evaluated in clinical trials (Belman et al., 1996).

Interleukins

Interleukins have been known for a long time for their immunomodulatory

activities but their role in angiogenesis is just becoming a hot topic in cancer research.

Some of the interleukins have anti-angiogenic properties (interleukin-10, -12, -18) while

others seems to be pro-angiogenic (interleukin-1, -6, -8, -15), some may even have both

effects (interleukin-4). The mechanisms by which interleukins achieve their effects on

endothelial cells are quite different and not fully understood (El Awad et al., 2000; Huang

et al., 1996; Voest et al., 1995). Interleukins may be useful tools to treat angiogenesis-

related diseases including cancer. However, better understanding of their specific

functions, as well as their interactions is needed.

Anti-angiogenic Factors Summary

Among all these angiogenic regulatory factors, the most important pro-angiogenic

growth factors in cancer are VEGF and bFGF Both factors are found to stimulate

endothelial cell proliferation and migration (Leung et al., 1989; Plate et al., 1992;

Schweigerer et al., 1987). Indeed, the expression of VEGF has been related to

fundamental features of tumors, such as growth rate (Kim et al., 1993; Nagao and









Nishikawa, 1989), microvessel density (Toi et al., 1994; Straume and Akslen, 2002) and

vascular architecture (Drake and Little, 1999; Faridi et al., 2002) as well as the

development of tumor metastasis (Weidner et al., 1991; Faridi et al., 2002). A correlation

between VEGF and/or bFGF expression and survival has been noted in some cancer

patients (Gasparini et al., 1997; Yiangou et al., 1997; Dietz et al., 2000). Given the

importance of VEGF and bFGF in the angiogenic process of cancer, these two growth

factors were chosen as the targets for the presents investigations.

Antisense Oligodeoxynucleotides Technology

In order to inhibit the expression or function of specific gene products, such as

VEGF and bFGF, a strategy with promise is the use of antisense oligodeoxynucleotides

(AS-ODNs).

In 1977, it is first described that gene expression can be modified with exogenous

nucleic acids by using single strand DNA to inhibit translation of a complementary RNA

in a cell-free system (Paterson et al., 1977). Soon after, Zamecnik and Stephenson also

demonstrated that AS-ODNs targeted to 3' end of virus could inhibit viral replication in

vitro (Zamecnik and Stephenson, 1978). These initial findings showed that AS-ODNs

could inhibit gene expression in a sequence specific manner. One of the first studies

showing in vivo activities of AS-ODNs was published in 1991 (Whitesell et al., 1991).

Since then, particularly with the introduction of efficient methods for DNA sequencing

and ODNs synthesis, various targets have been analyzed in vitro and in animals with

encouraging results (Jansen et al., 1998; Tamm et al., 2001; Pawlak et al., 2000; Golden

et al., 2002; Braasch and Corey, 2002; Corey, 2002).









The essential steps in drug design are the identification of an appropriate target

responsible for a certain disease and the development of a drug with specific recognition

of and affinity to that target. For most conventional drugs the mechanism of fairly broad.

In contrast, the basis of the use of AS-ODNs is that the introduction of ODNs

complementary to target mRNA sequences into the cytoplasm can result in decreased

expression of the gene being targeted (Dias N and Stein, 2002). Since AS-ODNs inhibit

gene expression in a sequence dependent way, selective alteration of specific gene

expression is possible. The AS-ONDs approaches generally used are mainly to inhibit

oncogene expression, to induce apoptosis, to overcome multidrug resistance, or to inhibit

pro-angiogenic growth factors (Pawlak et al., 2000).

In 1998, the first antisense drug (fomivirsen) was approved by the US Food and

Drugs Administration (FDA) for the treatment of cytomegalovirus-induced retinitis in

AIDS patients (de Smet et al., 1999). Although fomivirsen is administrated locally, the

approval shows the feasibility of AS-ODNs as drugs for the treatment of human diseases.

With continuous development and understanding of tumor biology, inappropriate

expression of certain genes was found to be basic to the pathophysiology of cancer.

Consequently the use AS-ODNs as therapeutic strategies in the treatment of cancer has

attracted much attention and intensive investigation. The currently more than 8 ongoing

clinical trials illustrate the growing interest in AS-ODNs in the treatment of cancer

(Koller et al., 2000; Tamm et al., 2001). Besides these approaches, AS-ONDs may also

have a role in overcoming multidrug resistance in cancer. For example, the use of anti-

mdrl AS-ODNs has been shown to lead to reduction of the gene product gp170









expression, restore chemotherapy drug sensitivity, and even lead to eventual prolongation

of survival (Cucco and Calabretta, 1996; Kuss et al., 2002; Pan et al., 2001).

Oncogene over-expression is one of the most common molecular events that may

lead to cancer development. AS-ODNs are specific tools to inhibit expression of certain

oncogenes and so can be used as potential drugs to reverse the harmful effects of

dysregulated gene expression. AS-ODNs against ras. myc, myb, bcr-abl, as well as viral

oncogenes, such as E6, E7 or HBx have been evaluated (Gray et al., 1993; Szczylik et al.,

1996; Venturelli et al., 1990; Leonetti et al., 1996; Citro et al., 1994; Szczylik et al.,

1991; Beer-Romero et al., 1997; Lappalainen et al., 1996). Methods to regulate the

mechanisms of cell death and survival in order to shift the balance toward apoptosis are

of great interest in the treatment of cancer. Studies have focused on targeting the vital

anti-apoptotic genes, such as bcl-2, p53 and CRIPTO and MDM-2 proteins using AS-

ODNs (Ziegler et al., 1997; Campbell et al., 1998; Normanno et al., 1999; Chen et al.,

1998).

The application of AS-ODNs to target the pro-angiogenic growth factors is a new

and promising strategy in cancer management. Studies with AS-ODNs against VEGF

showed such treatment can significantly impair tumor angiogenesis and lead to tumor

growth inhibition in VEGF dependent tumors (Masood et al., 2001; Masood et al., 1997).

AS-ODNs directed at inhibiting the expression of bFGF also showed both anti-

angiogenic and anti-tumor efficacy (Wang and Becker, 1997). The strategy of using AS-

ODNs against both VEGF and bFGF as therapeutic intervention of cancer was explored

in detail in studies described in this dissertation.









To date proof of clinical efficacy of AS-ODNs in oncology is very limited (de

Smet et al., 1999). However, data providing proof of principle exist. Future development

of AS-ODNs holds considerable promise in the treatment of cancer. Further development

of new various new targets, assessment of combination treatments with several PS-

ODNs, and investigations focused on strategies targeting tumor mechanisms should

improve their therapeutic activities.

Renal Cell Carcinoma

Renal cell carcinoma (RCC) is the sixth leading cause of cancer death in the

United States, accounting for 3% of adult malignancies. There were an estimated 30,800

RCC cases diagnosed in 2001, with approximately 12,100 deaths in the United States

(Jemal et al., 2002). The incidence of RCC deaths in the United States has been steadily

increasing during the past 25 years, possibly partly because of increased sensitivity and

greater use of various imaging modalities (Chow et al., 1999; Homma et al., 1995; Jayson

and Sanders, 1998).

Although the etiology of RCC is unknown, several risk factors, including obesity,

smoking, hypertension, diuretic use, consumption of fired meat, asbestos exposure,

petroleum exposure, and frequent analgesic use have been consistently implicated (Dhote

et al., 2000). Renal transplantation, with its associated immunosuppression, acquired

cystic kidney disease also increase the risks of developing RCC (Hoshida et al., 1999).

RCCs are clinically, histologically, and genetically a very heterogeneous group of

tumors. Clear cell RCC is the most common type of RCC, accounting for over 70%/ of the

cases. It is a highly vascularised neoplasm demonstrating clear evidence of abundant

angiogenesis and abnormal blood vessel development (Yoshino et al., 2000).









Clinically, RCC patients can present with a multiplicity of manifestations ranging

from the classic presenting triad of hematuria, pain and palpable renal mass to more

obscure symptoms, such as those of paraneoplastic syndromes. Unfortunately, the classic

triad usually indicates patients with far advanced disease, and it is seen in less than 10%

of patients at presentation (Gibbons et al., 1976). More commonly, renal tumors are

discovered incidentally during the course of various diagnostic studies.

The treatment of choice for RCC is surgical removal. Radical nephrectomy is

accomplished by early ligation of the renal artery, renal vein, and en bloc removal of the

kidney with the surrounding Gerota's fascia (Robson et al., 1969). Despite the

remarkable improvements and response rates of single and combination chemotherapy in

some solid tumors, RCC remains a chemoresistant tumor. The most common agents,

vinblastine and floxiuridine, have response rates of 7%/ and 16%, respectively (Yagoda et

al., 1995). In a review of 72 agents, evaluated in 3500 patients, between 1983 and 1992,

an overall objective response rate of only 5.6% was found, mostly of short duration

(Yagoda et al., 1995). Hormone therapy has been found to be equally ineffective

(deKemion and Lindner, 1982). Consequently, there currently is no role for

chemotherapy or hormone therapy in the treatment of RCC. Unlike chemotherapy,

radiotherapy has been shown to provide some benefits to patients with RCC (Rost and

Brosig, 1977). Preoperative radiation can reduce the risk of tumor dissemination at the

time of nephrectomy; reduce primary tumor size; increase resectability and reduce tumor

vascularity. Postoperative radiation therapy has the theoretical benefit of providing local

control of tumor in patients with positive surgical margins, incompletely rejected primary

tumors, or lymph node involvement (Riches, 1966; Mantyla et al., 1977). Still, currently,









radiation therapy in RCC is generally reserved for palliation, most often for symptomatic

bony metastases (Halperin and Harisiadis, 1983).

Taken together, five-year survival rates after radical nephrectomy for stage 1 RCC

is approximately 94%, and stage II 79%. Patients with renal vein or inferior vena caval

involvement have a survival rate of 25-50%, and patients with regional lymph node

involvement or expracapsular extension have a survival rate of 12-25%. Five year

survival rate for patients with stage IV disease is less than 5%.

The unsatisfactory management of RCC with conventional anticancer therapies

warrants novel approaches to augment the tumor response and treatment outcome.

Histopathological studies of RCC reveal it to be a highly vascularised neoplasm

demonstrating clear evidence of abundant angiogenesis and abnormal blood vessel

development (Figure 1-3) (Yoshino et al., 2000). Therefore it may provide an excellent

target for anti-angiogenic therapeutic approaches. Basic fibroblast growth factor (bFGF)

and vascular endothelial growth factor (VEGF) are of particular interest. Both factors

have been shown to be expressed in renal cell carcinoma tissues and renal cell carcinoma

cell lines (Mydlo et al., 1993; Gospodarowicz et al., 1986; Mydlo et al., 1988; Sato et al.,

1999; McLaughlin and Lipworth, 2000; Ferrara and Keyt, 1997). Serum levels of VEGF

and bFGF often are elevated in RCC patients (Nguyen et al., 1994b; Fujimoto et al.,

1991; Wechsel et al., 1999; Tomisawa et al., 1999; Paradis et al., 2000) and renal cell

carcinoma VEGF and bFGF mRNA levels have been reported to be much higher than

those found in surrounding normal tissues (Tricarco et al., 1999; Thelen et al., 1999;

Tsuchiya et al., 2001; Eguchi et al., 1992). In addition, elevated serum/urine bFGF levels

have been shown to associated with malignant progression and poor treatment outcome









(Song et al., 2001; Jacobsen et al., 2000; Rasmuson et al., 2001; Edgren et al., 1999;

Dosquet et al., 1997; Fujimoto et al., 1995; Plunkett and Hailey, 1990; Miyake et al.,

1996; Nguyen et al., 1994a; Huang et al., 1996). Taken together, these findings suggest

that VEGF and bFGF are key factors involved in the angiogenic process of RCC. For

these reasons, RCC was the tumor of choice for the current investigation of anti-

angiogenic therapeutic approaches with the use of VEGF and bFGF AS-ODNs. An RCC

cell line (Caki-1) was used. This cell line is a human clear cell renal cell carcinoma

originally derived from a 49 year-old Caucasian male patient (Fogh, 1978). It grows in

vitro as an anchored cell culture and also forms tumor in athymic nude mice. The

histology of the Caki-l xenograft displays many of the features of clinical samples of

RCC (Figure 1-5).

Significance

Angiogenesis is unique process that contributes to a variety of pathologic

processes and is especially critical to tumor growth and metastasis. It therefore has been

proposed and utilized as both a prognostic indicator as well as a possible target for

therapeutic intervention in certain malignant states. In the present studies, the feasibility

of inhibiting tumor angiogenesis by utilizing AS-ODNs directed against VEGF and bFGF

were investigated. A human renal cell carcinoma cell line grown in vitro or as solid tumor

xenografts in nude mice was used as the tumor model. Through the investigations, a

viable means of intervening with the angiogenic process and in situ growth of renal

carcinoma cells was developed. It is further believed that this approach will not be

confined to RCC, but will be applicable to other neoplasms and may even provide a basis

for selective intervention in other diseases characterized by angiogenesis.













Anti-angiogenic

Pro-angiogenic


Figure 1-1. The angiogenic process. Tumor cells or host cells secrete pro-anglogenic
growth factors, which then bind to specific receptors on endothelial cells. This ligand-
receptor interaction leads to endothelial cell proliferation, migration, invasion and.
eventually, capillary tube formation (Fidler et al, 2000)

















Lo3 ritem M g

4 EC ativloum






WMigr d

7 ECM rmodong



Mp Liallrmage
8 rb fniat




acliar Stabizaion


Figure 1-2. Cascade of events in tumor anglogenesis. (Source: The Angiogenesis
Foundation, www.anglo.org)











pro-angiogenic
I I- to

Iv- t- an I ,






I .






anti-angiogenic



Figure 1-3. Major regulators of angiogenesis and their receptors. On the upper part of the
cell pro-angiogenesis regulators are shown, on the lower part are inhibitory molecules.
(Hagedom and Bikfalvi, 2000)



































Figure 1-4. Resin cast of RCC tumor microvascular network. (Gerwins et al., 2000)









































Figure 1-5. H&E staining histology section ofCaki-1 xenograft tumor.














CHAPTER 2
CELLULAR DELIVERY OF ANTISENSE OLIGONUCLEOTIDES

Introduction

Many of the limitations of current cytotoxic therapies of cancer result from a lack

of specificity of the anti-cancer agents. The advances in molecular biology over the past

The application of two decades have made possible the concept of genetically based,

targeted treatment. Antisense oligonucleotides (AS-ODNs) is one approach to

specifically inhibit gene expression (Stein and Cheng, 1993; Wagner, 1994). These

molecules, usually 18-20 bases in length, can undergo Watson-Crick hybridization to

target mRNAs, ultimately resulting in decreased expression of the gene products. AS-

ODNs technology has attracted great interest and shown great promise as agents to

inhibit the expression of specific genes that regulate physiological functions or mediate

various diseases (Harrison, 1993; Wagner, 1994; Wagner, 1995; Pan etal., 2002; Mani et

al., 2002; Morris et al., 2002).

However, one of the biggest challenges in the application of AS-ODNs as

therapeutic agents is the development of ways to maximize their cellular uptake (Wagner,

1995; Crooke, 1993). AS-ODNs are negatively charged molecules that behave as

polyanions. In general this property leads to poor cellular uptake and intracellular

distribution. Furthermore, the commonly used phosphorothioate modified AS-ODNs

have quite high affinity for proteins, especially heparin-binding proteins (Fennewald and

Rando, 1995). In addition, some new cell surface ODNs-binding proteins have also been

identified recently (Beltinger et al., 1995; Hawley and Gibson, 1996). Even after









internalization into the cell through endocytosis (Yakubov et al., 1989; Crooke et al.,

1995), naked AS-ODNs are localized to endosomes or lysosomes, that are topologically

still "outside" of the cell (Tonkinson et al., 1994). ODNs may then either be released

from the cell via exocytosis or may be partially digested (Tonkinson et al., 1994). Still,

exists a growing literatures of antisense effects after naked AS-ODNs delivery (Anfossi

et al., 1989; Gewirtz and Calabretta, 1988). Nontheless, it should be recognized that these

effects were achieved at high AS-ODNs concentration and may be the result of AS-

ODNs release into the cytoplasm through spontaneous endosomal/lysosomal rupture.

Many techniques have thus been then used to enhance AS-ODNs uptake, the most

widely used being based on the application of cationic lipids. Cellular uptake as well as

the activity in cell cultures can be improved greatly by cationic liposomes (Bennett et al.,

1992; Lappalainen et al., 1994; Zelphati and Szoka, Jr., 1996b; Zelphati and Szoka, Jr.,

1996a). Highly polar, water soluble molecules including AS-ODNs can be entrapped in

the internal aqueous space of the liposme, while the lipids form into bilayers. Cationic

liposomes spontaneously bind the negatively charged AS-ODNs and protect them against

degradation. The macromolecular complexes have a positive charge at the surface, this

results in a high affinity for most cell membranes, which are negatively charged under

physiological conditions. Following the attachment to the membrane, the complexes are

taken up via endocytosis. To help facilitate the release from endosomes and lysosomes, a

helper lipid such as 1,2-dioleoyl-3-sn-phosphatidylethanolamine (DOPE) is often used in

the liposome preparation. This inverted-cone-shaped lipids thought to facilitate cytosolic

release through the fusion and disruption of endosomal membranes (Farhood et al.,

1992). The flip-flop of anionic phospholipids in the endosome membrane, lead to









neutralization of the cationic lipid charge, displacement of the bound oligonucleotides,

and release form the endosome (Figure 2-1) (Koltover et al., 1998; Lebedeva et al.,

2000). Studies have demonstrated that ODNs can readily dissociate from the liposome

complexes and are in bioavailable form within the cells (Tari, 2000; Abe et al.,

1998).Cationic liposomes facilitated delivery of AS-ODNs has proven to be effective in

many different cell lines and to be of general utility (Bennett et al., 1992).

However, studies in mammalian cell lines have demonstrated that AS-ODNs

efficacy varies with different cationic lipids and lipid complexes, cationic lipid/DNA

ratio, and cell type (Flanagan and Wagner, 1997; Lappalainen et al., 1997). This suggests

that as a general rule, the best liposome composition and optimal liposome/ODNs ratio

needs to be established for each cell lines to achieve best results.

In the present studies, fluorescein isothioicyanate (FITC) labeled ODNs were

used to study the cellular uptake of ODNs using a cationic liposome delivery system

(Noonberg et al., 1992). Although the biological activity of AS-ODNs against the

molecular target is highly sequence-dependent, this is typically not the case for

pharmacokinetics and toxicology. Indeed, the pharmacokinetics and toxicology of AS-

ODNs of widely differing sequences directed against vastly disparate gene products have

proven surprising similar (Srinivasan and Iversen, 1995). Thus, the same ODNs sequence

was used for the present cellular uptake and toxicity studies.

Materials and Methods
Cell Culture

The clear cell RCC cell lines Caki-l, Caki-2 and A498 were gifts from Dr. Susan

Knox (Stanford University). These cells were grown in Dulbecco's modified minimum

essential medium (DMEM, Invitrogen, Grand Island, NY) supplemented with 10% fetal









bovine serum (FBS, Invitrogen, Grand Island, NY), 1% penicillin-streptomycin

(Invitrogen, Grand Island, NY) and 1% 200 mmol/L L-glutamine (Invitrogen, Grand

Island, NY).

FITC Labeled Phosphorothioate Oligodeoxynucleotides

The 20-mer ODNs, sequence: 5'- CAC CCT GCT CAC CGC ATG GC -3' (20-

mers) were customer synthesized by Geno Mechanix (Alachua, FL). The entire backbone

was phosphorothioate modified and FITC was labeled at the 5' end of the ODNs. The

ODNs were suspended in sterile and endotoxin-free water at a concentration of 1 mM,

aliquoted and stored at -20C.

Liposome Preparations

Cationic liposomes of different lipid compositions were obtained from Dr. Jeffrey

Hughes' lab (University of Florida, Gainesville, FL). DOTAP:DOPE is composed of

cationic lipid 1,2-dioleoyloxy-3-(trimethylammonium) propane (DOTAP) and a helper

lipid 1,2-dioleoyl-3-sn-phosphatidylethanolamine (DOPE) at a molar ration of 1:1.

DS3DOPC is composed of 1,2-Dioleoyl-sn-Glycero-3-Phophoserine-N-Citraconyl and

dioleoyl-phosphatidylcholine (DOPC) at a molar ratio of 1:1. DOGSDSO is composed of

l',2'-dioleoyl-sn-glycero-3'-succinyl-2-hydroxyethyl disulfide omithine conjugate (Tang

and Hughes, 1998). CHDTAEA is composed of cholesterol hemidithiodiglycolyl

tris(aminoethyl)amine (Tang and Hughes, 1999). PEG-PE is composed of DOTAP,

DOPE and polyethylene glycol distearoyphospatidylethanolamine (PEG-PE) at a molar

ration of 25:25:3 (Meyer et al., 1998). All the lipids were obtained from Avanti Polar-

Lipids (Alabaster, Al).









Briefly, the lipid mixture was evaporated to dryness in a round-bottomed flask

using a rotary evaporator at room temperature. The resulting lipid film was dried by

nitrogen for an additional 10 min to evaporate any residual chloroform. The lipid film

was re-suspended in sterile water to a final concentration of 1 mg/ml based on the weight

of cationic lipid. The resultant mixtures were shaken in a water bath at 35C for 30 min.

The suspensions then were sonicated using a Sonic Dismembrator (Fisher Scientific,

Pittsburgh, PA) for 1 min at room temperature to form homogenized liposomes. The

particle-size distribution of liposomes was measured using a NICOMP 380 ZLS

instrument (Santa Barbara, CA). The average particle diameter was 144.0 77.0 nm.

Liposomes were stored at 40C and used within 3 months.

Antisense Treatment

Caki-1, Caki-2, A498 cells were set at 1 x 105 in 60 mm dishes and allowed to

attach overnight. For comparison of delivery efficiency by different liposomes, FITC

labeled ODNs were mixed with different liposomes in serum free medium or 10% FBS

medium and incubated at room temperature for 30 min. For other studies, only

DOTAP:DOPE liposome was used. The medium of cells was then changed with that

containing ODNs-liposome complex at a ODNs concentration of 1 gM/ml and incubated

at 370C for 3 hours. Equal amounts of 20% FBS medium were added to dishes and

continued to incubate for a total of 24 hr, except for the time course studies, in which,

cells were incubated for different lengths of time.

Fluorescence Microscope

After FITC labeled ODNs treatment, the medium containing ODNs was removed

and cells were washed 4 times with PBS. The cells were then fixed in 1% p-









formaldehyde for 1 hr. Fluorescent microscope pictures were then taken using a Zeiss

Axioplan 2 Florescence Microscope (Zeiss, Thomwood, NY) made available by the

Optical Microscopy Facility at Brain Institute, University of Florida.

Flow Cytometry Analysis

After FITC labeled ODNs treatment, the medium containing AS-ODNs was

removed and cells were washed 4 times with PBS. The cells were then collected by

trypsin digestion. After fixation in 1% p-formaldehyde for I hr, cells were re-suspended

in PBS at the concentration of I x 106 cells/ml and kept in the dark. Green fluorescent

intensities of the cells derived from FITC were then analyzed by FACS on a Becton

Dickinson flow cytometer made available through the University Cole Facility for Flow

Cytometry at the University of Florida.

Toxicity Studies

Caki-1 cells were set in 96-well dishes at 1 x 104 cells per well and allowed to

attach overnight. The culture medium was then changed to 100 pl serum free or 10%

FBS medium containing various doses of DOTAP:DOPE. The cells were incubated for

24 hr at 37C. The viable cells after treatment were measured using a

CellTiter96AQueous Assay System (Promege, Madison, WI). Briefly, 100 ld of

phenylmethasulfazone (PMS) solution was added to 2 ml 3-(4,5-dimethylthiazol-2-yl)-5-

(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tet razolium (MTS) solution and mixed

by gentle swirl. 20 pl of the combined MTS/PMS solution was added into each well and

incubated at 37C for 2 hr. After incubation, 25 il of 10% SDS was added into each well

to stop the reaction. The absorbance was then measured at 490 nm using a microplate

reader.









Results

Uptake of FITC labeled ODNs measured by flow cytometry and fluorescence

microscopy, allowed the assessment of cellular delivery of ODNs. Delivery of naked

ODNs resulted in poor internalization into the cells, either in serum free or 10% FBS

conditions. Only a small portion of cells treated showed moderate increases in

fluorescence intensity derived from FITC (Figure 2-2). However, the use of cationic

liposomes (DOTAP:DOPE) was found to significantly improve the up-take efficiency of

ODNs in Caki-1 cells, in both serum free and 10% FBS conditions (Figure 2-2).

In order to optimize the delivery system for ODNs, other cationic liposome

compositions were evaluated for their delivery efficiency of ODNs in Caki-1 cells. The

cells were treated with FITC labeled ODNs delivered by different cationic liposomes

(DOTAP:DOPE, DS3DOPC, DOGSDSO, CHDTAEA and PEG-PE) for 24 hr at a AS-

ODNs concentration of 1 uM/ml. Both serum free and 10% FBS conditions were studied

and compared. Significant enhancements of ODNs cellular up-take were observed in all

liposome treatment groups, with DOTAP:DOPE and DS3DOPC being most efficient and

with minimal serum resistance (Figure 2-3). Since DOTAP:DOPE is of simple

composition and easy to prepare, it was chosen as the delivery vehicle for the rest of

studies.

Studies to optimize the delivery efficiency of ODNs by DOTAP:DOPE were then

carried out. One major factor that determines the delivery efficiency is the liposome to

ODNs charge ratio. In order to determine the optimal charge ratio in Caki-1 cells, the

cells were treated for 24 hr with I iM/ml FITC labeled ODNs prepared with different

amount of DOTAP:DOPE to achieve various charge ratios (1, 1.25, 1.5, 1.75, 2, 2.5). The









fluorescent intensities of Caki-l cells after the treatment were then compared (Figure 2-

4). Increases in delivery efficiency and as well as resistance to serum were observed as

the charge ratio increased. A plateau in these effects occurred at a charge ratio of about 2.

This charge ratio was used in all subsequent studies evaluating the delivery of AS-ODNs

in Caki-1 cells.

The delivery efficiency ofODNs by DOTAP:DOPE as the function of time was

then evaluated. Caki-l cells were treated with 1 uM/ml FITC labeled ODNs delivered by

DOTAP:DOPE at the pre-determined optimal charge ratio of 2. The fluorescent intensity

of Caki-l cells after different lengths of treatment were determined by FACS (Figure 2-

5). Very fast up-take of ODNs by Caki-l cells were observed when delivered by

DOTAP:DOPE liposomes. The up-take of ODNs by Caki-1 cells reached a plateau

within 24 hr of incubation (Figure 2-5).

With the optimized delivery system based on cationic liposomes (DOTAP:DOPE)

very efficient cellular up-take of ODNs was achieved in Caki-1 cells. Significant up-take

of FITC labeled ODNs was observed in Caki-I cells after 24 hr treatment of 1 pM ODNs

delivered by DOTAP:DOPE at a charge ratio of 2. ODNs were delivered into -100% of

the cells with uniform cellular distribution and enhanced nuclear concentration (Figure 2-

6). This result was achieved in both exponential phase and plateau phase Caki-1 cells

(Figure 2-7).

This optimized delivery system was then further tested for ODNs delivery

efficiency in two other RCC cell lines. The results showed that similar highly efficient

cellular up-take of ODNs also could be achieved in Caki-2 and A498 cells (Figure 2-8)









ODNs as fragments of DNA sequences have very low toxicity (Rubenstein et al.,

1997; Agrawal et al., 1997). The toxicity of the cationic liposome delivery vehicle was

also examined in both serum free and 10% FBS conditions. No significant toxicity of

DOTAP:DOPE in Caki-I cells was observed with doses up to 150 mg/ml in 10% serum

and 100 mg/ml in serum free conditions. At doses higher than 100 mg/ml some

cytotoxicity was observed using DOTAP:DOPE in serum medium. The DOTAP:DOPE

dose used in the delivery studies was only 10 mg/ml, which resulted in no toxicity in

Caki-l cells either in serum free or 10% FBS conditions.

Discussion

AS-ODNs can block the expression of specific target genes involved in the

development of human diseases. Therapeutic applications of antisense techniques are

currently under investigation in many different fields. In order for AS-ODNs to down-

regulate gene expression, it must penetrate into the target cells. Phospholipid bilayers

represent a strong barrier to the movement of ions. Studies on the diffusion of ODNs

through model membranes have led to the general conclusion that it is of little

importance. Meanwhile, internalization of ODNs into cells has been clearly

demonstrated, implying the existence of other mechanisms other than passive diffusion

(Garcia-Chaumont et al., 2000). To date, the precise mechanisms involved in ODNs

penetration are still not totally clear. Though it has been found that up-take of AS-ODNs

can occur through receptor mediated active transport (Wu-Pong et al., 1994), which

depends on temperature (Loke et al., 1989; Yakubov et al., 1989), the structure and the

concentration of ODNs (Vlassov et al., 1994), and the cell lines. At the present time, it is

generally believed that adsorptive endocytosis and fluid phase endocytosis are the major

mechanisms of ODNs interalization (Dias N and Stein, 2002; Garcia-Chaumont et al.,









2000). At relatively low ODNs concentrations, it is likely that intemalization occurs via

interaction with a memberane-bound receptors (Loke et al., 1989; Yakubov et al., 1989;

de Diesbach et al., 2000). While at higher ODNs concentrations, these receptors are

saturated, and the endocytotic process assumes larger importance. Numerous reports have

demonstrated that naked ODNs are internalized poorly by cells (Gray et al., 1997; Stein

et al., 1993)(Figure 2-2). More importantly, naked ODNs tend to localize in

endosomes/lysosomes, where they are unavailable for antisense purposes. As has been

demonstrated in numerous experiments, the sine qua non of antisense activity appears to

be nuclear localization (Dias N and Stein, 2002).

To improve cellular uptake and ODNs spatial and temporal activity, delivery

vehicles were developed. Among them, cationic liposomes are most commonly used.

Cellular up-take of ODNs, as well as their activity in cell cultures, can be improved by

cationic liposomes (Bennett et al., 1992; Lappalainen et al., 1994; Zelphati and Szoka, Jr.,

1996b, Zelphati and Szoka, Jr., 1996a). Cationic liposomes are safe, simple and easy to

produce on a large scale (Nabel et al., 1993; Nabel et al., 1994a; Nabel et al., 1994b).

They have been approved by FDA for clinical use. However, when tested for their ability

to promote delivery of FITC-ODNs in mammalian cell lines, their efficacy varied

significantly (Lewis et al., 1996). Important variables include cationic lipsome/ODNs

ratio, composition of lipid and cell type tested (Flanagan and Wagner, 1997; Lappalainen

et al., 1997; Vellon et al., 2002). This suggests that even though cationic liposomes have

been proven to be effective and to be of general utility, a best reagent and optimal

liposome/ODNs ratio could and should be empirically established for each cell type.









In the present studies, delivery efficiencies of FITC tagged ODNs by different

liposome compositions were first evaluated. All cationic liposome compositions

significantly enhanced cellular up-take of ODNs. However, the incorporation of PEG-PE

significantly inhibited the delivery efficiency of cationic liposomes (DOTAP:DOPE).

PEG-PE, when incorporated into conventional liposomes, can provide a steric barrier at

the surface of lipsomes that inhibits opsonization, and therefore can extend the residence

time of liposomes in the blood (Webb et al., 1998). However, even though the use of

PEG-PE in liposomes have minimal effect on the binding and subsequent endocytosis of

lipid/DNA complexes, it did severely inhibit the endosomal release of AS-ODNs into the

cytoplasm (Figure 2-3) (Song et al., 2002).

Given the efficiency of delivery, minimum serum resistance and simplicity of

liposome composition and preparation, DOTAP:DOPE was chosen as the delivery

vehicle for AS-ODNs in Caki-1 cells (Figure 2-3). A major problem associated with

cationic liposomes is low transfection efficiency due to inactivation of cationic liposomes

by serum (Feigner et al., 1987). Much effort has been devoted to resolving this problem.

The charge ratio of liposome to DNA has been proven previously to be critical for high

efficiency oflipofection and serum resistance (Yang and Huang, 1997; Yang and Huang,

1998). The optimal transfection efficiency and minimal serum inactivation was achieved

at a DOPTA:DOPE/ODNs charge ratio of about 2 (Figure 2-4). Using this charge ratio,

cellular up-take of ODNs by Caki-l cells reaches a plateau within 24 hr (Figure 2-5).

ODNs in Caki-1 cells were evenly distributed in the cytoplasum with enhanced nuclear

concentration (Figure 2-6). This distribution is believed to best facilitate antisense

function of AS-ODNs (Wagner, 1994; Hogrefe, 1999). With the optimized delivery







39

system, efficient cellular up-take of ODNs were achieved in both plateau and exponential

phases Caki-l cells (Figure 2-7) as well as two other RCC cell lines (Caki-2, A498)

(Figure 2-8).

Cytotoxic evaluation of DOTAP:DOPE confirmed that cationic liposomes are

safe unless very high doses are used (Porteous et al., 1997; Gao and Huang, 1995). The

dose used in the present studies (10 mg/ml) was far below the doses that resulted in

cytotoxic effects in Caki-1 cells (>100 mg/ml)(Figure 2-9).

In conclusion, cationic liposomes (DOTAP:DOPE) enhanced the cellular up-take

of ODNs in RCC cell lines. The simplicity of preparation, efficiency of up-take and

safety features have rendered the cationic liposome (DOTAP:DOPE) an attractive vehicle

for AS-ODNs therapy.










.._a_.



m......Ic o


& 1A' arM aby lrar ar
wrrt, ,4
u hinofs ha h a
..ene or--- --Lm "
It".' *v.aa h, p U ,


Figure 2-1. Proposed mechanisms of internalization of cationic liposome into cells and
release ofODNs in cytoplasm. Modified from Lebedeva (Lebedeva et al., 2000).







41



Control Naked PS-ODNS with IOTAP:DOPE







Serum Free"


:IT






10% FBS

Figure 2-2. Flow cytometry histogram of Caki-1 cell fluorescece intensity after
treatments of 1 LiM FITC labeled ODNs either with or without DOTAP:DOPE liposome.












i0000 -
100 Serum Free
SD10% FBS




S
Q 100-
o








Con, DOTMIOPE DS3DOPC DOOSOSO CHDTAMA PEPE

Liposomes


Figure 2-3. Efficiency of cellular up-take of FITC labeled ODNs delivered by different
liposmes in Caki-I cells. Caki-1 cells were treated with 1 pM FITC labeled ODNs in
serum free or 10% FBS conditions for 24 hr. Each bar represents the mean + S.E. of 3
independent experiments.













10000
S Serum Free
0 10% FBS


1000-
C


S100,



o
10
LL 1



0.5 1 1.25 1.5 1.75 2 2.5

Charge ratio


Figure 2-4. Effect of liposome to ODNs charge ratio on the delivery efficiency of ODNs
in Caki-l cells. Caki-l cells were treated for 24 hours with I pM FITC labeled ODNs
delivered by different amounts ofDOTAP:DOPE liposomes at the charge ratio indicated.
Each bar represents the mean S.E. of 3 independent experiments.












1000




I W
100



L io,
0 Serum free
0 E10% FBS



0 24 48 72
Time (hr)


Figure 2-5. Time course of up-take of ODNs delivered by DOTAP:DOPE in Caki-1 cells.
Caki-1 cells were treated with 1 p.M FITC labeled ODNs delivered by DOTAP:DOPE
liposome at a charge ratio of 2.0. Results are the mean S.E. of 3 independent
experiments.













































Figure 2-6. Fluorescent microscopic pictures of Caki-I cells after FITC labeled ODNs
treatment. Caki-1 cells were treated with FITC labeled ODNs for 24 hr at a dose of 1 pM.
A) low magnification, 5x; B) high magnification, 20x


































Figure 2-7. Fluorescent microscopic pictures showing cellular uptake of FITC labeled
ODNs in Caki-l cells. Caki-1 cells were treated with FITC labeled ODNs at a dose of I
pM for 24 hr. A) bright field picture of confluent Caki-1 cells; B) fluorescent field
picture of confluent Caki-I cells: C) bright field picture of exponential phase Caki-1
cells; D) fluorescent field picture of exponential phase Caki-1 cells.











A498 Caki-2









Control









1 iM FITC-AS-ODNs

Figure 2-8. Delivery efficiency of FITC labeled ODNs by DOTAP:DOPE liposome in
other RCC cell lines (A498, Caki-2).The cells were treated with FITC labeled ODNs at a
dose of 1 p|M for 24 hr.













120-


100-


so-

so


s 40-

> 20
S. 10%FBS
2 D Serum Free




0 50 100 150

DOTAP:DOPE concentration (mglml)


Figure 2-9. Caki-1 cell viability after 24 hr treatment with different concentrations of
DOTAP:DOPE liposomes in serum free or 10% FBS conditions.














CHAPTER 3
AS-ODNS DESIGN AND IN VITRO ASSESSMENT

Introduction

Angiogenesis, a complex multi-step process involving the formation of new blood

vessels from pre-existing ones, is tightly regulated by both positive and negative

regulatory factors (Risau, 1997). These regulators, which include pro-angiogenic factors

such as basic fibroblast growth factor (bFGF) (Montesano et al., 1986), angiogenin (Gho

and Chae, 1997) and vascular endothelial growth factor (VEGF) (Leung et al., 1989;

Saleh et al., 1996; Asano et al., 1995; Borgstrom et al., 1996; Cheng et al., 1996), as well

as angiostatic peptides such as endostatin (O'Reilly et al., 1997; Perletti et al., 2000),

angiostatin (O'Reilly et al., 1994a; O'Reilly et al., 1994b; O'reilly et al., 1994) and

thrombospondin (Folkman and Shing, 1992; Folkman, 1995) are potential targets for

anti-angiogenic therapy of solid tumors (Smith et al., 1999; Bicknell and Harris, 1992;

Denekamp, 1999; Bicknell R and Harris A.L., 1992; Folkman, 1971; Denekamp J.,

1999). Among all these factors, VEGF and bFGF are believed to be most important

regulators in tumor angiogenesis (Risau, 1997;Siemeister et al., 1998).

VEGF is an endothelial cell specific mitogen, secreted as a 45 kDahomo dimer

protein. There are five human isoforms derived from alternative splicing (VEGF 121,

145, 165, 189, 206) as illustrated in Figure 3-1 (Tischer et al., 1991; Houck et al., 1991;

Poltorak et al., 1997). VEGFi21 and VEGFi65 are the only soluble isoforms and also the

most abundant, with VEGF165 being the major isoform and most powerful stimulator of

endothelial cell proliferation (Houck et al., 1992; Soker et al., 1997). VEGFi65 is









commonly expressed in a wide variety of human and animal tumors (Hanahan and

Folkman, 1996) and has been shown to induce angiogenesis both in vitro and in vivo

(Leung et al., 1989; Plate et al., 1992a; Plate et al., 1992b). It is currently believed that

this diffusible molecule is probably a key mediator of tumor angiogenesis (Ferrara,

1999a; Ferrara, 1999b). Indeed, the expression of VEGF has been related to fundamental

features of tumors, such as growth rate (Kim et al., 1993), microvessel density (Toi et al.,

1994) and vascular architecture (Drake and Little, 1999) as well as the development of

tumor metastasis (Weidner et al., 1991). A correlation between VEGF expression and

survival has been noted in some cancer patients (Gasparini et al., 1997; Maeda et al.,

1996).

Basic fibroblast growth factor is a prototype of a large family of 13 structurally

related, heparin-binding growth factors. It affects the growth, differentiation, migration

and survival of a wide variety of cell types (Bikfalvi et al., 1997). BFGF was originally

purified from the bovine pituitary gland as a 146-amino acid protein with a molecular

weight of 15 kD (Gospodarowicz, 1975). It was later found to represent a proteolytic

product of the primary 18 kD form (Bikfalvi et al., 1997). The amino acid sequence of 18

kD bFGF is highly conserved among species with 89-95% identity among human, bovine

and rat (Abraham et al., 1986a; Abraham et al., 1986b). This low level of divergence

suggests that there may be functional importance for all regions of bFGF. Larger forms of

bFGF have also been identified resulting from alternative CUG translation starting sites

(Florkiewicz et al., 1991a; Florkiewicz et al., 1991b). The use of different in-frame CUG

codons upstream of the conventional AUG start codon allows translation of several bFGF

isoforms with different molecular weight (Figure 3-6) (Okada-Ban et al., 2000). In









addition to the 18 kD isoform, alternative translation of 22, 22.5, 24 and 34 kD isoforms

are also possible (Araud et al., 1999). The main structural feature of the four high

molecular weight forms of bFGF is the presence of nuclear localization sequence which

directs the growth factors to the nucleus, whereas the 18 kD bFGF isoform initiated from

AUG start codon is essentially cytosolic.

BFGF is a multifunctional growth factor which has various effects in a large panel

of cells and tissues. It plays key role in development, remodeling and disease states in

almost every organ system (Bikfalvi et al., 1997). One of best characterized activities of

bFGF is its ability to regulate the growth and function of vascular cells such as

endothelial cell and smooth muscle cells. BFGF also regulates the expression of several

molecules thought to mediate critical steps during angiogenesis. These include interstitial

collagenase, urokinase type plasminogen activator (uPA), plasminogen activator inhibitor

(PAI-1), uPA receptor, and pi integrins (Montesano et al., 1992; Mignatti and Rifkin,

1993; Klein et al., 1993).. It is a potent angiogenic factor involved in tumor angiogenesis

and metastasis (Basilico and Moscatelli, 1992). Up-regulation of bFGF and its receptors

have been found in tumor tissues compare to normal tissues (Smith et al., 1999;

Dellacono et al., 1997; Arbeit et al., 1996). Clinically, associations between serum/urine

bFGF and cancer outcome have been shown in several tumor systems, including RCC

(Wechsel et al., 2000; Edgren et al., 1999; Nanus et al., 1993; Fujimoto et al., 1991),

breast cancer (Yiangou et al., 1997), head and neck cancer (Dietz et al., 2000), cervical

cancer (Sliutz et al., 1995), liver cancer (Poon et al., 2001), pancreas cancer (Ohta et al.,

1995), thyroid cancer (Sasaki et al., 2001) and glioma, neuroblastoma (Bredel et al.,

1997; Komuro et al., 2001).









In light of their important roles in tumor angiogenesis, VEGF and bFGF may be

attractive targets for anti-angiogenic therapeutic interventions applied to the treatment of

cancer. Attempts to abrogate the angiogenic activity of VEGF and bFGF have focused on

inactivating VEGF/bFGF through the use of antibodies against VEGF/bFGF or their

receptors (Mordenti et al., 1999; Kim et al., 1993; Aonuma et al., 1999; Lu et al., 2002;

Brekken et al., 2000) and VEGF soluble receptors (Lin et al., 1998), inhibiting

VEGF/bFGF receptor tyrosine kinases (Hennequin et al., 1999; Laird et al., 2002;

Solorzano et al., 2001; Ning et al., 2002) or suppressing VEGF and bFGF messages

(Smyth et al., 1997; Ellis et al., 1996a; Ellis et al,, 1996b; Nguyen et al., 1998a; Nguyen

et al., 1998b; Inoue et al., 2000). The latter relied on antisense oligonucleotides (AS-

ODNs) or antisense RNA (Eguchi et al., 1991; Mercola and Cohen, 1995a; Mercola and

Cohen, 1995b) to modulate gene expression by disrupting RNA expression. AS-ODNs

technology provides an approach for inhibiting gene expression with target specificity as

a particular advantage (Stein and Cheng, 1993; Engelhard, 1998a; Engelhard, 1998b).

AS-ODNs are also easy to produce in large quantities which make them potentially more

practical than antisense RNA vector delivery approaches.

In the present studies, AS-ODNs against VEGF and bFGF were designed and

their efficacy tested in vitro in the model of human RCC (Caki-1).

Materials and Methods

Cell Culture

The clear cell RCC cell lines Caki-l, Caki-2 and A498 were gifts from Dr. Susan

Knox (Stanford University, CA). Caki-1 cells were grown in Dulbecco's modified

minimum essential medium (DMEM, Invitrogen, Grand Island, NY) supplemented with









10% fetal bovine serum (FBS, Invitrogen, Grand Island, NY), 1% penicillin-streptomycin

(Invitrogen, Grand Island, NY) and 1% 200 mmol/L L-glutamine (Invitrogen, Grand

Island, NY).

Antisense Phosphorothioate Oligodeoxynucleotides (AS-ODNs)

Antisense and control ODNs (20-mers) were custom synthesized by Geno

Mechanix (Alachua, FL). AS-ODNs V515 was complementary to 5' UTRjust up-stream

of the translation start site (AUG codon) of VEGF mRNA: 5' CTC ACC CGT CCA

TGA GCC CG 3'. Scramble sequence: 5'- CAC CCT GCT CAC CGC ATG GC 3';

sense sequence: 5' CGG GCT CAT GGA CGG GTG AG 3' and an inverted sequence:

5'-GCC CGA GTA CCT GCC CAC TC 3', were used as controls ODNs. AS-ODNs

B460 was complementary to the translation start site (AUG codon) of bFGF mRNA: 5'

TCC CGG CTG CCA TGG TCC CT 3', AS-ODNs B471 was complimentary to the

coding region of bFGF mRNA: 5' CGT GGT GAT GCT CCC GGC TG 3'; AS-ODNs

B931 was complimentary to the 3' UTR: 5' GAT GTG GCC ATT AAA ATC AG 3'.

Scramble sequence: 5' GCC TGG ACC CTG GCT CTC TC 3'; sense sequence: 5' AGG

GAT GGC TGC CGG GA 3' and an inverted sequence: 5' TCC CTG GTA CCG TCG

GCC CT 3' were used as controls. All AS-ODNs were suspended in sterile and endotoxin

free water at a concentration of I mM, aliquoted and stored at -200C.

DOTAP:DOPE Liposome Preparation

Cationic liposomes were prepared using the method described by Tang (Tang and

Hughes, 1999). Briefly, cationic lipid 1,2-dioleoyloxy-3-(trimethylammonium) propane

(DOTAP) was dissolved in chloroform and mixed with a helper lipid 1,2-dioleoyl-3-sn-

phosphatidylethanolamine (DOPE) (Avanti Polar-Lipids, Alabaster, Al) at a molar ratio









of 1:1. The mixture was evaporated to dryness in a round-bottomed flask using a rotary

evaporator at room temperature. The resulting lipid film was dried by nitrogen for an

additional 10 min to evaporate any residual chloroform. The lipid film was re-suspended

in sterile water to a final concentration of I mg/ml based on the weight of cationic lipid.

The resultant mixtures were shaken in a water bath at 35C for 30 min. The suspensions

then were sonicated using a Sonic Dismembrator (Fisher Scientific, Pittsburgh, PA) for 1

min at room temperature to form homogenized liposomes. The particle-size distribution

of liposomes was measured using a NICOMP 380 ZLS instrument (Santa Barbara, CA).

The average particle diameter was 144.0 + 77.0 nm. Liposomes were stored at 4C and

used within 3 months.

VEGF Enzyme Immunoassay

Caki-l cells (1 x 105) were set in 60 mm dishes and allowed to attach overnight.

The medium then was removed and replaced with AS-ODNs in serum free medium with

liposome (DOTAP:DOPE) and incubated for 5 hr. Fresh medium containing 10% FBS

then was added. After 24 hr of incubation, or at different time points for the time course

studies, the VEGF concentration was determined in the medium using a human VEGF

ELISA kit (R &D Systems, Minneapolis, MN).

Enzyme Immunoassay of bFGF

Caki-l cells (I x 105) were set in 60 mm dishes and allowed to attach overnight.

The medium then was removed and replaced with AS-ODNs in serum free medium with

liposome (DOTAP:DOPE) and incubate for 5 hr. Fresh medium containing 10% FBS

then was added. Caki-1 cells were collected 72 hr later, washed and suspended 1 x 106

cells in 1 ml PBS containing protease inhibitors (100 pg/ml Phenylmethanesulphonyl









fluoride, 20 [pg/ml leupeptin, 3 lpg/ml aprotinin). The suspension was subjected to 3

freeze-thawing cycles, ultrasonication for 5 s (100 W) on ice, and was centrifuged at

14,000 g for 10 min. The supernatant containing the intracellular bFGF was used for the

bFGF concentration determination (human bFGF immunoassay kit, R & D Systems,

Minneapolis, MN).

VEGF and bFGF Relative Quantitative RT-PCR

Caki-I cells were set at 3 x 105 in 100 mm dishes and allowed to attach overnight.

The cells were then treated with 1 JiM VEGF antisense (V515), bFGF antisense (B460)

or control ODNs as described. 24 hr later the cells were collected and the total RNA was

isolated using RNeasy Mini Kit (Qiagen, Valencia, CA) and RNA concentrations were

determined by UV spectrophotometry. A 2 pg total RNA sample was used to reverse

synthesize cDNA using Superscript II reverse transcriptase (Invtrogen, Grand Island,

NY). A 2.5 pl aliquot of the reverse transcriptase reaction product then was used for the

PCR reaction. VEGF PCR reactions were carried out with a VEGF gene specific relative

RT-PCR Kit (Ambion, Austin, TX). BFGF PCR reactions were carried out using a

forward primer: 5'GCA GCC GGG AGC ATC ACC A 3' and reverse primer: 5' GCC

CAG TTC GTT TCG GTG CCC A 3' (Campbell et al., 1999). The PCR reactions were

run 22 cycles (denature 940C 30 s, anneal 600C 60 s, extension 720C 60 s) in aDNA

Engine 200 (MJ research, Waltham, MA). PCR products then were run in 2% agrose gel

and stained by ethidium bromide. The gels were visualized and analyzed (Gel Doc 2000

gel documentation system, Bio-Rad, Hercules, CA). All PCR preparations were carried

out in a laminar flow hood using aerosol resistant plugged pipette tips.









FGF Receptors 1-4 RT-PCR

Total RNA of exponential phase Caki-l cell was isolated using RNeasy Mini Kit

(Qiagen, Valencia, CA) and RNA concentrations were determined by UV

spectrophotometry. A 2.5 [l aliquot of the reverse transcriptase reaction product then was

used for the PCR reaction. Primers for human FGFR 1-4 designed by Tartaglia etc were

used (Tartaglia et al., 2001). The PCR reactions were run for 30 cycles (denature 94C 30

s, anneal 60"C 60 s, extension 720C 60s) in a DNA Engine 200 (MJ research, Waltham,

MA). The specificity of the cDNA amplifications were then verified by endonuclease

restriction analyses (Tartaglia et al., 2001). All PCR preparations were carried out in a

laminar flow hood using aerosol resistant plugged pipette tips. Negative controls without

template DNA were included in each assay. 18S primer set (Ambion, Austin, TX) was

used as positive control.

Cell Cycle Assays

Caki-l cells were plated in 60 mm dishes at 2 x 105 cells per dish and allowed to

attach overnight. The cells were then treated with I pM B460 or control ODNs mixed

with DOTAP:DOPE as described above. 48 hr later, the cells were trypsinized, counted

and fixed in 50% ethanol overnight. Before analyzed by FACS, the cells were treated

with 1 mg/ml RNase (in PBS) for 30 min. The samples were then washed with PBS twice

and resuspended in 25 mg/ml propidium iodine (PI) in PBS at a concentration of 1 x 106

cells/ml. The cells were stained with PI in darkness for 15 min and were analyzed by

FACS for cell cycle distribution on a Beckman Dickinson flow cytometer made available

through the University of Florida Core Facility for Flow Cytometry.









Apoptosis Assays

Caki-1 cells were set in 2-well chamber slides and treated with 1 gM B460 or

control ODNs as described earlier. 48 hr later, the cells were washed and fixed in 4%

para-formaldehyde solution for Fluorometric TdT-mediated dUTP Nick-End labeling

(TUNEL) assay. Briefly, the cells were permeabilized in 0.2% Triton X-100 solution for

5 min. DNA strand breaks were then labeled with fluorescein-12-dUTP in TdT

incubation buffer at 370C for 1 hr. The samples were then counterstained with 1 pg/ml

PI in PBS, which binds to the A-T rich regions of DNA. Localized green fluorescence of

apoptotic cells (fluorescein-12-dUTP) in a red background (PI) was detected by

fluorescence microscopy. The percentage of apoptotic cells was obtained by dividing the

number of cells with green fluorescence by the total number of cells counted A

minimum of 300 cells were counted for each condition.

Results

VEGF AS-ODNs Design and Assessment

Since VEGF has multiple isoforms resulting from alternative splicing (Figure 3-

1), AS-ODNs design was targeted at the common region of all isoforms, regions around

the AUG start codon. After screening several different designs, AS-ODNs V515 which is

complimentary to the 5'UTR region just up-stream of the AUG start codon of the VEGF

gene, was found to be most effective. The results showed that after 24 hr treatment with 1

pM VEGF AS-ODNs (V515) delivered by cationic liposome (DOTAP:DOPE), the

medium VEGF levels were significantly reduced from a normal of 850 pg/ml/106 cells to

250 pg/ml/106 cells (p<0.05, student's I-test) (Figure 3-2). This antisense effect was

sequence, and target region specific. Treating Caki-1 cells with liposome vehicles









(DOTAP: DOPE) or control scramble ODNs did not affect VEGF levels. Similarly,

treatment with sense or inverted sequence ODNs failed to reduce VEGF expression.

Continued exposure of Caki- cells to the VEGF AS-ODNs (V515) resulted in a

constant repression of VEGF in the culture medium (Figure 3-3). However, if the culture

medium containing V515 was replaced with fresh medium 24 hr later, the VEGF levels

in the Caki-l cell medium gradually recovered, and reached about -80% of that found in

the untreated Caki-l cell medium in about 7 days (Figure 3-3).

This repression of VEGF expression by V515 was also dose dependent (Figure 3-

4). For example, a 24 hr treatment with 0.5 giM, reduced the medium VEGF level to 56%

of control (p<0.05, student's M-test) whereas a 1 gM dose down-regulated the VEGF level

to 22% of control (p<0.05, student's t-test).

VEGF mRNA levels in different AS-ODNs treatment groups also were

determined (Figure 3-5). The results indicated a marked inhibition of VEGF mRNA after

treatment with V515 which was absent in cells treated with scramble control ODNs. This

result indicated that RNase H plays an important role in the function of V515.

BFGF AS-ODNs Design and Assessment

Alternative translation utilizing CUG start codons other than the AUG start codon

leads to different isoforms bFGF gene products (Figure 3-6). In order to target all the

bFGF isoforms using the same AS-ODNs sequence, the AS-ODNs were designed to

target the common regions of all isoforms, especially around the AUG start codon.

Effective AS-ODNs sequences against bFGF were identified and Caki-l cells treated

with them showed bFGF levels significantly lower than those normally observed (720

pg/ml/106 cells) (Figure 3-7). This effect was sequence and target region specific. The









AS-ODNs complimentary to the start codon (AUG) region (B460) was found to be the

most effective. For example, the cellular bFGF level of B460 treated Caki-1 cells was

found to be about 41% of that found in control or untreated cells (p<0.05, student's t-

test). By comparison, the AS-ODNs complimentary to the 3' UTR (B931) or coding

region (B471) were less effective at down regulating bFGF expression (57% and 65% of

control values respectively, p<0.05, student's t-test). Since B460 had the most prominent

inhibitory effect, it was used in all subsequent studies. Treating Caki-1 cells with control

scramble ODNs or liposome vehicles did not affect bFGF levels in Caki-1 cells.

Similarly, treatment with sense or inverted sequence ODNs failed to reduce bFGF

expression. This inhibitory effect was also found to be AS-ODNs dose dependent (Figure

3-8). While a low dose of 0.5 pM B460 reduced the cellular bFGF level to about 80% of

control, a high dose of 5 [.M B460 led to a reduction by 65%.

BFGF mRNA levels in different PS-ODNs treatment groups also were determined

(Figure 3-9). The results indicated a marked inhibition of bFGF mRNA after treatment

with B460 which was absent in cells treated with scramble ODNs. Again, this suggests a

role for RNase H in the efficacies of bFGF AS-ODNs B460.

Because FGF can have mitogenic effects in renal cells (Gospodarowicz et al.,

1986; Issandou and Darbon, 1991), the influence of antisense and control ODNs

treatment on Caki-1 cell growth was investigated. Control ODNs or liposome vehicles

showed no effect on Caki-1 cell growth (Figure 3-10). In contrast, Caki-1 cell growth was

significantly inhibited by AS-ODNs targeted against different regions of bFGF mRNA.

B460 was found to be the most effective while AS-ODNs targeting the 3' UTR (B931) or

coding region (B471) showed less cell growth inhibition. When comparing the data of









Figures 3-7 and 3-10, it also is apparent that the extent of Caki-1 cell growth inhibition

by different AS-ODNs was closely related to their potency in down regulating bFGF

expression.

In order to gain a better understanding of the underlying mechanisms of the

observed growth inhibitory effect, FGF receptor expression was determined in Caki-1

cells (Figure 3-11). It was found that 3 out of 4 FGF receptors involved in the bFGF

signal transduction pathway were expressed by Caki-1 cells, indicating that bFGF may

play an autocrine role in Caki-l cells. Additional studies indicated that B460 treatment

had small but significant effects on Caki-l apoptosis and cell cycle distribution (Figures

3-12 and 3-13). However, clonogenicity studies showed no significant difference between

B460 treated and control cells indicating that B460 treatment had no direct cell killing

effect on Caki-I cells (data not shown). These findings suggest that blocking the bFGF

signal transduction pathway may affect Caki-I cell growth through cell cycle inhibition

and induction of apoptotic cell death.

Discussion

Although dramatic advances have been made in the treatment of cancer, the

development of efficacious anticancer agents still lags behind the rapid strides in our

understanding of cancer biology, especially with the advent of molecular biology. The

continued progress in our knowledge of the biology of neoplasm and in the identification,

cloning and sequencing of genes critical to tumor cell function permits the exploitation of

this information to develop specific agents that may directly modulate the function of

these genes or their protein products. One methodology that takes direct advantage of









molecular sequencing data involves the use of antisense oligonucleotides (Ho and

Parkinson, 1997).

The antisense-mediated gene inhibition was first introduced in 1978 by

Stephenson and Zamecnik (Zamecnik and Stephenson, 1978; Stephenson and Zamecnik,

1978). The underlying concept is relatively straightforward: the use of a sequence,

complementary by virtue of Watson-Crick basepair hybridization, to a specific mRNA

can inhibit its expression and then induce a blockade in the transfer of genetic

information from DNA to protein.

The selection of an appropriate target sequence is the first step in the process of

AS-ODNs drug development. As a matter of fact, the hybridization between AS-ODNs

and the target mRNA, which has a particular three dimensional shape resulting from

secondary and tertiary structures, depends on the accessibility of the target sequence.

Only limited stretches of mRNA sequences are actually available for heteroduplex

formation with AS-ODNs. Still there is no sure way to determine a prior which AS-

ODNs sequence would work best (Cohen, 1989;Woolf et al., 1992; Brysch and

Schlingensiepen, 1994). The region surrounding the start codon (AUG) is probably the

most popular target, followed by 5'UTR, coding regions or splicing sites. In the present

studies, AS-ODNs sequence design and selection focused mainly on the start codon

region of VEGF and bFGF genes. This region is also common to all isoforms of these

two growth factors (Figure 3-1 and 3-6).

The most commonly used AS-ODNs are 18-20 bases in length. According to

statistical calculations, a particular sequence of 13 bases in RNA and of 17 bases in DNA

should be found only once in the entire human genome, thus representing unique









elements within the cell (Helene and Toulme, 1990). It also has been found that the

activities of AS-ODNs increased with AS-ODNs length, but the increased

thermodynamic stability of hybridization observed with AS-ODNs binding to non-target

mRNA sequences that may be similar, but not identical to the target sequence, results in

reduced specificity (Monia et al., 1992). Thus, an AS-ODNs length of 20-mer is usually

considered optimal and therefore was used in the present studies for all the designed AS-

ODNs. To ensure the specificity of the target sequence, all designed AS-ODNs sequences

were checked for global sequence comparison using the Basic Local Alignment Search

Tool (BLAST) from the National Center for Biotechnology Information (NCBI,

http.//www.ncbi.nlm.nih.gov/BLAST/). Only sequences specific for the VEGF/bFGF

gene and having at least 4 miss-match bases with other genes were used. In addition,

polyguaosine (GGGG), which is known to exert non-antisense effects was avoided

(Benimetskaya et al., 1997). Lastly, all the AS-ODNs designed were examined for

secondary structures such as hairpins, self-dimers, and cross-dimers using Netprimer

(PREMIER Biosoft International, Palo Alto, CA).

Since cells contain a variety of exo- and endonucleases that can degrade ODNs,

nucleotide modifications have been made to make the AS-ODNs more resistant to

nuclease digestion than the native ODNs that have phosphodiester linkages in their

backbones. The most widely explored analogues have been the phosphorothioates, in

which one of the nonbridging oxygen atoms in each intemucleoside phosphate linkage is

replaced by a sulfur atom. This modification is easily adapted to automated synthesis and

confers metabolic stability because its resistance to degradation by DNase (Stein et al.,

1988). In addition, these analogues retain water solubility and permit RNase H mediated









hydrolysis of the target mRNA strand. This modification has been successfully used in a

variety of investigations (Galderisi et al., 1999; Ho and Parkinson, 1997; Crooke, 1998).

Based on these favorable properties and extensive information available for its

application, phosphorothioate modification was used in the current investigations of

VEGF/bFGF AS-ODNs.

AS-ODNs may exert biological activity through a variety of mechanisms.

Although some of the mechanisms of inhibition have been characterized, rigorous proof

for others is still frequently lacking. Two classes of AS-ODNs can be discerned: (a) the

RNase-H dependent ODNs, which induce the degradation of mRNA; and (b) the steric-

blocker ODNs, which physically prevent or inhibit the progress of splicing or

translational machinery (Crooke, 1992; Dias N and Stein, 2002).

The most commonly implicated antisense mechanisms relate to RNase H

mediated hydrolysis of the target mRNA. This is also the case for most of the antisense

drugs investigated in the clinic. RNase H is a ubiquitous enzyme that hydrolyzes the

RNA strand of the RNA/DNA hybrid. Thus the binding of AS-ODNs to its target mRNA

may induce digestion of the message. AS-ODNs assisted RNase H dependent reduction

of target RNA expression can be quite efficient, reaching 80-95% down-regulation of

protein and mRNA expression (Dias N and Stein, 2002). Furthermore, in contrast to the

steric-blocker ODNs, RNase H dependent ODNs can inhibit protein expression when a

much wider region of the mRNA is targeted. Thus, unlike most steric-blocker ODNs that

are efficient only when targeted to certain 5' UTR or AUG start codon regions, RNase H

dependent AS-ODNs can exert effects when targeted to widely separated areas in the

coding region as well (Dean and McKay, 1994; Larrouy et al., 1992). In the present









studies, mRNA levels of VEGF and bFGF were significantly down-regulated after the

AS-ODNs treatment (Figures 3-5 and 3-9). These findings suggest that current AS-ODNs

function mainly through the RNase H mechanism.

Successful design and evaluation of AS-ODNs relies on their efficient delivery of

into the cytoplasm, where they can exert their antisense effect. A cationic liposome

(DOATP:DOPE) based delivery system was used to deliver AS-ODNs in the present

studies. The evaluation and optimization of this delivery system has been discussed in

detail in Chapter 2.

Following the aforementioned guidelines in AS-ODNs design and selection,

effective AS-ODNs against VEGF and bFGF have been identified. Treatment of Caki-1

cell with the AS-ODNs led to significant repression of VEGF and bFGF expression

levels (Figures 3-2 and 3-7). These effects were ODNs sequence specific, dose dependent

and could be achieved at low, non-toxic doses. These results indicate that AS-ODNs can

be used to efficiently modulate the specific target gene expressions.













VEGF mRNA


VEF 121 mm 121 3 I 4

VEOF 145 1 2 3 4

VEGF 165 Im 1 2 3 1 4 //

VEGF 189 1 2 3 4

VEGF206 1 M 2 3 I4 IS //
V515





Figure 3-1. VEGF mRNA structure showing all isoforms derived from alternative

splicing.














100

0



0
I.
LJ 20
> 0


Untreaed DOTAP Scrarble Snse Invrted V515

Treatment




Figure 3-2. VEGF levels in culture medium of Caki-1 cells treated with different AS-
ODNs. Caki-1 cells were untreated, treated with vehicle (DOTAP) only, treated with I
gM control ODNs or VEGF AS-DONs (V515) for 24 hr. The 100% VEGF expression
level of the untreated group corresponds to -800 pg/ml/106 cells. Each bar represents the
mean S.E. of 3 independent experiments. The star indicates a statistically significant
difference from the untreated group (p<0.05, student's t-test).













120


S100





ua
S so,


s*


U 20.
> 20,


O Continued V515 treatment
* 24 Ih V515 treatment


0-*
0 1 2 3 4 5 6 7 8

Time (day)



Figure 3-3. VEGF levels in the culture medium of Caki-1 cells at different times after
V515 treatment. Media containing V515 were either unchanged (1) or replaced with
fresh medium after 24 hr (0). Each datum point represents the mean S.E. of 3
independent experiments.













120-





8 80-

a0-
So

40
240-
u. *
20- *
^0---- --------------a


0 1 2 3 4 5

Antisense PS-ODNs dose (pM)




Figure 3-4. VEGF levels in culture medium of Caki-1 cells after treatment with different
doses of V515. The 0 dose represents the Caki-1 cells treated with 5 giM control scramble
ODNs. Each datum point represents the mean S.E. of 3 independent experiments. Stars
indicate statistical significance compared to the control ODNs treated (p<0.05, student's
t-test).











A Untreated Scramble V515

18W

VZi/F


B 200-

S180-



1 120

so

S60-
E 40
IL
0 20
> 0-


Untreated
Untreated


Scramble V515


Treatment


Figure 3-5. Message RNA levels in Caki-l cells either untreated, treated with a 1 JM
dose of scramble ODNs or VEGF AS-ODNs (V515). A) Representative relative RT-PCR
results; B) Relative VEGF mRNA levels of Caki-l cells after the treatment. The star
indicates a statistical significance compared to untreated control (p<0.05, student's t-test).







70





bFGF mRNA
CUG CUG CUGCUG AUG STOP
86 319 346 361 486 951
s' I I I I I I
5'

bFGF proteins
S18 kD
I 22 kD
I I 22.5 kD
I I I 24 kD
SI 34kD
SNuclear localization signal




Figure 3-6. Message RNA structure and protein products of bFGF gene. Modified from
Okada-Ban (Okada-Ban et al., 2000).

















= 000
600



400
2soo *
L. 300
U-
200

100

0
Untreated DOTAP Scramble Sense Intted B460 B471 B31

Treatment





Figure 3-7. Cellular bFGF levels of Caki-l cells after AS-ODNs treatment. Caki-1 cells
were either untreated, treated with liposome alone, I IM control ODNs or bFGF AS-
ODNs for 24 hr. Each bar represents the mean S.E. of 3 independent experiments. Stars
indicate statistically significant difference from the untreated control group (p<0.05,
student's t-test).
















100




60



2 \

0 20 -
0-


0 1 2 3 4 5

B460 dose (pM)



Figure 3-8. Cellular bFGF levels in culture medium of Caki-1 cells after treatment with
different doses of B460. The 0 dose Caki-1 cells treated with 5 ltM control scramble
ODNs. Each datum point represents the mean S.E. of 3 independent experiments. Stars
indicate statistically significant differences compared to the control ODNs treated group
(p<0.05, student's t-test).










A Untreated DOTAP


18s
bFGF


amur a" me i- h man A* 1-


o Iso


Sloo *


z
E
0
U.
0
Untreated DOTAP Scramble B460
Treatment

Figure 3-9. Message RNA levels in Caki-l cells either untreated, treated with a 1 gM
dose of scramble ODNs or bFGF AS-ODNs (B460). A) Representative relative RT-PCR
results; B) Relative bFGF mRNA levels of Caki-1 cells after the treatment. The star
indicates a statistically significant difference compared to the untreated control group
(p<0.05, student's t-test).


Scramble B460















T 100-

s 80-
08 *


U 4-


S20-

0
Untreated DOTAP Scramble Sense Inverted 480 B471 BM31

Treatment




Figure 3-10. Effect of AS-ODNs treatment on Caki-1 cell growth. Caki-I cells were
either untreated, treated with liposome alone, 1 gM control ODNs or bFGF AS-ODNs for
4 days and the number of cells determined. Each bar represents the mean S.E. of 3
independent experiments. Stars indicate statistically significant differences compared to
untreated control group (p<0.05, student's t-test).
































Figure 3-11. FGF receptor expression in Caki-l cells. RT-PCR results of FGF receptors
1-4 and control 18s expression in Caki-l cells.















































Figure 3-12. Apoptotic cell death in Caki-1 cells either untreated or treated with bFGF
AS-ODNs. A) Representative picture of untreated Caki-1 cells stained with DeadEndTM
Fluorometric TUNEL System: B) representative picture of Caki-1 cells after treatment
with B460 stained with DeadEndTM Fluorometric TUNEL System.







77




10-
9
a 8
S 7
6.* -6



S 3-
< 4


0 2
C I
0
Untreated DOTAP Scramble B460

Treatment



Figure 3-13. Apoptotic cell death in Caki-1 cells after AS-ODNs treatment. Caki-l cells
were either untreated, treated with liposome alone, treated with I iM scramble control
ODNs or bFGF AS-ODNs. Each bar represents the mean S.E. of 3 independent
experiments. The star indicated a statistically significant difference from the untreated
control group (p<0.05, student's 1-test).














Untreated
C Control
60 [O B460
0


4 45


C 30

1 -




GO-G1 S G2-M

Cell cycle distribution



Figure 3-14. Effect of bFGF AS-ODNs treatment on the cell cycle distribution of Caki-l
cells. Each bar represents the mean S.E. of 3 independent experiments. Stars indicate
statistical significant differences compared to the untreated control group (p<0.05,
student's -test).














CHAPTER 4
ANTI-ANGIOGENIC EFFICACY STUDIES

Introduction

The greater understanding of the process of tumor angiogenesis, coupled with the

notion that tumors require a blood supply to grow and metastasize, has fueled the

research for strategies that block or disrupt the angiogenic process. Moreover, because

normal vascular endothelial cells turn over so slowly, conventional wisdom suggests that

an anti-angiogenic approach to cancer therapy should offer improved efficacy and

reduced toxicity, with much less potential for drug resistance.

Angiogenesis is a complex process with multiple, sequential and interdependent

steps (Fidler, 1999). This complexity creates many potential targets for inhibition. Key

characteristic of the immature vasculature of tumors have allowed the development of

several categories of anti-angiogenic agents (Kerbel, 2000). Preclinical studies have

identified agents that (i) inhibit endothelial cell activation (ii) inhibit endothelial

proliferation/migration (iii) inhibit basement membrane degradation and (iv) inhibit

integrin receptor activation.

Angiogenesis can be qualitatively and quantitatively measured in a large variety

of in vitro and in vivo model systems. As mentioned before the angiogenic cascade can be

dissected into different sequential steps so that can be studied separately in vitro.

Research has mainly focused on the proliferation and migration of endothelial cells as

key elements for angiogenic potential in vitro. For this research, different endothelial cell

sources can be utilized. For human tumor research most laboratories make use of









HUVECs. Although readily available, a major advantage, the major drawback of these

cells is their macrovascular origin, which makes them less suitable for studies on

angiogenesis, a microvascular process. In recent years, microvascular endothelial cells

derived from different organs have been established and become commonly available.

Protocols for isolating purified tumor endothelial cells also have been developed (St

Croix et al., 2000).

Assays to study proliferation of endothelial cells are based on cell counting, radio-

labeled thymidine incorporation, or on colorimetric assays for measurement of

mitochondrial activity. Detection of cell death also is used to determine cell growth

effects. To measure endothelial cell migration, Boyden chambers are primarily used.

Though an easier system based on wounding of a confluent monolayer of endothelial

cells and measuring wound width or invading cells as a function of time is also available,

it maybe physiologically less relevant to tumor angiogenesis. In the present studies, we

established a co-culture system based on modified boyden chambers to evaluate

endothelial cell proliferation and migration potential (Figure 4-1 and 4-4). In order to

mimic the in vivo interaction of tumor and endothelial cells, co-cultured tumor cells were

the primary source ofpro-angiogenic growth factors for the endothelial cells.

The advantage of the in vitro assays is clearly the control that can be exerted over

selected parameters. However the angiogenic cascade consists of multiple steps in their

entirety, in vivo investigations are needed. The most frequently used in vivo assay

systems are the chicken chorioallantoic membrane assay (Nguyen et al., 1994a), the

comeal pocket assay (Conrad et al., 1994), transparent chamber preparations such as the

dorsal skin-fold chamber (Algire G.H., 1943; Lichtenbeld et al., 1998), the cheek pouch









window (Shubik et al., 1976) and polymer matrix implants (Mahadevan et al., 1989;

Plunkett and Hailey, 1990). A simpler system using intradermal implantation of tumors

cells is also available to study the tumor angiogenic process in vivo (Sidky and Auerbach,

1976). This assay has been validated for evaluating tumor-induced angiogenesis in vivo

using a variety of different tumor models and treatment interventions (Lindner and

Borden, 1997; McMillan et al., 1999; Danielsen and Rofstad, 1998) and was applied in

the present studies.

Through the utilization of both the co-culture system in vitro and the intradermal

angiogenesis assay in vivo, objective and reasonable assessments of the anti-angiogenic

efficacy of treatment interventions can be achieved. In the present studies, the anti-

angiogenic efficacy of VEGF and bFGF AS-ODNs treatments were evaluated using these

in vitro and in vivo models.

Materials and Methods

Cell Culture

The clear cell RCC cell line Caki-1 was a gift from Dr. Susan Knox (Stanford

University, CA). Caki-1 cells were grown in Dulbecco's modified minimum essential

medium (DMEM, Invitrogen, Grand Island, NY) supplemented with 10% fetal bovine

serum (FBS, Invitrogen, Grand Island, NY), 1% penicillin-streptomycin (Invitrogen,

Grand Island, NY) and 1% 200 mmol/L L-glutamine (Invitrogen, Grand Island, NY).

The mouse heart endothelial cell line (MHE) was a gift from Dr. Robert Auerbach

(University of Wisconsin, WI). MHE cells were grown in DMEM supplemented with

10% heat inactivated FBS, 1% penicillin-streptomycin and 1% 200 mmol/L L-glutamine.

Human microvascular endothelial cell from the lung (HMVEC-L) cells were obtained









from Clonetics (San Diego, CA). HMVEC-L cells were grown in EBM-2-MV (Clonetics,

San Diego, CA) supplemented with 5% FBS.

Antisense Phosphorothioate Oligodeoxynucleotides (AS-ODNs)

Antisense and control ODNs (20-mers) were custom synthesized by Geno

Mechanix (Alachua, FL). The entire backbone of all ODNs was phosphorothioate

modified. AS-ODNs V515 was complementary to 5' UTR just up-stream of the

translation start site (AUG codon) of VEGF mRNA: 5' CTC ACC CGT CCA TGA

GCC CG 3'. Scramble sequence: 5'- CAC CCT GCT CAC CGC ATG GC 3'; sense

sequence: 5' CGG GCT CAT GGA CGG GTG AG 3' and an inverted sequence: 5'-

GCC CGA GTA CCT GCC CAC TC 3', were used as ODNs controls. AS-ODNs B460

was complementary to the translation start site (AUG codon) of bFGF mRNA: 5' TCC

CGG CTG CCA TGG TCC CT 3'. Scramble sequence: 5' GCC TGG ACC CTG GCT

CTC TC 3'; sense sequence: 5' AGG GAT GGC TGC CGG GA 3' and an inverted

sequence: 5' TCC CTG GTA CCG TCG GCC CT 3' were used as controls. All ODNs

were suspended in sterile and endotoxin free water at a concentration of I mM, aliquoted

and stored at -200C.

DOTAP:DOPE Liposome Preparation

Cationic liposomes were prepared using the method described by Tang (Tang and

Hughes, 1999). Briefly, cationic lipid 1,2-dioleoyloxy-3-(trimethylammonium) propane

(DOTAP) was dissolved in chloroform and mixed with a helper lipid 1,2-dioleoyl-3-sn-

phosphatidylethanolamine (DOPE) (Avanti Polar-Lipids, Alabaster, Al) at a molar ratio

of 1:1. The mixture was evaporated to dryness in a round-bottomed flask using a rotary

evaporator at room temperature. The resulting lipid film was dried by nitrogen for an









additional 10 min to evaporate any residual chloroform. The lipid film was re-suspended

in sterile water to a final concentration of 1 mg/ml based on the weight of cationic lipid.

The resultant mixtures were shaken in a water bath at 35C for 30 min. The suspensions

then were sonicated using a Sonic Dismembrator (Fisher Scientific, Pittsburgh, PA) for 1

min at room temperature to form homogenized liposomes. The particle-size distribution

of liposomes was measured using a NICOMP 380 ZLS instrument (Santa Barbara, CA).

The average particle diameter was 144.0 77.0 nm. Liposomes were stored at 4C and

used within 3 months.

Co-culture Assay

Transwell (Coming, Coming, NY) 6-well dishes with a membrane pore size of

0.4 uM were used. Caki-1 cells were seeded at 5 x 104 in the transwell inserts and MHE

or HMVEC-L cells were plated at 5 x 104 per well in the 6-well dishes and allowed to

attach overnight. The Caki-l cell medium then was replaced with serum free medium

containing 1 pM V515 or B460 AS-ODNs or control ODNs delivered with liposome

(DOTAP:DOPE). After a 5 hr of treatment, medium containing 10% heat inactivated

FBS was added to yield a final FBS concentration of 2.5%. The transwells containing

treated Caki-l cells were assembled with 6-well dished containing MHE and HMVEC-L

cells and incubated at 370C for 72 hr at which time the numbers of MHE or HMVEC-L

cells were determined by haemocytometer count (Figure 4-1).

Migration Assay

Caki-1 cells were set at 1 x 105 per well in 24-well dishes and allowed to attach

overnight. The Caki-1 cells then were treated with 1 .M V515 or B460 AS-ODNs or

control ODNs for 24 hr. HTS FluoroBlok inserts (Becton Dickinson, Franklin Lakes, NJ)









with a pore size of 8.0 urm were assembled into the 24-well dish with the Caki-l cells.

MHE or HMVEC-L cells were grown in T-150 flasks to about 80% confluence. The

endothelial cells were stained in medium containing 10 p.g/ml Di-l (Molecular Probes,

Eugene, OR) for 24 hr, washed 4 times with PBS, collected and added into the

FluoroBlok inserts (5 x 104 MHE or HMVEC-L) and incubated for another 24 hr. The

number of migrated endothelial cells then was determined by direct measurement of the

fluorescence in the bottom well using a CytoFluor 4000 plate reader (Perceptive

BioSystems, St. Paul, MN). (Figure 4-4)

Intradermal Angiogenesis Assay

Caki-1 cells were treated with AS-DONs for 5 hr in vitro as described before. The

cells were then collected and inoculated intradermally (5 x 104) in a volume of 10 .1 at 4

sites on the ventral surface of nude mice. One drop of 0.4% trypan blue was added to the

cell suspension before injection, which making it lightly colored, simplified subsequent

location of the sites of injection. Three days later the mice were killed, the skin was

carefully separated from the underlying muscle and the number of vessels counted using

a dissecting microscope (Sidky and Auerbach, 1976). Scoring of all of the reaction areas

was carried out at the same magnification (5x) and only vessels readily detected at this

magnification were counted. The sites of injection, recognized by local swelling and blue

staining, were exposed by carefully removing fat or other tissue covering the area. All

vessels that touched the edge of the tumor inoculates were counted. All the animals in the

experiments were pre-coded and vessel counts in each animal were scored twice. The

resultant data points for each treatment group were pooled for statistical analysis

(Wilcoxon rank sum test).









Results

Anti-angiogenic efficacy of VEGF/bFGF AS-ODNs treatments was first

evaluated in vitro. The transwell co-culture system was used to examine the effect of AS-

ODNs treatment of Caki-1 cells on the proliferation of co-cultured endothelial cells

(Figure 4-1). This setting allowed the constant exchange of growth factors without direct

tumor-endothelial cell-cell interaction and mimicked the paracrine interaction between

tumor and endothelial cells. Since a human RCC tumor cell line and xenograft in nude

mice were used in these studies, both human (HMVEC-L) and mouse (MHE) endothelial

cells were studied. Caki-1 tumour cells were grown in transwells with 0.4 pmr membrane

pores. The effects of pretreating Caki-1 tumor cells with VEGF or bFGF AS-ODNs on

endothelial cell proliferation then were determined (Figure 4-2 and 4-3). The results

showed that compared to untreated Caki-l cells, Caki-l cells pre-treated with V515 or

B460 significantly inhibited both HMVECV-L and MHE cell proliferation. Once again,

treating Caki-1 cells with a variety of control PS-ODNs had no effect on HMVEC-L or

MHE cell growth.

To test whether a reduction in VEGF or bFGF expression by tumor cells could

affect endothelial cell migration, HMVEC-L or MHE cells were stained with 10 gg/ml

Di-I for 24 hours and added into Fluoroblok inserts placed into 24 well dishes containing

Caki-l tumour cells treated with V515. The number of pre-labelled endothelial cells

which migrated through the 8 mu pore size membranes in a 24 hr period were quantified

by determining the fluorescence intensity in the bottom well (Figure 4-4). The results

showed (Figure 4-5) that 24 hr after co-culturing the two cell populations -45%

(p<0.05, student's t-test) and 37% (p<0.05, student's t-test) fewer MHE or HMVEC-L









cells respectively migrated through the membrane in the presence of V515 treated Caki-1

cells compared to untreated or scramble control AS-ODNs treated Caki-1 cells.

Similarly, -37% (p<0.05, student's t-test) and -33% (p<0.05, student's t-test) fewer

MHE or HMVEC-L cells respectively migrated through the membrane in the presence of

B460 treated Caki-I cells compared to untreated or scramble treated control AS-ODNs

(Figure 4-6).

Although, the in vitro studies indicated that treating Caki-l tumor cells with

VEGF or bFGF mRNA targeted AS-ODNs down-regulated VEGF/bFGF protein

production sufficiently to affect the proliferation and migration of endothelial cells, it was

important to demonstrate that such treatments also could affect Caki-1 cell induction of

angiogenesis in vivo. To examine this possibility Caki-1 cells that had been treated with

V515, B460 or control ODNs were injected intradermally and the number of vessels

induced were counted 3 days later (Figure 4-7). While untreated Caki-l cells and control

ODNs treated Caki-l cells had very similar angiogenic potency in vivo (both groups

induced -44-46 new vessels in the assay period), the angiogenic potential of Caki-I cells

that had been pre-treated with VEGF AS-ODNs (V515) was found to be significantly

impaired; only -25.5 (p<0.05, Wilcoxon rank sum test) new blood vessels were observed.

Similarly, bFGF AS-ODNs (B460) treated tumor cells also induced less vessels, -27 new

blood vessels (p<0.05, Wilcoxon rank sum test). More importantly, the most significant

inhibition of formation of new blood vessels was observed when the Caki-1 cells were

treated with both V515 and B460, only -20 new vessels developed (p<0.05, Wilcoxon

rank sum test).









Discussion

Anti-angiogenesis treatment strategies represent a new approach to cancer

management. Given that solid tumors cannot progress effectively without the generation

of new blood vessels, various tacks have been taken to interfere tumor angiogenesis. One

possible target which has received considerable attention is the pro-angiogenic factor

VEGF. VEGF can induce endothelial cell proliferation and migration in vitro (Soker et

al., 1997; Hanahan and Folkman, 1996a; Hanahan and Folkman, 1996b) and

angiogenesis in vivo (Leung et al., 1989; Plate et al., 1992a; Plate et al., 1992b). Its

expression level has been associated with a variety of tumors and correlated to treatment

outcome (Gasparini et al., 1997; Maeda et al., 1996a; Maeda et al., 1996b). Basic

fibroblast growth factor (bFGF) is another important pro-angiogenic factor (Yoshida et

al., 1996; Hoying and Williams, 1996). It also has been found to associated with different

tumors and correlated to treatment outcome, especially RCC (Nguyen et al., 1994b;

Nanus et al., 1993; Miyake et al., 1996). Antisense oligodeoxynucleotide technology

provides an approach for inhibiting gene expression with target specificity as a particular

advantage (Stein and Cheng, 1993; Engelhard, 1998a; Engelhard, 1998b). Effective AS-

ODNs against VEGF and bFGF have been identified and tested in vitro (Chapter 3).

To evaluate the VEGF/bFGF suppression through the use of AS-ODNs on Caki-1

tumor angiogenesis, both in vitro and in vivo efficacies were studied. The co-culture

system, which allows constant exchange of growth factors between tumor and endothelial

cells was used to study the treatment on endothelial proliferation. The tumor cells also

served as primary source of pro-angiogenic growth factors for endothelial cells. This

provides a setting that closely mimics the in vivo situation. Significantly impaired









proliferation potential of both human and mouse microvascular endothelial cells were

observed after VEGF/bFGF AS-ODNs treatment (Figure 4-2 and 4-3). In order to

evaluate the AS-ODNs treatment on endothelial cell migration potential, a very similar

co-culture system (FluoroBlok system) utilizing 24 well dishes and larger pore sized

membranes (8 pM) was used. The results showed that suppression of VEGF/bFGF

expression by AS-ODNs was sufficient to inhibit the migration of endothelial cells in

response to pro-angiogenic growth factors produced by tumor cells (Figure 4-5 and 4-6).

VEGF/bFGF AS-ODNs treatment of Caki-1 cells led to inhibition of endothelial

cell proliferation and migration in both human and mouse microvascular endothelial cells

(Figure 4-2, 4-3, 4-5 and 4-6) suggesting that such treatments should also exert their

effects in a mouse model. In deed, the anti-angiogenic efficacy of VEGF/bFGF AS-

ODNs was readily demonstrated in vivo using the intradermal angiogenesis assay (Figure

4-8).

These results not only support the role of VEGF and bFGF as important pro-

angiogenic growth factors in Caki-1 cell induced angiogenesis, but also clearly suggest

that inhibition of cancer cell VEGF or bFGF expression may ultimately impact tumor

growth. The use of AS-ODNs against VEGF/bFGF was sufficient to illicit anti-

angiogenic effects both in vitro and in vivo.

Taken together, these findings suggest that AS-ODNs targeted to VEGF and

bFGF are effective in inhibiting Caki-l tumor cell induced angiogenesis. They further

implying that such a treatment strategy may have utility in the treatment of RCC.







89












I Endothellal cells


Figure 4-1 Transwell co-culture system for evaluating VEGF/bFGF AS-ODNs treatment
of Caki-1 cells on endothelial cell proliferation.







90




4.0x10 -
0 Untreated
:S0_ DOTAP
3.5x10 0 Cortrol
O V515
3.0x10s

S2.5x10

U 2.Ox0O.

O 1.5x10



5.OxtO


MHE HMVEC-L
Cell lines



Figure 4-2. Treatment of Caki-l cells with VEGF AS-ODNs on the growth of co-
cultured endothelial cell. Caki-1 cells were either untreated, treated with liposome alone,
treated with 1 tM scramble control ODNs or VEGF AS-ODNs. Each bar represents the
mean S.E. of 3 independent experiments. Stars indicate statistical significance
compared to untreated control group (p<0.05, student's t-test).












4.0x105
T Urtreated
3.5x10s DOTAP
Control
3.ox0 B46

S2.5x10l

2.sxlo"

O 1.5x10s

1.0x10.

5.0x10


MHE HMVEC-L

Cell lines



Figure 4-3. Treatment of Caki-1 cells with bFGF AS-ODNs on the growth of co-cultured
endothelial cell. Caki- cells were either untreated, treated with liposome alone, treated
with 1 vM scramble control ODNs or bFGF AS-ODNs. Each bar represents the mean
S.E. of 3 independent experiments. Stars indicate statistical significance compared to
untreated control group (p<0.05, student's 1-test).








92





Endothellal cells

Sabed with D-I- k




Fluorescence
opaque
membrane

Tumor cells










Figure 4-4. Transwell co-culture system for evaluating VEGF/bFGF AS-ODNs treatment
of Caki-1 cells on the migration potential of endothelial cells.












900
00 Untreated
800. DOTAP
SScramble
S700 V515




0 400-

300

20000
100


MHE HWMEC-L

Cell line


Figure 4-5. Treatment of Caki-l cells with VEGF AS-ODNs on the migration potential of
co-cultured endothelial cells. Caki-1 cells were either untreated, treated with liposome
alone, treated with pLM scramble control ODNs or VEGF AS-ODNs. Each bar
represents the mean S.E. of 3 independent experiments. Stars indicate statistical
significance compared to untreated control group (p<0.05, student's t-test).




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EEFTIE0CQ_IGKHAL INGEST_TIME 2012-09-24T14:25:28Z PACKAGE AA00011835_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES