EQUILIBRATIVE-SENSITIVE NUCLEOSIDE TRANSPORTER FUNCTION AND REGULATlON IN GEMCITABINE SENSlTlVîTY

AND RESISTANCE:

IS THERE A POTENTIAL THERAPEUTRC BENEFIT FOR PANCREATIC CANCER?

David R. Rauchwerger

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

Graduate Department of Pharrnacology, University of Toronto

@ Copyright by David R. Rauchwerger (1999)

National Library Bibiiothèque natiorra!e du Canada

Acquisitioris and Acquisitioris et Bibliogtaphic Services services bibiiog raphiques

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

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

L'auteur a accordé une licence non exclusive permettant à la Bibiiotheque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/jïlm, de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thése ni des extraits substantiels de celle-ci ne doivent ê e imprimés ou autrement reproduits sans son autorisation.

Equilibrative-Sensitive Nucleoside Transporter (eeM) Function and Regulation in Gemcitabine Sensitivity and Resistance: b lhere a Potential

Therapeutic Benefit for Pancreatic Cancer? David R. Rauchwerger

ûepartment of Pharmacology University of Toronto

Abstract

Salvage of preformed nucleosides requires transport across the plasma

membrane by specific transport proteins and subsequent conversion to their ribo-

and deoxyribonucleotide foms. In mammalian cells, plasma membrane

transport occurs by sodiumdependent (concentrative) and sodium-independent

(equilibrative) mechanisms. These transport systems are also the route of

cellular uptake for many synthetic nucleoside analogue agents used in cancer

treatment, including gemcitabine (a novel deoxycytidine analogue). This thesis

examines the in vitro effects of gemcitabine on cytotoxicity and the modulation of

these effects by the es-nucleoside transporter and two DNA synthesis inhibitors

(5-fluorouracil and tomudex), with a specific focus on pancreatic cancer.

Cytotoxicity was assessed by clonogenic assay in one human bladder

(MGH-U1 ) and three human pancreatic cancer cell lines (PANC-1, HS-766T, PK-

8). Basal levels of es-NT were quantified in al1 four cell lines by flow cytometric

analysis. Combination experiments were cawied out to determine if upregulating

the es-NT modulates increases in sensitivity to gemcitabine. In two pancreatic

cell lines (PANC-1, HS-766T), treatment with 5-FU followed by gemcitabine

yielded increased cytotoxicity. This effect was also seen in the HS-7661 cell line

when pre-treated with tomudex. For these concentrations of 5-FU and tomudex,

es-NT content was found to be increased over basal levels.

Acknowledgements

I would first like to thank my supervisor, Dr. Malcolm Moore for his constant encouragement. guidance and support during the past two yean. Your comments. criticisms and suggestions were instrumental in my development as an independent, critical, research scientist.

I would also like to thank my advisor. Dr. Gerald Goldenberg and the members of my thesis defense cornmittee, Dr. David Riddick, Dr. David Hedley, Dr. Ted lnaba and Dr. Patricia Harper for their feedback and suggestions.

Special thanks go to Juliet Sheldon for her help with flow cytornetry and Pat for ail her assistance, expertise and. of course, patience through my stay at OCI.

Finally, many thanks go to rny family for their support, love and encouragement that I have always had and will continue ?O enjoy.

. . Abstract ............................................................................................................. II ... Acknowiedgernents ........................................................................................... 111

Table of Contents ............................................................................................. iv List of Tables .................................................................................................... vi List of Figures .................................... ... ........................................................... vii . . ... List of Abbreviations ........................................................................................ vu1

CHAPTER 1 : INTROOUCTION

Introduction ........................................................................................... -2 Pancreatic Cancer ................................................................................ -2 Antirnetabolites ..................................................................................... -6 Cytidine Analogues ................................................................................ 6

. . 1.4.1 Gemc~ta bine ............................................................................... -7

1 .4.1 -1 Mechanism of Action of Gemcitabine ........................... 7 1.4.1.2 Drug Resistance and Gemcitabine ............................ -13

................................. 1 .4.1.3 Phase I Studies of Gemcitabine 15 ...................... 1 .4.1 -4 Phase II and lil Studies of Gerncitabine 16

Thymidylate Synthase Inhibitors .................................................. 1 8 1 . 5. 1 5-Fluorouracil ............................................................................ 19

t 5.1 -1 Mechanisrn of Action of 5-fluorouracil ....................... -21 1 .5 . 1 -2 Dnig Resistance and 5-Fluorouracil ........................... 23

1 .5.2 Tomudex (Raltitrexed) ............................................................. -24 1 S.2.1 Mechanism of Action of Tomudex ............................. -25 1 .5.2.2 Drug Resistance and Tornudex ................................. -27

Nucleoside Transporters ...................................................................... 27 7 .6.1 Gemcitabine and Nucleoside Transport ................................... -34 5-(SAENTA-x8)-Fluorescein Binding Assay ......................................... 35 Combination Chemotherapy ............................................................... -37

CHAPTER 2: RATIONALE. HYPOTHESIS AND OBJECTIVES

2.1 Rationale Behind the Project .............................................................. -41 2.2 Hypotheses ........................................................................................ -42 2.3 Objectives .......................................................................................... -42

CHAPTER 3: MATERIALS AND METHODS .

3.1 Cell Lines and Reagents ...................................................................... 45 3.2 Cell Cytotoxicity Assay ........................................................................ 46

3.2.1 Single Drug Exposures ............................................................ -46 3.2.2 Drug Corn bination Exposures .................................................. -47

3.3 5-(SAENTA-x8)-Fluorescein Binding Assay ........................................ -47

CHAPTER 4: RESULTS

.................................................................................... 4.1 Growth Curves 50 ............................................................................. 4.2 Cytotoxicity Assays -50

............................................................. 4.2.1 Single Dnig Exposures 56 ........................................................................... 4.3 Combination Studies 56

4.4 Equilibrative-Sensitive Nucleoside Transporter Quantitation ................ 70

CHAPTER 5: Discussio~

5.1 Single Dnig Cytotoxicity ...................................................................... -81 ........................................................................ 5.1 1 Gemcitabine ..... : 82

............................................................................ 5.1 -2 5-FI uorouracil 84 ....................................... .................... 5.1 -3 Tomudex ............ ... ... 86

.............................................................................. 5.2 Dnig Combinations 86 ............................................... 5.2.1 Gemcitabine and 5-Fluorouracil.. 88

..................................*................... - 5.2.2 Gemcitabine and Tomudex -88 5.3 EquilibrativeSensitive Nucleoside Transporter Expression ................ -89 .

CHAPTER 6: CONCLUS~ONS AND FUTURE DIRECTIONS

........................................................................................ 6.1 Conclusions -93 6.2 Future Directions ................................................................................. 94

Appendix A . Statistical Analysis ......................................................................... 97

................................................................................................... . References - 1 00

CHAPTER 1:

Table 1.2a: Randomized Trials in Advanced Pancreatic Carcinoma ................ 5 Table 1.4a: Phase II Trials of Gerncitabine ... . . .. . . . . . ........... . .... ........ ... ..... .. .... ... 1 7 Table 1.6a: Characterized Human Nucleoside Transporter Systems ..... .. ...... 31

CHAPTER 4:

Table 4.2a: lCso Values. .. . . .. . . . . .. ... .. . . . . .. ... ............... . . . . . .. . . . . . .. .... . .. . . . . . . .. . . .... . ..... 61 Table 4.26.- iCm Values ... .. . . .. .... ..... . ... ........... .. ........... . .. .-. ..... .. ....... .. . . . . . .. ...... ...... 61 Table 4.3a: Dnig Concentrations Used in Combination Stucfies . . .. . . . . . . . .. . . . . . . . -62 Table 4.4a Es Nucleoside Transporter Content of Four Human Cell Lines and

their Gemcitabine Sensitivities .. . . . . . . . . . . .. . . . . . . . . . .. . . . . . .. . . . . . . . ... . . .. ... .. -76 Table 4.46 Es Nucleoside Transporter Content After Drug Treatment ... . ....... 78

CHAPTER 5:

Table 5.1 a: Human Bladder, Colon, Ovarian, Pancreatic and Squarnous Cell Carcinoma Cell Line Doubling Times and Gemcitabine Sensitivities . . . . . . . . . . . . . . . . . . .. . . . . . . .. . . . . . .. . .. . .. . . . . . . . . . . . . .. . . . . . .. . . . . . . . . .. .. .. . . ... -83

Table 5.1 6: Human Bladder, Colon and Pancreatic Tumour Cell Line Doubling Times and 5-FU and Tomudex Sensitivities ....... . . . . .. . .. . .. .. .... ........ 85

Append ix:

Table 1 : Staastical Differences Between Basal and Drug-Treated Levels of es Nucleoside Transporter in the MGH-U1 cell line .......... ............. 97

Table 2: Statistical Differences Between Basal and Drug-Treated Levels of es Nucleoside Transporter in the PANC-1 cell line ........................ 97

Table 3: Statistical Differences Between Basal and Dnig-Treated Levels of es Nucleoside Transporter in the HS-766T cell line . . . . . . . . . . . . . . . . . . . . . . -98

CHAPTER 1 :

Figure 1.4a: Figure 1.46: Figure 1.4~: Figure 1.5a: Figure 1.5b: Figure 1.5~: Figure i . 5d. Figure 1.6a: Figure 1.7a

CHAPTER 4:

Figure 4.1a: Figure 4.1 b Figure 4.7 c: Figure 4.ld Figure 4.2a: Figure 4-26: Figure 4.2~: Figure 4.2d: Figure 4.3a: Figure 4.36: Figure 4.3~: Figure 4.3d: Figure 4.3e: Figure 4.3f Figure 4.4a Figure 4-46 Figure 4 . 4 ~ Figure 4.4d Figure 4.4e

Chernical Structures of Deoxycytidine. Ara-C and Gemcitabine .... 8 Metabolic Scheme of Gemcitabine .............................................. 10

............................ Self-Potentiation Mechanisms of Gemcitabine 12 ......................... Chernical Structures of Uracil and 5.Fluorouracil 20

Mechanism of Action of 5-Fluorouracil ......................................... 22 Chernical Structure of Tomudex .................................................. 24 Mechanism of Action of Tomudex ................................................ 26

................ Structures of Nucleosides and Nuclecside Analogues 29 ..................... Chemica f Structure of 5-(SAENTA-x8)-Fluoresœin 36

Growth Curve for MGH-U1 Cell Line .......................................... 5 2 Growth Cuwe for PANGl Cell Line .................................... ... . 5 3 Growth Cuwe for HS-766T Cell Line ................... .. ................... 54

.................................................. Gmwth Cuwe for PK-8 Cell Line 55 Cell Survival Cuwes for MGH-Ul Cell Line ................................ 57 Cell Survival Cuwes for PANGI Cell Line .................................. 58 Celi Suwival Cuntes for HS-766T Cell Line ................................. 59 Cell Survival Cuwes for PK-8 Celi tine ....................................... 60 Combination Studies of Gemcitabine and 5-FU in MGH-U1 ........ 64 Combination Studies of Gerncitabine and Tomudex in MGH-U1 . 65 Combination Studies of Gemcitabine and 5-F U in PANC-1 ......... 66 Combination Studies of Gemcitabine and Tomudex in PANGI .. 67 Combination Studies of Gemcitabine and 5-FU in HS-766T ....... -68 Combination Studies of Gemcitabine and Tornudex in HS-766T . 69 Binding Cuwes of 5x8 in MGH-U1 Cell Line +/- NBMPR ............. 71 Binding Cunres of 5x8 in PANCI Cell Line +/- NBMPR .............. 72 Binding Cuwes of 5x8 in HS-766T Cell Line +/- NBMPR ............. 73 Binding Cuwes of 5x8 in PK-8 Cell Line +/- NBMPR ................... 74 Sample Calibration Curve For RCP-30-5 Rainbow Beads .......... -75

vii

a-MEM 2CdA 5-FU 5x8 ara-C cib

cif

CDP CTP dCK dCMP (d)CMPD dCTP dFdC dFdCDP dFdCMP dFdCTP dFdU dFdUMP DNA dTMP dTrP dUMP ei es es-NT FdUMP FdUTP FITC FPGS FUMP GST HEPES G o G o

a-modification of minimal essential medium 2-chlorodeoxyadenosine 5-Fluorouracil 5-(SAENTA--)-Fluorescein Cytosine arabinoside Concentrative and insensitive to NBMPR, accepts broad array of nucleosides as permeants Concentrative and insensitive to NBMPR, accepts fomycin 6 as a pemeant Concentrative and insensitive to NBMPR, accepts thymidine as a pemeant Concentrative and sensitive to NBMPR Concentrative and sensitive to NBMPR, accepts guanosine as a pemeant Cytidine diphosp hate Cytidine triphosphate Deoxycytidine kinase Deoxycytidine monophosphate (0eoxy)cytidine monophosphate deaminase Deoxycytidine triphosphate 2',2'dlluorodeoxycytidine (gemcitabine) 2'2'-difluorodeoxycytidine diphosphate 2',2'-âifluorodeoxycytidine monophosphate 2',2'difluorodeoxycytidine triphosphate 2',2'difluorodeoxyundine 2',2'difluorodeoxyuridine monophosphate Deoxyribonucleic acid Z'deoxythymidine-5'-monophosphate Deoxythyrnidine triphosphate Z'deoxyuridine-5'-monophosphate Equilibrative and insensitive to NBMPR Equilibrative and sensitive to NBMPR Equilibrative-sensitive nucleoside transporter 5'-fluoro-2'deoxyuridine-5'-monophosphate 5'-f luorodeoxyuridine triphosphate Fluorescein-54sothiocyanate Folyl polyglutamate synthetase Fluorouridine monophosphate Glutathione S-transferase N-(2- hydroxyeth yl(piperazine)-N'-(4-ethanesfonic acid)) Concentration required to inhibit cotony formation by 50% Concentration required to inhibit colony formation by 90%

viii

LV MDR MESF MRP NBMPR PBS p-gp RFC RNA RR SAENTA TS TTP

Leucovorin Multi-drug resistance Molecules equivalent soluble fluorescein MultMrug resistance associated protein S-(p-nitrobenzy1)-64hioinosine; nitrobenzylmercaptopurine n'boside Phosphate-buffered saline P-glycoprotein Reduced folate carrier Ribonucleic acid Ri bonucleotide reductase 5 ' - ~ - ( 2 - a m i n o m e t h y l ) - ~ ~ - ( 4 - n i t r o b e n q l ) ~ i n e Thymidylate synthase Thymidine triphosphate

1.1 Introduction

There exist two pathways of DNA synthesis within human cells: the de

novo and salvage pathways. Salvage of prefomed nucleosides requires their

transport across the plasma membrane and subsequent conversion to their ribo-

and deoxyribonucleotide forms. The efficiency of the salvage pathway of DNA

synthesis can have major effects on cellular sensitivity ta a wide variety of

antirnetabolite drugs because most of these dnigs, including nudeoside

analogues, undergo a generic two-stage metabolism: transpoct into the cell

('"salvage"), followed by phosphorylation to their active triphosphate foms.

Nucieoside analogues taken up by the salvage pathway include cytosine

arabinoside (ara-C), 2-chlorodeoxyadenosine (2CdA) and gemcitabine (dFdC).

Therefore, low activity of this pathway may be a potential mechanism of drug

resistance to these agents. This paper examines the in vitro effects of the novel

nucleoside analogue gemcitabine as a single agent and in combination with two

DNA synthesis inhibitors, 5-FU and tomudex, lod<ing at cytotoxicity and the

effects on the equilibrative-sensitive nucleoside transporter (es-NT). In addition,

schedule dependence of bmg administration was examined in order to provide a

rationale for future clinical trials protocol design.

1.2 Pancreatic Cancer

Adenocarcinorna of the pancreas is the fifth leading cause of cancer-

related deaths in North America exceeded only by lung, colorectal, prostate and

breast cancers (Moore, 1996). Surgery is the only curative treatment currently

available; however, greater than 80% of patients with carcinoma of the exocrine

pancreas are metastatic at diagnosis. Chemotherapy and radiation therapies

rnost commonly play a palliative role in pancreatic cancer care and have not

shown a significant impact on 5-year survival rates (Clark et al., 1996). At

present, pancreatic cancer has the worst 5-year survival rate of any cancer - less

than 5% of al1 pancreatic cancer patients survive five yean (Clark et al., 1996;

Moore 1996).

A variety of drug resistance rnechanisms have been described in

pancreatic tumours. These include P-glycoprotein (P-gp), mufti-drug resistance-

associated protein (MRP), glutaaiione S-transferases (GSTs) and metallothionein

(Collier et el., 1994; Dietel, 1996; Holm et al., 1994; Miller et al., 1 996; Ohshio et

al., 1996; Verovski et al., 1996). P-gp is an integral membrane-bound efflux

pump capable of removing a variety of structurally and functionally unrelated

cytotoxic drugs including the anthracyclines, epidophyllotoxins, alkylating agents

and vinca alkaloids, from tumour cells (Dietef, 1 996; Holm et al., 1 994; Verovski

et al., 1996). Holm et al. (1 994) found that by decreasing the level of mdrl

mRNA expression in a human pancreatic carcinoma cell line resistant to

doxorubicin, one could inhibit the formation of P-gp and reduce the cell'a

resistance to doxorubicin. GSTs function as enzymes of detoxification. By

catalysing the conjugation of potentially mutagenic electrophilic compounds with

reduced glutathione, cells may be protected from toxic cell damage (Dietel,

1996). The Pi class of GSTs (those with acidic isoelectric points) are expressed

by a vanety of human tumours at higher than normal values. It is this isozyme

that is more prominentiy expressed in the majority of pancreatic

adenocarcinornas (Collier et a/., 1994). Because a variety of cytotoxic drugs are

metabolised by GSTs, elevated levels of this enzyme may potentially lead to

tumour dnig resistance.

Over the past ten yean, many new dnigs have been tested for activity

against pancreatic cancer. Over forty phase II studies of new single agents and

combination regimens have been reported (Moore, 1994). None has shown a

response rate greater then 20%, the usual standard that mus! be met for further

testing to be warranted (table 1.2a). In randomized trials, gemcitabine was the

first and only chemotherapeutic agent that has been shown to have any

meaningful impact on either survival or disease related symptoms in pancreatic

adenocarcinorna. Much interest now exists in maximizing the benefit of

gemcitabine against pancreatic cancer and other solid tumours.

Best supportive care Gemcitabine

5-FU

5-FU

5-FU

5-FU, doxorubicin, cisplatin 15

5-FU, methyl CCNU 10

5-FU, methyl CCNU, streptozotoxin 7

5-FU, Mi-C, methotrexate, cyclophospharnide, vincnstine Mi-C, 5-FU

FAM

FAM

Melphalan

SMF

SMF

SMF

SMF (modified)

SMF 10 10

Cisplatin, Ara-C, caffeine 5 5

'The rnajority of patients in these studies did not have measurable disease.

5-FU, 5-Fluorouracil; CCNU, lomustine; Mi-C, mitomycin C; FAM, 5-FU, doxorubicin, mitomycin C; SMF, streptozotoxbin, mitomycin C, 5-FU; N/S, not significant.

Table 1.2s Randomized trials in advanced pancreatic carcinoma (adopted from Moore, 1 996; Buris et al., 1 997)

1.3 Antirnetabolites

The antirnetabolite classes of chemotherapeutic agents generally function

as inhibitors of macromolecule biosynthesis that block cell replication. These

antirnetabolites compete with endogenous substrates in cellular metabolic

processes feading to what can be thought of as 'competitive stawation' (Kinsella

and Smith, 1998). 5FU, tomudex and gemcitabine are al1 members of this class

of oncolytic agents which act at different points in the cellular metabolic process.

Three important classes of antirnetabolites are: Nucleoside analogues (ara-C,

gemcitabine), anti-folates (methotrexate) and thymidylate synthase (TS)

inhibitors (tomudex, 5-FU).

1.4 Cytidine Analogues

Cytidine analogues are compounds that closely resemble the endogenous

nucleoside cytidine. These agents require transport into the cell and

phosphorylation to their active, triphosphate fonns at which point they compete

with deoxycytidine triphosphate (dCTP) for incorporation into DNA (White et a!.,

1987; Wiley et al., 1985). These drugs may also exert their cytotoxic effects by

acting at various points along the DNA synthesis pathway. Cytosine arabinoside

(ara-C) for example, is a competitive inhibitor of DNA polymerase. Ara-C has

been the prototypical drug for this class of chemotherapeutic agents for some

time.

1.4.1 Gemcitabine

Gerncitabine (2',2'difluorodeoxycytidine (dFdC); eemzarTM) is a cell cycle

dependent (S-phase specific) deoxycytidine analogue and oncolytic agent of the

antimetabolite class. It has two fluorine atoms substituted for the hydrogens of

the P-carbon atom in the sugar moiety of its deoxycytidine parent (figure 1.4a).

Gemcitabine was initially synthesized as a potential anti-viral agent and

proved to be quite active in vitro against both DNA and RNA viruses. Its

unfavourable therapeutic index when administered on a daily schedule, however,

precluded any fumer development of gemcitabine as an anti-viral dnig

(Guchelaar et al., 1996).

Gemcitabine is currently used in the palliative care of non-small cell lung

cancer and is the only agent that has shown any clinical benefit toward

pancreatic cancer. As pancreatic cancer has one of the poorest prognoses of al1

types of cancer, considerable interest lies in discovenng ways to improve the

efficacy of gemcitabine, both as a single agent and in combination therapy.

1.4.1.1 Mechanism of Action of Gerncitabine

Gemcitabine, like most antimetabolite drugs, must first be transported into

the cell and then be phosphorylated to its active, triphosphate fon . Transport of

gemcitabine occurs via the nucleoside transporter, of which there exist multiple

foms (section 1.6, Heinemann et al., 1995). Transport of nucleosides into the

ce11 becomes saturated at extracellular nucleoside concentrations of 1 pfUl (Wiley

et al., 1985; White et al., 1987; Pressacco et al., 1995). Once inside the cell,

Oeoxycytidine Cytosine Arabinoside Gemcitabine

Figure 1 Chernical structures of deoxycytidine, cytosine arabinoside ( Ara-C) and gerncitabine.

numerous enzymatic reactions lead to the formation of gemcitabine triphosphate

(dFdCTP) (figure 1.4b). Gerncitabine is phosphorylated to its monophosphate

fom 2',Zdifluorodeoxycytidine monophosphate (dFdCMP) by the enzyme

deoxycytidine kinase (dCK). Phosphorylation of gemcitabine (by dCK) is also a

saturable process, however, uptake of gemcitabine into the cell is the rate-

limiting step of its rnetabolic process because when gemcitabine transport is

saturated, dCK phosphorylation is not (Gucheiaar et al., 1996). The specific

enzyme that phosphorylates dFdCMP to its diphosphate fom (dFdCDP) has not

yet been identified, but is assurned to be a base-specific (deoxy)cytidine

monophosphate ((d)CMP) kinase that lacks specificity for the cabhydrate

(Plunkett et al., 1 995). dFdCDP is converted to the triphosphate fom (dFdCTP)

by the ubiquitous non-specific enzyme, nucleoside diphosphate kinase (Plunkett

et al., 1 995; Heinemann et al., 1995; Guchelear et al., 1 996). Incorporation of

dFdCTP into DNA is most likely the major mechanism by which gemcitabine

exerts its cytotoxic actions. After incorporation of dFdCTP on the end of the

elongating DNA strand, one additional deoxynucleotide is added before the DNA

polymerases are unable to continue DNA synthesis. This action, termed

"rnasked-chain temination", locks the drug into place within the DNA as

proofreading enzymes are unable to remove dFdCTP from this position (Plunkett

et al., 1995; Huang et al., 1991).

Elirnination of gemcitabine occurs via three routes (figure 1.46): (1)

transport out of the cell via equilibrative nucleoside transporters. (2)

Deamination of gemcitabine by cytidine deaminase into 2',2'-difluorodeoxyuridine

DNA

dCMP deaminase dFdCMP * dFdUMP

cvtidine deaminase dFdC dFdU

Uracil + difluorodeoxvribose

Figure 1.46. Metabolic scheme of gemcitabine. dFdC, gemcitabine; dFdCDP, gemcitabine diphosphate; dFdCDP, gemcitabine diphosphate; dFdCTP, gemcitabine triphosphate; dCMP, deoxycytidine monophosphate; dFdU, difluorodeoxyuridine; dFdUMP, dlluorodeoxyuridine monophosphate.

(dFdU). (3) Deamination of dFdCMP by dCMP deaminase into 2',2'-

difluorodeoxyuridine monophosphate (dFdUMP). Both dFdUMP and dFdU are

ultirnately broken down into uracil and difluorodeoxyribose (Plunkett et a/., 1995;

Guchelear et al., 1996).

Gerncitabine, abng with its metabolites, interact with a variety of cellular

me!abolic regulatory processes. These interactions serve to enhance the overall

inhibitory actions of gerncitabine on cell viability. Termed 'self-potentiation',

these interactions are not evidenced to such an extent in other anticancer drugs

(Plunkett et al., 1995; Heinemann et al., 1992; Gandhi et al., 1991). The

pathways of gemcitabine self-potentiation and the sites of action of gemcitabine

and its metabolites are summarized in figure 1.4~. The six major mechanisms of

gemcitabine self-potentiation are as follows: (1) dFdCDP is an inhibitory

substrate for ribonucleotide reductase (RR), the enzyme that is responsible for

producing deoxynucleotides required for both DNA synthesis and repair. (2) The

resulting decrease in cellular deoxynucleotides, in particular deoxycytidine

triphosphate (dCTP), is essential in that dFdCTP directly cornpetes with dCTP for

incorporation into DNA by DNA polymerases. This decrease in cellular dCTP,

therefore, increases gemcitabine nucleotide incorporation into DNA. (3) dCK

phosphorylation of gemcitabine is inhibited by deoxycytidine and dCTP

(Heinemann et a!., W88; Bouffard et al., 1993). Therefore, when cellular dCTP

pools are lowered, the rate of gemcitabine phosphorylation is increased resulting

in greater dFdCTP accumulation for incorporation into DNA. These higher levels

of dFdCTP would also act to rnaintain inhibition of ribonucleoside diphosphate

DNA

UTP

--a-- synthetase ,, 1 +

CTP CDP

Gemcitabine dFdU

Figure 1.4~. Self-potentiating mechanisrns of gerncitabine. The numbered pathways are discussed sequentially in the text. The dashed lines indicate inhibitory reactions. RR, ribonucleotide reductase. ---a indicates an inhibitory pathway.

reductase. (4) dCTP is also a required cofactor in the activity of dCMP

deaminase, the rate-limiting enzyme of gemcitabine nucleotide elimination

(Heinemann et al., 1992; Xu et al., 1992). When the cellular levels of dCTP

decline as a result of ribonucleotide reductase inhibition by dFdCDP, dCMP

deaminase activiîy also decreases accordingly. This results in a lower rate of

removal of gemcitabine nucleotides from the celf and most likely contributes to

the retention of the active nucleotides in tumour cells. (5) dFdCTP also functions

to inhibit the dCMP deaminase reaction, adding an additional mechanism that

contributes to the prolonged retention of gemcitabine nucleotides found in tumour

cells (Heinemann et al., 1 992; Gninewald et al., 1 992). (6) dFdCTP inhibits CTP

synthetase at high concentrations, thereby blocking the synthesis of dCTP as

well (Heinemann et al., 1995). This action decreases the available supply of

CDP as a substrate for ribonucleotide reductase, which puts further stress on the

other normal pathways of dCTP formation.

1.4.1.2 Drug Resistance and Gemcitabine

Resistance to gemcitabine may be due to several mechanisms. dCK

deficiency, leading ultimately to decreased dFdCTP formation; increased

elimination of gemcitabine or its monophosphate through increased deamination;

increased dCTP pools, resulting in increased feedback inhibition of dCK; and

finally decreased influx or increased efflux of gemcitabine (Van Haperen et al.,

1995; Van Haperen et al., 1994).

Van Haperen et al. (1995) developed a gemcitabine-resistant human

ovarian cancer cell line in which they found a 10-fou reduction in dCK activity as

compared to the normal, parent cell line. This resistant cell line was found to be

cross-resistant to antimetabolites known to depend on dCK for activation (ara-C,

2-CdA), dFdU (the deamination product of gemcitabine) as well as to cisplatin,

doxorubicin and vincristine (Van Haperen et al., 1995). This may have

implications in future gemcitabine combination Vierapy.

Various lines of evidence have shown an association between increased

cytidine deaminase activity and cellular resistance to cytidine analogues. Firstly,

cytidine deaminase irrevenibly catalyses the deamination of its physiologie

substrates (cytidine and deoxycytidine) as well as nucleoside analogue agents,

including gemcitabine (Camiener, 1967; Chabot et al., 1983; Bouffard et al..

1993). This deamination causes a marked decrease in anti-tumour activity

(Creasey et al., 1966; Muller and Zahn, 1979). Second, in leukaemia patients,

high levels of cytidine deaminase activity have been correlated with resistance to

treatment with ara-C (Steuart and Burke, 1971 ; Onetto et al., 1987). Third,

inhibition of cytidine deaminase in experimental tumours and cefl lines has been

associated with an increased susceptibility to killing by ara-C (Honma et al..

1991 ; Riva et al., 1992). None of these studies, however, has demonstrated a

direct association between cytidine deaminase activity and drug resistance.

Tobias and Blau (1996) performed in vitro studies to determine whether forced

expression of a gene encoding for the enzyme cytidine deaminase can confer

resistance to the cytidine analogues ara4 and gemcitabine. They found that by

overexpressing cytidine deaminase they couM confer at least a 2-fold resistance

to a ra4 and gemcitabine as compared to cells expressing normal levels of this

enzyme.

Sliutz et al. (1996) linked a heat shock protein (hsp70) overexpression

with gemcitabine resistance. Heat shock protein expression is induced in

response to adverse changes in the cellular environment (Lindquist and Craig,

1 988). Hsp70 overexpression has been repoited to induœ cytoprotection in vivo

and in vitro under a wide variety of adverse conditions (Sliutz et al., 1996). In

this study, the cytoprotective effect of hsp70 against gemcitabine was moderate,

with a factor of 2-3 fold. It was shown that by depbting cellular levels of hsp70

using quercetin, a natural flavonoid known to inactivate the heat shock

transcription factor, sençitivity to gemcitabine increased. This increase was

quercetin dosagedependant.

Mackey et al. (1998) showed that nucleoside transporter activity was a

prerequisite for growth inhibition by gemcitabine in vitro. They evidenced two

reasons for this: 1) nucleoside transportdeficient cells were highly resistant to

gemcitabine and 2) treatment of cells that exhibited only equilibrative nucleoside

transporter activity with nitrobenzylthioinosine (NBMPR) or dipyridamole

(equilibrative transport inhibitors, section 2.6) increased gerncitabine resistance.

1.4.1 -3 Phase I Studies (Toxicities)

Various dosing schedules of gemcitabine were evaluated during many

phase I studies. Dose-lirniting toxicities of gerncitabine have k e n found to be

schedule dependent (Eckardt et al., 1995). Poplin et al. (1992) studied two

different therapy regimens: 5-90 mg/m2 twice weekly as a 30-minute infusion and

30-1 50 mg/m2 twice weekly as a bolus injection dunng 5 minutes. The maximum

tolerated doses of gemcitabine for these schedules were 65 and 100 mg/m2

respectively. The dose limiting toxicity was myelosuppression for both schedules

and non-haematological toxicities, including nausea, vomiting and malaise, were

mild (Poplin et al., 1 992).

From other phase I studies, gemcitabine (1 0-1000 mg/m2 as a 30-minute

intravenous infusion weekly for 3 weeks eveiy 4 weeks) was founâ to have

myelosuppression as its major dose-limiting toxicity. Other haematological

toxicities included anaemia and thrombocytopenia. Non-haematological

toxicities found were nausea, vomiting and malaise, which were all mild. The

maximum tolerated dose of gemcitabine was assessed to be 790 mg/m2 for this

dosing schedule (Abbniuese et al., 1991 ; Rosso and Martin, 1994; Pollera et al.,

1994). In studies of previously untreated patients with pancreatic

adenocarcinorna, high-dose gemcitabine (1200, 1500 or 1800 mglm2 given

weekly for 3 weeks every 4 weeks at a constant infusion rate of 10mglm2/min)

elicited fever, myelosuppression, nausea, vomiting and confusion as its dose-

limiting toxicities (Tempo et al., 1994).

1 -4.1 -4 Phase II and III Studies

Phase II studies of the weekly schedule have demonstrated that

gemcitabine has activity against non-small cell lung cancer, bladder cancer,

ovarian cancer, breast cancer and pancreatic cancer (table 1.4a). In pancreatic

cancer, the efficacy of gemcitabine (1000 mglm2 weekly for 7 weeks followed by

1 week of rest and thereafter weekly for 3 weeks every 4 weeks) has been

studied (Rothenberg et al., 1995). 17 of 63 patients (27%) in the study showed a

positive clinical benefit (defined as at least 50% reduction in pain or at least 50%

reduction in daily consumption of analgesics). Clinical benefit, defined as a

composite measure of pain, analgesic consumption, performance status and

weight gain (Hidalgo et al., 1 999). has only recently been incorporated into the

investigation of novel therapies in pancreatic cancer care and recognized as a

valid end point for dnig approval. In a phase III study (Moore, et al., 1995) in

previously untreated patients, gemcitabine was compared to 5-FU. Here, 1 26

patients with advanced pancreatic cancer were randomized to gemcitabine at the

aforementioned schedule or to 5-FU (600 mg/m2 over 30 minutes, weekly). With

gemcitabine, 24% of patients showed a clinical benefit versus 5% of those in the

5-FU a n of the study. Median suwival of the two groups was also reported, and

was found to be 5.65 and 4.41 months respectively, with 18% of gemcitabine

patients alive at 1 year as cornpared to 2% of those who received 5-FU.

1.5 Thymidylate Synthase (TS) Inhibitors

Thyrnidylate synthase (TS) is the rate-limiting enzyme involved in the de

novo synthesis of thymine nucleotides. Inhibition of this enzyme limits the

formation of thymidine triphosphate (TTP), resulting in inhibition of DNA

synthesis. TS has been an attractive target for anticancer drugs since the

Stud y Dooing Scheduk Prior Response Therapy Rate(%)

Non Small-Ce11 Lung Cancer

Anderson et a1 (1 994) 800-1 000 mglm2/wk x 3 wk No 20 (1 6/79)

Abratt et al (1 994) 1000-1 250 mg/m2/wk x 3 wk No 20 (15/76)

Shepherd et al (1 993) 1250 mglm21wk x 3 wk No 20 (1 9/93)

Fosella et a1 (1 993) 1000-1 750 mglm2/wk x 3 wk No 21 (4/19)

Negoro et al (1 994) 1 000-1 250 mg/m2/wk x 3 wk Not stated 30 (1 1/37)

Negoro et al (1 994) 1 000-1 250 mg/m2/wk x 3 wk Not stated 24 (9137)

Lund et al (1 992) 90 mglm2 Nice weekly No 13 (5/40)

Ovarian Cancer

Lund et al (1 994) 800 mg/m2/wk x 3 wk Yes 19 (8142)

Breast Cancer

Carnichael et a1 (1993) 800 mg/m21wk x 3 wk Yes 29 (9135)

Pancreatic Cancer

Casper et ai (1 991 ) 800-1 250 mglm21wk x 3 wk No 13 (5139)

Carnichael et al (1 993) 800-1 000 mg/m2/wk x 3 wk No 9 (2123)

Table i.4a. Phase II trials of gemcitabine.

introduction of the fluoropyrimidines 5-FU and 5-fluorodeoxyuridine by

Heidelberger et al. (Heidelberger et al., 1957). TS inhibition by these compounds

is dependent on their conversion to fluorodeoxyuridine monophosphate (FdUM?)

and the subsequent formation of a temary complex between the enzyme,

FdUMP and the folate cofactor (Langenbach et al., 1972; Danenberg et al.,

1974). In addition to their effects on TS, fluoropyrimidines can also inhibit cell

growth by incorporation into RNA and DNA (Heidelberger et al.. 1957). An

alternative approach to the inhibition of TS has b e n directed toward the

synthesis of analogues of the fobte cofactor. These folate-based TS inhibitors

are specific in their actions and do not possess any of the non-specific actions of

the fluoropyrimidines (Cunningham et al., 1 996; Blackledge, 1 998).

Other factors also favour TS as a chemotherapeutic target. TS levels are

increased in neoplastic cells, implying a greater dependency of tumour cells on

de novo pyrimidine biosynthesis and therefore, enhanced sensitivity to TS

inhibitors relative to normal cells (Hashimoto et al., 1988). In addition, folate-

based inhibiton of TS may also gain selectivity because of the increased

reduced folate transport and polyglutamation capacity in tumour cells (Sirotnak et

al., 1984). These factors should result in increased concentration and retention

of the inhibitors in tumour cells.

1.5.1 5-Fluorouracil

~-FIuo~-c~~..(~-Fu) is a fluonnated pyrimidine antirnetabolite belonging

to the class of anti-metabolites known as TS inhibitors. It was originally

synthesized in 1957 (Heidelberg et ai., 1957). The fluorine substitution occurs at

carbon 5 of the pyrimidine ring in place of a hydrogen (figure 1 5a) .

5-FU is the therapeutic mainstay for colorectal cancer (Schnall and

MacDonald, 1991). It is also used clinically in the treatment of breast, pancreatic

and stomach cancers and squarneous cell carcinoma of the head and neck.

Response rates to 5-FU monotherapy in patients with advanced colorectal

cancer are typically less than 20% (Bleiberg, 1997). Therefore, considerable

interest exists in combining 5-FU with other chemotherapeutic agents in order to

improve therapeutic outcornes. Thus far, efforts to improve efficacy have

included modifying the route or schedule of administration and combining 5-FU

with biochemical modulating agents, such as folinic acid. Gastrointestinal

toxicities (diarrhea, mucositis) are dose-limiting in clinical use.

U raci l 5-FU

Figure 1.5a. Chemical structures of uracil and 5- Ffuorouracil.

1 S.1.1 Mechanism of Action of SIFU

5-FU, like gemcitabine, is an S-phase specific agent. Therefore, dnig

cytotoxicity requires active DNA synthesis and is related to exposure tirne. As 5-

FU is a moâified nucleobase, it must be transported into the cell before it can

exert its cytotoxic actions. This uptake occurs via various nucleoside transport

proteins located at the cell surface (section 1.6; Wang et al., 1997). Once it

enters the cell, three mechanisms exist by which 5-FU exeitç its cytotoxic effects

(figure 1.56): (1) Inhibition of thymidylate synthase (TS) (Heidelberg et aL,

1960a. 19606). This is the primary mechanism of action of 5-FU. Thymidylate

synthase catalyzes the methylation of Tdeoxyuridine-5'-monophosphate

(dUMP) to 2'4eoxythymidine-5'-monophosphate (dTMP) in a reaction which

uses the reduced folate 5,1 O-methylene tetrahydrofolate as a cofactor. 5-FU is

converted intracellularly to 5luoro-2'-deoxyuridine-5'-monophosphate (FdUMP).

FdUMP fonns a covalent temary bond with thymidylate synthase and its reduced

cofactor ~~*'~-methylene tetrahydrofolate. Binding results in enzyme inhibition

and subsequent depletion of deoxythymidine triphosphate (dTTP), thereby

preventing DNA synthesis and celf growth. (2) 5-FU or FdUMP may be

converted to fluorouridine monophosphate (FUMP) which is further

phosphorylated to the triphosphate, FUTP. FUTP may be incorporated into RNA

producing non-functional RNA aiid defects in protein synthesis. (3) FdUMP may

be phosphorylated to the 5'-fluorodeoxyuridine triphosphate (FdUTP) and

incorporated directly into DNA, altering DNA stability (Cheng and Nakayama,

UK UMPK UDPK 5-FU FUMP - FUDP FUTP RNA (2)

ADP

Nucleoside Mono phosphate Nucleoside Diphosphate FdUMP FdUDP Kinase ) FdUTP

dUMP ,---f-f+ dTMP dTïP Thymidylate

S ynthase (1) DNA Polymeme

DNA * Uracil-DNA G 1 ycos ylase

Figure 1.5h Mechanism of action of 5-FU. UK, uridine kinase; UMK, UMP kinase; UDK, UDP kinase; TK, thymine kinase. Numbered pathways correspond to those discussed in the text.

1 983). The DNA repair enzyme uracil DNA glycosylase functions to remove

uracil from DNA and also excises FdUTP resulting in DNA single strand breaks

and DNA fragmentation (Lonn and Lonn, 1 986).

It has been speculated that many of the toxic effects of 5-FU (and its

modulators) are due to the lack of specificity in their actions on thymidylate

synthase (Mead, 1996) and thus a search for more specific TS inhibitors has

been undertaken.

1.5.1.2 Drug Resistance and ÇFU

5-FU resistance has been primarily associated with insufficient inhibition

of thymidylate synthase (Peters et al., 1 995). In vitro and in vivo 5-FU resistance

has also been associated with the following: (1) Decreased accumulation of

FdUMP due to decreased 5-FU activation or increased 5-FU inactivation

(Mul kins and Heidelberger, 1 982). (2) lncreased activity of thyrnidylate synthase

(Chu et aL, 1993; Johnston et al., 1995) and amplification of the thymidylate

synthase gene (Clark et al., 1987; Berger et al., 1985). (3) Changes in

nucleotide pools (Aronow et al., 1984; Kaufmann et al., 1984). (4) Point

mutations in the gene encoding for thymidylate synthase, resulting in an enzyme

with an altered structural fom and lower affinity for both FdUMP and the cofactor

~~*'~-rnethylene tetrahydrofolate. Binding of 5-FU in the absence of this cofactor

results in an unstable binary cornplex. FdUMP then becomes a weak inhibitor of

thymidylate synthase.

1.5.2 Tomudex (Raltbeulcl)

The biosynthesis of thymidine monophosphate (TM P) requires 5,100

methylene tetrahydrofolate, which serves as a cofactor in the TS-catalysed

transfer of a one-carbon unit to dUMP. Due to the Iimited success of

fluoropyrimidine substrate analogues (such as 5-FU) in the treatment of

colorectal and other cancers (Rusturn and Creaven, 1988), analogues of the

folate cofactor of TS were developed. Non-specifk, non-TS effects of 5-FU (and

its modulators) on RNA are believeâ, as previously noted. to account foi some

aspects of toxicity seen during therapy with these antimetabolites, such as

mucositis and rash. A specific TS inhibitor, which does not require modulation

and does not possess non-specific actions on RNA, presents an attractive

research goal. Tomudex, a quinazoline folate analogue (figure 1.5~). is a new

chemotherapeutic agent that aims to rneet this research target. It entered phase

I evaluation in Europe in 1991 (Jackman et al., 1995). Tomudex is a folate-

based TS inhibitor and is an analogue of the folate cofactor for TS. It is a highly

specific, laboratorydesigned TS inhibitor that does not require modulation and

which does not have non-specific effects on RNA (Cunningham et a/., 1996;

Blackledge, 1 998).

Figure 1 .Sc. Chemical structure of tornudex.

24

Tomudex has been approved for the treatment of advanced colorectal

cancer and is currently k i n g studied in a wide vanety of malignancies including

squameous cell head and neck carcinoma, prostate and gastric cancers and

pancreatic cancer. In patients with advanced colorectal cancer, Tomudex has a

response rate similar to 5-FU (Blieberg, 1997; Cunningham et al., 1996) when

used as a single agent. In pancreatic cancer, studies have shown that Tomudex

has a response rate of 12% (Cunningham et al.. 1996). As with 5-FU. there is

considerable interest in combining Tomudex with other chemotherapeutic agents

in an attempt to improve therapeutic outcomes. Adverse effects of Tomudex

include fatigue, leukopenia, thrombocytopenia and diarrhea.

1 S.2.1 Mechanism of Action of Tomudex

Tomudex is a mixed noncornpetitive inhibitor of TS (Duch et al., 1993).

The mode of action of tomudex is specific and is based on cornpetition with 5 1 0-

methylene tetrahydrofolate for TS. Tomudex is transported into cells via a

reduced folate carrier (RFC) and is rapidly and extensively polyglutamated by the

enzyme folyl polyglutamate synthetase (FPGS) (figure 1.5d). Polyglutamation

enhances TS inhibitory potency and increases the duration of TS inhibition in

cells, which may improve antitumor activity (Cunningham et al., 1996). This

polyglutamation could also contribute to increased toxicity due ta drug retention

in normal tissues

Tornudex

f Tomudex (glu),,

FdUMP

Figure 1.5d Mechanism of action of Tomudex. RFC, Reduced Folate Carrier; 5,1 0-CH2FH4, 5,l O-methylene tetrahydrofolate; mhd, deoxythymine; TMP, thymine monophosphate; TTP, thymine triphosphate; TK, Thymidine Kinase.

DNA

1.5.2.2 Drug Resistance and Tomudex

Acquired resistance to tomudex may occur through induction of TS,

disturbed RFC function leading to reduced drug uptake, reduction in FPGS

activity and alterations in intracellular folate homeostasis (Van Triest et al., 1999;

Lu et al., 1 995; Drake et al., 1 996; Jackman et al., 1995). In the latter case, cell

lines having expanded folate pools showed decreased polyglutamation of

tomudex thus cuitailing its bioiogical acüvity (Van Triest et al.. 1999). This may

indicate that a high folate status may preclude activation of folate-based TS

inhibitors such as tomudex, to Meir active polyglutarnate foms, whereas a low

folate status may be a cause of toxicity observed in patients.

1.6 Nucleoside Transporters

Cells can synthesise nucleotides either through de novo synthesis, or via

reutilisation of nucleotides and nucleobases derived from either the intracellular

turnover of nucleic acids and nucleotides, or from extracellular sources

(Wohlhueter and Plagemann, 1980; Muller et al., 1983). In the latter case,

known as the salvage pathway, nucleosides and nucleobases must first be

transported across the cell membrane by specific transport proteins. These

transport systems are of considerable pharmacological interest. Transport

inhibitors, such as dipyridamole and dilazep, can enhance the effectiveness of

various chemotherapeutic agents, including 5-FU and methotrexate, by

modulating either drug influx or efflux and interfering with the salvage pathways

of DNA synthesis (Lonn et aL, 1989; Kang and Kimball, 1984; Marz et al., 1977;

Gaffen et al., 1989; Cabral et aL, 1984; Chan, 1989; Chan and Howell, 1985;

Plagemann and Kraupp, 1986). The therapeutic effects of combining a transport

inhibitor with an inhibitor of de novo nucleotide synthesis in order to develop a

chemotherapy regimen in which both the de novo and sahrage pathways of DNA

synthesis are blocked has been widely explored. The combination of 5-FU and

dipyridamole as well as methotrexate and dipyridamole are two such examples.

Studies have shown that dipyridamole (7.7 mgkg/day as a continuous infusion

over three days) en hances both methotrexate and 5-FU cytotoxicity (Remick et

al., 1 990; Willson et al., 1 989).

In addition to the endogenous nucleosides, some dmgs belonging to the

nucleoside analogue class of oncolytic agents are also taken up into the cefl via

these specific nucleoside transport proteins (figure 1.6a). Therefore, combining

a nucleoside analogue with agents that increase nucleoside transporter

expression may lead to greater cell kill.

Two carrier mediated transport mechanisms for purine and pyrimidine

bases have been described. In mammalian cells, plasma membrane transport

occurs by both sodium-dependent (concentrative) and sodium-independent

(equilibrative) mechanisms (Sundaram et al., 1998; Cass, 1995; Griffith and

Jarvis, 1996). These protein mediated transport systems are also the route of

cellular uptake for many synthetic nucleoside analogue agents including a ra4

and gemcitabine (Wiley et aL, 1985; White et al., 1987; Clumeck, 1993; Burke et

al., 1998). In addition, through its effect on adenosine concentration at the cell

surface, nucleoside transport plays an important role in a variety of physiological

Guanosine Adenosine fnasine

Thymidine Cytidine Uridine

NUCLEOSIOE ANALOGUES

Cl

2-chlorodeoxyadenosine Cytosine Arabinoside

Figure 1.68. Structures of nucleosides and nucleoside analogues.

29

processes including coronary vasodilation, renal vasoconstriction,

neurotransmission, piateiet aggregation and Iipolysis (Sundaram et al., 1998;

Belardinelli et al.. 1989; Jacobson et al., 1990). There is also growing evidence

that nucleoside transport processes of intracellular membranes play an important

role in the intracellular distribution of nucleosides and their analogues (Cass et

al., 1998; Hogue et a/., 1996; Mani et al., 1998).

Sodiumdependent mechanisms of nucleoside transport are limited to

specialized cells such as intestinal and renal epithelia, choroid plexus, liver,

splenocytes, macrophages and leukaemic cells (Belt et al., 1992; Cass, 1995;

Griffith and Jawis, 1996; Huang et al., 1994; Fang et al., 1996; Yao et al., 1996a;

Yao et al., 1 996b; Ritzel et al., 1997; Che et al., 1 995; Wang et al., 1 997a; Patil

and Unadkat, 1997). These sodium-dependent, concentrative nucleoside

transporters generally mediate influx only and act via active transport processes,

thus depending on cellular ATP for their function. These transportes use the

physiologie sodium gradient (extracellular to intracellular) generated by the

ubiquitous Na*K+-ATPase to move nucleosides intracellulady against their

concentration gradient (Griffith and Jawis, 1996; Cass. 1995). Based on

functional studies, six subtypes (NbN6) of sodium-dependent nucleoside

transporters have been characterised in human cells and tissues (table 1.6a).

They are subclassified according to substrate specificity and numbered

sequentially in order of discovery (Wang et ab, 1997; Cass et al., 1998). Cif (NI)

are purine-selective and also accept uridine and formycin 6 as substrates. Cit

(N2, N4) prefer pyrimidine nuclecsides and also accept adenosine (N2 and N4

NT Permeant Selectivity Sensitivity Na'- Refe re mes

Process to NBMPR dependency

Concentra tive transmtters

cif (N 1 )

cit (N2)

cit (N4)

cib ( N 3 )

CS (N5)

csg (W

PurÏne nucleosides,

uridine. formycin B

Pyrimidine nucleosides,

adenosine

Pyrimidine nucleosides,

guanosine, adenosine

Purine and pyrimidine

nucleosides

Adenosine analogues,

formycin B

Guanosine

Eauilibra tive transmrters

es Purine and pyrimidine

nucleosides

ei Purine and pyrimidine

nucleosides

+ Ritzel et al. 1998; Wang et ai. 1997b

+ Ritzel et al. 1997

+ Gutierrez et ai. 1 992; Gutierrez and Giacomini 1993

+ Paterson et al. 1993

+ Flanagan and Meckling-GiII 1 997

- Griffiths et al. 1 997a

O Crawford et al. 1 998; Griffiths et al. 1 9976

Table 1.68. Characterized human nucleoside transporter systerns.

31

subtypes) and guanosine (subtype N4) as substrates. Ci& (N3) is non-selective

and accepts a broad range of purine and pyrimidine nucleosides as substrates

(Cass, 1995; Belt et al., 1993; Cass et al.. 1998). All of the above isofoms are

resistant to inhibition by S-(pnitrobenzy1)-6-thioinoçine (NBMPR). Two

additional isofomis, CS (N5) and wg (N6), that are sensitive to NBMPR inhibition,

have recently been identified (Paterson et al., 1993; Flanagan and Meckfing-Gill.

1 997; Burke et al., 1998). These transport proteins accept adenosine analogues

and guanosine respectively as substrates. The histological prevalence of these

isofoms has yet to be identified.

Sodium-independent, equilibrative nucleoside transport processes

mediate the facilitated diffusion of nucleosides across plasma membranes and

are widely distributed in different cell types (Hammond, 1991). They also

function bidirectionally in the transmembrane flux of nucleosides in accordance

with the concentration gradient. This class of transporter displays low substrate

specificity and affinity, and high substrate capacity. Sodium-independent

nucleoside transporten are classified into two subtypes based on their

sensitivities to inhibition by NBMPR and dipyridamole (table 1.6a). NBMPR

sensitive (equilibrative sensitive (es)) transporters bind NBMPR with high affinity,

whereas NBMPR insensitive (equilibrative insensitive (eu) transporters are

unaffected even by micromolar concentrations of NBMPR. 60th display broad

substrate specificity for purine and pyrimidine nucleosides. Equilibrative

transporters are pharmacological targets for coronary vasodilator cornpounds

such as dipyridamole and dilazep (Cass, 1995; Griffith and Jarvis. 1996) that

compte with permeant for the substrate-binding site (Janris et al., 1982; Jarvis,

1986). It has been previously shown that es transporter expression can be

increased by depleting the endogenous intracellular nucleotide pools using DNA

synthesis inhibitors such as 5-FU (Pressaco et al., 1 995). The cell süll has to

maintain nucleotide levels in order to preserve the DNA synthesis process and

must now scavenge those nucleotides f rom extracellular pods. By increasing

the number of nucleoside transporters at the cell surface, it would follow that a

greater number of nucleosides would enter the cell.

The transport process is followed by phosphorylation of the nucleosides

by kinases which, contraiy to the nucieoside transporters. have a high substrate

affinity and low substrate capacity. Computer and kinetic analyses have

suggested that the transport of nucleosides is rate-limiting at low (less than 1 phil)

concentrations of extracellular nucleosides (Wiley et a/., 1985; White et al.,

1987). Nucleoside levels of this magnitude occur in human senim (Nottebrock

and Then, 1977), indicating that nucleoside transport (and nucleoside transporter

expression) may be an important, rate-limiting step in the utilisation of

nucleosides by cells for the salvage pathway of DNA synthesis.

The recent cloning of the four major human nucleoside transporters (es,

ei, cit (N2), cil) (Cass et al., 1 998; Griffiths et al., 1 997a; Crawford et al., 1 998;

Griffiths et al., 1 9976; Ritzel et al., f 997; Wang et al., 1 997; Belt et al., 1 993) has

led to the construction and discovery of new molecular and immunological

probes with which we are able to study nucleoside transporter distribution. This

has the potential to lead to novel, transport-based strategies aiming to improve

the therapeutic indices of the nucleoside analogue class of chemotherapeutic

drugs.

1.6.1 Gemcitabine and Nuckoside Transport

Gemcitabine has been shown to be a substrate for four of the nucleoside

transporters found in humans (es, ei. cit, cib) (Mackey et al., 1998). The major

mediators of gemcitabine uptake, howeve:, are most probably the equilibrative

n ucleoside transporters because human cit activity has only been demonstrated

in kidney, liver and intestinal epithelium (Ritzel et al., 1997; Patil and Unadkat,

1997) and that for human cib in myeloid leukemic ce11 lines, freshly isolated

myeloblasts and the CaCo-2 colon cancer cell line (Belt et al., 1992; Belt et al.,

1 993).

To date, in vitro experimental evidence has shown that the efficiency of

uptake of nucleoside analogues, such as gemcitabine, varies among cell lines

that express only single nucleoside transporters. The efficiency of gemcitabine

uptake in these cell lines as assessed through transport kinetics (W, ratio)

was as follows: es = cit :. ei > cib »> cif (Mackey et al., 1998). Mackey et al.

(1 998) also demonstrated that gemcitabine resistance increased when the

various cellular nucleoside transport processes were inhibited or removed.

1.7 5-(SAENTA-x8)-Fluorescein Binding Assay

The recent synthesis of SAENTA, a derivatizable analogue of NBMPR

that is a tightly-bound inhibitor of the es transport process (Agbanyo et al., 1 990),

has enableâ the development of a method to allow the measurement of cell-

surface es-NT content that requires 10-fold fewer cells than conventional

[ 3 ~ ] ~ ~ ~ PR radioligand binding assays. Conjugates of SAENTA where the

nudeoside analogue is linked to fluorescent reporting moieties, in this case

fluorescein-5-isothiocyanate (FITC), have proven to be specific stains for cellular

NBMPR binding sites (Jamieson et aL, 1993). The interaction of fluorescein

derivatives of SAENTA at the es-NT has been shown through mutual inhibition of

NBM PR and SAENTA-fluorescein at binding sites in es-expressing cells

(Buolamwini et al., 1994). Jarnieson et al. (1993) showed that, when converted

to molecules of equivalent soluble fluorescein, the fluorescence signal arising

f rom 5-(SAENTA-x8)-f luorescein specif ically bound to leu kaemic myeloblasts

correlated well (r = 0.98) with the number of es-NT sites assayed for concurrently

by [ 3 ~ ] ~ ~ ~ ~ ~ equilibrium binding analysis. This enabled the cell-bound

fluorescence output of 5-(SAENTA-x8)-fluorescein to be expressed in nurnben

of cell-surface es-NT sites per cell. The current use of 5-(SAENTAox8)-

fluorescein (figure 1.8a) in flow cytometric equilibriurn binding assays, therefore,

provides a simple and accurate means of detemining the es-NT content of celk.

Figure 1 .Ta. Chernical structure of 5-(SAENTA-x8)-fluorescein showing tha fiuorescein and SAENTA moieties joined by an 8-atom linking structure.

1.8 Combination Chemotherapy

Combination chemotherapy aims to improve therapeutic effects verrus

single agent treatment. These improved effects can include an increased

therapeutic index, increased dnig efficacy and decreased dnig toxicity.

Combination chemotherapy may also play a role in the reduction or the delay of

developrnent of dnig resistance in tumour cells. These cells are less likely to be

resistant to multiple agents with differing mechanisms of action. Also, combining

dnigs that have different toxicity profiles may allow the use of each agent at its

maximum tolerated dosage, thereby maximizing the effectiveness of each drug in

the combination regimen.

When drugs are combined, we aim to determine if the effect of the

combination treatrnent is greater to, equal to or less than what would be

expected from the single agents alone. There are three possible outcornes of

any combination when each drug used has an effect on its own: synergy,

additivity or antagonism. Several different mathematical models exist to quantify

these effects. If, however, one drug exhibits no effect on its own yet enhances

the effect of another drug, the outcome is referred to as potentiation.

Several factors must be considered when evaluatîng dnig interactions as

each drug in the combination has its own effect (concentration-response curve)

and is influenced by vanous pharmacokinetic and phamacodynamic factors. To

study the effect of drug combinations, it is advisable to Vary one variable (eg.

concentration, exposure time) and measure one outcome (eg. cell kill) at a time.

Following the quantitative determination of the synergy or antagonism of a drug

combination, biochemical or molecular biology techniques may be used to

ascertain the mechanism behind the interaction seen ( Rideout and Chou, 1 99 1 ) .

There have been several methods developed for the quantification of dnig

interactions including the isobdagram method, fractional-product method and

median-effect principle. The classical isobdogram method is only valid for dnigs

whose effects are mutually exclusive. The fractional-product method is valid only

for those dnigs whose effects are rnutually non-exclusive and have a hyperbolic

concentration-response curve. Because of these limitations, these methods

cannot be randomly used (Chou and Talalay, 1984). The median effect principle,

however, is a more basic methodology that requires fewer assumptions and can

be generalized to a greater number of situations including cancer treatrnent.

Analysis with this method is based on the law of mass action and does not

require prior knowledge of the specific mechanisms of action of the drugs

involved .

Combining drugs afso allows for the biochemical modulation of one drug's

effects by the other. An example of this would be leucovorin (LV) given in

combination with 5-FU which is now a standard treatrnent approach in colorectal

cancer (Bfaszkowsky, 1998; Labianca et al., 1996). For this paper, we will

examine the modulating effects that 5-FU and tomudex have on gemcitabine

toxicity. It is known that gemcitabine has activity against pancreatic cancer

(section 1.2; Heinemann, et al., 1998; Peters et al., 1995; Bergman, 1996). It

has also k e n previously shown that the equilibrative-sensitive nucleoside

transporters can be upregufated by administering a DNA synthesis inhibitor such

as 5-FU or tomudex (Pressacco et al., 1995). We wish to examine the

combination of 5-FU and tomudex with gemcitabine to detemine if gemcitabine

toxicity can be modulated. It is also important to note that the modulating agents

used here also exhibit their own cytotoxicity through inhibition of the de novo

pathway of DNA synthesis. Therefore, cytotoxic actions of the two drugs are

acting on the different pathways of DNA synthesis, de novo and salvage

pathways. These diffenng mechanisms of action make for an appealing

combination approach.

Previous studies examining the interaction of gemcitabine and 5-FU have

been conducted to date. These studies used clinical benefit response as an end

point of efficacy in addition to objective tumour response. Hidalgo et al. (1999)

and Cascinu et al. (1998) found that 50% of patients achieved a clinical benefit

response. Half of the patients who achieved clinical benefit in the latter study did

not achieve an objective tumour response. In a dose-escalation study of

gemcitabine (starting at 700 mg/m2/week for 3 weeks of a 4 week cycle) and

fixed dose 5-FU (200 mg/m2/day) by Cortes-Funes et al (1 998), 55% of patients

displayed a cliniczl benefit response and a median survival of 10.4 months was

reported.

2.1 Rationale Behind Project

Pancreatic cancer is the f i leading cause of cancer death in North

America (Clark, et al., 1996; Schnall and Macdonald, 1996; Moore, 1996;

Rothenberg, 1996; Robertson et al., 1996) and also has the worst five-year

survival rate of any cancer (Clark et al., 1 996; Moore 1 996). For these reasons,

as well as the resistance of pancreatic cancer to current chemotherapy, research

to improve therapy is desperately needed. Recently. gemcitabine, a nucleoside

analogue, has demonstrated more efficacy towards pancreatic tumours Man

other dnigs, such as 5-FU, cisplatin and cytosine arabinoside (arac)

(Heinemann, et al., 1 988; Peters et al., 1 995; Bergman, 1 996).

As gemcitabine has been shown to exhibit efficacy in the treatment of

pancreatic cancer, examining methods to modulate this effect would be of great

interest and of potential clinical benefit. One method to accomplish this would be

to manipulate cells so that greater amounts of the dnig are taken up by

increasing the number of nucleoside transport sites. More transporters should

lead to increased uptake of the dnig and possibly to increased efficacy.

Therefore, the purpose of the work reported in this thesis was to define a new

regimen of combination chemotherapy using gemcitabine and agents that

upregulate nucleoside transporters, that has the potential to be used in the clhic

for the treatment of pancreatic cancer.

2.2 Hypothesis

(1) The sensitivity of human pancreatic and bladder carcinoma cell lines to

gemcitabine is related to their basal equilibr&tive-sensitive nudeoside

transporter (es-NT) expression.

(2) Upregulation of es-M in human pancreatic and bladder cancer cell lines

by pre-treatment with the thyrnidylate synthase inhi bitors 5-FU or tomudex

will increase the efficacy of gemcitabine therapy.

2.3 Objectives

The experimental approach addresses the potential use of geincitabine in

combination with either 5-FU or tomudex and provides insight into the role of

drug transport in cytotoxicity when the drugs are combined.

(1) Detemine the sensitivity of one human bladder cancer (MGH-U1) and

three human pancreatic cancer (PANC-1, PK-8, HS-766T) cell lines to

gemcitabine, 5-FU and tornudex given alone and in combination. The

MGH-U1 cell line was used as a method control as previous experiments

demonstrating es-NT upregulation by 5-FU and tomudex were perfonned

on this cell Iine.

(2) Detemine the basal levels of the es-NT in al1 four human tumour cell

lines.

(3) Detennine the relationship between drug sensitivity and es-NT

expression.

(4) Explore and gain insight into the relationship between sequence of drug

administration, es-M expression and cytotoxicity.

3.1 Cell Lines and Reagents

The cell lines used for these experiments were the human PANC-1, HS-

766T and PKS pancreatic carcinoma cell lines and the human MGH-Ul bladder

carcinoma cell line. PANC-1, HS-766T and MGH-U1 cell lines were onginally

purchased from the American Type Culture Collection (Rockville, MD, USA).

PK-8 cells were provided by Dr. D. Hedley (Ontario Cancer Institute, Toronto,

ON, Canada). Cells were maintained as monolayer cultures as foilows: PANC-1

and HS-7661 were cultured in Dulbecco's Modified Eagle's Medium, (Media

Department, Ontario Cancer Institute, ON, Canada) in a humidified 10% CO2

incubator at 37OC. PK-8 cells were cultured in RPMl 1640 growth medium,

(Media Departrnent, Ontario Cancer Institute, ON Canada). MGH-U1 cells were

cultured in a-MEM growth medium (Meâia Departrnent, Ontario Cancer Institute,

ON, Canada). 60th PK-8 and MGH-U1 cells were maintained in a humidified 5%

CO2 incubator at 37OC. All growai media were prepared at pH = 7.25, contained

penicillin (1 00 mg/mL) and streptornycin (1 00 mg/mL) and was supplemented

with 10% fetal Bovine Senim (Whitaker Bioproducts Inc., Walkersville, MD.

USA). RPMl 1640 medium was also supplemented with HEPES buffer (10 mM).

Cells were trypsinized (0.2% trypsin (Difco Laboratones, Detroit, MI, USA) in

phosphate buffered saline (PBS)), washed and replaced when necessary.

Unless otherwise specified, al1 reagents were purchased from the Sigma

Chernical Company (St. Louis, MO, USA). Gemcitabine was provided by Eli Lilly

Phamaceuticals (Indianapolis, IN, USA). Tomudex was provided by Zeneca

Pharrna (Mississauga, ON, Canada). Stock solutions of gemcitabine in PBS, 5-

FU in hydrochloric acid (0.1 M) and tomudex in PBS were prepared and stored at

-20°C.

3.2 Cell Cytotoxicity Assay

Cytotoxicity was assessed by clonogenic assay. Cells were harvested

during the logarithmic phase of cell growth, trypsinized, washed, counted and

resuspended at concentrations of 1-5x1 o5 celldmL 1 mL of cell suspension was

plated in 25cm2 flasks (Nunc, Roskilde, Denmark) containing 4 mL of the

appropriate growth medium. Following a 24 hour incubation, 5 pL of dnig

solution was added to each flask for a 1 : 1 000 dilution. PBS or hydrochloric acid

(0.1 M) were added as vehicle controls for the appropriate experiments.

3.2.1 Single Agent Exposures

Foflowing 24 hour exposures. cells were washed with PBS, trypsinized,

centrifuged (1200 rpm) (Beckman Instruments Inc., Palo alto, CA, USA),

resuspended in the appropriate growth medium, counted and 100 - 10,000 cells

were plated in 60 mm2 petri dishes. Cells were incubated for 7 days (MGH-Ut),

10 days (PANC-1, PK-8) or 14 days (HS-766T). The resulting colonies were

fixed and stained with a 1% (wlv) methylene blue solution (BI0 RAD, Hercules,

CA, USA).

Visible colonies (>40 cells/colony) were counted to detemine the colony

forming efficiency for each drug concentration. Fractional survival was

calculated by dividing the number of colonies foned in dmg-treated plates by

the number of colonies formed in the control plates. The ICW and Cm, defined

as the dnig concentration that inhibited cloning efficiency (colony formation

efficiency) by 50% or 900/0 respectively, were estimated for each drug in al1 cell

lines by plotting fractional survival versus drug concentration.

3.2.2 Drug Combination Exposures

Cells were exposed to gerncitabine prior to, at the same time as. or

following treatment with either 5-FU or tomudex for 24 hours. For concurrent

drug combinations, the procedure was the same as for single drug exposures.

For sequential drug combinations, following the first 24 hour exposure, medium

was removed by aspiration, cells were washed twice with PBS and 5 rnL of fresh

medium was added prier to the addition of new drug, to which the cells were

exposed for a further 23 hours.

3.3 5-(SAENTA-x8)-Fluorescein (5x8) Binding Assay

Es-specific binding of 5x8 was rneasured in al1 cell lines using a flow

cytometric assay as described by Gati et a l (1997). Briefly, cells were

trypsinized, counted and resuspended in growth medium at a concentration of at

least 2x1 o6 cellslml. Cell suspensions were then diluted with an equal volume of

either PBS (for assay of total binding of the 5x8 probe) or PBS containing 10 pM

NBMPR (for assay of non-specific binding of the probe) and kept at room

temperature for at least 30 minutes. These suspensions were then added to

graded concentrations of 5x8 (0-20 nM) in the absence or presence of 5 pfvl

NBMPR. 5x8 wwas allowed to equilibrate with cellular binding sites for 15 minutes

before flow cytometric measurement of cell-associated fluorescence was carried

out. Sarnples were analysed on an Epics XL II flow cytometer (Coulter

Corporation, Hialeah, FL, USA). Data were gated on light scatter before

recording of a fluorescence histogram composed of at least 10,000 cells.

Fluorescence histograms for each condition were generated. Mean fluorescence

intensity values for the fluorescence histograms were detemiined using Epics XL

II software (Coulter Corporation, Hialeah, FL, USA). Es specific binding of 5x8

was calculated as the difference in mean fluorescence intensities between total

and non-specific binding. To allow for day-today variation in the settings and

performance of the flow cytometer, fluorescent beads containing known nurnbers

of fluorescein molecules (rainbow calibration particles, RCP-30-5) (Spherotech,

Libertyille, ILL, USA) were analyzed by the cytometer prior to each experiment.

Calibration curves of molecules of equivalent soluble fluorescein (MESF) against

mean fluorescence intensity were constnicted following each assay to enable

specffic binding of the 5x8 probe to be converted to MESF.

Es transport sites were quantified using a modification of the above-

described procedure. Three cell suspensions were useâ for each control or drug

treatment. Cell suspensions were added to a single, es site-saturating

concentration of 5x8 (detennined from above assay) in the absence and

presence of 5 pM NBMPR to assay for total and non-specific binding of the

probe, and to PBS containing no 5x8 or NBMPR for assay of background

autofluorescence of the cells and growth medium.

48

4.1 Growth Cumes

Figures 4.1 a d show that the MGH-U1 , PK-8, PANC-1 and HS-766T cell

lines exhibited the characteristic growth pattern of immortalized ce11 lines. This is

an initial lag period, followed by a period of exponential growth followed by a

stationary phase where cells are confluent and no cell division occurs and finally

a loss in cefl number and cell death. Doubling times were calculated using the

linear portion of the cuwe where celk were growing exponentially. The doubling

times of the MGH-U1, PK-8, PANC-1 and HS-766T cell lines were 14,23,28 and

40 hours respectively.

4.2 Cytotoxicity Assays

Cells growing exponentially were tre ,ated with various concentrations of

gemcitabine, 5-FU and tomudex for 24 hours and survival measured by the

clonogenic assay in order to determine the ICy, and ICgO values. Following

treatment, cells were washed free from drug and re-plated with fresh growth

medium supplemented with 10% ÇBS. Due to the different growth rates, the

MGH-UI, PK-8, PANC-1 and HS-766T cell lines were stained 7, 10, 10 and 14

days following re-plating, respectively. The cloning efficiencies of the MGH-U1,

PK-8, PANC-1 and HS-766T cells were 21 -60%, 24-44%, 36-59% and 18950%

respectively. Experiments in which there were less than 20 colonies on the

control plates were discarded. Cell cytotoxicity was expressed as percent colony

foning efficiency, calculated as the ratio of the number of colonies (240

cells/colony) on the drug treated plate divided by the number of colonies on the

control plate.

O 100 150 200 250 300

Time (hrs)

Figure 4.1a Growth Curve for the MGH-UI cell line. Each point represents the mean f standard deviation of three or more independent experiments.

. O 50 100 3 5 0 200 250 300

Time (hrs)

Figure 4.1 6. Growth Curve for the PANC-1 cell line. Each point represents the mean f standard deviation of three or more independent experiments.

O 50 100 150 200 250 300

Time (hm)

Figure 4.1~. Growth curve for the HS-766T cell line. Each point represents the rnean f standard deviation of three or more independent experiments.

100 150 200

Tirne (hrs)

Figure 4.1 d. Growth Curve for the PK-8 cell line. Each point represents the mean f standard deviation of three or more independent experiments.

4.2.1 Singk Drug Exposures

Log concentration-response cuwes of al1 four cell lines following exposure

to either gemcitabine, 5-FU or tomudex are illustrated in figures 4.2a-d. As

indicated, al1 fits have a regression coefficient r 0.90. Relative sensitivities to

individual agents at the ICS0 versus the lCoo levels differed due to the dope of the

log concentration-response curves and are shown in tables 4.2a-6. With the

exception of the PANC-1 cell line, where a plateau was reached, an increase in

dnig concentration resulted in an increase in cell cytotoxicity. As illustrated in

table 4.2a. at the lCso concentration, sensitivity to gemcitabine for al1 cell lines

was as follows: MGH-U1 (1.5 nM) w HS-766T (4 nM) > PANC-1 (40 nM) >z PK-8

(35 pM). At lCso concentrations of 5FU. cellular sensitivity was as follows:

PANC-1 (230 nM) > MGH-U1 (500 nM) > HS-766T (2.75 pM) > PK-8 (4 pM).

Sensitivity to tomudex at lCso concentrations was as follows: PANC-1 (0.02 nM)

> pK-8 (0.04 nM) > HS-766T (3 nM) > MGH-U1 (20 nM).

4.3 Combination Studies

The effect of combining gemcitabine with either 5-FU or tomudex at a

range of concentrations was detemined by exposing cells to the combination

either concurrently for 24 hours, or sequentially, with each exposure lasting 24

hours. Table 4.3a illustrates the cytotoxicity of the agents used when acting

alone (data estimated from log concentration-responçe curves, figures 4.2a-c).

Concentration [pM]

Figure 4.28. Log concentration-response curves following 24 hour exposure of the MGH-Ul cell line to (i) gemcitabine. (a) 5-FU and ( A ) tomudex. Cytotoxicity was detennined by the colony-forming assay. Each point represents the rnean of three independent experiments; each line has an ? value 2 0.90.

Concentration [pM]

Figure 4.26. Log concentration-response curves following 24 hour exposure of the PANC-1 cell line to (m) gemcitabine, (e) 5-FU and ( A ) tomudex. Cytotoxicity was determined by the colony-fonning assay. Each point represents the mean of three independent experiments; each line has an 8 value 2 0.90.

Concentration pM

Figure 4 .2~ . Log concentration-response curves following 24 hour exposure of the HS-766T cell line to (a) gemcitabine, (a) 5-FU and (A) tomudex. Cytotoxicity was detennined by the colony-fonning assay. Each point represents the mean of three independent experiments; each line has an P va!ue 2 0.90.

Concentration [pJM]

Figure 4.2d. Log concentration-response curves following 24 hour exposure of the PK-8 cell line to (m) gemcitabine, (a) 5-FU and (A) tomudex. Cytotoxicity was deterrnined by the colony-foming assay. Each point represents the mean of three independent experiments; each line has an ? value 2 0.90.

Table 4.2a Dnig concentrations required to inhibit colony formation by 50% following 24 hour dnig exposures. Values were estimated from the log concentration-response cuwes.

Table 4.2b. Dnig concentrations required to inhibit colony formation by 90% following 24 hour dnrg exposures. Values were estimated from the log concentration-response curves. - indicates ICgO concentration was not achieved.

:wm-<p--l . . ..- ..- 0.8pM(tO%) 1 @A (20%) 0.4 pM (8%)

Table 4.3a. Concentration of agents used in combination expenments. Numbers in parentheses indicate cell survival when used as a sing!e agent (estimated from log concentration-response CU ives in figures 4.2a-c).

As shown in figures 4.3af pre-treatment of the HS-766T cell line with

either 5-FU or tomudex augmented the effects of sole gemcitabine treatment

along with the PANC-1 cell line when pre-treated with 5-FU. The concurrent and

gemcitabine pre-treatment regimens showed no added benefit over single agent

gemcitabine treatrnent in these cell lines. In the MGH-Ul cell line, no

augmentation in cytotoxicity was seen when 5-FU or tornudex was combined in

any manner with gemcitabine.

Figure 4.38. Combination studies of gemcitabine (0.8 pM) and 5-FU in the MGH-U1 cell line following both concurrent (24 hour) and sequential (24 hours per exposure) exposures. Results are expressed as % colony foning efficiency and are the mean f S.O. of three or more independent experiments.

100 Gemcitabine 4 Tomude

100 Gemcitabine + Tomude

.I .

.I .

Figure 4.36. Combination studies of gerncitabine (0.8 pM) and tomudex in the MGH-U1 cell line foilowing both concurrent (24 hour) and sequential (24 hours per exposure) exposures. Results are expressed as % coiony forrning efficiency and are the mean f S.D. of three or more independent experiments.

Gemdtaôine + 5Çl.i -r- -

Figure 4.3~. Combination studies of gemcitabine (1 phd) and 5-FU in the PANC-1 cell line following boai concurrent (24 hour) and sequential (24 hours per exposure) exposures. Results are expressed as % colony foning efficiency and are the mean f S.D. of three or more independent experiments.

Gemdtabine Tomudex

T T - T

- -

1 Lt

10 - 4

100 1OOO

Gemcitabine + Tomudex t

Figure 4.3d. Combination studies of gemcitabine (1 pM) and tomudex in the PANC-1 cell line following both concurrent (24 hour) and sequential (24 hours per exposure) exposures. Results are expressed as % colony forming efficiency and are the mean f S.D. of three or more independent experiments.

Figure 4.3e. Combination studies of gemcitabine (0.4 pM) and 5-FU in the HS- 766T cell line following both concurrent (24 hour) and sequential (24 hours per exposure) exposures. Results are expressed as % colony forming efficiency and are the mean t S.D. of three or more independent experiments.

100 Gemcitabine Tomude

100 Gerncitabine + Tomudex

a

I

100 Tomudex a Gemcitabi

a a

10 b 7- -, - - - - . - - - - -

9

1 7

- . . . a . - . . . . , . . . . . . .

9 '. * . . . . . . . . . . - ; - ; < a . *

O- 1 . 8

- 1 . 1 8

O 1 10 100 1000

Figure 4.3t Combination studies of gemcitabine (0.4 pM) and tomudex in the HS-766T cell Iine following both concurrent (24 hour) and sequential (24 houn per exposure) exposures. Results are expressed as % colony foming efficiency and are the mean & S.D. of three or more independent experiments.

4.4 Equilibrative-Sensitive Nucleoside Transporter (es-NT)

Quantitation

Flow cytometric analysis was perfomed on al1 four cell lines to quantitate

basal cell-surface levels of the es-NT and to ensure that these celis did indeed

contain es-NT. The total and non-specific binding cuwes of 5-(SAENTA-%)-

fluorescein (5x8) in al1 four cell lines are illustrated in figures 4.4a-d. In every

case there was found to be binding of the probe that was saturable and

revenible when exposed in conjunction with the es transport inhibitor NBMPR.

An es-NT saturating concentration of 5x8 was defined as that concentration

w here no further increase in cell-associated fluorescence occurred. This point,

at the plateau of the binding curves, was where measurements of basal leveis of

es-NT were taken. All measurernents from the flow cytometric analysis were

reported as mean fluorescence intensities. These were converted into molecules

equivalent soluble f luoresœin (MESF) f rom calibration curves de nved from

beads containing known numbers of fluorescein molecules fun through the flow

cytometer before each experiment (figure 4.4e). As 5x8 binds to the es-NT on a

1 :1 ratio, the MESF values for the specific binding of 5x8 correspond to the

number of cell-surface es transport sites par cell. The cellular qwntity of these

cell surface es-NT (expressed as MESF) for the four ceIl lines used was as

~OIIOWS: PANC-1 (1 41 48) > PK-8 (4792) - MGH-U1 (4281) > HS-766T (121 6).

This is illustrated in table 4.4a.

O 5 10 15

[SAENTA] (nM)

Figure 4.4s Binding cuwes of 5x8 in the (i) absence and (a) presence of NBMPR in the MGH-U1 cell fine. Specific binding may be calculated by subtracting non-specific binding (in the presence of NBMPR) from total binding (in the absence of NBMPR).

O 5 10 7 5

[SAENTA] (n M)

Figure 4.46. Binding curves of 5x8 in the (i) absence and (a) presence of NBMPR in the PANC-1 cell line. Specific binding may be calculated by subtracting non-specific binding (in the presence of NBMPR) from total binding (in the absence of NBMPR).

O 5 10 15 20

[SAENTA] nM

Figure 4.4~. Binding curves of 5x8 in the (i) absence and (4 presence of NBMPR in the HS-766T cell line. Specific binding may be calculated by subtracting non-specific binding (in the presence of NBMPR) from total binding (in the absence of NBMPR).

[SAENTA] nM

Figure 4.4d. Binding curves of 5x8 in the (i) absence and (a) presence of NBMPR in the PK-8 cell line. Specific binding may be calculated by subtracting non-specific binding (in the presence of NBMPR) from total binding (in the absence of NBMPR).

MESF

Figure 4Ae. Sample calibration cuwe for flow cytornetric analysis using the rainbow calibration beads (RCP-30-5). 6 > 0.99.

Table 4.48. Specific binding of SAENTA (molecules equivalent soluble fluorescein (MESF) per cell) and gemcitabine sensitivities for four human cancer cell lines prior to treatment with dnig. These correspond to basal levels of cell surface es-NT and lCso values respectively.

Following treatment of the MGH-UI, PANC-1 and HS-766T cell lines with

varying concentrations of either gemcitabine, tornudex or 5-FU. cells were

analyzed for es-NT content. As shown in table 4.46, treatment of the PANC-1

cell line with 1 pM dMC and 30 pkî 5-FU for 24 houn showed a significant (p c

0.05) increase in cell-surface es-NT content over basal levels. In addition,

treatment of the HS-766T pancreatic carcinoma cell line with 30 pM and 100 pM

5-FU and 100 nM and IWO nM tomudex afl resuited in a significant (p < 0.05)

increase in es-NT sites versus control levels. AH other treatment groups for

these three cell lines showed no difference in es-NT content compared with

basal levels.

T m n t Factor Incmu6 In iT MESF ww oontrd-

20 5-FU 0.93 20 jN5-FU x 48 hr 0.70

40 PM 5-FU 1 .IO 1 O0 5-FU 1.57

100 nM Tomudex 1 -66 1 pM Tomudex 1.41

10 nM Gemcitabine 0.96 800 nM Gemcitabine 1 .O3

30 5-FU 1 O 0 W5-FU

100 nM Tomudex 1 pM Tomudex

40 nM Gemcitabine 1 pM Gemcitabine

30 ~J.M 5-FU 1 O 0 @l5-FU

100 nM Tomudex 1 phi Tomudex

Table 4.4b. Factor increase over control in specific binding of SAENTA after treatment with either 5-FU, tomudex or gemcitabine. All incubations were for 24 hours unless otherwise indicated. ** denotes statistically significant difference (p c 0.05, t-test) behnreen treatment group and control.

Characteflzing the various factors that affect cellular differences in

cytotoxicity to a chemotherapeufc agent can provide insight into how one might

modulate these effects. This thesis focused on the nucleoside transporter.

Specifically. the role that the es-NT plays in gemcitabine cytotoxicity. We

hypothesized that the sensitivity to gemcitabine would be influenced by basal

levels of es-NT. If this proved tnie, it would follow that manipulation of the

amount of es-NT couM alter cellular sensitivity to gemcitabine.

DNA synthesis inhibitors such as 5-FU and tornudex have been previously

shown to upregulate the number of cell surface es-NTs (Pressacco et a/..

1995a). These results point towards the usefulness of these agents as possible

modulators of gemcitabine tuxicity. We hypothesized that if we could increase

the amount of es-NT on the cell surface, then greater amounts of gemcitabine

would enter the cell leading to increased cytotoxicity. In examining this

hypothesis, it is important to realize that both 5-FU and tomudex are cytotoxic

antirnetabolite drugs that will exert their own cytotoxic actions on de novo DNA

synthesis in addition to any nucleoside transporter-related effects. These

cytotoxic effects m u r through different mechanisms than those caused by

gemcitabine; TS inhibition versus incorporation into DNA. We therefore are

interested in these drugs' effects on the es-NT and in their combined cytotoxic

effects so that we may gain additional insight into this novel combination

regimen.

5.1 Single Drug Cytotoxicity

The cytotoxicity of gemcitabine, 5-FU and tomudex in the MGH-U1,

PANC-1, HS-766T and PK-8 cell lines was initially detemined. Growoi inhibition

was found to be concentration dependent for all three drugs in al1 four cell lines.

The log concentration-response curves for tomudex in the MGH-U1, PANC-1

and HS-766T cell lines as well as that for 5-FU in the HS-766T cell line appeared

to reach a plateau. In addition, the log concentration-response curves for al1

three drugs in the PANC-1 cell line seemed to reach a plateau. The log

concentration-response curves for gerncitabine and 5-FU continued to decline for

al1 other cell lines, an indication that maximal cytotoxicity has not yet been

attained. Due to this plateau effect mentionad earlier, lCw values were not

obtained for al1 drugs in every cell line.

All the drugs used in this study belong to the anti-metabolite class of

oncolytic agents. These agents ail act as inhibitors of macromolecule

biosynthesis which block cellular replication. Antimetabolites generally compete

with endogenous substrates in cellular metabolic processes to accomplish this

goal. Gemcitabine is in cornpetition with dCyd, 5-FU with dUMP and tomudex

with the endogenous cofactor for TS (5,lO-methylene tetrahydrofolate). For the

TS inhibitors 5-FU and tomudex, it is not surprising to observe a plateau in the

log concentration-response curves. Once TS is completely inhibited, cytotoxic

effects are at a maximum and additional drug present in the cell will not

significantly enhance these effects. Gemcitabine must first be taken up into the

cell via a saturable transport process. Once transport has reached saturating

levels, any additional dnig present will not serve to increase cytotoxicity.

Therefore, we would again expect a plateau to occur in the log concentration-

response plots for gemcitabine. In those cell lines having greater amounts of

functional transporters we would expect this plateau effect to occur farther along

the log concentration-response curve. This trend, however, was not observed

and could be accounted for by non-functional transporter present in the cells or

by decreased dCK eff iciency .

5.1 -1 Gemcitabine

The log concentration-response cuntes for gemcitabine are shown in

figures 4.2ad. The MGH-U1 cell line was the most sensitive to gemcitabine at

its lCso concentration (1.5 nM), having an IC50 value 2.7, 26.7 and 2.3 x 1 o4 times

lower than for the HS-766T, PANC-1 and PK-8 cell lines respectively. No lCso

value was reached for the PK-8 cell line. For the other cell lines, the relative

sensitivities to gemcitabine changed so that an lCgo value of 250 nM for the HS-

766T cell line was 3 and 16 fold lower than those for the MGH-U1 and PANC-1

cell lines respectively (tables 4.2a,b).

Gemcitabine has also been tested in several other human cell lines

including ovarian (A2780, OVCAR-3), colon (WiDr, C26-10) and squarnous cell

carcinoma (UM-SCC-14C, UM-SCC-226) cell fines. Table 5.la compares

gemcitabine sensitivities (24 hour exposure) in these cell lines with those in our

cell Iines at the ICW level. Aside from Our gemcitabine resistant PK-8 cell line,

the IC5* values al1 fall in the nanomolar range.

. :-aiII Une

Table 5.18. Cell doubling times and sensitivities to gemcitabine after 24 hour exposure in several human tumour cell lines. Cell lines defined in text. Data for cell lines not from this work taken from Ruiz Van Haperen et al., 1994.

Gemcitabine is a cell cycle specific agent, with maximal cytotoxicity

occurring during S-phase. We would therefore expect cell lines with shorter

doubling times to be more sensitive to the cytotoxic actions of gemcitabine. The

MGH-U1 cell line has the shortest doubling time, followed by the PK8, PANC-1

and HS-766T cell lines. This, however, does not correspond with cell

sensitivities

at the lCsd values. This agrees with previous data from other human carcinoma

cell Iines mentioned eariier which also shows no correlation between cell

doubling time and gemcitabine sensitivity (Ruiz Van Haperen et ai.. 1994) (Table

5.1 a).

5.1.2 SIFluorouracil

With the exception of the PANC-1 cell Une, the log concentration-response

curves of 5-FU continue to decrease at the highest drug concentrations used,

indicating that maximal cell cytotoxicity has not yet b e n reached (figures 4.2a-

d). 5-FU was most sensitive at its ICx, value of 230 nM in the PANC-1 cell line.

This was 2.2, 12 and 17.4 fold lower than the lCso values in the MGH-U1, HS-

766T and PK-8 cell lines respectively. An ICgO value was only reached in the HS-

766T and PK-8 cell lines and was 55 pM and 65 pM respectively (table 4.2ab).

5-FU has b e n well studied in human colon tumour cells. A cornparison of

our bladder and pancreatic cell line sensitivities with several colon carcinoma cet1

lines may be seen in table 5.1 b. The pancreatic cell lines used for this work are

more sensitive to 5-FU than almost al1 of these colon tumour cell lines.

Table 5.1 b. Cell doubling times and sensitivities to 5-FU and tomudex following 24 hour exposure in several human tumour cell lines. Cell fines defined in text. Data for cell lines not from this work taken from Van Triest et al., 1999.

5-FU is also a cell cycle, S-phase specific cytotoxic agent and like

gemcitabine, dnig sensitivity in the four cell lines used and those listed in table

5.1 b could not be explaineâ by doubling times.

5.1.3 Tomudex

The log concentration-response curves for tomudex reached a plateau in

ail cell lines except for PK-8, in which maximal cytotoxicity was never attained

(figures 4.2a-d). At its lCso concentration, tomudex was most sensitive in the

PANC-1 cell line (0.02 nM), which had an lCso value 2, 150 and 1OOO fou lower

than for the PK-8, HS-766T and MGH-U1 cell lines respectively. Only in the PK-

8 cell line was an IC90 value reached (150 nM) (tables 4.2a.b).

Like 5-FU, tomudex has been studied extensively in human colon cancer

cells. Table 5.1 b shows that al1 of the pancreatic cancer cell lines used in our

study were more sensitive to tomudex than the colon carcinoma cells. This

indicates that tomudex may prove to be a beneficial agent in chemotherapy of

the pancreas. Once again, cell cytotoxicity for a cell cycle specific agent did not

correlate to cell doubling times.

5.2 Drug Combinations

In designing a clinically useful combination therapy regimen, it is desirable

that the dmgs have different, non-overlapping toxicities and different

mechanisms of action. Having non-overiapping toxicities will ideally allow us to

use the maximum tolerated dose of each individual drug when designing the

combination dosing schedule. If the drugs had similar toxicities, we would have

to roll back the dosage of each drug when combining them in order to make the

regimen tolerable to the patient. Possessing different mechanisms and targets of

action prolongs the development of resistance to the therapy because there exist

two separate rnechanisms and cellular processes to which resistance must

develop against For the combination regimens studied in this work (5-FU &

gemcitabine; tomudex & gemcitabine) the dmgs involved possessed non-

overfapping toxicities and differing mechanisms of action. This is one reason

that 5-FU and tomudex make good candidates for combination therapy with

gemcitabine.

In order to establish a combination regimen showing increased efficacy

against pancreatic tumours over gemcitabine alone, gemcitabine was combined

with either 5-FU or tornudex bath concurrently and sequentially in three human

tumour cell lines (MGH-U 1, PANC-1 and HS-766T). Ceil cytotoxicity following

these combination exposures was dependent upon the cell line, single agent

concentrations used and sequence of drug administration. For al1 combination

experiments, a constant concentration of gemcitabine was used within each cell

line while varying the concentration of the DNA synthesis inhibitor (5-FU or

tomudex) (Table 4.3a). This was done to detemine at what concentration(s), if

any, would 5-FU or tomudex increase cellular sensitivity to gemcitabine.

Pressacco et al. (1 9956) have previously reported that DNA synthesis inhibitors,

including 5-FU and tomudex, will upregulate es-NT expression in the MGH-U1

cell line. Based on this observation, we hypothesized that if we could upregulate

the es-NT'S by pre-treatment with either 5-FU or tomudex, we could increase

gemcitabine cytotoxicity.

5.2.1 Gemcitabine and 5-FU

A single concentration of gemcitabine (ICao, ICW, lCse concentration in the

PANC-1, MGH-U1 and HS-766T cell lines respectively) was combined with a

range of 5-FU concentrations from 0-100 pM (Table 4.3a). In ths MGH-U1 cell

line, no increased effects were seen during any schedule of treatment with 5-FU

and gemcitabine, as the cell survival remained approximately the same as for

gemcitabine treatment alone (10%). At higher concentrations of 5-FU in the

PANC-1 and HS-766T cell lines (30, 100 pM), there appeared to ôe an

interaction between the two drugs when they were administered in sequence,

where 5-FU was given first (figures 4.3a,cte).

5.2.2 Gemcitabine and Tomudex

A single concentration of gemcitabine (as above) was combined with

tomudex concentrations ranging from 0-1000 nM. In the MGH-U1 and PANC-1

cell lines, no increased effects were seen versus single agent gemcitabine

treatment with any of the schedules. At the upper range of concentrations of

tomudex in the HS-766T cell line (0.1, 1 PM), increased cell kill was observed

over gemcitabine alone when the cells were pre-treated with tomudex (figures

4.3b,d,q

5.3 Equilibrative-Sensitive Nucieoside Transporter Expression

One way in which gemcitabine entes the cell is via the es-NT

(Heinemann et al., 1995). Through flow cytometric measurements, we quantifid

the amount of cell-surface esNT present per cell in each of the four cell lines

and showed that: PANC-1 > PK-8 = MGH-U1 > HS-766T. We hypothesized that

cell lines containing greater numbers of es-NT'S would show increased

cytotoxicity to gemcitabine. This, however, did not prove to be the case,

indicating that es-NT levels on their own do not provide a good measure for

predicting gemcitabine sytotoxicity. Other factors are most probably playing a

greater role in the detemination of gemcitabine cytotoxicity than are the es-NT'S.

These factors may include: another nucleoside transporter is playing a major role

in gemcitabine uptake in some or al1 of the four cell lines used in this study; efflux

of gemcitabine (through the es-NT andor other mechanisms) may be occurring

at different rates in the four cell lines us&, with one cell line k i n g more efficient

than the others; and finally, the number of es transport sites may not necessanly

correspond to their functionality. Functional studies of the es-NT in these cell

lines are necessary to fully understand and characterize the relationship between

nucleoside transport and gemcitabine cytotoxicity.

In addition to transport-rslated factors, other cellular determinants may be

influencing gemcitabine sensitivity. lntracellular CTP pools - increased pools

lead to increased cornpetition with gemcitabine for incorporation into DNA;

gemcitabine phosphorylation efficiency by deoxycytidine kinase - increased

phosphorylation leads to increased gemcitabine triphosphate concentrations;

gemcitabine elimination efficiency by (d)CMP deaminase - increased elimination

leads to decreased intracellular gemcitabine concentrations. Assaying for these

cellular properties may lend us further insight into the mechanisms detennining

gemcitabine sensitivity.

Cell surface es-NT content in the MGH-U1, PANC-1 and HS-766T cell

lines was also quantified after treatrnent with varying concentrations of 5-FU,

tornudex or gemcitabine to ascertain whether increased cytotoxictty was a

function of increased numbers of cell-surface es-NTs and also to determine

which, if any, of these drugs best modulates gemcitabine cytotoxicity. When

gemcitabine was administered as a single agent, cytotoxicity among the four cell

lines did not correspond to the relative basal arnounts of es-NT in these cell

lines. When gemcitabine was administered in a combination regimen where it

sequentially followed either 5-FU or tomudex, the es-NT contribution to the

modulation of gemcitabine toxicity became more pronounced. We found that in

one of the pancreatic carcinoma cell lines (HS-766T) the amount of es-NT

increased by a factor of 1.7 over basal levels when these cells were pre-treated

with 30 uM and 100 pM 5-FU. When pre-treated with 100 nM and 1000 nM

tornudex, the same cell line showed a 1.9 and 1.4 fold increase respectively in

cell surface es-NT over basal levels. In the other pancreatic tumour cell line

(PANC-1 ), the cell surface es-NT content increased by a factor of 1.6 over basal

levels when the cells were pre-treated with 30 pM 5-FU, but was unaffected by

tomudex pre-treatment (Table 4.4b). When pre-treated with 100 pM 5-FU, the

PANC-1 cells did exhibit increased cell kill as compared to gemcitabine

monotherapy, but did not show a significant increase in cell surface es-NT. All of

these increases were found to be statistically significant (p < 0.05). These

concentrations of 5-FU and tomudex were analysed for increased es-NT content

because the cytotoxic effects seen during their use in combination stuâies

showed increased cell kill over single agent gemcitabine administration. This

evidence lends support to our hypothesis that an increase in cell surface es-NTs

can positively modulate tha in vitro cytotoxicity to gemcitabine in the pancreatic

tumour cell lines tested.

6.1 Conclusions

Following single dnig treatments for 24 hour exposures, one log of cell kill

was not reached in al1 cell lines and the sensitivity to 5-FU, gemcitabine and

tomudex varied in each cell line. In wmbination treatrnents, cell cflotoxicity was

unaffected in the human bladder cancer cell line as compared to gemcitabine

rnonotherapy. In the pancreatic carcinoma cell lines, cytotoxicity was influenced

by the sequence of drug administration. Pre-treatment of these cell Iines with

high concentrations of either 5-FU or tomudex yielded increased cell kill as

compared to gemcitabine alone.

Gerncitabine sensitivity among cell lines was found to be independent of

cell surface es-NT content. Es-NT content was found to be increased in

pancreatic cells that were treated with high concentrations of either 5-FU or

tomudex for 24 hours. Human bladder carcinoma cells did not show this

increase.

In pancreatic tumour cells, we have demonstrated that combinations of 5-

FU and gemcitabine as well as tomudex and gemcitabine produce increased cell

kill versus gemcitabine rnonotherapy in a sequence dependent fashion. where 5-

FU or tomudex is administered prior to gemcitabine. This effect was observed in

the PANC-1 cell line for 5-FU and in the HS-766T pancreatic carcinoma cell line

for both DNA synthesis inhibitors used.

While es-NT content seems to play some role in the determination of

gemcitabine toxicity, our results suggest that there are other major factors

involved. Our studies of the equilibrative-sensitive nucleoside transporter lend us

insight, but do not provide us with the complete picture. The combination

regimen of a DNA synthesis inhibitor such as 5-FU or tomudex. followed by

gemcitabine seems to show promise in in vitro pancreatic cell line studies.

Fuither exploration and analysis of these treatment regimens may lead to a

refined dosing schedule that may be used in the clinic.

6.2 Future Directions

The dmgs used for this work were al1 Sphase specific anti-metabolites

and thus required that the cells be actively synthesizing DNA to exert their

cytotoxic actions. If the duration of drug exposure is increased, more cells may

be in S-phase and thus be susceptible to the actions of gemcitabine, 5-FU and

tomudex. Additionally, a longer exposure pend to these anti-metabolite dnigs

will reduce the concentration of drug needed to produce a fixed effect.

Cytotoxicity was only assayed for 24 hour (maximum) dnig exposures and it may

thus be beneficial to examine longer exposures such as 36, 48 and possibly 72

hours.

For the studies in which we quantified the es-NT, we only examined 24-

hour dnig exposures. Studying longer drug exposures may give us insight into

any time-dependence that exists in nucleoside transporter upregulation.

Fur~ctional, radiolabel studies of the es-NT in the pancreatic cell lines to

characterize parameters such as transport rates and efficiencies would also add

important pieces to the nucleoside transport picture. Functional and rnolecular

studies to determine what other nucleoside transporter types are present and

expressed in pancreatic cancer cells would enable us to gain greater knowledge

into the drug-transporter relationship. This rnay be accomplished in part by

utilizing the fact that cDNAs encoding four human nucleoside transportes have

recently been cloned and functionally expressed: es, ei and two that show cit

activity (Mackey et al., 1998). Gemcitabine sensitivity (and resistance).

therefore, has the potential to be studied in pancreatic tumour cell lines that have

been manipulated so they only express a single type of nucleoside transport

protein.

This project focussed on the quantitative aspects of the equilibrative-

sensitive nucleoside transporter. Studies examining other cellular processes

including intracellular CTP pools, dCK and (d)CMPD expression and efficiency

may provide added insight into the underlying mechanisms of gerncitabine

cytotoxicity among pancreatic cell lines.

Appendix A

Statistical Analysis:

Treatment P Value -

4û 5-FU 0.3998

1 O0 W5-FU 0.1 153

1 00 nM Tomudex - - 0.1 007 . - - -- - - - - - - --.. . - -- - - . . .i. - . - - - - o - l - ~ - . ~ ; . - "

1 pMTomudex ---;--.;:*. . - - - . . - . ..- - - , * . .

10 nM dFdC . 0;4388 7,- -+ - . i

800 nM dFdC OA410 '

Table 1: Is there a statistically significant increase in the number of cell surface es nucleoside transporters after treatment as compared to basal levels in the MGH-U1 cell line. Bolded figures indicate statistically significant differences.

Treatment P.Valus

100 pM 5-FU 0.0562

1WnMTomudex 0.1027 '

1 pkl Tomudex 0.0867

40 nM dFdC 0.0561

1 pM dFdC 0.01 68

Table 2: Is there a statistically significant increase in the number of cell surface es nucleoside transporters after treatment as compared to basal levels in the PANC-1 cell line. Bolded figures indicate statistically significant differences.

1 Treatment P-V-ùe 1 30 pM )FU

100 5-FU O.ml7 100 nM Tornudex

Table 3: Is there a statistically significant increase in the nurnber of cell surface es nucleoside transpoiters after treatment as compared to basal levels in the HS- 766T cell Iine. Boldad figures indicate statistically significant differences.

References

Abbrunese, J. L., Gninewald, R., Weeks, E. A., Gravel, D., Adams, T., Nowak, B., Mineishi, S., Tarassoff, P., Satterlee, W. and Raber, M. N. (1991). A phase I clinical, plasma, and cellular phamacology study of gemcitabine. J Clin Oncol 9(3) : 491 -498.

Abratt, R. P., Bezwoda, W. R., Falkson, G., Goedhals, L., Hacking, D. and Rugg, T. A. (1994). Efficacy and safety profile of gemcitabine in non-smalf-cell lung cancer: a phase II study. J Clin Oncol l2(8): 1 535-1 540.

Agbanyo, F.R., Vijakalakshmi, D., Craik, J.D., Gati, W.P., McAdam, D R , Asakura, J., Robins, M.J., Paterson, ARP. and Cass, C.E. (1990). SAENTA, a novel ligand with high affinity for polypeptides associated with nucleoside transport: Partial purification of the nitrobenzylthioinosine-binding protein of pig erythrocytes by aff inity chromatography. Biochern J 270(3): 605-61 4.

Anderson, H., Lund, B., Bach, F., Thatcher, N., Walling, J. and Hansen, H. H. (1994). Single-agent activity of weekly gemcitabine in advanced non-smallcell lung cancer: a phase II study. J Clin Oncol 12(9): 1821 -1 826.

Aronow, 8., Watts, T. and Lassetter, J. (1984). Biochemicaf phenotype of 5- f l uorouracil-resistant mu rine T-lymphoblasts with geneticalfy altered CTP synthetase activity. J Bi01 Chem 259(14): 9035-9043.

Belardinelli, L., Linden, 3. and Berne, R. M. (1989). The cardiac effects of adenosine. Pr- Cardiovasc Dis 32(1): 73-97.

Belt, J. A., Harper, E. H., Byl. J. A. and Noel, L. D. l1992). Sodiumdependant nucleoside transport in human myefoid leukemic cell lines and freshly isolated myeloblasts. Proc Am Assoc Cancer Res 33: 20.

Belt, J. A., Marina, N. M., Phelps, D. A. and Crawford, C. R. (1 993). Nucleoside transport in normal and neoplastic cells. Adv Enzyme Regul33: 235-252.

Berger, S. H., Jenh, C.-H., Johnson, L. F. and Berger, F. G. (1985). Thymidylate synthase overproduction and gene amplification in fluorodeoxyuridine-resistant human cells. Mol Phannacol 28(7): 461 -467.

Bergman, A. M., Ruiz van Haperen, V. W., Veennan, G., Kuiper, C. M. and Peters, G. J. (1 996). Synergistic interaction between cisplatin and gemcitabine in vitro. Clin Cancer Res 2(3): 521 -530.

Blackledge, G. (1998). New developments in cancer treatment with the novel thymidylate synthase inhibitor raltitrexed (Tomudex'). Br J Cancer f7 Suppl 2: 29-37.

Blaszkowsky, L. (1998). Treatment of advanced and metastatic pancreatic cancer. Front Biosci 3: €21 4-225.

Bleiberg, H. (1997). Colorectal cancer - is there an alternative to 5-FU? Eur J Cancer 33(4): 536-541.

Bouffard, D. Y., Laliberte, J. and Momparler, R. L. (1993). Kinetic studies on 2',2'- difluorodeoxycytidine (Gerncitabine) with purified human deoxycytidine kinase and cytidine deaminase. Bichem Phamcol 45(9): 1857-1 861.

Buolamwini, J.K., Craik, J.D., Robins, M.F., Gati, W.P., Cass, C.E. and Paterson, A.R.P. (1994). Conjugates of fluorescein and SAENTA (5'-S-(2-aminoethy1)-N9 (4-nit robenzyl)-5'-thioadenosine) : Fiow cytometry probes for the es nucleoside transporter elements of the plasma membrane. Nudeosides and Nucleotides 13: 737-751 .

Burke, T., Lee, S., Ferguson, P. J. and Hammond, J. R. (1998). Interaction of 2',2'-difluorodeoxycyüdine (gemcitabine) and forrnycin B with the Nadependent and -inde pendent nucleoside transportes of Ehrlich ascites tumor cells. J Phannacol Exp Ther 286(3): 1 333- 1 340.

Burns, H. and Stomiolo, A. M. (1997). Assessing clinical benefit in the treatment of pancreas cancer: gemcitabine compared to 5-fluorouracil. Eur J Cancer 33 Suppl 1 : S I 8-22.

Cabral, S., Leis, S., Bover, L., Nembrot, M. and Mordoh, J. (1984). Dipyridamole inhibits reversion by thymidine of methotrexate effect and increases drug uptake in sarcoma 180 cells. Proc Natl Acad Sci USA 81 (1 0): 3200-3203.

Camiener, G. W. (1967). Studies of the enzymatic deamination of cytosine arabinoside. II. Properties of the deaminase of the hurnan liver. Biochem Pharmacol 1 6(9): 1 681 -1 689.

Cascinu, S., Silva, R. R., Bami, S., Labianca, R., Frontini, L., Pancera, G., Zaniboni, A., Ferretti, B., Giordani, P., Pessi, M. A., Fusco, V., Luporini, G., Catalano, G. and Cellerino, R. (1998). Gemcitabine and 5-fluorouracil in advanced pancreatic cancer, a GISCAD phase II study. Proc Natl Am Soc Clin Oncol 17.

Cass, C. E., Young, J. D. and Baldwin, S. A. (1998). Recent advances in the molecular biology of nucleoside transporters of mammalian cells. Bimhem Ce11 Bi01 76(5): 761 -770.

Cass, C. E. (1995). Nucleoside transport. Antirnicrobial and Anticancre Chemothera~y. N. H. Georgopapadakou ed. New York, Marcel Dekker, Inc.: 403-451 .

Chabot, G. G., Bouchard, J. and Momparler, R. L. (1983). Kinetics of deamination of 5'-azad'deoxycytidine and cytosine arabinoside by human liver cytidine deaminase and its inhibition by B'deazauridine, thymidine or uracil arabinoside. Biochem Phannacol 32(7): 1 327- 1 328.

Chan, T. and Howell, S. B. (1985). Mechanism of synergy between N- phosphonacetyl-L-aspartate and dipyridamole in a human ovarian carcinoma cell line. Cancer Res 45(8): 3598-3604.

Chan, T. C. (1 989). Augmentation of 1 -beta-D-arabinofuranosyfcytosine cytotoxicity in human tumor cells by inhibiting drug efflux. Cancer Res 49(10): 265602660.

Che, M., Orüz, O. F. and Arias, 1. M. (1 995). Primary structure and functional expression of a cDNA encoding the bile canalicular, purine-specific Na(+)- nucleoside cotransporter. J Bi01 Chem 270(23): 1 35964 3599.

Cheng, Y. C. and Nakayama, K. (1983). Effects of 5-fluoro-2'deoxyuridine on DNA metabolism in HeLa cells. Mol Phannacol P3(l): 171 -1 74.

Chou, T.-C. and Talalay, P. (1984). Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regulation 22: 27-55.

Chu, E., Koeller, D. M., Johnston, P. G., Zinn, S. and Allegra, C. J. (1993). Regulation of thymidylate synthase in human colon cancer cells treated with 5- f l uorou racil and inte rfe ron-gamma. Mol Pharmacol 43(4): 527-533.

Clark, J. L., Berger, S. H., Mittelman. A. and Berger, F. G. (1987). Thymidylate synthase gene amplification in a colon tumor resistant to fluoropyrimidine chemotherapy. Cancer Treat Rep 71 (3): 261 -265.

Clark, J. W., Glicksman, A. S. and Wanebo, H. J. (1996). Systemic and adjuvant therapy for patients with pancreatic carcinoma. Cancer 78(3 Suppl): 688-693.

Clumeck, N. (1993). Current use of anti-HIV drugs in AIDS. J Antimicrob Chemother 32 suppl A: 1 33-1 38.

Crawford, C. R., Patel, D. H., Naeve, C. and Belt, J. A. (1998). Cloning of the hurnan equilibrative nitrobenzylmercaptopuMe riboside (N5MPR)-insensitive nucleoside transporter ei by functional expression in a transport-deficient cell line. J Bi01 Chem 273(9): 5288-5293.

Creasey, W. A., Papac, R. J., Markiw, M. E., Calabresi, P. and Welch, A. O. (1 966). Biochemical and pharmacological studies with 1 -beta-D- arabinofuranosylcytosine in man. Biochem Phamacol15(10): 141 7-1 428.

Cunningham, O., Zalcberg, J., Smith, I., Gore, M., Pazdur, R., Bums, H., 3rd, Meropol, N. J., Kennealey, G. and Seymour, L. (1 996). 'Tomudex' (201 694): a novel thymidylate synthase inhibitor with clinical antitumour activity in a range of solid tumours. Tomudex' International Study Group. Ann OncoI 7(2): 179-1 82.

Danenberg, P. V., Langenbach, R. J. and Heidelberger, C. (1974). Structures of revenible and irrevenible complexes of thymidylate synthetase and fluorinated pyrimidine nucleotides. Bimhemistry l3(5): 926-933.

Drake, J. C., Allegra, C. J., Moran, R. G. and Johnston, P. G. (1996). Resistance to tomudex (ZZD1694): multifactorial in human breast and colon carcinoma cell lines. Biochem PharmacolSl(10): 1349-1355.

Duch, D. S., Banks, S., Dev, 1. K., Dickerson, S. H., Ferone, R., Heath, L. S., Humphreys, J., Knick, V., Pendergast, W. and Singer, S. (1993). Biochemical and cellular pharrnacology of 1843U89, a novel benzoquinazoline inhibitor of thymidylate synthase. Cancer Res 53(4): 81 0-81 8.

Eckardt, J., Eckhardt, G., Vi!lalona-Calero, M., Drengler, R. and Von Hoff, D. (1 995). New Anticancer agents in clinical development. Oncol 9(12): 1321 -1 331.

Fang, X., Parkinson, F. E., Mowles, O. A., Young, J. D. and Cass, C. E. (1996). Functional characterization of a recombinant sodium-dependent nucleoside transporter with selectivity for pyrimidine nucleosides (cNTl rat) by transient expression in cultured mamrnalian cells. Biochem J 31 7(Pt 2): 457-465.

Flanagan, S. A. and Meckling-Gill, K. A. (1 997). Characterization of a novel Na+- dependent, guanosine specific, nitrobenzylthioinosine sensitive transporter in acute promyelocytic leukemia cells. J Bi01 Chem 272(29): 18026-1 8032.

Gaffen, J. D., Chambers, E. A. and Bennett, A. (1989). The effects of dipyridamole and indomethacin on methotrexate cytotoxicity in lovo human colon cancer cells. J Pham Pharmacol41(5): 350-352.

Gandhi, V., Huang, P., Xu, Y. Z., Heinemann, V. and Plunkett, W. (1991). Metabolisrn and action of 2,'2,'-difluorodeoxycfldine: self-potentiation of cytotoxicity. Adv Exp Med Bio4309A): 125-1 30.

Gati, W. P., Paterson, A. R., Larratt, L. M., Turner, A. R. and Belch, A. R. (1997). Sensitivity of acute leukemia cells to cytarabine is a correlate of cellular es nucleoside transporter site content measured by flow cytometiy with SAEMA- fluorescein. BIood SO(1): 346-353.

Griffith, D. A. and Jarvis, S. M. (1 996). Nucleoside and nucleobase transport systems of mammalian cells. Biochim Biophys Acta 1 U)8(3): 1 53-1 81 .

Griffiths, M., Beaumont, N., Yao, S. Y., Sandaram, M., Boumah, C. E., Davies, A., Kwang, F. Y., Coe, I., Cass, C. E., Young, J. D. and Baldwin, S. A. (1997a). Cloning of a human nucleoside transporter implicated in the cellular uptake of adenosine and chemotherapeutic drugs. Nat Med 3(1): 89-93.

Griffiths, M., Yao, S. Y., Abidi, F., Phillips, S. E., Cass, C. E., Young, J. D. and Baldwin. S. A. (19976). Molecular cloning and characterization of a nitrobenzylthioinosine-insensitive (ei) equilibrative nucleoside transporter from human placenta. Biochem J 328(Pt 3): 739-743.

Grunewald, R., Kantajian, H.. Du, M., Faucher, K., Tarassoff, P. and Plunkett, W. (1992). Gemcitabine in Ieukemia: a phase I clinical, plasma, and cellular phamacology study . J Clin Oncol 1 O(3): 406-41 3.

Guchelaar, H. J., Richel, O. J. and van Knapen, A. (1996). Clinical, toxicological and phannacological aspects of gemcitabine. Cancer Treat Rev 22(1) : 1 5-31 . Gutierrez, M. M., Brett, C. M., Ott, R. J., Hui, A. C. and Giacomini, K. M. (1992). Nucleoside transport in bmsh border membrane vesicles frorn human kidney. Biochim Biophys Acta 1 1 OS(1): 1 -9.

Gutierrez, M. M. and Giacomini, K. M. (1993). Substrate selectivity, potential sensitivity and stoichiornetry of Na+-nucleoside transport in brush border membrane vesicles from human kidney. Biochem Biophys Acta 1149(2): 202- 208.

Hammond, J. R. (1991). Comparative phamacology of the nitrobenzylthiguanosine-sensitive and -resistant nucleoside transport mechanisms of Ehrlich ascites tumor cells. J Phannacol Exp Ther 259(2): 799- 807.

Hashimoto, Y., Shiotani, T., Eble, J. N., Gfover, J. L. and Weber, G. (1988). lncreased thymidylate synthase (EC 2.1 -1.45) activity in nomal and neoplastic proliferation. Cancer Biochem Biophys 10(1): 1-1 0.

Heidelberger, C.. Chauduri, N. K. and Danenberg, P. (1957). Fluorinated pyrimidines: a new class of tumour inhibitory compound. Nature 179: 663-66.

Heinemann, V.. Hertel, L. W., Grindey, G. B. and Plunkett, W. (1988). Comparison of the cellular pharmacokinetics and toxicity of 2',2- difluorodeoxycytidine and 1 -beta-D-arabinofuranosykytosine. Cancer Res 48(14): 4024-4031 .

Heinemann, V., Schulz, L., Issels, R. D. and Plunkett, W. (1 995). Gemcitabine: a modulator of intracellular nucleotide and deoxynucleotide metabolism. Semin Oncol22(4 Suppl 11): 11-18.

Heinemann, V., Xu, Y. Z., Chubb, S., Sen, A., Hertel, L. W., Grindey, G. B. and Plunkett, W. (1 992). Cellular elimination of 2',2'-difluorodeoxycytidine 5'- triphosphate: a mechanism of self-potentiation. Cancer Res 52(3): 533-539.

Hidalgo, M., Castellano, Dml Paz-Ares, L., Gravalos, C., Diaz-Puente, M., Hitt, R., Alonso. S. and Cortes-Funes, H. (1999). Phase 1-11 study of gemcitabine and fluorouracil as a continuous infusion in patients with pancreatic cancer. J Clin Oncol 1 7(2): 585-592.

Hogue, D. L., Ellison, M. J., Young, J. D. and Cass, C. E. (1996). Identification of a novel membrane transporter associated with intracellular membranes by phenotypic complementation in the yeast Saccharomyces cerevisiae. J Bi01 Chem 271 (1 6): 9801 -9808.

Honma, Y., Onozuka, Y., Okabe-Kado, J., Dasukabe, T. and Hozurni, M. (1991). Hemin enhances the sensitivity of erythroleukemia cells to 1-beta-D- arabinofuranosylcytosine by both activation of deoxycytidine kinase and reduction of cytidine deaminase activity. Cancer Res 51 (1 7): 453504538.

Huang, P., Chubb, S., Hertel, L., Grindey, G. B. and Plunkett. W. (1991). Action of 2l.2'-difluorodeoxycytidine on DNA. Cancer Res 51 (22): 61 10-61 17.

Huang, Q. Q., Yao, S. Y. M., Ritzel, M. W. L., Paterson, A. R. P., Cass, C. E. and Young, J. D. (1994). Cloning and functional expression of a cornplementary DNA encoding a mammalian nucleoside transport protein. J Biol Chem 269(27): 1 7757-1 7760.

Jackman, A. L., Farmgia, O. C., Gibson, W., Kimbell, R., Hanap, K. R., Stephens, T. C., Azab, M. and Boyle, F. T. (1995). 201694 (Tomudex): a new thymidylate synthase inhibitor with activity in colorectal cancer. Eur J Cancer 31A(7-8): 1277-1282.

Jacobson, K. A., Daly, J. W. and Manganieflo, V., Eds. (1990). Purines in Cellular Sianalfina: Taraets for New Gnias. New York, Springer Veriag, Inc.

Jamieson, G.P., Brocklebank, A.M., Snook, M.B., Sawyer, W.H., Buolamwini, J.K. and Paterson, A R P . (1 994). Flow cytometric quantitation of nucleoside transporter sites on human leukemic cells. Cytometry l4(l): 32-38.

Jarvis, S. M. (1 986). Nitrobenzylthioinosine-sensitive nucleoside transport system : mechanism of inhibition by dipyridamole. Mol PhamacoI30(6): 659-665.

Jarvis, S. M., McBride, O. and Young, J. O. (1 982). Erythrocyte nucleoside transport: asymmetrical binding of nitrobenzylthioinosine to nucleoside penneation sites. J Physiol324: 31 46 .

Johnston, P. G., Lenz, H.-J., Leichman, C. G., Danenberg, K. O., Allegra, C. J., Danenberg, P. V. and Leichman, L. (1 995). Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res 55(7): 1407-1 41 2.

Kang, G.J. and Kimball, A. P. (1984). Dipyridamole enhancernent of toxicity to LI 21 0 cells by deoxyadenosine and deoxycofomycin combinations in vitro. Cancer Res 44(2): 461 -466.

Kauf man, E. R. (1 984). Resistance to 5-fluorouracil associated with increased cytidine triphosphate levels in V79 Chinese hamster cells. Cancer Res 44(8): 3371 -3376.

Kinsella, A. R. and Smith, O. (1998). Tumor resistance to antimetabolites. Gen Phannacol 3O(5) : 623-626.

Labianca, R., Pessi, A., Facendola, G., Pirovano, M. and Luponni, G. (1996). Modulated 5-fluorouracil (5-FU) regimens in advanced colorectal cancer: a critical review of comparative studies. Eur J Cancer 32A Suppl5: S7-11.

Langenbach, R. J., Danenberg, P. V. and Heidelberger, C. (1972). Thymidylate synthetase: mechanism of inhibition by 5-fluoro-2'-deoxyuridylate. Biochem Biophys Res Commun 48(6): 1 565-1 571 .

Lindquist, S. and Craig, E. A. (1988). The heat shock proteins. Annu Rev Genet 22: 631 -677.

Lonn, U. and Lonn, S. (1986). DNA lesions in human neoplastic cells and cytotoxicity of 5'-fluropyrimidines. Cancer Res 46(8): 3866-3879.

Lonn, U., Lonn, S., Nylen, U. and Winblad, G. (1989). 5-fluoropyrirnidine induced DNA damage in human colon carcinoma and its augmentation by the nucleoside transport inhibitor dipyridamole. Cancer Res 49(5): 1085-1 089.

Lu, K., Yin, M. B., McGuire, J. J.. Bonmassar, E. and Rustum, Y. M. (1995). Mechanisms of resistance to N-[5-[N-(3,4dihydro-2-methyl-4-oxoquin~ ylmethyf)-N-methylamino)2-thenoyl1-l-glutamk acid (ZD1694), a folate-based thymidylate synthase inhibitor, in the HCT-8 human ileocecal adenocarcinorna cell line. Biochem Pharmacol 50(3): 39 1 -398.

Lund, B., Hansen, O. P., Theilade, K., Hansen, M. and Neijt, J. P. (1994). Phase II study of gemcitabine (2',2'-difluorodeoxycytidine) in previously treated ovarian cancer patients. J Nat1 Cancer lnst 86(20): 1530-1 533.

Mackey, J. R., Mani, R. S., Selner, M., Mowles, D., Young, J. D., Belt, J. A., Crawford. C. R. and Cass, C. E. (1 998). Functional nucleoside transporters are required for gemcitabine influx and manifestation of toxicity in cancer cell lines. Cancer Res 58(19): 4349-4357.

Mani, R. S., Hamrnond, J. R., Majan, J. M. J., Graham, K. A., Young, J. D., Baldwin, S. A. and Cass, C. E. (1 998). Demonstration of equilibrative nucleoside transporters (hENT1, hENT2) in intracellular membranes of cultured human choriocarcinoma (BeWo) cells by functional reconstitution in proteoliposome. J Bi01 Chem 273(46): 3081 8-30825.

Man, R., Wohlhueter, R. M. and Plagemann, P. G. (1977). Growth rate of cultured Novikoff rat hepatoma cells as a function of the rate of thymidine and h ypoxanthine transport. J Membr Bi01 34(2-3): 277-288.

Mead, G. M. (1996). Raltitrexed, a new drug for advanced colorectal cancer. Lancet 347(9015): 1568-1 569.

Moore, M. (1996). Activity of gerncitabine in patients with advanced pancreatic carcinoma. A review. Cancer 78(3 Suppl): 633-638.

Moore, M. J. (1994). Current status of chemotherapy in advanced pancreatic cancer. Curr Oncol(1): 21 2-21 6.

Moore, M. J., Andersen, J. and Burris, H. (1995). A randomized trial of gemcitabine (gem) versus 5-FU as first-line therapy in advanced pancreatic cancer. Proc Am Soc Clin Oncol 14: 199.

Mulkins, M. A. and Heidelberger, C. (1982). Biochemical characterization of fluoropyrimidine-resistant murine leukemic cell Iines. Cancer Res 42(3): 965-973.

Muller, M. M., Kraupp, M., Chiba, P. and Rumpold, H. (1983). Regulation of purine uptake in normal and neoplastic cells. Adv Enzyme Regulation 21: 239- 256.

Muller, W. E. and Zahn, R. K. (1979). Metabolisrn of 1-beta-D- arabinofuranosyluracil in mouse LW78 cells. Cancer Res 39(3): 1 102-1 107.

Oettle, H., Hochmuth, K., Schmidt, C. A., Vogl, T. J., Keck, H., Schmidt-Woif, I., Neuhaus, P., Riess, H. and Huhn, D. (1998). Phase I triai of Gemzar and 24 hour infusion of 5-fluorouracil, folinic acid, in patients with inoperable pancreatic cancer. Proc Am Soc Clin Oncol 17.

Onetto, N., Momparler, R. L., Momparler, L. F. and Gyger, M. (1987). In vitro biochernical tests to evaluate the response to therapy of acute leukemia with cytosine arabinoside or 5'-aza-2deoxycytidine. Semin Oncol 1 4(2 Suppl 1 ) : 231 - 237.

Paterson, A. R., Gati, W. P., Vjayalakshmi, D., Cass, C. E., Mant, M. J., Young, J. D. and Belch, A. R. (1993). Inhibitor-sensitive, Na+-linked transport of nucleoside analogs in leukemia cells from patients (Abstract). Pra: Am Assoc Cancer Res 34: 14.

Patil, S. D. and Unadkat, J. D. (1997). Sodium-dependent nucleoside transport in the hurnan intestinal bnish-border membrane. Am J Physiol272(6 Pt 1 ): 6 1 31 4- 1 320.

Peters, G. J., van der Wilt, C. L., van Triest, B., Codacci-Pisanelli, G., Johnston, P. G., van Grogeningen, C. J. and Pinedo, H. M. (1 995). Thymidylate synthase and drug resistance. Eur J Cancer 31 A(7-8): 1 299-1 305.

Plagemann, P. G. W. and Kraupp, M. (1986). Inhibition of nucleoside and nucleobase transport and nitrobenzylthioinosine binding by dilazep and hexobendine. Biachem Pharmacol 35(15): 2559-2567.

Plunkett, W., Huang, P., Xu, Y. 2.. Heinemann, V., Grunewald, R. and Gandhi, V. (1 995). Gemcitabine: metabolism, mechanisms of action, and self-potentiation. Semin Oncol 22(4 Suppl 1 1 ): 3-1 0.

Pollera, C. F., Ceribelli, A., Crecco, M. and Calabresi, F. (1994). Weekly gemcitabine in advanced bladder cancer: a preliminary report from a phase I study. Ann Oncol 5(2): 1 82-1 84.

Poplin, E. A., Co-tt, T., Flaterty, L., Tarasoff, P., Redman, B. G., Vaklivieso, M. and Baker, L. (1 992). Diflurodeoxycytidine (dFdC)- gemcitabine: a phase I study. lnvest New Dmgs 1 O(3): 1 65-1 70.

Pressacco, J., Mitrovski, B., Hedley, D. W., Tsang, R. and Erlichman, C. (1 995a). Biochernical modulation of iododeoxy u ridi ne b~ N6-[4- (morpholinosulfonyl)be~l]-N6-methyl-2,6diaminonz[cdindole glucuronate (AG-331 ) leading to enhanced cytotoxicity. Biochem Phannacol 50(1): 55-60.

Pressacco, J., Mitrovski, B., Erlichman, C. and Hedley, D. W. (1 9956). Effects of t hymid ylate synthase in hi bition on thymidine kinase activity and nucleoside transporter expression. Cancer Res 55(7): 1 505-1 508.

Remick, S. C., Grern, J. L, Fischer, P. H., Tutsch, K. D., Alberti, D. B., Nieting, L. M., Tombes, M. B., Bnrggink, J., Wilson, J. K. V. and Trurnp, D. L. (1990). Phase I trial of 5-fluorouracil and dipyridamole administered by seventy-two-hour concurrent continuous infusion. Cancer Res 50(9): 2667-2672.

Rideout, D.C. and Chou, T.C. (1 991 ). Synergism, antagonism and potentiation in chemotherapy: an ovewiew. Svneraism and Antaaonism in Chemotheraw. T.- C. Chou and D.C. Rideout eds. New York, Academic Press Inc.: 61 -1 02.

Ritzel, M. W., Yao, S. Y., Huang, M. Y., Elliott, J. F., Cass, C. E. and Young, J. D. (1997). Molecular cloning and functional expression of cDNAs encoding a human Na+-nucleoside cotransporter (hCNT1). Am J Physiol272(2 Pt 1): C707- 71 4.

Ritzel, M. W. L., Yao, S. Y. M., Ng, A. M. 1.. Mackey, J. R., Cass, C. E. and Young, J. D. (1 998). Molecular cloning, funtional expression and chromosomal localization of a cDNA encoding a human Na+-nucleoside cotransporter (hCNT2) selective for purine nucleosides and uridine. Mol Membr Bi01 1 S(4): 203-21 1.

Riva, C., Barra, Y., Carcassonne, Y., Cano, J. P. and Rustum, Y. (1992). Effect of tetrahydrou ridine on metabolism and transport of 1 -beta-D- arabinofuranosylcytosine in human cells. Chemotherapy 38(5): 358-366.

Robertson, J. M., Shewach, D. S. and Lawrence, T. S. (1 996). Preclinical studies of chemotherapy and radiation therapy for pancreatic carcinoma. Cancer 78(3 Suppl): 674-679.

Rosso, R. and Martin, C. (1994). Effect of patients age on gemcitabine toxicity profile and efficacy in advanced non-small cell lung cancer (NSCLC). Proc Am Soc Clin Oncol 13: 366.

Rothenberg, M. L. (1 996). New developments in chemotherapy for patients with advanced pancreatic cancer. Oncology (Huntingt) lO(9 Suppl): 1 8-22.

Rothenberg, M. L., Burns, H. A., Andersen, J. S., Moore, M., Green, M. R., Portenoy, R. K., Casper. E. S., Tarassoff, P. G.. Stomiolo, A. M. and von Hoff, O. D. (1 995). Gemcitabine: effective palliative therapy for pancreas cancer in patients failing 5-FU. Proc Am Soc Clin Oncol 14: 198.

Ruiz van Haperen, V. W., Veeman, G., Boven, E., Noordhuis, P., Vennorken, J. B. and Peters, G. J. (1994). Schedule dependence of sensitivity to 2',2'- difluorodeoxycytidine (Gemcitabine) in relation to accumulation and retention of its triphosphate in solid tumour cell lines and solid tumours. Biochem Phamacol 48 (7) : 1 327- 1 339.

Ruiz van Haperen, V. W., Veeman, G., Eriksson, S., Boven, E., Stegmann, A. P., Hemsen, M., Verniorken, J. B., Pinedo, H. M. and Peters, G. J. (1994). Development and molecular characterization of a 2',2'- difluorodeoxycytidine- resistant variant of the human ovarian carcinoma cell line A2780. Cancer Res 54(15): 41 38-4143.

Ruiz van Haperen, V. W., Veeman, G., Eriksson, S., Stegmann, A. P. and Peters, G. J. (1 995). Induction of resistance to 2',2'-difluorodeoxycytidine in the human ovarian cancer cell line A2780. Semin Oncol 22(4 Suppl 1 1 ): 35-41 .

Rustum, Y. M. and Creaven, P. J. (1988). Symposium on the expanding role of folates and fluoropyrimidines in cancer chemotherapy. J Nat1 Cancer lnst 80(19): 1524-1 525.

Schnall, S. F. and Macdonald, J. S. (1991). Adjuvant therapy in colorectal carcinoma. Seminar Oncol 18: 560-570.

Sirotnak, F. M., DeGraw, J. I., Schmid, F. A., Goutas, L. J. and Moccio, D. M. (1 984). New folate analogs of the 1 O-deaza-aminopterin series. Further evidence for markedly increased antitumor efficacy compared with methotrexate in ascitic and solid mu rine tumor models. Cancer Chemother Phannacol 1 2(1): 26-30.

Sliutz, G., Karlseder, J., Tempfer, C., Orel, L., Holzer, G. and Simon, M. M. (1996). Dwg resistance against gemcitabine and topotecan mediated by constitutive hsp7O overexpression in vitro: implication of quercetin as sensitiser in chemotherapy. Br J Cancer 74(2): 172-1 77.

Steuart, C. D. and Burke, P. J. (1971). Cytidine deaminase and the development of resistance to arabinosyl cytosine. Nat New Bi01 233(38): 1 09-1 1 0.

Sundaram, M., Yao, S. Y. M., Ng, A. M. L., Griffiths, M., Cass, C. E., Balâwin, S. A. and Young, J. D. (1998). Chimeflc constructs between human and rat equilibrative nucleoside transporters (hENT1 and rENT1) reveal h E M l structural dornains interacting with coronary vasoactive drugs. J Bi01 Chem 273(34): 21 51 9-21 525.

Tempera, M., Capadano, M. and Tarassoff, P. (1994). Dose escalation of gemcitabine in previously untreated patients with pancreatic adenocarcinorna. Proc Am Soc Clin Oncol 1 3: 21 8.

Tobias, N. and Blau, A. C. (1996). Forced expression of cytidine deaminase confen resistance to cytosine arabinoside and gemcitabine. Exp Hemafology 24: 1340-1 346.

van Triest, B., Pinedo, H. M., Hensbergen, Y., Smid, K., Telleman, F., Schoenmakers, P. S., Wilt, C. L., Laar, J. A. M., Noardhuis, P., Jansen, G. and Peters, G. J. (1999). Thymidylate synthase levef as the main predictive parameter for sensitivity to 5-fluorouracil, but not for folate-based thymidylate synthase inhibitors, in 13 nonselected colon cancer cell lines. Clinical Cancer Res 5: 643-654.

Wang, J., Schar:er, M. E., Thornassen, S., Su, S.-F., Piquette-Miller, M. and Giacomini, K. M. (1997a). Functional and molecular characteristics of Na+- dependent nucleoside transporters. Pharmaceutical Res l 4 ( l l ) : 1 524-1 532.

Wang, J., Su, S. F., Dresser, M. J., Schaner, M. E., Washington, C. B. and Giacomini, K. M. (19976)- Na(+)-dependent purine nucleoside transporter from human kidney: cloning and functional chracterization. Am J Physiol273(6 Pt 2): FI 058-1 065.

White, J. C., Rathmell, J. P. and Capiui, R. L. (1987). Membrane transport influences the rate of accumulation of cytosine arabinoside in human leukemia cells. J Clin lnvest ig (2 ) : 380-387.

Wiley, J. S., Taupin, J., Jarnieson, G. P., Snook, M., Sawyer, W. H. and Finch, L. R. (1985). Cytosine arabinoside transport and metabolisrn in acute leukemias and T cell lymphoblastic lymphoma. J Clin lnvest 75(2): 632-642.

Wilson, J. K. V., Fischer, P. H., Remick, S. C., Tutsch, K. O., Grem, J. L., Nieting, L., Alberti, D., Bniggink, J. and Trump, O. L. (1989). Methotrexate and dipyridamole combination chemotherapy based upon inhibition of nucleoside salvage. Cancer Res 49(7): 1866-1 870.

Wohlhueter, A. M. and Plagemann, P. G. W. (1980). The roles of transport and phosphorylation in nutrient uptake in cultured animal cells. h t Rev Cytol64: 171 - 240.

Xu, Y. Z. and Plunkett, W. (1992). Modulation of deoxycytidine deaminase in intact human leukemia cells. Action of 2',2'-dHuorodeoxycytidine. Biochem Phannacol 44: 1 8 1 9-1 827.

Yao, S. Y., Cass, C. E. and Young, J. D. (1996a). Transport of the antiviral nucleoside analogs 3'-azido-3'-deoxythymidine and 2',3'-dideoxycytidine by a recombinant nucleoside transporter (CM) expressed in Xenopus laevis oocytes. Mol Phannacol 50(2): 388-393.

Yao, S. Y., Ng, A. M., Ritzel, M. W., Gati, W. P., Cass, C. E. and Young, J. 0. (1 9966). Transport of adenosine by recombinant purine- and pyrimidine-selective sodium/nucleoside cotransporten f rom rat jejunum expressed in Xenopus laevis oocytes. Mol Pharmacol SO(6): 1 529-1 535.