€¦ · · 2004-11-29demonstrated the ARA-C resistance phenotype by clonogenic assay. ... Long-term expression of the CD transgene by enzyme activity assay was ... 4.2.2 Advantages

Universi* de M o n W F a d t 1 des etudes supirieures

Cette t h k intitulk

VALUATION ~ L I ~ ~ I Q U E DU G- DE LA CYTXDII'iE D~IMXNASE POUR LA CHIMIOPROTECTION DAlPS LA ~ R A P I E DU CANCER

PRECLI14ICAL EVALUATION OF THE CPIlDINE DEAMINASE GENE FOR CHEMOPROTECTION IN CANCER THERAPY

Present& par

Nicoletta Eliopoulos

a Cti ivalule par un jury compost des personnes suivantes:

Prisident-rapporteur Dr. Teresa Kus Directeur de recherche: Dr. Richard L. Momparler Membre du jury: Dr. Andre De Lian Examinateur externe: Dr. Moulay A. Alaoui-Jamali Reprisentant de la F.E.S.: Dr. Guy Sauvageau

Thtse acceptee le: 14 '. ' '' f

Patients with advanced metastatic cancer do not survive very long after treatment with conventional chemotherapy. New antineoplastic drugs and novel methods to significantly enhance the curative potential of chemotherapy should be investigated. Increasing the dose of anticancer agents has the potential to increase the therapeutic response. However, augmenting the intensity of the chemotherapy is often impeded by the severe bone marrow toxicity produced by most antineoplastic agents. An approac.1 to resolve this problem of dose-limiting myelosuppression would be to confer chemoresistance to normal hematopoietic cells. Studies have shown that marrow cells can be rendered resistant to certain classes of antineoplastic drugs by insertion of drug resistance genes.

The chemotherapeutic efficacy of cytosine nucleoside analogs, such a s cytosine arabiioside (ARA-C), 2',2'-difluorodeoxycytidine (dFdC), and 5- aza-2'-deoxycytidine (5-AZA-CdR), is limited by their dose-dependent myelotoxicity. The main objective of this work was to confer drug resistance against these cytosine analogs in normal blood cells by transfer of the gene for human cytidine deaminase (CD). CD catalyzes the deamination of cytosine nucleoside analogs resulting in a loss of their antineoplastic activity. The cDNA for CD was introduced into the MFG retroviral vector. Ecotropic murine packaging cells were transfected with the pMFG-CD retroviral construct and clones producing the recombinant virus were isolated. NIH 3T3 mouse fibroblast cells were transduced with these retroviral particles and clones of CD gene modified cells were isolated after ARA-C selection. These CD-transduced fibroblast cells displayed elevated expression of CD enzyme activity and augmented mRNA levels. In addition, it was established that transfer of the CD gene into fibroblast cells renders them resistant to the cytotoxicity of ARA-C, dFdC, and 5- AZA-CdR. This drug resistance to cytosine nucleoside analogs was shown to be reversed by tetrahydrouridine (THU), a competitive inhibitor of CD. THU could also diminish the enhanced CD enzyme activity observed in the CD-transduced cells. This c o n f i e d that the acquired drug resistance phenotype in CD gene-modified cells was due to the expression of the proviral CD. Primary murine bone marrow cells were also transduced with retroviral particles containing the human CD and subsequently demonstrated the ARA-C resistance phenotype by clonogenic assay.

To illustrate the feasibility of in uivo investigations with the CD gene, murine bone marrow cells f h m male mice were transduced ex uivo with the CD virions and then transplanted into female syngeneic mice. The recipient mice were subsequently treated with ARA-C. Analysis of the marrow, spleen, and blood of CD-recipient mice revealed the presence of proviral CD, detected by PCR, at over 10 months after transplantation. Long-term expression of the CD transgene by enzyme activity assay was also noted in these hematopoietic tissues.

These results suggest that gene therapy with the CD gene has the potential to increase the antitumor effectiveness of cytosine nucleoside analogs by preventing myelosuppression after dose intensification.

Pnkentement, en clinique, le taux de survie de patients atteints de certaines formes de cancer avan* n'est pas significativement augmente par la chimiotherapie conventionnelle. De nouveaux agents chimiothirapeutiques ainsi que de nouvelles approches exp6rimentales sont donc requis a h d'amiliorer la *ponse therapeutique. En thlorie, I'escalation des doses de medicaments antinhplasiques est une des solutions pour accroitre la destruction de cellules tumorales et consckluemment, atteindre une therapie curative. Cependant, la toxicite indesirable sur les cellules normales telles que les cellules htmatopoietiques, un effet secondaire courant de la chimiotherapie, est elle aussi augmentee avec les doses, limitant ainsi lritensification de la chimiothtrapie et donc I'efficacite.

Rlcemment, les nouvelles possibiiites offertes par la therapie genique nous permettent de tenter de resoudre le probleme de la myelosuppression en conferant une chimioresistance aux cellules hematopoiitiques normales par le transfert de genes de resistance midicamenteuse. A ce jour, plusieurs de ces genes ont ite identifies et ce contre nombreuses classes d'agents anticancereux. Des etudes ont demontre que Irisertion, par transfert genique, de ces genes dans des cellules peut rendre celles-ci rkistantes a certains medicaments. Plusieurs chercheurs ont introduit par exemple, le gene pour la resistance a une multitude de medicaments lipophilliques ("multidrug resistance"), a l'aide d'un vecteur *troviral, dans des cellules souches hematopoietiques normales dorigine marrimifere, incluant les cellules humaines. Une expression de ce gene a ete obtenue in uitro dans les cellules infectees et aussi in uivo chez la souris surtout Des essais cliniques avec ce gene sont presentement en cours chez des patients atteints d'un cancer du sein, de I'ovaire ou du cerveau mktastatique. Bgalement, une forme mutante du gene de la dihydrofolate reductase, l'enzyme cible du methotrexate, a aussi Cti insiree, par le moyen d'un vecteur rltroviral, dam des cellules souches hematopoietiques normales. Des souris transplantees avec ces cellules sont ainsi devenues rc5sistantes aux effets toxiques du methotrexate. Une etude recente a revtll que chez des souris avec adtnocarcinome mammaire transplantees avec des cellules hematopoiltiques infectees avec

le gene de la dihydrofolate rductase mutk, une -ion totale de la tumeur a it6 obtenue avec le mithotrexate, dans 44% des cas.

L'efficaciti chimiothirapeutique des analogues de la dhxycytidine. tels que le cytosine arabiioside (ARA-C), le 2',2'-dinuorodhxycytidine (dFdC), et le 5-aza-2'-dhxycytidine (5-AZA-CdR), est limitee par la toxicite himatopoiitique. La my~losuppression, surtout la neutropinie, restreint la dose de ces agents et donc par cons5quent leur potentiel curatif. Ces analogues de la dhxycytidine ont dtmontri des activit* antitumorales prometteuses chez des patients atteints d'un cancer avance. L'approche de la chimioprotection par thirape ginique a le potentiel de produire une meilleure eponse au traitement.

Le principal objectif de ce travail itait de conferer aux cellules sanguines normales, une n5sistance contre les analogues de la dlsoxycytidine par transfert du gene humain de la cytidine disaminase (CD). La CD catalyse la d6smhation des nucleosides de la cytosine. Cet enzyme catabolique pourrait jouer un rde important dans la chimion5sistance puisqu'il catalyse igalement la dtsamination et donc lriactivation des analogues de la disoxycytidine. L'ADN complementaire (ADNc) de la CD humaine a it6 introduit dans le vecteur ritroviral MFG. Des fibroblastes 'packaging" (lignee cellulaire d'encapsidation) GP+E86 de souris ont it6 transfectes avec le vecteur pMFG-CD et des clones produisant le virus ricombiant ont it6 isolis. Des fibrob1as:es NIH 3T3 de souris ont et i infectis avec ces particules virales et des clones de cellules modifiies par le gene de la CD ont i t i isoles apres silection a

I'M-C. Ces fibroblastes infectis avec la CD ont n5vili par dosage enzymatique une hausse significative de l'activite enzymatique de la CD. Par analyse 'Northernw des niveaux augmentes de I'ARN messager de la CD ont aussi eti reviles. De plus, le transfert du gene de la CD dans les cellules fibroblastes a rendu celles-ci risistantes a des concentrations cytotoxiques d'ARA-C, de dFdC et de 5-AZA-CdR. Cette risistance aux analogues de la desoxycytidine a i t i inhibie par le 3,4,5,6- tetrahydrouridine (THU), un inhibiteur cornpetitif de la CD. Le THU a aussi attinue la hausse de l'activiti de la CD observie dans les cellules infectees par la CD. Ce resultat a confirmi que le phenotype de la fisistance mtdicamenteuse acquis dans les cellules modiflies avec le gene de la CD itait dd a I'expression de I'ADN provirale de la CD. Des cellules htmatopoiCtiques primaires de souris ont aussi t te infecties avec des

vii

particules virales contenant I'ADNc de la CD. Une rkistance a I'ARA-C a it6 dimon* in vitru dans ces cellules par test de clonoginiciti.

Le prochain but itait d'accomplir le transfert et I'expression in vivo du gene de la CD. Des cellules himatopoiitiques provenant de la moelle osseuse de souris mide ont it6 infect& ex vivo avec des particules r&rovirales contenant I'ADNc de k CD. Ces cellules infectees ont i t t transplant& chez des souris femelles syngentiques qui par la suite ont

des traitements a 1'ARA-C. Le sang de ces souris, examine a plusieurs intervalles de temps apes la transplantation, a r&ele par analyse 'PCRn (hction en chaine de la polymerase) la p e n c e de I'ADN proviral de la CD. Plus de 10 mois apes la transplantation, I'ADN proviral de la CD a i t i detect& par PCK dans la mode osseuse, la rate et le sang des 'souris CD". L'ADN proviral de la CD a aussi ete detect6 par PCR dans des colonies hematopoiitiques formks a partir de cellules preletrees de la moelle osseuse des souris CD. Par test de clonogenicite q~~elques souris ont aussi montre de la resistance a I'ARA-C in vitro. L a surexpression a

long terme de la CD a ite notie par analyse de l'activitt enzylnatique de la CD dans ces tissus himatopoiitiques. Les cellules hinlatopoietiques infecties peuvent donc survivre in vivo, exprimer la CD pour une ptriode de temps prolong& et former des colonies in vitro.

Ces resvltats indiquent que le gene de la CD a une application future importante dans le dornaine de la thirapie genique. La thtrapie genique avec le gene humain de la CD aura le potentiel d'augmenter l'efficacite clinique des analogues de la disoxycytidine en reduisant leur my~losuppression et permettant alors I'usage de doses plus ilevees pour une meilleure activite antinbplasique.

viii

TABLE OF CONTENTS

... ............................................................................................... SUMMARY UI

................................................................................................... RGSUM~ v

................................................................................... LIST OF TABLES xiv

.................................................................................. LIST OF FIGURES x v

................................................................................. ABBREVIATIONS .xvii

DEDICATION .......................................................................................... xk

............................................................................ ACKNOWLEDGMENTS %x

ANNEX1 ................................................................................................. X X ~

PART ONE: INTRODUCTION

CHAPTER 1: Foreword ............................................................................. 1

CHAPTER 2: Cytosine nucleoside analogs ................................................ 2

.................................................................... 2.1 Cytosine arabiioside 2

2.2 2',2'-Difluorodeoxycytidine ........................................................... .3

........ CHAPTER 3: Hematopoietic toxicity produced by anti-cancer drugs 8

................................................................................... 3.1 Leukopenia 8

........................................................................................ 3.3 Anemia 1 1

CHAPTER 4: Chemoprotcction against the hematopoietic toxicity produced by anti-cancer drugs. using gene therapy ............ 12

..................................................................................... 4.1 Rationale 12

4.2 Vectors for gene transfer .............................................................. 13

4.2.1 Replication cyde of retroviruses and production of viral ............................................. particles using packaging cells 15

4.2.2 Advantages and disadvantages of retroviruses a s vectors .................................................................. for gene transfer 18

....................................................... CHAPTER 5: Drug-resistance genes 20

........................................................... 5.1 Multidrug-resistance gene 20

5.1.1 Mechanism .......................................................................... 20

5.1.2 In vitm studies in hematopoietic cells ................................. 20

5.1.3 In vivo studies in hematopoietic cells ................................... 22

..... 5.1.4 Clinical studies on chemoprotection with the MDRl gene 24

5.2 Dihydrofolate reductase gene ....................................................... 26

....................................................... 5.2.1 Biochemical role in cells 26

................................... 5.2.2 In vitro studies in hematopoietic cells 26

5.2.3 In vivo studies in hematopoietic cells ................................... 28

5.3 Methylguanine-DNA-methyltransferase gene ............................... 30

....................................................... 5.3.1 Biochemical role in cells 30

X

5.3.2 In vitro studies in hematopietic cells .................................. 31

5.3.3 In vivo studies in hematopietic cells ................................... 32

5.4 Glutathione-S-transferase gene ................................................... 34

5.4.: Biochemical role in cells ....................................................... 34

5.4.2 In vitro studies in cells .......................................................... 34

5.5 Aldehyde dehydrogenase gene ...................................................... 35

5.5.1 Biochemical role in cells ....................................................... 35

5.5.2 In vitro studies in cells ......................................................... 35

.............................. 5.6 Multidrug resistance-associated protein gene 37

5.6.1 Biochemical role in cells ....................................................... 37

5.6.2 In vitro studies in cells ......................................................... 37

5.7 Cytidiie deaminase gene ............................................................ 38

5.7.1 Description .......................................................................... 38

5.7.2 Inhibitors of cytidine deaminase .......................................... 42

CHAPTER 6: Objectives of the present investigation .............................. 45

PART TWO: ARTICLES

CHAPTER 7: Article 1 Resistance to cytosine arabiioside by rctrovirally mediated gene transfer of human cytidine deaminase into murine fibroblast and hematopoietic cells ....................................... 46

Abstract ............................................................................................ 47

Introduction .................... .. ............................................................. 48

Materials and Methods ....................................................................... 50

............................................................................................... Results 55

Discussion .......................................................................................... 65

............................................................................... Acknowledgments 67

.......................................................................................... References 68

CHAPTER 8: Article 2 Drug resistance to 5.aza.2'.deoxycytidine, 2'.2'.difluorodeoxycytidine. and cytosine arabinoside conferred by retroviral-mediated transfer of human cytidiie deaminase cDNA into murine cells ........................ 73

Abstract ............................................................................................. 74

....................................................................................... Introduction 75

Material and Methods ......................................................................... 77

.............................................................................................. Results -79

.......................................................................................... Discussion 86

............................................................................... Acknowledgments 88

.......................................................................................... References 89

CHAPTER 9: Article 3 Retroviral transfer and long-term expression of human cytidine deaminase cDNA in hematopietic cells following transplantation in mice ................................ 94

............................................................................................. Abstract 95

....................................................................................... Introduction 96

Materials and methods ...................................................................... 97

............................................................................................ Results -101

Discussion ........................................................................................ 111

Acknowledgments ............................................................................. 112

References ........................................................................................ 113

PART THREE: DISCUSSION

CHAPTER 10: DISCUSSION ................................................................. 117

10.1 Evalua5on of experimental results ........................................... 117

10.2 Advantages of using the CD gene for chernoprotection in cancer therapy with cytosine nucleoside analogs ...................... 124

10.3 Approaches to facilitate the use of gene therapy for ....................................................................... chemoprotection 125

10.3.1 Modification of retroviral vectors to improve gene transfer and expression ................................................................ 125

10.3.2 Hernatopoietic cells as targets for gene therapy ................. 126

10.3.3 Ways to overcome the potential problems of retroviral vectors ............................................................................. 126

10.4 Future studies ....................................................................... 127

10.4.1 Bicistronic vectors ............................................................ 127

10.4.2 Animal studies ................................................................. 128

10.4.3 Clinical studies on chemoprotection for tumor therapy ..... 129

REFERENCES ....................................................................................... 132

ANNEX 1: Article: Ransfection of murine fibroblast cells with human cytidine deaminase cDNA confers resistance to cytosine arabinoside ............................................... xxi

LIST OF TABLES

Table 4.1 Characteristics of gene delivery systems ................................. 14

Table 7.1 CR deaminase assay in murine cells ....................................... 57

Table 7.2 Effect of ARA-C on colony formation by murine fibroblasts ...... 58

Table 7.3 Effect of ARA-C on clonogenic assay for murine hematopoietic cells ........................................................................................ 59

Table 8.1 Effect of tetrahydrouridine (THU) on the antineoplastic action of 5-AZA-CdR, dFdC, and ARA-C on CR deaminase- transduced 3T3-CD3-V5 cells ................................................. 81

Table 8.2 Effect of different concentrations of THU on CR deaminase activity in murine cells ............................................................ 82

Table 9.1 PCR analysis for detection of CD proviral DNA in blood cells from CD mice collected after different ARA-C treatment cycles .......................................................................... 105

Table 9.2 PCR analysis of bone marrow-derived colonies from mice transplanted with MFG-CD transduced marrow cells ............ 106

Table 9.3 Cytid'ie deaminase (CD) activity in blood cells ...................... 107

Table 9.4 Effect of ARA-C on clonogenic assays .................................... 108

LIST OF FIGURES

Figure 2.1 Chemical structure of deoxycytidine and related analogs ....... 6

Figure 2.2 Metabolism of cytosine arabinoside (ARA-C), a cytosine nucleoside analog .................................................................. 7

Figure 3.1 Example of the hematopoietic toxicity in a patient with metastatic lung cancer following treatment with an 8 hour infusion of 5-aza-2'-deoxyctidine (5-AZA-CdR) at a total dose of 660 mg/ml ...................................................... 10

Figure 4.1 Schema of gene transfer into a packaging cell line for the production of recombinant retroviral virions for the transduction of target cells ............................................. 17

Figure 5.1 Conversion of 5-aza-2'-deoxyctidine and 2',2'-difluorodeoxycytidine to inactive metabolites

............. by enzymatic deamination with cytidine deaminase 4 1

Figure 5.2 Chemical structure of specific inhibitors of cytidiie deaminase ........................................................................... 44

Figure 7.1 Northern blot hybridization of 10 pg total RNA from each of the indicated cell limes ............................................ 60

Figure 7.2 Southern blot analysis of 10 pg of purified genomic ......................................... DNA from the indicated cell lines 61

Figure 7.3 Inhibition of DNA synthesis by ARA-C ................................. 62

Figure 7.4 Growth inhibition by ARA-C ................................................ 63

Figure 7.5 PCR analysis for the presence of proviral DNA in murine marrow cells ............................................................ 64

Figure 8.1

Figure 8.2

Figure 8.3

Figure 9.1

Figure 9.2

xvi

Effect of 5-AZA-CdR and THU on colony formation by murine fibroblasts .......................................................... 83

Effect of dFdC and THU on colony formation by murine fibroblasts ............................................................... 84

Effect of ARA-C and THU on colony formation by murine fibroblasts .............................................................. -85

Detection by PCR of Y-chromosome, CD provirus, and 5-globin in various tissues a t 11-13 months after

.................................................................. transplantation 109

Expression of human CD enzyme activity in bone marrow ...... and spleen cells at 11-13 months after transplantation 110

Figure 10.1 Molecular design of the retrovkal vector for the expression of cytidine dearninase cDNA (CD CDS) in target cells ......... 118

................................ Figure 10.2 Experiment for gene transfer in mice 123

Figure 10.3 Illustration of a hypothetical clinical protocol for chemoprotection .......................................................... 130

Figure 10.4 Hypothetical hernogram to illustrate chemoprotection from the hematopoietic toxicity produced by cytosine

............................................................ nucleoside analogs 131

ABBREVIATIONS

ALDH m - C ARA-CDP m - C M P ARA-CTP ATP 5-AZA-CdR 5-AZA-dCTP BCNU

bp CD CD34+

cDNA cfu, CFU CFU-C CFU-GM

CPG CR dCDP dCMP dCTP dFdC DHFR D-MEM FBS BSA 5-FU (2418 G-CSF GFP GM-CFC

aldehyde dehydrogenase cytosine arabiioside diphosphate form of ARA-C monophosphate form of ARA-C triphosphate form of ARA-C adenosine triphosphate 5-aza-2'-deoxycytidiie triphosphate form of 5-AZA-CdR 1,3-bis (2-chloroethy1)-1-nitrosourea base pair cytidine deaminase presence of cluster differentiation antigen #34 on hematopoietic cells complementary DNA colony forming unit colony forming unit - cell colony forming unit - granulocyte macrophage cytosine-phosphate-guanine cytidiie deoxycytidine diphosphate deoxycytidine monophosphate deoxycytidiie triphosphate 2',2'-difluorodeoxycytidine dihydrofolate reductase Dulbecco's modi!ied essential medium fetal bovine serum bovine serum albumin 5-fluorouracil neomycin granulocyte colony stimulating factor green fluorescence protein granulocyte macrophage colony forming cell

GM-CSF

GST IRES kb LacZ LTR MDR a-MEM MGMT

MoMLV mRNA MTX MRP pMFG

PBS PCR SCID SDS SSC THU TMTX 3T3-CD3-V5

VHL

granulocyte macrophage colony stimulating factor clone of GP+E86 packaging cells sansfected with pMFG-CD clone of GP+E86 packaging cells transfected with pMFG-Lac2 glutathione-S-transfemse internal ribosomal entry site kilobase P-galactosidase gene long terminal repeat multidrug resistance alpha minimal essential medium 06-methylguanine-DNA- methyltransferase Moloney murine leukemia virus messenger RNA methotrexate multidrug resistance-associated protein code name given by Richard Mulligan for this retroviral vector plasmid phosphate-buffered d i e polymerase chain reaction severe combined immunodeficiency sodium dodecyl sulphate standard saline citrate 3,4,5,6-tetrahydrouridine trimetrexate clone of 3T3 fibroblast cells transduced with MFG-CD virions clone of 3T3 fibroblast cells transduced with MFG-CD virions von Hippel Lindau

To my family,

with special dedication to my mother,

Panagiota Papageorgakopoulos Eliopoulos

I express my profound gratitude and appreciation to Dr. Richard L. Momparler for having accepted me into his laboratory, for his excellent supervision and constant guidance, his optimism, patience and encouragement, for the many opportunities he has offered me, and even more meaningful to me, for his remarkable kindness and good- heartedness.

I also gratefully acknowledge:

Louise F. Momparler for her indispensable assistance, and mostly for her friendship, her encouragement, and her goodness and caring nature which greatly contribute to creating a very pleasant atmosphere in the laboratory.

Oanh N.L. Le for her valuable assistance and especially for her kindness, humor, and friendship.

Christian Beausejour, Veronica Bovenzi, Sylvie Ccte, and Khanh Vu for their considerate actions and helpfulness, and particularly for their kindness and friendship.

David Bouffard, Li Caravechia, Benoit Dore, and Josee Laliberte for their help and experience when I joined the laboratory.

Dr. Denis Cournoyer and Sylvain Litourneau for their collaboration, informative meetings and helpful comments.

The Departement de Pharmacologie de I'Universite de Montreal and the Centre de recherche de lhdpital SteJustine for their contribution to my research training.

My parents, my grandparents, my brother and his beautiful family, for their love and support, and all they have done for me.

PART ONE

INTRODUCTION

CHAPTER 1: Foreword

Present day chemotherapy for advanced metastatic cancer is not very effective (Wingo et al., 1995). The long-term survival of patients with metastatic lung, breast, prostate, and colon cancer treated with chemotherapy is less than 10%. New approaches to improve the therapeutic effectiveness should be explored. In general, escalation of the dose of antineoplastic agents, and/or frequency of treaaent used in patients may potentially lead to a curative response. However. intensification of chemotherapy is limited by its toxicity to the bone marrow. A possible way to overcome this problem would be to render normal hematopoietic cells resistant to chemotherapeutic drugs and hence reduce their adverse effects. Several investigators have demonstrated that gene transfer of drug resistance genes into marrow cells decreased their sensitivity to some classes of anticancer agents. For example, chemoprotection to different types of antineoplastic drugs has been well- documented for the multidrug resistance (MDR) and diiydrofolate reductase (DHFR) genes (Corey et al, 1990; Hanania et al, 1995).

The potential effectiveness in cancer therapy, of cytosine nucleoside analogs, such as cytosine arabiioside (ARA-C), 2',2'-diffluorodeoxycytidine (dFdC), and 5-aza-2'-deoxycytidine (5-AZA-CdR), is reduced by their dose- limiting myelosupression (Czaykowski et al., 1997; Fossella et d., 1997; Momparler et al., 1997). The cytidine deaminase (CD) gene which inactivates cytosine nucleoside analogs by deamination (Bouffard et al., 1993; Chabot et al.. 1983) may be an interesting gene to confer resistance to these drugs. Gene therapy with the CD gene may have the potential to circumvent the hematopoietic toxicity produced by chemotherapy with cytosine nucleoside analogs and consequently increase their clinical effcacy. This thesis contains data to support this hypothesis.

CHAPTER 2: Cvtosine nucleoside analogs

2.1 Cytosine arabinoside

Cytosine arabinoside (ARA-C) is one of the most studied nucleoside analogs. It differs structurally from the natural occurring, deoxycytidine, by the existence of a 2'-hydroxyl group in the trans position of the sugar moiety (Figure 2.1). For review see Grant, 1998.

A characteristic of all cytosine nucleoside analogs, ARA-C is a prodrug that must be activated by phosphorylation in order to be rendered cytotoxic (Furth and Cohen. 1968; Graham and Whitmore, 1970). It exerts its toxic effect on cells present in the S phase of the cell cycle (Karon and Shirakawa, 1969). Like the natural nucleosides, ARA-C penetrates into cells by carrier-mediated nucleoside transport (Plagemann et al., 1978a). In the cell, deoxycytidine kinase, the rate-limiting enzyme, catalyzes the conversion of this analog to its 5'-monophosphate form, ARA-CMP (Momparler and Fischer, 1968). Phosphorylation to the active triphosphate derivative, ARA-CTP, occurs following the sequential actions of dCMP kinase and dCDP kinase (Grant, 1998). The metabolism of ARA-C is summarized in Figure 2.2.

ARA-CTP acts on DNA polymerase as a competitive inhibitor of dCTP (Momparler, 1969) and can also function as a substrate for incorporation into growing DNA strands. The presence of this analog in an internucleotide position can slow down DNA replication whereas its presence a t the 3'-terminal position of the DNA strand can lead to chain termination (Grant, 1998; Mikita and Beardsky, 1988).

ARA-C is one of the most active drugs used in the treatment of patients with acute myeloid leukemia (Keating et al., 1982). It also possesses activity against lymphomas (Kantarjian et al., 1983; Shipp et al., 1984) and acute lymphoblastic leukemia (Stryckrnans et al., 1987). Although in vitro studies demonstrated antineoplastic activity of ARA-C against human tumor cell lines (Kern et al., 1988), this drug a t various dose-schedules tested has not been shown to be very effective in treating solid tumors (Chabner, 1996). However, in a recent report, an unexpected prolonged remission was achieved in a woman with advanced metastatic breast cancer subsequent to the accidental administration of high dose

3

ARA-C (Czaykowski et al. 1997). The main side-effect of ARA-C is bone marrow toxicity (Lie and Slerdahl, 1985).

2',2'-difluorodeoxycytidine (dFdC) is a new cytosine nucleoside analog developed in the late 1980s in the research laboratories of Eli L iy (Hertel et al., 1988). For review see Parkinson et al. (1996) and Plunkett et al. (1995). The structure of this compound can be distinguished from deoxycytidine by the presence of two fluorine molecules a t the 2' position of the sugar ring (Figure 2.1). The metabolism and cellular pharmacology of dFdC is similar to ARA-C (Figure 2.2). dFdC passes across the cell membrane by facilitated nucleoside transport (Plagemann et al., 1978a) and is activated through phosphorylation by deoxycytidine kinase and other kinases (Guchelaar et al., 1996; Heinemann et al.. 1988). DNA synthesis inhibition arises due to the incorporation of dFdC into DNA. An interesting occurrence with dFdC, termed "masked chain termination", is that after incorporation of this analog into DNA, DNA polymerase adds one additional deoxynucleotide. Consequently, since dFdC triphosphate occupies a non-terminal spot in the DNA, proof reading exonucleases cannot excise it, resulting in a block of both DNA repair and DNA replication (Huang et al., 1991; Noble and Goa, 1997; Plunkett et al. 1995).

dFdC diphosphate exerts an inhibitory effect on ribonucleotide reductase, the enzyme that converts ribonucleotides to deoxynucleotides. As such, the size of the competing dCTP pool is reduced by dFdC thus potentiating its antineoplastic effect (Heinemann et al., 1990; Plunkett et al., 1995). The lethality of dFdC is preferential to cells in the S phase of the cell cycle (Rockwell and Grindey, 1992).

In clinical trials, dFdC therapy has exhibited promising activity against many solid tumor types. These studies have shown favorable response rates for the treatment of the following cancers: breast (Carmichael and Walling, 1997), pancreatic (Burris et al., 1997; Casper et al., 1994), bladder (Stadler et al., 1997), ovarian (Kaufmann and Von Minckwitz, 1997; Lund et al., 1994). lung (Abratt et al., 1994; Cormier et al., 1994; Fossella et a]., 1997), and squamous cell carcinoma of the head and neck (Catimel et al., 1994). dFdC has received approval for the clinical treatment of cancer in the United States and in Europe (Von Hoff

et al., 1996). The main dose-limiting toxicity engendered by this anticancer agent is myelosuppression (Fossela et al., 1997).

The cytosine nucleoside analog 5-aza-2'-deoxycytidine (5-AZA-CdR) was first synthesized in 1964 (Plirnl and Sorm, 1964). The chemical modification of deoxycytidine involves the placement of a nitrogen atom at the 5 position of the pyrimidine ring in place of a carbon (Figure 2.1). This drug shows some chemical instability (Lm et al. 1981). The metabolism of 5-AZA-CdR is similar to that of ARA-C (Figure 2.2). For review see Momparler (1985).

This analog enters cells by a facilitated nucleoside transport mechanism (Plagemann et al., 1978b) and is activated through phosphorylation by deoxycytidine kinase (Momparler et al., 1984). The triphosphate form of this analog, 5-AZA-dCTP, is a good substrate for DNA polymerase and is readily incorporated into replicating DNA (Momparler, 1985). The presence of 5-AZA-CdR at specific positions in DNA produces a block in the methylation of cytosine residues in DNA (Jones and Taylor, 1980). This arises because of its presence at CpG methylation sites, 5- AZA-CdR does not act a s a methyl group acceptor resulting in an inactivation of DNA cytosine methyltransferase (Creusot et al., 1982; Santi et al., 1983). The consequences of DNA hypomethylation are the transcriptional activation of genes, silent due to methylation of CpG islands (Jones, 1996; Razin and Cedar, 1991), and the promotion of cellular differentiation (Jones and Taylor, 1980; Razin and Riggs, 1980). 5:AZA-CdR is S-phase specific since it exerts its cqrtotoxic effect primarily on S-phase cells (Momparler et al., 1984). This analog was shown to induce the differentiation of leukemic cells (Momparler et al., 1985a). Recently, Bender et al. (1998) demonstrated that 5-AZA-CdR had a growth inhibitory effect on several human tumor cell lines, possibly due to the reactivation, by demethylation, of growth regulatory genes.

In clinical studies, 5-AZA-CdR displayed favorable activity against childhood and adult leukemia (Momparler et al., 1985b; Richel et al., 1991; Rivard et al., 1981) and myelodysplastic syndrome (Zagonel et al., 1993). In a recent clinical trial by Momparler and collaborators (1997, evaluating the safety and efficacy of 5-AZA-CdR against metastatic lung

cancer, promising results were revealed. Notably, one patient, the one that received the most 5-AZA-Ca, survived over 6 years post-therapy. Responses achieved in this trial were comparable to those seen in patients with non-small cell lung cancer treated with dFdC (Abratt et al., 1994).

Considerable interest in 5-AZA-CdR was aroused by investigators reporting the activation, by this experimental agent, of tumor suppressor genes. Such genes shown to have their expression activated due to demethylation by 5-AZA-CdR include VHL in renal carcinoma (Herman et al., 1994), p16 in lung tumors (Merlo et al., 1995; Otterson et al., 1995) gliomas (Costello et al., 1996) and bladder cancer (Gonzalgo et al., 1998). retinoic acid receptor beta in colon cancer (Cdtc? & Momparler, 1995; 1997, and mammary-derived growth inhibitor in breast cancer (Huynh et al, 1996). In addition, 5-,424-CdR can activate the expression of the tumor cell invasion and metastasis suppressor gene, E-cadherin, in primary tumors of breast and prostate (Graff et al., 1995). The main side effect of 5-AZA-CdR that interdicts dose increments is hematopoietic toxicity. predominantly neutropenia (Momparler et al, 1997).

Deoxycytidine Cytosine Arabinoside

5 - Ara - 2' - deoxycytidine 2' - 2' - Diiuorodeoxycytidine

Figure 2.1 Chemical structure of deoxycytidine and related analogs.

ARA-CTP - DNA

ARA - CDP

ARA - CMP - ARA - UMP

T dCMP deaminase

Deoxycytidine kinase

ARA. - C -------.--t ARA - U Cytidine deaminase

Figure 2.2 Metabolism of cytosine arabinoside (ARA-C), a cytosine nucleoside analog.

CHAPTER 3: Hematovoietic toxicity produced by anti-cancer drugs

Hematopoietic toxicity is a major and common complication of cancer chemotherapy (Bodensteiner and Doolittfe. 1993). The

'

myelosuppression produced by most anti-cancer drugs is dose-limiting and prevents dose escalation to improve clinical efficacy. These chemotherapeutic drugs can suppress, in a dose-dependent manner and by equal magnitude, neutrophil, platelet and red blood cell precursors. The rate a t which a certain cytopenia is produced is related to the turnover of the blood cells. The life span of neutrophiis, platelets, and red blood cells is approximately 10 hours, 10 days, and 120 days, respectively. Anemia is usually moderate and not a major problem. Neutropenia is the most serious as compared to the thrombocytopenia due to the high risk of infections produced by the fonncr went. A s more cycles of treament with cytosine nucleoside analogs arc administered, the reserve of stem cells is progressively exhausted and consequently more severe cytopenias appear. Myelotoxicity is also influenced by patient characteristics such a s age, nutritional health, bone marrow function, and kidney and liver function.

Several types of chemotherapeutic drugs can produce pronounced granulocytopenia which can lead to severe bacterial, viral and fungal infections (Bodensteiner and Doolittle, 1993). An inverse correlation exists between the risk of serious infection and the granulocyte count. When the absolute blood neutrophii cell count falls below 500 per MI, the risk of severe infections is very high. In this latter situation, antibiotics arc usually used a s preventive or therapeutic agents for neutropenia-caused infections.

Hematopoietic growth factors are increasingly used for the purpose of accelerating hematopoietic recovery to prevent infections (Bodensteiner and Doolittle. 1993). G-CSF was reported to be effective in decreasing the neutropenia and the consequent infections provoked by chemotherapy (Crawford et al., 1991). However, Trillet-Lenoir et al. (1993) reported that, although this growth factor reduced the occurrence of serious neutropenia, it was unable to permit a considerable drug-dose intensification. A rapid

neutrophil recovery has also been demonstrated in studies employing GM- CSF (Linkesch et al., 1990; Michon et al., 1990). However, in some clinical studies, GM-CSF did not reduce the myelosuppressive effects of the chemotherapy (Schwartz et al.. 1996; Stone et al., 1995). In some cases, the co-administration of phasespecific drugs with G-CSF or GM-CSF increased the bone marrow suppression (Meropol et al., 1992; Shaffer et al., 1993), indicating that the scheduling of these agents is very important The infusion of both peripheral blood progenitor and bone marrow cells also ameliorated the recovery from neutropenia and thrombocytopenia after high-dose chemotherapy (Gianni et al, 1989; Peters et al., 1993).

S phase-specific drugs, such as cytosine nucleoside analogs, produce a reduction of the neutrophii count, with a return to normal values, which varies in duration depending on the analog used. ARA-C. administered a s a continuous infusion at the conventional dose for 5 days produced a leukopenia starting from about day 14 to 21 (Chabner, 1996). For dFdC, administered as i.v. infusions (30-60 min), a granulocytopenia occurred with a nadir at day 15 on average (Fossella et al., 1997). Treatment with an 8 hour infusion of 5-AZA-CdR reduced the white blood cell count from about day 15 to 30 (Figure 3.1) (Momparler et al., 1997).

HEMATOPOIETIC TOXICITY 5 - aza - 2' - deoxycytidine

Time (days)

660 mg/m2 8h iv infusion day 0

Figure 3.1 Example of the hcrnatopoictic toxicity in a patient with metastatic lung cancer following treatment with an 8 hour infusion of 5-aza-2'-deoxycytidine (5-AZA-CdRj at a total dose of 660 mg/mz (Momparler et al., 1997).

All antineoplastic drugs that produce leukopenia can also cause thrombocytopenia whic" is usually a less serious complication (Bodensteiner and Dmlittle, 1993). If platelet numbers drop below 20 000 per pl, transfusion of ate lets becomes strongly recommended, especially if the patient shows some signs of hemorrhage. Interleukins, because of their abiity to stimulate megakaryocytes, have been used to conserve an adequate platelet count post-chemotherapy. Therapy with cytosine nucleoside analogs usually does not produce hemorrhagic complications due to the reduction in the platelet count.

3.3 Anemia

Potentially dangerous anemia is an uncommon side effect of chemotherapeutic compounds due to the longer lie-span of red blood cells (Bodensteiner and Doolittle, 1993). Transfusion of red blood cells is rarely necessary in the treatment of patients with non hematological malignancies and is indicated when the hemoglobin falls under 80 g per Iitre or when bleeding manifestations occur. Recombinant human erythropoietin was shown to be potentially beneficial in improving the anemia arising from the malignancy (Ludwig et al., 1990, 1994) or from the chemotherapy and can eliminate the need for red blood cell transfusion. Cytosine nucleoside analogs usually do not produce anemia that requires additional treatment.

CHAPTER 4: Chemoprotection against the hematopoietic toxicity produced by anti-cancer drugs, using gene therapy

4.1 Rationale

The toxic effects of antineoplastic agents are not restricted to cancer cells, but also encompass normal cells such as bone marrow cells. When this toxicity is too severe, it may require reducing the dose, decreasing the treatment frequency or even discontinuing the chemotherapy (Rafferty et al., 1996). Any reduction in dose intensity can diminish the likelihood of cure through insufficient tumor destruction. One approach that has been used to overcome this problem is to administer high dose treatments to increase tumor kill followed by autologous transplantation of normal marrow cells to accelerate the recovery of hemopoiesis thus escaping the lethality of this aggressive drug treatment (Deisseroth and m m o , 1997). Since many cycles of therapy may have to be used to completely eradicate the malignant tumor, the autologous transplantation procedure will have to be performed many times. The repetitive use of this transplantation procedure is hindered by many technical, ethical and medical problems.

An innovative solution for guarding against the toxic side effects of chemotherapy would be to use gene therapy to render normal cells resistant to antineoplastic drugs hence permitting safe treatment intensiiication in the absence of myelosuppression, and most signifkantly, elevated capability of achieving a curative response (Bertino, 1990). One advantage of using chemoprotection is that only a single transplantation of gene-modified cells from ar. autologous bone marrow would be suflicient to confer protection against drug-induced hematopoietic toxicity for many subsequent cycles of drug treatment.

This goal of enhancing the tolerance of the hematopoietic system to the toxic effects of speciiic anticancer agents may be accomplished by using a vector for gene transfer and expression of drug resistance genes in normal hernatopoietic cells (Bertino, 1990). Investigations on this approach of chemoprotection include the genes for the multidrug resistance (MDR), mutant dihydrofoIate reductase (DHFR), and 0 6 -

methylguanine-DNA-methyltransferase (MGMT).

4.2 Vectors for gene transfer

Various kinds of vectors are used for gene transfer into target cells. Although there is considerable research on expression vectors, an ideal vector does not yet exist Each method of gene delivery has advantages and disadvantages a s summarized in Tabie 4.1. For our experimental work, we have chosen to use a retroviral vector for gene transfer. In the section below, the basic biology of retroviral vectors will be summarized followed by a short commentary on the advantages and disadvantages of this gene transfer system.

Table 4.1 Characteristics of gene delivery systems I

Vector Advantages Disadvantages

Retrovirus High transduction frequency Unstable ~nfects hematopietic cells and epithelial cells ~hromosomal integration

Adenovirus Infects epithelial cells at a high frequency Cellular proliferation not reqAred

virus Adeno-associated Integrates into nondividig

cells at a low frequency

Herpes simplex virus type I

'Naked" DNA

Liposome

Infects a wide range of cell types Very high titers Relativeiy prolonged expression of foreign genes

No viruses involved Easy to use and develop

No viruses involved

Low titer Infects only dividing cells 9- 12 kb limit Potential mutagen

Does not infect marrow Immunogenic May function transiently

Small capacity for DNA (Skb) Low titers

No integration into genome of infected cells Neurotropism Difficult to develop due to complexicity

Ineficient gene transfer Transient expression

Low frequency of transfection Cytotoldc to certain cells

kb~kilobase (Modified from Friedman. 1997; Hanania et al., 1995; Weichselbaurn and Kufe, 1997)

4.2.1 Replication cycle of retroviruses and production of viral particles using packaging cells

A main objective in gene therapy protocols is the ef!icient and safe delivery of therapeutic genes to target cells (Friedrnann, 1997; Runnebaum, 1997). As possible vehicles, viruses have received the greatest consideration since they can infect target cells in which they can

express their genes. The majority of clinical trials in gene therapy utilize retroviruses. The retroviruses were investigated as vectors in the beginning of the 1980s. The first human gene therapy study took place in 1989 and involved retroviral gene transfer of the neomycin gene into autologous tumor-infitrating lymphocytes in patients with advanced melanoma (Rosenberg et al., 1990).

The retroviral genome consists of two identical strands of RNA that are bound to viral proteins in a nucleoprotein structure that constitutes the core of the virus (Mulligan, 1991; Uckert and Walther, 1994). This nucleoprotein structure is enclosed by a membrane that contains cellular proteins and, on its outside surface, viral envelope glycoproteins. Retroviruses enter cells by the interaction of their specific envelope proteins with cellular membrane proteins, acting as receptors, on target cells. The virus particle is internalized following either fusion of viral and cellular membranes or receptor-mediated endocytosis. The core containing the viral genome and viral enzymes is discharged into the cellular cytoplasm. The retroviral RNA is then copied into double stranded DNA by the reverse transcriptase and primer tRNA that exist in the viral nucleoprotein structure (Boris-Lawrie and Temin, 1994; Mulligan, 1991; Uckert and Walther, 1994). Linear duplex DNA penetrates into the nucleus, is circularized and integrated into the chromosomal DNA of the host cell utilizing the viraJly encoded integrase.

The viral genome integrated into a host cell chromosome is referred to a s a provirus. This proviral DNA can act as a template for the transcription of the full-length RNA. This RNA will be utilized a s the RNA

genome and also for the generation of RNAs coding for the dierent proteins required to construct the retroviral particle or virion (Mulligan, 1991). Following assembly of the virion, it is expulsed from the ceil by budding of the cellular membrane.

Generally, a wild-type provirus comprises three structural genes, spxXcally the gag (group-spedfic antigen), pol (polymerase) and env (envelope) genes which code for the viral con proteins, reverse transcriptase and envelope, respectively (Uckert and Walther. 1994). The composition of the provirus also indudes two long terminal repeats (LTRs), one 5' and one 3: which are sequences implicated in the regulation and expression of the proviral DNA. Often, the gag and pol proteins are translated from the full-length genomic mRNA, and the envelope proteins from the shorter mRNA created by splicing. The constituents of a provirus that are essential and required are the sequences for encapsidation or packaging, for reverse transcription, and for integration of the retroviral genome (Mulligan, 1991). Therefore. what must be retained for the production of recombinant retroviral vectors are the LTRs, the tRNA provirus binding region, the psi (v) packaging signal, and an area upstream of the 3' LTR (Mulligan, 1991).

To generate retroviral vectors, the structural genes have been removed and replaced by the desired therapeutic gene, rendering the ,virus replication-incompetent (Whartenby et al., 1995; Mulligan, 1991). The recombinant vector is introduced, via transfection, into packaging cells (Figure 4.1). These specialized cells contain a helper virus that supplies the viral proteins required for virion production, but that lacks the w sequence for packaging its own RNA genome. However, these packaging cells can replicate, transcribe and package foreign retroviral RNA that was introduced into its cytoplasm in the plasmid DNA form. The packaging cell line that now produces replication-defective retroviral particles containing the recombinant vector DNA is referred to a s a producer cell line. V i particles released from producer cells into the cell culture media can serve to introduce the therapeutic gene, via transduction, into target cells. Murine retroviruses are termed ecotropic when capable of infecting only rodent cells, and amphotropic when they can infect various kinds of mammalian cells, including mouse and human ce!ls.

I Retroviral Vector I LTR~GENE'H LTR~

4 rransfecrion

Packaging Cell Line Y+ b

I GENE H-1 4 gal 1 pol ( e n v k + transcriprion

YA mRNA \Y + c + viral proteins * encapsidarion of

vecror RNA \ & + secretion of virus

containing gene

Virion

4 infecrion

F + reverse \ Target Cell transcription

Y+ + integration

~~dP~~ I : I~ I~ IS I I I I I I ~ GENE H . . ~(IIIII

+ z;pgr,","ion

RNA + \ Protein b

Figure 4.1 Schema of gene transfer into a packaging cell line for the production of recombinant retroviral virions for the transduction of target cells (modified from Banerjee et al., 1994a).

4.2.2 Advantages and disadvantages of retroviruses as vectors for gene transfer

The advantages of employing retroviral vectors are the following:

A) The genetic architecture of these vectors is relatively simple (Weichselbaum and Kufe. 1997).

B) A gene of interest of up to 9kb can be inserted into the retroviral vector (Sandhu et al., 1997).

C) Gene transfer into target cells occurs efliciently (Whartenby et al., 1995).

D) Stable integration of the foreign gene takes place into the chromosomal DNA of the host cell and subsequently in progeny cells (Richter. 1997; Whartenby et al., 1995).

E) Retroviral vectors may evoke only a minor immune response in the host (Weichselbaum and Kufe, 1997).

The disadvantages and risks associated with the use of retroviral vectors are summarized below:

A) Considering that retroviral particles are generated in cell culture, some cellular contaminants may thus coexist (Whartenby et al., 1995).

B) Retroviral gene transfer occurs only in cells actively dividing, therefore the frequency of transduction of resting hematopoietic stem cells is modest (Richter, 1997; Weichselbaum and Kufe, 1997). To overcome this obstacle, it is possible in some systems to induce the cycling of quiescent cells (Tzeng et al., 1996).

C) Retroviruses are susceptible to inactivation by serum complement (Roth et al., 1997).

D) Certain retroviruses cannot infect certain cell types which may be due to the absence of specific receptors on these target cells (Mulligan. 1993; Roth et al., 1997). Modification of the viral envelope by proteins may allow infection of selective target cells (Friedmann, 1997).

E) Retroviral titers can be low often ranging from 1@ to 106 cfu/ml (Runnebaum, 1997; Tzeng et al., 1996). Very good packaging cell lines produce virions with titers reaching a maximum of 107 cfu/ml. The ability to concentrate viral supernatants by such means a s ultrafiltration or centrifugation, and the use of better packaging cell lines may generate higher viral titers.

F) Large-scale preparations of retroviral vectors may increase the risk of replication-competent virus arising (Roth et al., 1997; Ss~ldhu et al., 1997). The possibility of recombination between the vector and the helper viral genome present in packaging cells has been lowered by the utilization of the packaging cells in which the genome of the helper virus is divided on two separate plasmids (Markowitz et al., 1988a. 1988b). To further reduce the risk of contaminatioc, the supernatants are analyzed for the presence ofhelper viruses (Runnebaum, 1997).

G)The integration of the retrovirally transferrec! gene into the host cell genome is random. There is a potential danger of insertional mutagenesis and consequently malignant transformation occurring if the proviral DNA is inserted into an area where it can cause the activation of an oncogene or the silencing of a tumor suppressor gene (Richter, 1997; Sandhu et al.. 1997). Donahue et al. (1992) reported that lymphoma developed in monkeys due to insertional mutagenesis originating from the infection with retroviruses that were replication- competent. To this date, over 2000 patients have been entered into clinical gene transfer studies since 1989, and none of the risks related to retroviral v3ctcr ;ise have eventualized (Richter, 1997; Sandhu et al.. 1997). As kn=~!edge from the Human Genome Project is obtained, it may lead to a way of achieving integration into specific areas of genornic DNA hence decreasing the potential danger accompanying random gene insertion (Uckert and Walther, 1994).

CHAPTER 5: Dm-r-

5.1 Multidrug-resistance gene

5.1.1 Mechanism

The human multidrug resistance gene-1 (MDRl) codes for the multidmg transporter or P-glycoprotein, an ATP-dependent drug eillux pump located in the plasma membrane (Koq et al., 1996; Roninson, 1992a). A diversity of lipophillic polycyclic anticancer drugs with different chemical structures act a s substrates for MDR1. These compounds include anthracyclines, vinca alkaloids, epipodophyllotoxins, paclitaxel, and actinomycin D. It is postulated that the MDRl transport protein pumps these drugs out of the cytoplasm through the cell membrane using a common channel (Raviv et al, 1990;. lowering intracellular drug concentrations, thus conferring drug resistance. MDRl expression has been detected in various tumors (Fojo et al., 1987), in addition to numerous normal tissues such as bone marrow and peripheral blood cells where expression, although low, is highest in early stem cells (Chaudhary and Roninson, 1991; Chaudhary et al., 1992).

A direct correlation was proposed to exist between the amount of MDRl transporter molecules per cell and the extent of drug resistance (Roninson, 1992b). Competitive inhibitors of P-glycoprotein, which include verapamil and cyclosporin A, may restore drug sensitivity (Kessel, 1986). Many studies have focused on the retroviral transfer of the cDNA for MDRl for the purpose of offering considerable resistance in hematopoietic cells to MDR-responsive drugs (Banerjee et al., 1994; Koq et al., 1996).

5.1.2 In dtro studies in hematopoietic cells

Human KB carcinoma cells transfected or transduced with the MDRl cDNA in a retroviral vector, acquired drug resistance to MDRI- sensitive agents, such as colchicine, doxorubicin and vinblastine (Pastan e t al., 1988; Ueda et al., 1987). Using a clonogenic assay, McLachlin et al. (1990) demonstrated overexpression of MDRl in murine marrow-derived hematopoietic cells following retroviral transfer of the MDRl cDNA, conferring resistance to colchicine and vinblastine. They noted that short-

term drug exposure of cells following transduction increased the percentage of chemoresistant hematopoietic colonies. Also using a retroviral vector containing MDRl cDNA, DelaFlor-Weiss et al. (1992) observed chemoprotection in MELC mouse erythroleukemia cells and showed enhanced mRNA and protein expression of MDRl which was further augmented by step-wise increase in colchicine concentrations.

Gene transfer of MDRl cDNA into human hematopoietic stem cells has also been accomplished, but generally with a lower frequency of transduction than murine marrow progenitor cells (Licht et al., 1997). Several investigators demonstrated retrovirus-mediated transfer of the MDRl gene into human CD34+ enriched progenitor cells from bone marrow, peripheral blood and cord blood, and expression of the MDR phenotype (Bertolini et al., 1994; Boesen et al., 1993; Hanania et al., 1995a; Hegewisch-Becker et al., 1995; Ward et al., 1994; Ward et al., 1996). For instance, Bertolini et al. (1994) transduced human hematopoietic progenitor cells from cord blood and bone marrow, and noted in 17-25% of CD34+ cells, expression of P-glycoprotein, and resistance to MDR drugs in 14-31% of hematopoietic colonies.

Ward et al. (1994) showed in MDRl transduced CD34+ cells purified from human marrow, P-glycoprotein expression which was higher in paclitaxel selected cells compared to non-selected. These investigators later observed efficient MDRl gene transfer and expression in peripheral blood CD34+ cells, similar to that achieved in bone marrow cells, a s well as increased protection from the cytotoxicity of paclitaxel (Ward et al, 1996). Several groups noted that MDRl gene-transduced CD34+ cells from human marrow and blood emitted lower fluorescence when exposed to the fluorescent and non-toxic P-glycoprotein substrate, rhodamine 123, and showed lower sensitivity to paclitaxel, as compared to untransduced cells (Hanania et al., 1995a; Hegewisch-Becker et al., 1995). Fruehauf et al. (1995) also reported transfer and expression of the MDRl gene in CD34+ peripheral blood cells.

Novel retroviral vectors are currently under investigation. For example, the Friend mink cell focus fonning/murine embryonic stem cell virus and myeloproliferative sarcoma virus/murine embryonic stem cell virus hybrid vectors, were utilized for MDRl gene transduction of human bone marrow progenitors. Modified cells revealed high level resistance to paclitaxel suggesting that these novel vectors may provide better

chemopmtection than the commonly used Moloney murine leukemia virus derived vectors (Eckert et al., 1996). A recent study by Rund et al. (1998) showed that an MDRl-conraining SV40 pseudoviral vector had a high gene transfer efficiency and considerable transgene expression in several mouse and human cell lines, including human hematopoietic cells.

5.1.3 In vioo studies in hematopoietic cells

Several investigators have observed in vivo chemoresistance following transplantation of MDR1-transduced mouse marrow into murine recipients (Hanania and Deisseroth, 1994; Hanania et al., 1995b; Podda et al., 1992; Sorrentino et al., 1992). For example, Sorrentino et al. (1992) transplanted mice with murine hematopoietic cells transduced with MDRl retroviral particles. Paclitaxel administration in these mice produced an enhancement of the copy number of MDRl cDNA, as well as a reduction in the chemotherapy induced neutropenia, compared to mice that received the neomycin-resistance gene. These researchers noted MDRl provirus integration in bone marrow, peripheral blood, and thymus cells, and suggested that gene transfer into pluripotent progenitors had occurred.

Podda et al. (1992) employed an identical kind of retroviral vector, but a different packaging cell line, GP+E86 cells, to introduce the MDRl cDNA into bone marrow cells which were then infused into irradiated syngeneic mice. The MDRl transgene persisted for over 8 months in some mice and, in one mouse analyzed, significantly augmented MDRl expression was noted in about 14% of the granulocytes in a marrow sample. In vivo selection by drug treatment was also shown. One year post-transplantation, paclitaxel administration provoked the expansion of MDR1-containing peripheral blood c e k Successful transduction of stem cells was postulated from these data. These two studies showed that the increased expression of the MDRl gene offered the modified bone marrow cells a selective advantage (Podda et al., 1992; Sorrentino et al., 1992).

One group reported similar in vivo fmdings of transfer of the MDR phenotype using Harvey murine sarcoma virus LTR and Moloney murine leukemia virus LTR-containing retroviruses. They transferred the MDRl gene into murine bone marrow cells and conducted serial transplantations into 6 successive cohorts of mice (Hanania and Deisseroth, 1994). All these mice showed protection from the Ieukopenia produced by paclitaxel

administration, signifying the successful transduction of immature stem cells. In a later study, they serially transplanted m w cells transduced with the MDRl cDNA into 6 successive cohorts of mice that were treated with high doses of paclitaxel after each transplantation (Hanania et al.. 1995b). Chemoresistant mice were generated, in comparison to the control mice, suggesting that genetically altered marrow cells can eficiently repopulate bone marrow. Expression of the MDRl gene was maintained for more than 12 months from the first marrow transplantation, indicating that in vivo selection produced prolonged transgene expression.

In general, marrow stem cells utilized for MDRl gene transfer are collected from mice previously treated with 5-fluorouracil to induce the entry of stem cells into the cell cycle (KOC et al., 1996). Bodine et al. (1994) obtained with the MDRl gene a higher transduction of murine peripheral blood pluripotent stem cells when donor mice were treated with G-CSF and stem cell factor, a s compared to marrow stem cells harvested from mice treated only with 5-FU. The MDRl proviral DNA was detected in the peripheral blood cells 4 months after their transplantation in mice. Richardson and Bank (1995) inserted the MDRl gene into murine marrow cells and showed that preselection of MDR1-modifled cells by flow cytometry produced, after transplantation in mice, higher and longer expression of the MDRl transgene as compared to non-selected cells.

Licht and collaborators (1995a) showed that ex vivo drug selection of transduced bone marrow cells prior to transplantation lead to enhanced MDRl expression and drug resistance in mice. In another study, Licht et al. (1995b) transplanted hematopoiedc stem cells transduced with the MDRl gene into SCID mice and noted successful engraftment of these cells into different lineages, and expression of this transgene in bone marrow. Following transplantation into a second generation of mice, the presence of the MDRl proviral DNA was detected in the hematopoietic cells.

Recently, Schwanenberger et al. (1996) performed retrovirus- mediated transfer of the MDRl gene into MO-7e cells, a growth factor- dependent human CD34+ hematopoietic progenitor cell line, conferring an increased expression of MDRl and a chemoresistance phenotype to these cells. These transduced human cells were transplanted into immunodeficient non-obese diabetic (NOD) SCID mice which subsequently revealed decreased sensitivity to paclitaxel. Mickisch and Shroeder (1994)

reported the use of a Moloney relmviral construct for inserting the MDRl gene into CD34+ enriched cells derived from the bone marrow of rhcsus monkeys. This led to enhanced and long-term MDRl gene expression after transplantation into recipient monkeys.

Aksentijevich et al. (1996) showed that mice reconstituted with MDRl gene-transduced bone marrow cells were better protected from the leukocytopenia caused by bisantrene, an experimental anticancer drug. than control mice. Selection for MDR1-modified cells in vivo was also shown to occur with bisantrene treatment. Scaradavou et al. (1997) transferred the human MDRl gene into murine fetal peripheral blood cells which were then transplanted into sublethally irradiated syngeneic adult mice. Some recipient mice showed long-term engraftment, greater than 8 months, of the MDRl transgene-containing cells.

5.1.4 Clinical studies on chernoprotection with the MDRl gene

The preclinical investigations described above provided the foundation for phase I clinical trials on chemoprotection in patients with cancer. Hanania et al. (1996) utikzed two methods for MDRl transduction of hematopoietic cells. These investigators reported that the MDRl transgene was not detected in any of the 10 patients transplanted with cells transduced by Lie 'suspension" protocol in contrast to the 'stromal growth factor transduction" which yielded 518 positive patients for the MDRl proviral DNA. Various clinical studies have commenced on MDRl gene transfer to hematopoietic cells in patients with advanced breast cancer, ovarian cancer. and brain tumors (Deisseroth et al.. 1994, 1996; HesdorfTer et al., 1994; OShaughnessy et al., 1994, 1996). The objective of these clinical trials is to investigate if MDR1-transduced normal hematopoietic cells are more resistant to the toxicity of chemotherapy to hence allow more intensive drug treatment following autologous transplantation of these cells (KOC et al., 1996; Licht et al., 1997). These clinical studies will reveal if this chemoprotection strategy is possible, safe, and efficient. Moreover, some patients transplanted with MDRl gene- modixied cells will be administered paclitaxel to determine if enrichment of these cells occurs by selection.

Results from the phase I clinical trial revealed the safety of the MDRl gene transfer procedure that was aimed to reduce the myelotoxicity

of intensive drug therapy. In this study, cancer patients were transplanted postchemotherapy with autologous marrow cells which had undergone MDRl gene transfer (Hesdorffer et al.. 1994, 1998). The safety of the technique was demonstrated. The MDRi gene was detected in the bone marrow in 2 of the 5 patients transplanted and for only a maximum of 10

' weeks (Hesdorffer et al., 1998). The short duration of detection of the MDRl transgene may be explained by non-transduced cells outgrowing the gene-modified cells after transplantation. Briefly, the gene transfer procedure consisted of harvesting bone marrow and/or peripheral blood progenitor cells, transducing these target cells with MDRl retroviral particles and lastly transplanting these modified cells back into the patients.

Recently, Devereux et al. (1998) reported results from a pilot investigation of MDRl gene transfer in lymphoma patients. The CD34+ cells purified from the peripheral blood of these patients were remvirally transduced with the MDRl gene and then reinfused back after these patients were submitted to marrow ablation with antineoplastic drugs. When peripheral blood and marrow were assayed following transplantation, the MDRl proviral DNA was not detected by PCR, indicating a limited in vivo survival of the transgene.

5.2 Dihvdrofolate rednctase gene

5.2.1 Biochemical role in cells

Dihydrofolate reductase (DHFR) catalyzes the reduction of dihydrofolate to tetrahydrofolate (Schimke et al., 1978; Werkheiser, 1961). Tetrahydrofolate is a cofactor that is essential for the biosynthesis of glycine, purine nucleotides and thymidylate, important precursors of proteins and nucleic acids. Methotrexate (MTX), a folate analog, exerts its cytotoxic action by inhibition of the enzyme activity of DHFR The antineoplastic potentia! of MTX is limited by the development of tumor drug-resistance (Bertino et al., 1963) and its bone marrow toxicity (Bertino, 1979). In uitro studies on rodent cells have revealed that drug resistance to MTX may occur by an enhancement of DHFR enzyme activity (Alt et al., 1978; Littlefield, 1969), decreased transport of MTX into the cell (Sirotnack et al., 1968) and structural changes in DHFR (Flintoff et al., 1976).

Mutant variants of the DHFR gene can arise spontaneously or by site-directed mutagenesis. Resistant cells have been characterized and shown to contain modifications in the DHFR structure, leading to a diminished affmity for MTX (Dicker et al., 1990; Flintoff et al., 1976; McIvor and Simonsen, 1990; Srimatkandada et al., 1989). Mutation in codon 22 of DHFR resulting in a change of Leu to Arg confers very high chemoresistance to MTX. The codon 31 mutation of Phe to Ser also produces MTX resistance and a more efficient enzyme for folate metabolism (Moms and McIvor, 1994). Additional DHFR mutants in codon 22 are being investigated, such as a replacement of Leu by Tyr, to augment resistance to MTX and still retain good catalytic activity (KOC et al., 1996). Certain DHFR mutants have been utilized a s dominant selectable markers in in uitro studies on cells (Hussain et al., 1992; McIvor and Simonsen, 1990; Simonsen and Levinson, 1983).

5.2.2 In vitro studies in hematopoietic cells

Several investigators have performed studies on the retrovirus- mediated transfer of the murine Leu 22 to Arg 22 modified DHFR gene into primary hematopoietic cells of mouse, dog and human origin, and

demonstrated its expression in these cell types (Banejn et al., 1994b; Hock and Miller, 1986; K w et al., 1996; Kwok et al., 1986). Hock and Mier (1986) reported in vitro resistance to MTX in human bone marrow progenitor cells following retroviral transduction with the modified DHFR gene. In a study by Kwok et al. (1986). hematopoietic progenitor cells from canine bone marrow were transduced by retroviral particles carrying the mutant DHFR gene. Subsequent colony assays showed that transfer of this DHFR gene conferred the drug resistance phenotype in the canine progenitor cells.

Banerjee et al. (1994b) utilized a human Phe 31 to Ser 31 modified DHFR gene to retrovirally transdun murine bone marrow progenitors and obtained a level of MTX resistance comparable to what they observed with the murine Leu 22 to Arg 22 modified DHFR. These investigators also reported in vitro MTX resistance conferred following transfection of mouse marrow progenitor cells with the cDNA for a murine Trpl5 mutant DHFR (Bane jee et al., 1994~).

Flasshove and colleagues (l995a, 1995b) transduced CD34+ purified hematopoietic cells, derived from human peripheral and umbilical cord blood, with the Phe 31 to Ser 31 modified DHFR gene. The altered hematopoietic cells manifested the MTX resistance phenotype. This group very recently reported the successful transduction of human umbilical cord blood CD34+ cells with the cDNA for the Ser 31 DHFR (Flasshove et al., 1998). They noted augmented drug resistance to MTX in transduced CFU-GM colonies.

Another form of modifled DHFR cDNA, the Leu 22 to Tyr 22 human variant, was demonstrated to confer in vitro protection, in murine marrow precursors, from the cytotoxicity produced by MTX and the novel antifolate, trimetrurate (TMTX) (Spencer et al., 1996). Braun et al. (1997) compared the efficacy of Arg 22 or Tyr 22 mutant DHFR genes for conferring resistance to MTX and TMTX in K562 human hematopoietic cells. They proposed that the Tyr 22 modified DHFR may be more useful clinically since it offers resistance to a wider range of concentrations of folate analogs. In a recent investigation, Patel et al. (1997) compared the antifolate chemoresistance produced by the wild type human DHFR gene and 20 of its mutant variants. Following transduction of CEM human lymphoblastoid cells, the highest level of resistance to TMTX was produced by the Phe 31 to Arg 31 single amino acid \ u i u l t of DHFR, and for its

double amino acid variants, the Leu 22 to Tyr 22 plus Phe 31 to Arg 31 and the Leu 22 to Qr 22 plus Phe 31 to Ser 31 showed the most activity.

5.2.3 In vim studies in hematopoietic cells

Several studies have established that transfer and expression of the mutated DHFR gene in murine hematopoietic cells confers in uivo protection from the myelosuppression produced by MTX treatment in mice. In early work, Cline et al. (1980), transplanted mice with hematopoietic cells transfected with DNA containing the Leu 22 to Arg 22 modified DHFR gene. These recipient mice were treated with MTX which selected for transduced marrow cells expressing the DHFR transgene and the drug resistance phenotype. Williams et al. (1987) demonstrated that mice transplanted with bone marrow cells transduced with a Moloney retroviral vector containing the Arg 22 mutant DHFR cDNA were resistant to doses of MTX which were lethal to control mice.

A low level of gene expression was noted in canine recipients of autologous marrow transduced with the Arg 22 variant form of the DHFR gene (Schuening et al., 1988; Stead et al., 1988). In one dog, following MTX in uivo selection, investigators noted 0.1% MTX-resistant hernatopoietic colonies at 3 weeks after transplantation and 0.03% after 5 weeks. Vinh and McIvor (1993) performed retrovirus-mediated gene transfer of Arg 22 modified DHFR into mouse bone marrow cells which subsequently served to reconstitute irradiated syngeneic mice. They observed DHFR enzyme activity solely in recipients treated with MTX, even though DHFR proviral DNA was detected in all transplanted mice, implying that In vivo selection of the transgene-expressing cells had occurred.

Cony et al. (1990) showed chemoprotection from MTX in mice following serial transplantation of marrow cells containing the Arg 22 modified DHFR gene, indicating successful transduction of hematopoietic stem cells. The primary and secondary transplant recipient mice displayed resistance to MTX and increased survival time versus control mice, due to elevated expression of mutant DHFR. Zhao et al. (1994) used a retroviral vector containing the cDNA for a Leu 22 to Arg 22 modified DHFR to transduce murine marrow cells, and performed serial transplantations into three generations of mice. Primary, secondary and tertiary recipient mice were conferred chemoprotection since they were shown to have

consequently acquired the abiity to survive lethal treatment with MTX, which control mice did not Therefore, these investigators also demonstrated the efficient gene transfer and expression of the mutant DHFR in the hematopietic stem cells

Li et al. (1994) demonstrated that the human Ser 31 mutant DHFR '

can also confer chemoprotection from MTX in vivo. They showed that retrovirus-mediated gene transfer of human Ser 31 modified DHFR into murine hematopietic cells resuked in significantly decreased leukocytopenia and deaths in mice transplanted with these cells and administered elevated doses of MTX.

Blau et al. (1996) proposed that MTX and TMTX are cytotoxic to relatively mature non-clonogenic progenitors and spare clonogenic precursor cells, suggesting that folate analogs are unsuited for in uivo selection of transduced early progenitors of blood cells. In order to investigate chemoresistance to TMTX, Spencer et al. (1996) studied various mutant forms of the human DHFR gene and observed the highest resistance with a Tyr 22 variant in murine hematopoietic cells. Mice transplanted with marrow cells transduced with retroviral virions containing the Tyr 22 DHFR cDNA displayed protection from TMTX induced myelosuppression.

In a recent investigation by Zhao et al. (1997), mice with mammary adenocarcinoma were administered a lethal dose of cyclophosphamide and then transplanted with Ser 31 DHFR-transduced hematopoietic cells. Total tumor regression was observed after intense dose MTX therapy in 44% of the mice reconstituted with marrow cells containing the DHFR transgene. Control mice which were not transplanted with DHFR gene- modified marrow did not survive high dose MTX.

5.3 Methy-e-DNA-methvltransferase gene


The 06-methylpanine-DNA-methyltransferase (MGhlT) is a DNA repair enzyme that protects mammalian cells from the DNA damage produced by certain antineoplastic drugs (Baum et al., 1996; Kw et al.. 1996; Pegg, 1990). The 06-alkylating agents, such as the nitrosoureas and related methylating compounds, cause DNA damage which can have toxic, mutagenic, transforming and carcinogenic consequences. Use of these chemotherapeutic agents in patients comprises a risk of secondary cancer development due to their mutagenic activity (Boffetta and Kaldor. 1994). MGMT repairs the 06-alkylguanine DNA adducts produced by these anticancer compounds (Bogdcn et al., 1981; Pegg, 1990). In this enzymatic reaction, the methyl group at the 0 6 position of guanine in DNA is transferred to the cysteine residue in the active site of MGMT, resulting in the irreversible inactivation of the enzyme. It is interesting to note that myelotoxicity is the major dose-limiting side-effect of 06-alkylating drugs (Baum et al., 1996; Rafferty et al., 1996). The overexpression of a DNA repair system may impart protection in normal tissues such as bone marrow, from the toxic effects of these chemotherapeutic agents.

Several researchers have investigated the possibility of MGMT gene transfer for rendering primary hematopoietic cells resistant to the toxicity of nitrosoureas permitting safe intensification of therapy (Rafferty et al., 1996). In humans, MGMT activity was reported to be high in liver and low in bone marrow cells (Gerson et al.. 1985).

06-benzylpanine is a potent competitive inhibitor of MGMT e w e activity (Dolan et al., 1990; Pegg et al., 1993). Mutant forms of MGMT have been generated that are insensitive to inhibition by this compound (Crone and Pegg, 1993; Crone et al., 1994). Employing the mutated MGMT would allow the inhibition of its repair enzyme activity in tumor cells and simultaneously enhance the protection of normal cells from the toxic action of allcylating agents, leading to a higher therapeutic effectiveness of these drugs (Baum et al.. 1996).

5.3.2 In vitro studies in hematopoietic cells

M a y e t al. (1995) demonstrated efficient transduction of human K562 and murine primary hematopoietic cells, utilizing a myeloproliferative sarcoma virus derived retroviral vector containing the human MGMT cDNA. The gene-modified K562 cells showed considerable MGMT expression a s well as signscant 1,3-bis (2-ch1oroethyI)-1- nitrosourea (BCNU) resistance, a s compared to non-transduced cells. Primary murine bone marrow cells transduced with this vector also displayed the BCNU resistance phenotype. Moritz et al. (1995), with a retroviral vector comprising the MGMT gene, obtained in transduced murine hematopoietic cells, a n increase in MGMT expression and a greater resistance to nitrosourea-caused toxicity, compared to control cells.

Jelinek et al. (1996) introduced the human MGMT cDNA by retroviral transduction into mouse hematopoietic cells, and subsequently noted higher transgene expression in cells from the long-term bone marrow cultures. The CFU-GM colony-forming cells from MGMT- transduced marrow cell cultures showed three-fold higher resistance to N- methyl-N-nitrosourea than the mock-transduced cultures. Wang et al. (1996) exposed a murine multi-potent hematopoietic stem cell l i e to retroviral particles containing the human MGMT cDNA. The transduced cells exhibited elevated expression of MGMT activity a s well a s greater tolerance to the toxicity of several alkylating agents. This chemoprotection was found to be reversed by the MGMT inhibitor 06-benzylguanine.

Reese et al. (1996) transduced K562 cells and normal human CD34+ cells with a n MFG vector containing a gene for a modified human MGMT enzyme which is insensitive to inhibition by 06-benylguanine. Transduced cells showed augmented MGMT expression a s well a s resistance to BCNU combined with 06-bcnzylguanine. Over 30% of the gene-modified CD34+ cells displayed the chemoresistance phenotype. Likewise, Hickson et al. (1998) used a retroviral vector containing the cDNA for an 06-benylguanine-resistant form of the human MGMT, to transduce human CD34+ hematopoietic cells. By clonogenic assays, these gene-modified cells exhibited in vitro resistance to 06-alkylating drugs in the presence of 0-benzylguanine.

5.3.3 In vim studies in hematopoietic cells

Maze et al. (1994) performed retrovirus-mediated gene transfer of human MGMT into murine bone marrow cells. Mice transplanted with these cells showed MGMT overexpression and protection h m the myelosuppression of alkylating agent BCNU. These investigators later reported higher survival a t 5 weeks post-transplantation in mice which received MGMT-transduced marrow stem cells and treated with several doses of BCNU (Maze e t al. 1996). MGMT overexpression protected mice from the pancytopenia and lethal dose of the drug. In addition, bone marrow cells from these mice showed strong in uitro resistance to BCNU.

Long-term transgene expression was reported by Allay et al. (1995) for MGMT-modified marrow cells transplanted into lethally irradiated mice. A signscant increase in MGMT expression that lasted over 23 weeks since transplantation, was detected in various tissues. Moreover, bone marrow cells harvested from high MGMT-expressing transplant recipients, manifested a decreased sensitivity to BCNU cytotoxicity. In a more recent investigation, Allay et al. (1997) demonstrated in uivo selection of MGMT- expressing murine marrow cells by repeated BCNU treatments. Several weeks post-transplantation, mice given BCNU showed a n increase in MGMT provirus-positive bone marrow progenitors and a n augmentation in the ICso of BCNU. These investigators also noted that the MGMT enzyme activity was much higher in myeloid cells of BCNU-treated transplanted mice than in reconstituted mice not administered BCNU. Moritz et al. (1995) demonstrated hematopoietic chemoprotection in mice treated with BCNU and retransplantated every two weeks with MGMT-transduced bone marrow cells. Over two months after the fvst treatment, colony assays of the marrow cells from some recipient mice indicated elevated drug resistance to BCNU.

Using a mutant form of MGMT that is resistant to inhibition by 0 6 -

benzylguanine, Davis e t al. (1997) transferred this gene, using the MFG ntroviral vector, into murine hematopoietic progenitors. These transduced marrow cells were transplanted into lethally irradiated mice. Simultaneous treatments with BCNU and 06-benzylguanine led to in uiuo resistance to the myelosuppression produced by BCNU. About 3 months after transplantation, 30% of bone marrow cells showed MGMT expression and this value increased to 60% following one treatment with the drug

combination BCNU and 06-benzylguanine. In these experiments, the proportion of CFU-C hematopoietic colonies positive for the MGMT transgene increased from 67% to 100%. These researchers concluded that drug selection took place in this study. Most of the MGMT recipient mice were resistant to a lethal dose of BCNU in combiition with 0 6 -

' benzylguanine. Chinnasamy et al. (1998) used a retroviral vector containing the cDNA for an 06-benzylguanine-insensitive human MGMT to transduce murine marrow cells which were subsequently transplanted into mice. The hematopoietic cells of the recipient mice showed chemoresistance to the toxic effects of the alkylating agent, temowlomide, administered concomitantly with 06-benzylguanine.

5.4.1 Biochemical role in ceiIs

Glutathione-S-transferase (GST) enzymes exist a s several isoforms containing two monomers (Schecter and AlaouiJamali. 1992). The main role of GSTs is to catalyze the conjugation of the thiol glutathione to a diversity of xenobiotics and a s such the detoxification of possibly harmful molecules. These proteins are encoded by various gene families such a s alpha, mu and pi. In rats, genes in the alpha family are referred to as Ya and Yc, in the mu family they are referred to a s Ybl, Yb2 and Yb3, and in the single gene family pi, it is referred to as Yp. Mice, rats and humans have the same GST gene families and with similar homologies (Coles and Ketterer, 1990). Different cytotoxic agents can enhance the expression of these enzymes (Schecter and AlaouiJamali, 1992). Several studies have shown a correlation between enhanced GST activity and resistance to the cytotoxicity of alkylating drugs (Buller et at, 1987; Lewis et al.. 1988; Robson et al, 1987). Myelosuppression constitutes an important side- effect of alkylating agents.

5.4.2 In vitru studies in cells

Litourneau et al. (1996) demonstrated that GST gene transfer provided protection against alkylating agent-induced myelosuppression. They successfully transduced human K562 myeloid leukemia cells and mouse primary marrow progenitor cells with a Moloney-based retroviral vector containing the cDNA for the Yc isofom of rat GST. GST-Yc modified cells showed lower sensitivity to the alkylating drugs mechlorethamine and chlorambucil, than control cells that were transduced with the antisense construct The same team of investigators had previously shown that transduction of NIH 3T3 fibroblast cells with the rat GST Yc isoform resulted in drug resistance to alkylating agents (Greenbaum et al., 1994).

5.5 Aldehyde dehvdrovenase vene

5.5.1 Biochemid role in cells

The aldehyde dehydrogenase (ALDH) enzyme catalyzes the conversion of antineoplastic agents containing the oxazaphosphorine moiety, such as cyclophosphamide and maphosphamide, to the non-toxic metabolites (Koq et al., 1996). Overexpression of the class 1 ALDH isozyrne provided cyclophosphamide drug resistance in mouse leukemic cells L121O/CPA (Radii et al., 1991). It was also reported that the human class 3 ALDH isozyme, highly expressed in a human breast cancer cell line, appeared to be responsible for the resistance to the cytotoxicity of oxazaphosphorine agents (Sreerama et al., 1993). The highest expression of ALDH in different types of blood cells was determined to be in the human CD34+ cells (Kastan et al., 1990), but not high enough to provide chemoprotection. Gene transfer of ALDH to normal hematopoietic cells may provide higher expression of the ALDH gene to confer clinical resistance to cyclophosphamide, which would reduce the myelosuppression produced by this anticancer drug and permit more intensive drug treatment (Koc et al., 1996).

5.5.2 In vitro studies in c e b

Bunting et al. (1994) noted that MCF7 breast tumor cells were rendered resistant to oxazaphosphorine drugs following transfection with the rat class 3 ALDH gene. These researchers later used several retroviral constructs to transfer class 1 ALDH (ALDH-1) cDNA into different cell lines (Bunting et al., 1997). They observed that only one clone of transfected cells expressed ALDH-1 enzyme activity. In transduced cells, they detected minimal or absent ALDH-1 mRNA and protein expression, respectively. These investigators postulated that these results were due to the instability of the vector-derived ALDH mRNAs.

Magni et al. (1996) showed that retroviral vectors containing the cDNA for human ALDH-I conferred increased transgene expression and resistance to cyclophosphamide in L1210 murine and U937 human hematopoietic cells. In addition, human peripheral blood progenitor cells that were transduced with ALDH-1 also exhibited in vitro

cyclophosphamide drug resistance (Magni et al., 1996). Using a retroviral vector, Moreb et al. (1996) transfected the human ALDH-1 cDNA into K562 human leukemic cells and showed that these gene-modified cells were protected against 4-hydroperoxycyclophosphamide. the active form of cyclophosphamide. More recently, this same group of investigators retroviral-transduced K562 leukemic cells with the human ALDH-1 gene and observed increased ALDH-1 expression as well as enhanced resistance to 4-hydroperoxycyclophosphamide (Moreb et al., 1998). Higher levels of ALDH-1 expression were observed after in vitro drug exposure which caused the selection of transduced leukemic cells that contained a greater number of copies of this transgene integrated into the genome.

5.6 Multidrug resistance-associated ~rotein gene


The multidrug resistance-assodated protein (MRP) is a membrane protein which functions as a drug efflux pump and uses glutathione a s a cofactor (D'Hondt et al., 1997; Koq et al., 1996). MRP, whose function is similar to MDR1, is another gene that has the potential to confer multiple drug resistance. However, the mechanism of action of MRP differs from that of MDRl (Barrand et al., 1994; Cole et al., 1994). MRP may provide resistance to many MDR-responsive drugs, such a s vinca alkaloids, colchicine and the anthracyclines doxorubicin and daunorubicin, but not to all MDR-drugs. In addition, drugs utilized for inhibiting MDRl P- glycoprotein activity, such as cyclospo~e A and verapamil, are not able to reverse MRP-induced drug resistance. Therefore, MRP gene transfer into normal hematopoietic cells could be a way to overcome the problem of drug resistance in tumor cells due to high expression of MDR-1. For example, inhibitors of MDR-1 activity would increase the sensitivity of tumor cells expressing the endogenous MDRl phenotype, to MDR drugs. The normal blood cells would escape the toxicity of these drugs if they are conferred drug resistance by the MRP gene (D'Hondt et al., 1997).

5.6.2 In vitro studies in cells

Studies have shown that human cells may be rendered multidrug- resistant by transfection with the cDNA for MRP (Cole et al., 1994; Grant et al., 1994; Zarnan et al., 1994). D'Hondt et al. (1997) utilized retroviral vectors containing the cDNA for MRP to transduce murine fibroblast cells. These gene-modified cells exhibited increased MRP expression and decreased sensitivity to the toxic effects of doxorubicin, vincristine or etoposide. In addition, MRP-transduced cells were rendered more drug resistant following in uitro drug selection with the antineoplastic agent doxorubicin.

5.7 -dine deaminase gene

5.7.1 Description

cytidine deaminase (CD) converts, by irreversible hydrolytic ' deamination, cytidine and deoxycytidine to uridine and deoxyuridine,

respective!^ (Carniener and Smith, 1965) (Figure 5.1). The principal tissue involved in deamination is the liver since it was found to contain the highest CD activity (Camiener and Smith. 1965; Ho, 1973). High levels of enzyme activity are also present in human spleen, lung, kidney, intestinal mucosa (Ho, 1973), placenta (Cacciamani et al., 1991; Lalibertt and Momparler, 1994; Vita et al., 1991) and mature granulocytes (Chabner et al, 1974). It was shown that CD levels rise in direct correlation with the differentiation status of normal and leukemic granulocytes (Chabner et al. 1974; Coleman et al, 1975). Nygaard and Sundstrtim (1987) reported the presence of low CD activity in immature human hematopoietic cells. Shrtider et al. (1996) also noted that human CD34+ peripheral blood progenitors contain low CD activity. In accordance with these observations is that the induction of differentiation of the human leukemic cell line HL-60 by 5-AZA-CdR produced a significant augmentation in CD activity (Momparler and Lalibertt , 1990).

The CD, purified from human placenta by Lalibertt and Momparler (1994), was estimated to have a molecular weight of 48.7 kDa and contain several identical subunits of about 16 kDa. These findings are similar to those obtained following purification of this enzyme from human normal and leukemic granulocytes (Chabner et al., 1974), leukemic myeloblasts (Cheng et al., 1983), spleen (Vita et al., 1989), and placenta (Cacciamani et al., 1991; Vita et al.. 1991).

A cDNA clone for CD was isolated from a human liver cDNA library, sequenced and revealed to consist of 910 base pairs (Laliberti and Momparler. 1994). It consisted of a 5' nontranslated region, a 438 bp coding region for a 146 amino acid polypeptide and a 3' nontranslated region with a polyadenylated tail. Following ligation into a bacterial expression vector, this cDNA expressed a functional protein of 16.3kDa (Lalibertt and Momparler, 1994). An incomplete cDNA sequence for CD was obtained from human U937 leukemic cells by Kiihn et al. (1993). The sequence of the open reading frame and deduced amino acid composition

showed 99% homology with that reported by Laliberte and Momparler (1994).

Vincenzetti et al. (1996) isolated the CD cDNA from human peripheral blood polymorphonuclear leukocytes and discovered a nucleotide sequence equivalent to that earlier stated by Laliberte and

'

Momparler (1994). The enzyme was shown to be a tetramer containing 1

zinc atom per subunit (Vincenzetti et al., 1996, 1997). The zinc in the active site of the CD may function in the catalytic reaction by activating a water molecule crucial for the hydrolytic attack on carbon 4 of the substrate's pyrimidine ring (Carter, 1995; Vincenzetti et al., 1996. 1997). Recently, Gran et al. (1998) isolated from a blood cDNA library the cDNA for human CD. The nucleotide sequence they obtained from the coding region corresponded to that earlier reported by Laliberte and Momparler (1994). These investigators showed that CD exerts a growth inhibitory effect on granulocyte-macrophage colony-forming cells (GM-CFC). One speculated mechanism may be through the depletion of cytidiie and deoxycytidiie stores needed for DNA replication.

In addition to the natural substrates, cytidine and deoxycytidine, CD also catalyzes the deamination of cytosine nucleoside analogs includiig the antineoplastic agents ARA-C, dFdC and 5-AZA-CdR, causing a loss of pharmacologic activity (Bouffard et al., 1993; Camiener and Smith, 1965; Chabot et al., 1983). The affiity of the enzyme is greater for the natural substrates than for the related cytosine nucleoside analogs (Chabner et al., 1974; Chabot et al., 1983; Vita et al., 1989). CD may be involved in drug resistance to cytosine nucleoside analogs since elevated enzyme activity in leukemic cells was reported for some patients a t relapse after chemotherapy with these agents (Onetto et al., 1987; Steuart and Burke, 1971).

The high CD levels present in human liver explains the short half-life of these analogs (Ho, 1973). The chromosomal locus of the human CD gene was determined to be 1p35-36.2 (Saccone et al., 1994). Teng et al. (1975) reported a genetic polymorphism for CD in human granulocytes. In a pharmacokinetic study on ARA-C, Kreis et al. (1992) concluded that the patients could be classified into two different phenotypes, "slow" (70%) and "fast" (30%) dearninators. Kirch et al. (1998) reported the existence of two

natural variants of the gene for human CD that differ in codon 27. They observed that ARA-C is deaminated more rapidly by the natural variant of

CD caxrying at codon 27 a Iysine as compared to the enzyme carrying a glycine. These investigators concluded that this genetic polymorphism is implicated in the different phenotypes of ARA-C deamination observed in vivo.

deaminase HO HO

5 - aza - 2' - dmxycytidine active drug inactive metabolite

deaminase HO F HO F

2',2' - difluorodeoxycytidine active drug inactive metabolite

Figure5.1 Conversion of 5-aza-2'-deoxyctidine and 2'.2'- diffluorodeoxycytidiie to inactive metabolites by enzymatic deamination with cytidiie deaminase.

To aid overcome the problem of drug resistance and increase the therapeutic response, the use of CD inhibitors in combination with cytosine nucleoside analogs has been proposed (Camiener and Smith, 1965; Laliberti and Momparler, 1992). Another advantage in utilizing CD inhibitors is to ensure that leukemic cells present in tissues expressing high CD activity, such as the liver, will not escape the cytotoxic action of cytosine nucleoside analog drugs due to rapid inactivation by deamination (Ho et al., 1980). The human liver was reported to contain sufficient CD activity to deaninate 2 g of ARA-C per hour (Ho, 1973). The chemical structures of diierent CD inhibitors are presented in Figure 5.2.

The fust potent inhibitor of CD to be investigated was 3,4,5,6- tetrahydrouridine (THU) (Camiener, 1968). In one study, the loss of CD activity in human leukocytes was directly related to the concentration of THU (Chabner et al., 1974). This inhibition produced by THU was identified as competitive and reversible (Chabner et al., 1974; Wentworth and Wolfenden, 1975). THU was proposed to function a s a transition-state analog (Wentworth and Wolfenden, 1975). Chou et al. (1977) demonstrated that human leukemic cells containing high CD activity, when treated with ARA-C together with THU, showed an elevated formation of the pharmacologically active metabolite, ARA-CTP. This result suggests that THU can enhance the therapeutic potential of ARA-C.

Other CD inhibitors have been synthesized (Kim et al., 1986; Liu et al., 1981; Marquez et al., 1980; McCormack et al., 1980). Some of these inhibitors demonstrated higher potency than THU (Marquez et al., 1980; McCormack et al., 1980). Laliberte et al. (1992) examined and compared the effect of different inhibitors on the deamination of ARA-C and 5-AZA- CdR by human placental CD. They found that all were competitive inhibitors with the strongest being diazepinone riboside, followed by two equally potent agents, THU and 5-fluorozebularine, and lastly zebularine. Previously, Liu et al. (1981) obtained similar results showing that diazepinone riboside produced the most potent inhibition. McConnack et al. (1980) also reported that the inhibitory action of 5-fluorozebularine was similar to that of THU and greater than that exerted by zebularine. However, zebularine and 5-fluorozebularine as single agents showed some

antineoplastic activity in vitro and in animal models (Driscoll et al., 1991; McConnack et al.. 1980).

Kreis et al. (1988) noted in patients with solid tumors, a considerable rise in plasma ARA-C levels subsequent to the administration of ARA-C in combination with THU. In addition, these investigators later

' observed that when leukemic patients were given these two agents the dose of ARA-C required to obtain adequate plasma concentrations of this analog was reduced (Kreis et al., 1991).

3.4. 5, 6 - Tetrahydrouridine Zebularine

5 - Fluorozebularine Diazepinone Riboslde

F m 5.2 Chemical structure of specific inhibitors of cytidine deaminase.

45

CHAPTER 6: Obiectives of the present investigation

The goals of this study were the following:

--Construct the expression vector pMFG-CD by inserting the cDNA for human CD into the MFG retroviral vector derived from the Moloncy murine leukemia virus.

-Transfect murine ecotropic packaging cells with pMFG-CD and isolate clones of these cells that produce virions containing the CD transgene.

-Transduce with these retroviral particles mouse fibroblast cells and isolate clones by selection with the cytosine nucleoside analog, ARA-C.

-Determine if CD-transduced fibroblasts express higher levels of CD enzyme activity and CD proviral mRNA.

--Observe if retroviral gene +-sfer of CD into fibroblast cells confers drug resistance to the cytosine nucleoside analogs, ARA-C, dFdC, and 5-AZA- CdR.

-Establish if the drug resistance phenotype and increased CD enzyme activity can be reversed by the competitive inhibitor of CD, THU.

-Transduce murine primary bone marrow cells with CD retroviral particles in order to determine if they thus acquire drug resistance to

, ARA-C.

-Perform ex vivo transduction of murine marrow cells with CD virions followed by transplantation into recipient mice and subsequent treatment with ARA-C.

-Determine in recipient mice if the CD proviral DNA can be detected and expressed in different tissues long after transplantation.

PART TWO

ARTICLES

CHAPTER 7: Article 1

Resistance to cytosine arabinoside by retrovirally mediated gene transfer of human cytidine deaminase into murine fibroblast and hematopoietic cells

Richard L. Momparler,l Nicoletta Elioponlos,~ Veronica Bovenzi,l Sylvain L4toarnean,a Mona Greenbaum,a and Denis Coarnoyera IDCpartement de pharmacologie, Uriiversite de Montreal, Centre de recherche pidiatrique, HBpital Ste Justine, and =Departments of Medicine and Oncology, Montreal General Hospital, Montreal, Quebec, Canada.

Pnblished in Cancer Gene Therapy 3(5): 331-338, 1996.

Running title: Momparler et al. CR deaminase gene transfer confers ARA-C resistance.

Footnotes: Address of correspondence and reprint requests to: Dr. Richard L. Momparler, Centre de recherche pidiatrique, Hdpital Ste Justine, 3175 Cdte Ste-Catherine, Montreal, Quebec H3T 1C5, Canada.

Abstract

Dose-limiting hematopoietic toxicity produced by the cytosine nudeoside analogue cytosine arabiioside (ARA-C) is one of the major factors that limit its use in the treatment of neoplastic diseases. An interesting approach to

' overcome this problem would be to insert a gene for drug resistance to ARA-C in normal hematopoietic cells to protect them from drug toxicity. The deamination of ARA-C by cytidie deaminase results in a loss of its antineoplastic activity. The objective of this study was to determine if gene transfer of human cytidiie deaminase into murine fibroblast and hematopoietic cells would confer drug resistance to ARA-C. Retrovirally mediated transfer of the human cytidine dcaminase gene into 3T3 fibroblasts resulted in efficient expression of the proviral RNA for this gene and in increased cytidine deaminase activity in cytoplasmic extracts. These cells showed marked resistance to 4RA-C as determined by the effects of this drug on colony formation, r . : Z p.owth and DNA synthesis. The transfer of the human cytidiie dearnkase gene into murine bone marrow cells by the retroviral vector conferred a high level of drug resistance to ARA-C in clonogenic assays. These studies indicate that the cytidine deaminase gene could be used in cancer gene therapy by protecting normal hematopoietic cells against the cytotoxic effects of ARA- C and related cytosine nucleoside analogues.

Key words: cytidine deaminase, cytosine arabiioside, drug resistance, hematopoietic cells, retroviral vector

Introduction

One of the major problems in cancer chemotherapy is the sensitivity of normal cells, such as hematopoietic cells, to antineoplastic agents, which limits the dose and frequency of drug treatments that can be safely administered to patients. Strategies that would prevent dose-limiting drug toxicities might therefore have an important impact on tumor therapy. An attractive approach to this end may be to transfer a drug resistance gene into normal bone marrow cells and thus protect them from the toxicity of anticancer agents (1).

Replication-defective retroviruses have been efficiently used for transfer and expression of new genetic sequences into hematopoietic stem cells. Transfer of the dihydrofolate reductase gene into murine bone marrow cells has been shown to confer drug resistance to methotrexate (2- 5). Chemoprotection also has been afforded by the insertion of the human multidrug resistance (MDR) gene into murine and human hematopoietic cells (6-9). Likewise, transfer of the Yc isoform of the glutathione-S- transferase gene provided alkylating drug resistance to murine bone marrow cells (10).

Cytidine (CR) deaminase catalyzes the dearnination of cytidine or deoxycytidine to uridine or deoxyuridine, respectively. Furthermore, this enzyme can dearninate cytosine nucleoside analogues such as cytosine arabinoside (ARA-C) ( l l ) , a potent antileukemic agent, yielding inactive uracil derivatives (12). CR deaminase may be implicated in drug resistance since leukemic blasts from some patients with acute leukemia showed high levels of this enzyme at the time of relapse after treatment with ARA-C (13,14). Although cell lines derived from various other tumor types have been shown to be sensitive to ARA-C in uitro, the marked hematotoxicity of this drug has limited its use in nonhematologic malignancies (15). Retrovirally mediated transfer of the CR deaminase gene into normal hematopoietic cells would be expected to improve the patient's tolerance to ARA-C and other cytosine nucleoside analoges, thus improving the clinical usefulness of these drugs.

Human CR deaminase complementary DNA (cDNA) has been cloned and expressed in our laboratory and found to encode a l4dainino-acid protein (16). In a recent report, we used a plasinid expression vector containing the human CR deaminase cDNA promoted by the Moloney

murine leukemia virus (MoMLV) long terminal repeat (LTR) to transfect a murine fibroblast retrovirus-packaging cell line (17). Clones of this cell line were isolated and shown to express increased levels of CR deaminasemessenger (mRNA) and protein. The level of expression of CR deaminase in the clones correlated with the level of drug resistance to ARA-C. In the present study, we used the r e t r o d partides produced by one of these clones to transduce NIH 3T3 mouse fibroblasts. We observed an increased expression of thc human CR deaminase gene a s well a s the transfer of the ARA-C resistance phenotype in these transduced cells. Finally, we used the viral particles from the producer clone to introduce the human CR deaminase cDNA into mouse marrow progenitor cells and found that transduced cells showed in vitro protection from ARA-C toxicity.

IVIaterials and Methods

Cell Culture Techniques Cell lines were grown in Dulbecco's modXed essential medium (Canadian Life Technologies, Burlington, Ontario, Canada) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Wisent Technologies, St Bruno, Quebec, Canada) and 5 pg/ml Gentamicin (Canadian Life Technologies), and incubated at 37°C in 7% C02. GP+E-86 murine ecotropic packaging cells, which are derived from National Institutes of Health (NIH) 3T3 mouse fibroblasts, were obtained from Dr. A. Bank, Columbia University, New York (18).

Clonogenic assays were performed a s follows for fibroblast cells. Aliquots of 100 cells in 1 ml of medium were placed in 12-well Costar tissue culture dishes; 18 to 20 hours later, graded concentrations of ARA- C (Upjohn Canada Ltd) were added, and the cells were maintained in culture for an additional 12 to 14 days. The wells were then stained with 0.5% methylene blue in 50% methanol, and colonies were counted. For cell growth measurements, 10' cells per well were plated in Costar 12-well dishes, and the next day graded concentrations of ARA-C were added. After 5 days of culture, the cells were trypsinized and counted with a Coulter ZM electronic cell counter. For DNA synthesis assays, 105 cells per well were plated in 12-well Costar dishes. After an overnight incubation, ARA-C was added in graded concentrations along with 0.5 pCi of JH-thymidine (20 Ci/mrnol), and the incubation was continued for an additional 6 hours. The amount of radioactivity incorporated into DNA was determined after trypsinition as described previously (16)

Vector Design The expression vector containiig the human CR deaminase cDNA was constructed as described previously. Briefly, the plasmid pBluescript KSII containing the complete cDNA sequence for human cytidine deaminase was used a s the template for a PCR amplifying a 465-bp DNA fragment of the CR deaminase protein coding sequence using a 5'-oligonucleotide primer containing a Nco 1 linker and 3'-oligonucleotide primer containing a Barn HI linker (17).

The plasmid expression vector pMFG-tPA used in this study was obtained from R. Mulligan, Harvard University, Cambridge, Mass. (19).

The CR deaminase insert was cloned between the NwI and Barn HI sites of the vector. pMFG-Lac2 was used as a control vector.

The GP+E-86 ecotropic pack2ging cells were cotransfected with the purified plasmid DNAs pMFG-CD (or pMFG-LacZ) and pSV2-neo using the standard calcium phosphate precipitation method a s described previously (17). Clones of cells resistant to G418 were isokted by cloning. We selected GP+E-86-CD-3 clone for this study bec use it showed very high expression of CR dearninase and signscant drug resistance to ARA-C (17). The estimated viral titer of the GP+E-86-CD-3 packaging cells was about 4 x 10s infective particles/ml a s determined by colony formation by 3T3 cells in the presence of 106 mol/l ARA-C.

Transduction of marine fibroblast cells The medium of semiconfluent GP+E-86-CD-3 was replaced with fresh medium, and 18 hours later this media was removed, centrifuged at 500 X g for 10 minutes, and the supernatant containing the retroviral particles collected. One milliliter of this medium was placed in a 25 cm2 flask containing semiconfluent 3T3 fibroblast cells. Polybrene was added at a concentration of 4 pg/ml. After 4 hours incubation, the medium was replaced with fresh medium. The next day the cells were trypsinized, diluted 10-fold, and plated in dishes containing 104 moll1 ARA-C. After 2 weeks of selection in ARA-C, 3T3-CD3-V clones were isolated by ring cloning.

Transduction of marine bone marrow cells Exponentially growing virus-producing GP+E-86-CD-3 or GP+E-86-LacZ cells were sublethally irradiated (20 Gy, cobalt source), and 3 to 5 X 106 irradiated cells were plated in 100-mm diameter tissue culture dishes 6 to 24 hours before transduction. Seven- to 21-week-old C3H H d female mice (Jackson Laboratories, Bar Harbor, Me) weighing 15 to 30 g were injected intrapentoneally with 150 mg/kg 5-fluorouracil (David Bull Laboratories, Vaudreuil, Quebec, Canada) 48 hours before bone marrow harvest. Fresh marrow was collected from the femurs and tibias, and 3 to 5 X 106 nucleated bone marrow cells wen cocultivated with an equal number of irradiated virus-producing cells for 72 hours. The cocultivation medium for murine bone marrow cells consisted of alpha minimal essential medium (a-MEM) (Gibco, Grand Island, NY) supplemented with

20% FBS, 10 mg/ml bovine serum albumin (BSA) (Boehringer Mannheirn, Dorval. Quebec. Canada), 0.3 mg/ml iron-saturated human transferrin (Boehringer Mannheirn), 0.01% gentamicin, 4 p g / d polybrene, 10% conditioned medium from the murine myelomonocytic WEHI3B cells (20) and 10% coditioned medium from the human HTB-9 primary bladder carcinoma cells (2 1).

Drag sensitivity of clonogenic morw hematopoietic cells. Secsitivity of mouse bone marrow to ARA-C was assessed using an in vitm assay for clonogenic hematopoietic progenitor cells. Immediately after cocultivation with the virus producers, non-adherent bone marrow cells were plated at a final concentration of 1 to 2 X 10s cells/ml in alpha-MEM (with no nudeosides) containing 0.9% methylcellulose (Fisher Scientifc, Montreal, Quebec, Canada), 30% FBS, 1% pokeweed mitogen-stimulated spleen-conditioned medium (GIBCO), 1% BSA, lo4 mol/l 2- mercaptoethanol (Sigma), and 0.01% gentamicin. ARA-C was added to the assay mix in concentrations ranging from 10-8 to 10.5 mol/l. Cultures were plated in triplicate 1-ml aliquots in 35-mm diameter Petri dishes, and placed in 5% COz a t 37'C. The total number of hematopoietic colonies (>SO cells) was scored at days 10 to 14 of culture.

Enzpme Assay The enzymatic assay for CR deaminase was performed a s described previously (16). Briefly, 2 to 5 X 107 monolayer cells were trypsinized, centrifuged, and washed once in phosphate-buffered saline, centrifuged again, and resuspended in 100 ul of 5 mmol/l Tris-HC1, pH 7.4 and 5 rqmol/l dithiothreitol. The cell suspension was then subjected to three cycles of rapid freezing and thawing. The mixture was centrifuged a t maximum speed in a microcentrifuge at 5'C for 15 min. The molarity of the supernatant (cytosol) was then increased to 50 mmol/l Tris-HC1, pH 7.4. Different dilutions of the cytosol were used in 30-minute incubation a t 37'C to measure the conversion of JH-cytidine to 3H-uridine.

Northern Blot Analysis Total RNA was isolated from cells by a method modified from Chomaynski and Sacchi (22), using the Biotecx Laboratories Ultraspec I1 RNA Isolation System kit (Biotecx Laboratories, Houston, Tex). The cells were lysed with

a guanidine solution, and RNA was extracted with chloroform, precipitated with isopropanol, and purified with the RNA Tack resin. Total RNA was eluted from the resin with diethylpyrocarbonate-treated TE (10mmol/l Tris, lmmol/l EDTA, pH8) and stored a t -70°C. For Northern blot analysis, samples of 10 pg of RNA in loading buffer wen heated a t 65'C for 5 minutes and then electrophoresed in 1X borate buffer through a 1.5% agarose-borate-1.1% formaldehyde gel. After ultraviolet (UVJ photography of the gel, the RNA in the gel was blotted onto a Nytran Plus nylon membrane (Schleicher & Schuell, Keene, NH) using the Turbo Blotter device (Schleicher & Schuell) with 20X SSC for neutral downward transfer for 64 hours. The membrane was then baked a t 80°C for 2 hours and irradiated in a Bioslink UV linker with 0.3J/cm? The probe was prepared by labeling the complete cDNA for human CR deaminase by the random prime method using the kit from Boehringer Mannheirn and [llP]dCTP from ICN (Mississauga, Ontario). The membrane was prehybridiid in lOml Express Hyb solution (Clontech) a t 65°C overnight and hybridized a t 6S°C overnight in 5ml of Express Hyb solution containing 6.25 X 106 cpm of the DNA probe. The membrane was then washed once with 2X SSC and 0.1% SDS a t 65'C for 15 minutes, and then twice with 0.2X SSC and 0.1% SDS a t 65'C for 30 minutes. The blot was exposed to X-ray film (Kodak X- Omat) with two intensifying screens at -70°C for 5 days.

Southern blot analysis Genornic DNA was isolated from cells with the Stratagene DNA Extraction kit. About 2 X 107 cells underwent lysis with a mixture of 50mmol/l Tris- HCl, 20mmol/l EDTA, and 2% SDS, followed by digestion with 50 mg/rnl proteinase K. A saturated NaCl solution was then added, followed by centrifugation and RNAse treatment of the supernatant. DNA was precipitated with ethanol and air dried and suspended in TE (10mmol/l Tris. Imrnol/l EDTA, pH 8) buffer. For Southern blot analysis, 10 pg of DNA was digested with NcoI and BamHI and separated by electrophoresis on a 1% agarose gel. After W photography, the gel was immersed in denaturing solution (0.5mol/l NaOH; 0.15mol/l NaCI) for 30 minutes and then in neutralizing buffer (0.5mol/l Tris-HC1 pH7; 1.5mol/l NaCI) for 30 minutes. The DNA in the gel was transferred onto a Hybond-N nylon membrane (Arnersham, Oakville, Ontario] using the Turbo Blotter device with 1OX SSC for downward transfer for 40 hours. The membrane was

then baked a t 80'C for 2 hours and irradiated in a Bioslink UV linker with 0.3J/cm? The membrane was probed with XP-cDNA for human CR deaminase a s described above, except that the hybridization solution was 6X SSC + 2X Denhardt's soluton + 1% SDS. The membrane was exposed to X-ray f h with intensifying screens a t -70°C for 4 8 hours.

Polymerase Chain Reaction W R ) A PCR assay was used to verify the presence of the MFG-CD construct in transduced marrow cells. The oligonucleotides: 5'-GGT GGA CCA TCC TCT AGA CTG-3' (Pl) and 5'-AGC AGC TCC TGG ACC GTC ATG-3' (P2) were used as primers in presence of genornic DNA to amplify a specific 421-bp fragment as predicted by the DNA sequence of the pMFG-CD construct. The sense oligonucleotide P1 was -270 bp downstream from the splice acceptor region of MFG and 2 bp upstream from the start of the env coding region. The antisense oligonucleotide P2 was from positions 377 to 397 of the CR deaminase coding region. Genomic DNA was isolated from individual hematopoietic colonies in methyIcelluIose with the In ViSorb DNA Kit (ID Labs Biotechnology, London, Ontario) by cell lysis with guanidiie thiocyanate, DNA adsorption on silica gel, and elution with TE buffer. For the PCR reaction, -1 ng genomic DNA was denatured a t 95'C for 2 min and amplified for 35 cycles using I.D. Roof Taq DNA polymerase (ID Labs Biotechnology), each cycle consisting of denaturation for 1 minute a t 94°C. annealing for 1 minute a t 5 6 T , and extension for 1 minute a t 72°C with a terminal 5 minute extension a t 72'C. The reaction mixture was separated on 2% agarose electrophoresis.

Enzymatic assays for CR deaminase activity were performed on the nontransduced and transduced 3T3 cells (Table 7.1). CR deaminase activity for the 3T3 cells was low (2.5 units/mg). However, the enzyme activity was 105.1 units/mg for the 3T3-CD3-V5 and 31.2 units/mg for the 3T3-CD3-V6 cells. This represents a 42- and 12.5-fold increase in enzyme activity for the transduced cell lines, respectively, in comparison to the 3T3 control cells.

Northern blot analysis showed the efficient mRNA expression of the transduced CR deaminase gene in both the 3T3-CD3-V5 and 3T3-CD3-V6 cells, but not in the nontransduced parental 3T3 cells (Fig. 7.1). Two transcripts for CR deaminase of approximately 2.8 and 1.9 kb were seen in the 3T3-CD3-V5 and 3T3-CD3-V6 cell lines. These transcripts correspond to the predicted size of unspliced and spliced transcripts of MFG-CD. The higher level of expression of the 1.9-kb spliced transcript in the 3T3-CD3- V5 cells, as compared with the 3T3-CD3-V6 cells, may correlate with the lower enzyme activity in this latter clone.

In the Southern blot analysis, a DNA band of about 450 bp, similar to the size of the coding region of the CR deaminase cDNA, was observed in both the 3T3-CD3-V5 and 3T3-CD3-V6 cells, but not in the 3T3 cells (Fig. 7.2). In the 3T3-CD3-V5 cells, the DNA band was of higher intensity than that in the 3T3-CD3-V6 cells, indicating that the 3T3-CD3-V5 cells may possess more copies of the transduced CR deaminase gene.

The incorporation of radioactive thymidine into DNA was used to evaluate the inhibition of DNA synthesis by different concentrations of ARA-C in the 3T3 cell lines (Fig. 7.3). ARA-C a t concentrations of 104 moll1 and 10-5 moll1 produced much less inhibition of DNA synthesis for the 3T3-CD3-V5 and 3T3-CD3-V6 cells as compared with the 3T3 cells. This difference in sensitivity to ARA-C inhibition was more evident a t the concentration of 104 mol/l.

The effect of ARA-C on the growth of the 3T3 cell lines is shown in Fig. 7.4. For the 3T3-CD3-V5 and 3T3-CD3-V6 cells, ARA-C a t concentrations of 104 moll1 and 10-5 mol/l produced less growth inhibition than for the parental 3T3 cells. At the concentration of 104 mol/l, the decrease in sensitivity of the transduced cells to ARA-C growth inhibition was more evident

A colony assay was used to evaluate the effects of ARA-C on the clonogeniaty of the transduced 3T3 cells in comparison with the parental cells (Table 7.2). ARA-C a t a concentration of 10-6 moll1 reduced the colony formation of the 3T3 cells to less than 1%. At this same concentration, the colony survival for the 3T3-CD3-V5 and 3T3-CD3-V6 cells was 88.6% and 74.8%. respectively. At 10" moll1 ARA-C, only the 3T3-CD3-V5 cells showed residual colony survival (17.1%).

Murine bone marrow cells were transduced by the viral producer cells lines GP+E86-Lac Z or GP+E86-CD3. Using a clonogenic assay, we determined the effects of different concentrations of ARA-C on colony survival (Table 7.3). In the presence of ARA-C from 10-7 mol/l to 10-5 mcl/l the percentage survival was much greater for the marrow cells transduced by the GP+E86-CD3 cells as compared with the GP+E86-Lac Z cells. For marrow cells transduced by GP+E86-Lac 2, treatment with ARA-C at 104 mol/l reduced the survival to less than 3% as compared with untreated cells but did not reduce survival of the marrow cells transduccd by GP+E86-CD3 cells.

In order to verify the presence of the proviral DNA in marrow cells transduced by GP+E86-CD3 cells, individual hematopoietic colonies were isolated and the genomic DNA purified. This DNA was used in a PCR reaction with speciiic primers to detect the presence of the proviral DNA. A DNA band of the predicted size of 421 bp was amplified from eight of 10 colonies, indicating efficient transfer of the proviral DNA (Fig. 7.5). There were no specific MFG-CD bands detectable after PCR using DNA from marrow colonies transduced by the GP+E86-Lac Z cells (data not shown).

Table 7.1 CR deaminase assay in murine cells

Cell line Enzyme activity (units/ md

Values are means + SD; n=3-7. Unit, deamination 1 nmole CR/min.

Table 7.2 Effect of ARA-C on colony formation by murine fibroblasts

Concentration of ARA-C Cell line 10.7 moll1 10-6 moll1 10-smol/l

(% survival)

Values are means * SD; n = 4.

Table 7.3 Effect of ARA-C on clonogenic assay for murine hematopoietic cells

Concentration of ARA-C V i producer 104 mol/l 10-7 mol/l 104 mol/l 10-5 mol/l

Values are means * SEM; n = 5. 'Determined by colony assay for granulocytes-macrophages. iDifference between % survival by paired Student's t test (P < 0.33). *Difference between % survival by paired Student's t test (P < 0.001).

Figure 7.1 Northern blot hybridization of 10 pg total RNA from each of the indicated cell lines. After transfer to a nylon membrane, hybridization was performed with SP-labeled CR dcaminase cDNA probe as described in Methods.

CONCENTRATION OF ARA-C (M)

F i e 7.3 Inhibition of DNA synthesis by ARA-C. The cell lines were incubated with the indicated concentrations of ARA-C and radioactive thymidine for 6 hours. DNA synthesis was measured as described in Methods. The bar and the vertical line represent the mean value and SD, respectively, for five experiments.

CONCENTRATION OF ARA-C (M)

Figure 7.4 Growth inhibition by ARA-C. The cell lines were incubated with the indicated concentrations of ARA-C for 5 days. Cell counts were measured as described in Methods. The bar and the vertical line represent the mean value and SD, respectively, for three to five experiments.

Figure 7.5 PCR analysis for the presence of proviral DNA in murine marrow cells. Genomic DNA was isolated from colonies of murine hematopoietic cells transduced by GP+E86-CD3 cells. A sense primer downstream from the MFG splice acceptor region and an antisense primer from the CR deaminase coding region with purified genomic DNA from the colonies were used in the PCR as described in Methods. Amplification of a specific 421-bp band indicates the presence of MFG-CD proviral DNA Number. DNA from marrow colony assayed; M, molecular size marker. H. It0 used as template.

For most antineoplastic agents, toxicities to the hematopoietic system limit the dose of drug that can be safely administered. In the case of cytosine nucleoside analogues such a s ARA-C, marked myelotoxicity has S i t e d their use to the treatment of primary hematologic malignancies where severe marrow suppression is generally part of effective therapy.

However, primary cultures of human tumors showed that ARA-C is active against lung tumors and melanoma in a colony assay (23). In addition, the combination of ARA-C with cis-platinum showed antitumor synergy that was schedule dependent. If the optimal dose-schedule fcr ARA-C in combination with cis-platinum or other antineoplastic agents could be determined, there could be more use of this nucleoside analogue for the clinical treatment of nonhematologic malignancies, especially if its myelotoxicity could be circumvented.

One approach to circumvent ARA-C myelotoxicity would be to confer chemoresistance on the normal hematopoietic system. The feasibility of that approach has been demonstrated by conferring methotrexate or multidrug resistance in animal models following retrovirus-mediated gene transfer in hematopoietic stem cells (3,4,7). This has lead to the development of clinical trials in which MDR gene transfer will be used to confer hematopoietic chernoprotection to patients undergoing chemotherapy for breast cancer (24).

We have recently cloned the human cDNA for CR deaminase (16). CR deaminase eficiently inactivates intracellu!ar ARA-C by deamination (11,12). In order to evaluate the potential use of CR deaminase for chemoprotection, we constructed the retroviral vector pMFG-CD, and by transfection produced MFG-CD virions in GP+E-86 ecotropic packaging cells. We observed that these MFG-CD virus-producing cells showed markedly increased CR deaminase activity and were resistant to ARA-C

(17). In this study NIH 3T3 mouse fibroblasts were transduced with MFG-

CD virions produced by GP+E-86-CD3 packaging cells and selected in 10-6 mol/l ARA-C. Two of the clones isolated (3T3-CD3-V5 and 3T3-CD3-V6) were found to express increased levels of CR deaminase enzymatic activity (Table 7.1). This overexpression of CR deaminase was accompanied by a significant decrease in the degree of DNA synthesis inhibition produced by

ARA-C in the CD-transduced NIH 3T3 cells; this effect was particularly marked a t 104 mol/l of ARA-C (Fig. 7.3). Wtewise. CR deaminasc gene transfer in NIH 3T3 cells significantly reduced the inhibition of cell growth produced by ARA-C (Fig. 7.4) and confered drug resistance to this analogue a s shown by colony formation (Table 7.2). Again these effects of CR deaminase transduction were especially evident a t 1W moll1 ARA-C. The CD-transduced fibroblasts maintained their drug resistance to ARA-C for more than 6 months without any selecting agent in the culture medium indicating that the CR deaminase expression was stable.

To further evaluate the potential of CR deaminase gene transfer for chemoprotection of hematopoietic cells, we transduced primary mouse bone marrow cells with either MFG-CD or MFG-Lac2 retroviral particles produced by the GP+E-86-CD3 or GP+E-86-Lac2 packaging cells, respectively. PCR analysis of individual hematopoietic colonies grown from the transduced bone marrow cells in the absence of ARA-C showed a high efficiency of MFG-CD gene transfer into clonogenic hematopoietic progenitor cells (Fig. 7.5). We then determined the effect of ARA-C on in vitro colony formation by the populations of clonogenic hematopoietic progenitor cells. At concentrations of ARA-C ranging from 10.7 to 10-5

mol/l, colony formation by the hematopoietic cells transduced with the MFG-CD vector was practically unaffected by this drug, whereas colony formation by the control La&-transduced cells was gradually suppressed to reach less than 1% in presence of 10.5 mol/l ARA-C (Table 7.3). This indicates a very strong in vitro survival advantage in favor of CR deaminase-transduced hematopoietic cells in the presence of ARA-C.

The degree of hematopoietic chemoprotection confered in vivo by CR deaminase transduction will need to be evaluated directly in bone marrow tranplantation experiments in mice. However, the powerful in Yitro survival advantage of CR deaminase-transduced clonogenic hematopoietic progenitor cells suggests that the degree of in Yivo chemoprotection should be sufficient to confer drug resistance to ARA-C. The mouse marrow cells showed high survival at an ARA-C concentration of 10.6 mol/l (Table 7.3) which is in the same range a s the plasma concentration of this analogue reported for patients that received a continuous intravenous infusion of ARA-C a t the conventional dose (25). These observations suggest that gene transfer of CR deaminase into human hematopoietic cells could have the

potential to prevent ARA-C hematotoxicities produced by clinically effective doses of this drug.

This approach of using the CR deaminase gene for chemoprotection also could have application for other interesting deoxycytidine analogues such a s 2',2'-difluorodeoxycytidine (dFdC) and 5-aza-2'-deoxycytidine (5-

'

AZA-CdR), which are also deaminated by this enzyme (26-28). In phase I and I1 studies, dFdC has shown some promising antitumor activity (29.30). We and others have demonstrated that 5-AZA-CdR is an active antileukemic agent (3 1-33)

Recently, considerable interest in 5-AZA-CdR has been generated by the reports that this analogue activates the expression of different tumor suppressor genes by demethylation (34-37). In a preliminary study we observed that gene transfer of CR deaminase into murine fibroblasts conferred drug resistance to dFdC and 5-AZA-CdR (38).

In summary, the present in vitro studies suggest that CR deaminase gene transfer could be used to confer hematopoietic chemoprotection from ARA-C and related cytosine nucleoside analogues. Future experiments will aim at evaluating the degree of in vivo chemoprotection obtained after transplantation of CR deaminase-transduced bone marrow cells into lethally irradiated animals. Animals reconstituted with CR deaminase- transduced hematopoietic cells will also permit us to evaluate the feasibiIity of enrichment of these cells by in uivo selection with ARA-C, as reported for MDR (7).

Acknowledgments This work was supported by a grant from the Cancer Research Society Inc (Montreal). We thank Louise F. Momparler for her technical assistance.

1. Banejee D, Zhao SC, Li M-X, Schweitzer BI, Mineishi S, Bertino JR. Gene therapy utililizing drug resistance genes: A review. Stem Cells 1994; 12: 378-385.

2. Corey CA, DeSilva AD, Holland CA, W i a m s DA. Serial transplantation of methotrexate-resistant bone m w : Protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 1990; 75: 337-343.

3. Williams DA, Hsieh K, DeSiIva A, Mulligan RC. Protection of bone marrow transplant recipients from lethal doses of methotrexate by the generation of methotrexate resistant bone marrow. J. Exp. Med.1987; 166: 210-218.

4. Zhao SC, Li M-X, Banejee D, Schweitzer BI, Mineishi S, Giboa E, Bertino JR. Long-term protection of recipient mice from lethal doses of methotrexate by marrow infected with a double-copy vector retrovirus containing a mutant dihydrofolate reductase. Cancer Gene Ther. 1994; 1: 27-33.

5. Banejee D, Schweitzer BI, Volkenandt M, Li M-X, Waltham M, Mineishi S, Zhao S-C, Bertino JR. Transfection with a cDNA encoding a Ser31 or Ser34 mutant human dhydrofolate reductase into Chinese hamster ovary and mouse marrow progenitor cells confers methotrexate resistance. Gene 1994; 139: 269-274.

6. Sorrentino BP, Brandt SJ, Bodine D, Gottesman M, Pastan I, Cline A. Nienhuis AW. Selection of drug-resistant bone marrow cells in vivo after retroviral transfer of human MDR-1. Science 1992; 257: 99-103.

7. Podda S, Ward M, Himelstein A, Richardson C, de la Flor-Weiss E, Smith L, Gottesman M, Pastan 1, Bank A. Transfer and expression of the human multiple drug resistance gene into live mice. Proc. Natl. Acad. Sci. USA 1992; 89: 9676-9680.

8. Hanania EG, Fu S, Roninson I. Zu Z, Gottesman MM. Deisseroth AB. Resistance to tax01 chemotherapy produced in mouse marrow cells by safety-modified retroviruses containing a human MDR-I transcription unit Gene Ther. 1995: 2: 279-284.

9. Ward M, Richardson C, Pioli P. Smith L, Podda S, Goff S. HesdorEer C. Bank A. Transfer and expression of the human multiple drug resistance gene in human CD34+ cells. Blood 1994; 84:1408-1414.

10.Etourneau S, Greenbaum M. Cournoyer D. Retrovirus-mediated gene transfer of rat glutathione S-transferase Yc confers in vitro resistance to alkylating agents in human leukemia cells and in clonogenic mouse hematopoietic progrenitor cells. Human Gene Ther. 1996; 7: 831-840.

11.Camiener GW, Smith CG. Studies of the enzymatic deamination of cytosine arabinoside. I. Enzyme distribution and species specificity. Biochem. Phannacol. 1965: 14: 1405-1416.

12.Maer WEG, Zahn RK. Metabolism of l-f3-D-arabinofuranosyluracil in mouse L5 l78Y cells. Cancer Res. 1979: 39: 1 102- 1 107.

13.Steua1-t CD, Burke PJ. Cytidine deaminase and the development of resistance to arabinosyl cytosine. Nature New Biol. 1971; 233: 109-1 10.

14.0netto N, Momparler RL, Momparler LF, Gyger M. In Vitm biochemical tests to evaluate the response to therapy of acute leukemia with

, cytosine arabinoside or 5-aza-2'-deoqqtidine. Semin. Oncol. 1987; 14: 23 1-237.

15.Chabner 5. Cytidine Analogs. In: Cancer Chemotherapy: Principals and Practice (BA Chabner & JM Collins, eds.) 1990 (J.B. Lippincott Co., New York) pp. 154-179.

16.LaIiberte J , Momparler RL. Human cytidine deaminase: pudication of enzyme, cloning, and expression of Its complementary DNA. Cancer Res 1994; 54: 5401-5407.

17.Momparler RL, Lalibere J. Eliopoulos N, Beausejour C. Cournoyer D. Transfection of murine fibroblast cells with human cytidine deaminase cDNA confers resistance to cytosine arabinoside. Anti-Cancer Drugs 1996; 7: 266-274.

18.Markowia D, Goff S, Bank A. A safe packaging line for gene transfer: separating viral genes on two diierent plasrnids. J. Virol. 1988; 62: 1120-1124.

19.Radelain M, Wang CH, Antoniou M, Grosveld F, Mulligan RC. Generation of high-titer retroviral vector capable of'expressing high levels of human beta-globin gene. Proc. Natl. Acad. Sci. USA 1995; 92: 6728-6732.

20.Murphy WH, Urnovitz HB, Maryanski JL, Abrams GD. Characterization of transplantable myelomoncytic leukemia WEHI-3B in syngeneic BALB/c mice. Proc. Soc. Exp. Biol. Med. 1978; 157: 556-564.

21.Hoang T, McCullough EA. Production of leukemic blast growth factor by a bladder carcinoma cell line. Blood 1985; 66: 748-751 .

22.Chomcynski P. Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987; 162: 156-159.

23.Kem DH, Morgan CR, Hildebrand-Zanki SU. In vitm

pharmacodynamics of 1-6-D-arabinofuranosylcytosine: synergy of antitumor activity with cis-diamminedichloro-platinum (11). Cancer Res. 1988; 48: 117-121.

24.O'Shaughnessy JA, Cowan KH, Nienhuis AW, McDonagh KT, Sorrentino BP, Dunbar CE, Chiang Y, Wilson W, Goldspiel B, Kohler D, CottIer-Fox M, Leitman S, Gottesman M, Pastan 1, Denicoff A, Noone M, Gress R. Retroviral mediated transfer of the human multidrug resistance gene (MDR-1) into hematopoietic stem cells during

a~tologous transplantation after intensive chemotherapy for metastatic breast cancer. Human Gene Ther. 5(7): 89 1-9 1 1, 1994.

25.Slevin ML, Piall EM, Aheme GW, Johnston A. Lister TA. The pharmacokinetics of cytosine arabiioside in the plasma and cerebrospinal fluid during conventional and high-dose therapy. Med. Pediatric Oncol. 1982; lO(Supp1 1): 157-168.

26.Chabot GG, Bouchard J, Momparler RL. Kinetics of deamination of 5- aza-2'-deoxycytidine and cytosine arabiioside by human liver cytidine deaminase and its inhibition by 3-deazauridie, thymidine or uracil arabinoside. Biochem. Pharmacol. 1983; 32: 1327-1328.

27.Laliberte J , Marquez VE, Momparler RL. Potent inhibitors for the deamination of cytosine arabiioside and 5-aza-2'-deoxycytidine by human cytidine deaminase. Cancer Chemother. Pharmacol. 1992; 30: 7-11.

28.Bouffard DY, Laliberte J , Momparler RL. Kinetic studies on 2',2'- diiuorodeoxycytidine (Gemcitabiie) with puriiied human deoxycytidine kinase and cytidine deaminase. Biochem. Pharmacol. 1993; 45: 1857- 1861.

29.Lund B, Hansen OP, Theilade K, Hansen M, Neijt JP. Phase I1 study of gemcitabiie (2',2'-difluorodeoxycytidine) in previously treated ovarian cancer patients. J. Natl. Cancer Inst. 1994; 86: 1530-1533.

30.Abratt RP, Bezwoda WR, Falkson G, Goedhals L, Hacking D, Rugg TA. Eficacy and safety profde of gemcitabine in non-small-cell lung cancer: A phase I1 study. J. Clin. Oncol. 1994; 12: 1535-1540.

31.Momparler RL, Rivard GE, Gyger M. Clinical trial on 5-aza-2'- deoxycytidine in patients with acute leukemia. Pharmac. Ther. 1985; 30:277-286.

32.Rivard GE, Momparler RL, Demers J. Benoit P, Raymond R, Li KT, Momparler LF. Phase I study on 5-aza-2'-deoxycyiidine in children with acute leukemia. Le'dcemia Res. 1981; 5: 453462.

33.Wiemze R, Richel DJ, Arensten-Honders W, Colly LP. The antileukemic activity of 5-aza-2'-deoxycytidine (Aza-dC) in patients with relapsed and resistant leukemia. Br. J. Cancer 1991; 64: 144-148.

34.Herman JG, Latif F, Weng Y. Lerman MI, Zbar B, Liu S. Samid D, Duan D-S, Gnarra JR, Linehan WM, Baylin SB. Silencing of the VHL tumor- suppressor gene by DNA methylation in renal carcinoma. Proc. Natl. Acad. Sci. USA 1994; 91: 9700-9704.

35.Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, Baylin SB, Sidransky D. 5'CpG island methylation is associated with transcriptional silencing of the tumour suppressor pl6/CDKN2/MTSl in human cancers. Nature Med. 1995; 1: 686-692.

36.0tterson GA, KhIeif SN, Chen W, Coxon AB, Kaye FJ. CDKN2 gene silencing in lung cancer by DNA hypermethylation and kinetics of p16 INK4 protein induction by 5-aza-2'-deoxycytidine. Oncogene 1995; 11: 1211-1216.

3 7.Cdtt S, Momparler RL. Antineoplastic action of all-trans retinoic acid and 5-aza-2'-deoxycytidine on human DLD-I colon carcinoma cells. Cell. Pharmacol. 1995; 2: 221-228.

38.Eliopoulos N, Bovenzi V, Momparler LF, Cournoyer D, Momparler RL. Gene transfer of human cytidine deaminase cDNA into murine cells confers resistance to cytosine arabiioside, 5-aza-2'-deoxycytidine and 2',2'-dinuomdeoxycytidine. Ninth NCI-EORTC Symposium on New Drugs in Cancer Therapy. Ann Oncol. 1996; 7(Suppl 1): 58.

CHAPTER 8: Article 2

Drug resistance to 5-aza-2'-deoxycytidine, 2'3'- difluorodeoxycytidine, and cytosine arabinoside conferred by retroviral-mediated transfer of human cytidine deaminase cDNA into murine cells.

Nicoletta Eliopodos 1, Denis Coarnoyera, and Richard L. Momparler 1 'Dipartement de phannacologie, Universite de Montreal and Centre de recherche pidiatrique, Hdpital Ste-Justine 3175, Cdte Ste-Catherine, Montreal, Quebec H3T 1C5, and ZDepartments of Medicine and Oncology, Montreal General Hospital, Montreal, Quebec H3G 1A4, Canada

Published in Cancer Chemotherapy and Pharmacology 42(5): 373-378, 1998.

Correspondence to: Dr. Richard L. Momparler, Centre de recherche pidiatrique, Hepital Ste-Justine, 3175 Chemin Cdte Ste-Catherine, Montrial, Quibec H3T 1C5, Canada. tel: (514) 345-469 1; fax: (514) 345-4801

Supported by Grant MT-13754 from the Medical Research Council of Canada.

Rupose: The hematopoietic toxicity produced by the cytosine nucleoside analogs is a critical problem that limits their effectiveness in cancer therapy. One strategy to prevent this dose-limiting toxicity would be to insert a gene for drug resistance to these analogs into normal bone marrow cells. Cytidiie (CR) deaminase can deaminate and thus inactivate 5-aza- 2'-deoxycytidie (5-AZA-CdR), 2',2'-difluorodeoxycytidine (dFdC) and cytosine arabinoside (ARA-C). The aim of this study was to determine if gene transfer of CR deaminase into murine fibroblast cells confers drug resistance to these cytosine nucleoside analogs and if this resistance can be prevented by the CR deaminase inhibitor, 3,4.5.6-tetrahydrouridine (THU). Methods: NIH 3T3 murine fibroblast cells were transduced with retroviral particles containing the human CR deaminase cDNA. Assays measuring CR deaminase activity as well as the inhibitory action of 5-AZA- CdR, dFdC and ARA-C on colony formation, were performed in the presence of different concentrations of THU. Results: Retroviral-mediated transfer of the CR deaminase gene into 3T3 fibroblasts produced a considerable increase in CR deaminase activity. The transduced cells also showed significant drug resistance to 5-AZA-CdR, dFdC and ARA-C, as demonstrated by a clonogenic assay. This drug resistance phenotype and elevated CR deaminase activity were reversed by THU. Conclusions: These fmdings indicate that the CR deaminase gene can potentially be

used in cancer gene therapy for protecting normal cells against the cytotoxic actions of different cytosine nucleoside analogs. In addition, the CR deaminase-transduced cells can be used as a model for screening different CR deaminase inhibitors in ar? intact cellular system.

Kcy words. cytidiie deaminase, 5-aza-2'-deoxycytidine, 2',2'- diffluorodeoxycytidiie, cytosine arabimoside, 3,4,5,6' ~ydrouridine

Abbreviations: 5-AZA-CdR, 5-aza-2'-deoxycytidine; dFdC, 2',2'- dfiuorodeoxycytidine; ARA-C, cytosine arabiioside; THU, 3,4,5,6- tetrahydrouridine; CR, cytidine; MDR, multiple drug resistance; 3T3-CD3-

V5, NIH 3T3 cells transduced with MFG-CD virions

Introduction

For many antineoplastic agents the dose-limiting toxicity is myelosuppression. Thus, gene therapy strategies devised to prevent the severe granulocytopenia associated with chemotherapeutic drugs would

'

increase their curative potential 121. There have been several reports on the use of retroviral-mediated gene transfer of different drug resistance genes into normal hematopoietic stem cells for protection against drug- induced toxicity. For instance, transfer of a mutant dihydrofolate reductase gene into marrow precursors provided hematopietic chemoprotection from methotrexate in mice transplanted with these cells [6]. Likewise, drug resistance to alkylating agents in murine hematopietic cells has been realized using retroViral geny: transfer of 06-alkylguanine- DNA alkyltransferase [I] or glutathione S-transferase [14]. Extensive gene transfer studies using the human MDR gene have shown that this gene conferred protection from the toxicity of MDR-responsive drugs in murine hone marrow cells [31, 341. Administration of the cytotoxic drug tax01 to mice transplanted with MDR transduced hematopoietic cells resulted in the selective expansion of MDR expressing cells 1341.

CK deaminase catalyzes the deamination of cytosine nucleosides and their analogs such a s 5-AZA-CdR, dFdC and ARA-C resulting in a loss of their antineoplastic activity 13. 5, 251. CR deaminase may be involved in clinical drug resistance to 5-AZA-CdR or ARA-C in some patients with leukemia since a t the time of relapse after treatment the leukemic cells showed e!evated levels of this enzyme [27, 351. ARA-C is a very effective chemotherapeutic agent against acute myeloid leukemia [ll] whereas dFdC shows gromising clinical antitumor activity 19. 151. In clinical trials, 5-AZA-CdR has demonstrated antileukemic [19, 321 and interesting antitumor activity [24].

The primary toxicity of these cytosine nucleoside analogs, limiting their dose-intensity, is bone marrow suppression. Retroviral gene transfer of CR deaminase into normal hematopoietic cells would be expected to render them resistant to these analogs and consequently allow dose escalation to improve their clinical effectiveness.

We have cloned and expressed the human CR deaminase complementaxy DNA (cDNA) [13] and used a retroviral plasmid expression vector to transfect this gene into ecotropic packaging cells 1221. We

transduced normal mwhe hematopoietic and fibmblast cells and produced the ARA-C resistance phenotype in vib.o 1231. In the present investigation, which c o n f i s our preliminary study 181, we show that the CR deaminase-transduced fibroblast cells are not resistant uniquely to ARA-C, but also demonstrate cross resistance to other cytosine nudeoside analogs such as 5-AZA-CdR and dFdC. In addition, we determine that this drug resistance phenotype and enhanced CR deaminase activity can be reversed by THU, a competitive inhibitor of CR deaminasc [4].

Material and Methods

Ccll lines Cells were grown in monolayers in Dulbecco's modificd essential medium (Canadian Life Tech.ologies. Burlington, Ontario) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Wisent Technologies. S t Bruno, Quebec) and 5 pg/ml gentamich (Canadian Life Technologies), and incubated at 37°C and 7% CO2. Murine ecotropic retrovirus-packaging cells GP+E-86 were obtained from A. Bank (Columbia University, New York) [16].

Construction of retrooiral vector and generation of vinas-producing cell line The plasmid expression vector pMFG-CD containing the human CR deaminase cDNA was constructed as described recently [22]. Briefly, the coding cDNA sequence for human CR dearninase [13] was cloned between the NcoI and BamHI sites of the pMFG retroviral plasmid (R. Mulligan, Hanard University, Cambridge, Mass.) to give the pMFG-CD construct. The plasmid pMFG-CD was transfected with pSV2 neo into GP+E-86 ecotropic packaging cells. Clones of cells resistant to G418 were isolated and further analyzed. The clone GP+E-86-CD3 was chosen for gene transfer studies since it demonstrated very high levels of CR deaminase expression, signif~cant drug resistance to ARA-C, and a good viral titer

WI.

Transduction of mrvine fibroblast cells Supernatant containing MFG-CD virions from GP+E-86-CD3 cells was placed in a flask of 3T3 fibroblast cells with 4 pg/rnl polybrene. Three days following infection, the cells were plated in the presence of 10hM ARA-C and selected for 14 days. Clones of drug resistant cells were isolated by ring cloning and thereafter maintained in medium without ARA-C. The 3T3-CD3-V5 clone showed increased CR deaminase expression and marked drug-resistance to ARA-C [231.

Clonogenic colony assay Aliquots of 100 cells in 2 ml of medium were placed in 6-well Costar tissue culture dishes and 18-20 hours later, 10-7 M 5-AZ.4-CdR (Pharmachemie

B.V., Haarlem. Holland) or 10" M dFdC (Lilly Research Laboratories. IndianqoLis. Indiana) or 10-6 M ARA-C (Upjohn. Canada) were added alone or with different dilutions of THU (Calbiochem, La Jolla, California) for a 50 hour drug exposure. After drug removal and an additional incubation of 10 days, colonies were stained with 0.5% methylene blue in 50% methanol and then counted. The average plating efficiency was 40-60 %.

Enzpme ==Y CR deaminase activity was determined as described previously [21]. Briefly, monolayer cells (2-5 x 107) were trypsinizcd. washed with phosphate-buffered saline and suspended in 5mM Tris-HC1 (pH 7.4) and 5mM dithiothreitol. The cell suspension was freeze-thawed rapidly three times and centrifuged to obtain the cytosolic e.xtract (supernatant). For enzyme assay, dilutions of the cytosol were placed in a reaction mixture with 50mM Tris-HC1 and 0.5 pCi3H-cytidine (ICN Biomedicals, Irvine, California), with or without THU. The mixture was incubated 30 minutes a t 37'C and placed on Whatrnan P-81 phosphoceIIuIose discs. The amount of radioactivity bound to the discs was assessed by scintillation counting. One unit of enzyme activity was defined a s the amount of enzyme that catalyzed the deamination of 1 nmole of cytidine per minute a t 37'C. The protein concentration was measured using the BioRad dye method with bovine serum albumin a s the standard.

The sensitivity of non-transduced (3T3) and CR dearninase- transduced (3T3-CD3-V5) cells to cytosine nucleoside analogs 5-AZA-CdR, dFdC, and ARA-C, was evaluated by donogenic assay. To demonstrate a relationship between the drug .response on 3T3-CD3-V5 cells and the activity of the CR deaminase gene, the CR deaminase inhibitor THU was

also used. As shown in Fig. 8.1. 5-AZA-CdR at a concentration of 10-7 M

produced a substantial reduction in colony formation by the 3T3 cells, but had no considerable cytotoxic effect on the 3T3-CD3-VS cells. The addition of lo4 M THU restored drug sensitivity to 5-AZA-CdR by the transduced cells. In Table 8.1, the effect of different concentrations of THU on loss of clonogenicity by 3T3-CD3-V5 cells in the presence of 10-7 M 5- AZA-CdR is shown. THU at a concentration of 104 M increased the inhibitory action of 5-AZA-CdR on colony formation from 19.5 to 72 %.

dFdC at a concentration of 10.7 M pnctically abolished colony survival by the parental 3T3 cells (Fig. 8.2). However, a t this same concentration of dFdC, colony formation by the CR deaminase-transduced 3T3-CD3-V5 cells was not signscantly altered. The addition of 104 M THU completely restored drug sensitivity to dFdC by the transduced cells. The effect of different concentrations of THU on colony formation by 3T3- CD3-V5 cells in the presence of 103 M dFdC is shown in Table 8.1. THU at concentrations of 10-6 M and 104 M increased dFdC's inhibitory effect from 8.4 to 60.3 and 98.5 %, respectively.

Treatment with 104 M ARA-C suppressed almost entirely colony formation by the control 3T3 cells whereas the 3T3-CD3-V5 cells showed nearly complete drug resistance to this concentration (Fig. 8.3). The addition of 104 M THU completely restored ARA-C sensitivity by these latter cells. In Table 8.1, the effect of different concentrations of THU on colony formation by 3T3-CD3-V5 cells in the presence of 104 M ARA-C is also shown. THU at concentrations of 10-5 and 104 M increased the inhibitory action of ARA-C from 5 to 76.2 and 89.5 %, respectively.

Enzyme assays were performed on the cytosol obtained from the cells transduced with the MFG-CD retrovirus in order to compare the inhibitory activity of THU directly on the enzyme with the results obtained in intact cells. As revealed in Table 8.2, parental 3T3 cells showed low

(c3 units/mg) CR dcaminase activity. In contrast, the enzyme activity of 3T3-CD3-VS cells was 106.8 units/mg which represents an augmentation of over 30-fold in comparison with the 3T3 control cells. When the spedfic CR dearninase inhibitor THU was added to the cytosolic extract of 3T3- CD3-V5 cells, the lwel of enzyme activity declined in correspondence to the concentration of THU used. The CR dcaminase activiv in the transduced cells was inhibited by 52.7% with 104 M THU, 85.3% with 105 M THU, and >97% with 10- M THU. The correlation coefficient between THU inhibition of enzyme activity and its abiity in transduced cells to restore drug sensitivity was >0.96 for dFdC and ARA-C. and >0.87 for 5- AZA-CdR

Table 8.1 Effect of tetrahydrouridine 0 on the antineoplastic action of 5-AZA-CdR, dFdC, and ARA-C on CR deaminax- transduced 3T3-CD3-V5 cells.

Loss of donogenicity (YO) - THU Concentration 5-AZA-CdR dFdC ARA-C (MI (10-7 M) (107 M) (104 M)

Determined by colony assay for a 50 hr drug exposure. Data represent mean values * SD (n=4-10). Loss of clonogenicity in the presence of 104 M THU alone was <5%.

Table 8.2 Effect of different concentrations of THU on CR deaminase activity in murine cells

Cell Line THU CR deaminase Enzyme Concentration activity inhibition (MI (units/mg). (%I

CR deaminase activity was measured in cell extracts of different cell lines in the presence of the indicated concentrations of THU. *units defied as nmoles CR deaminated per min. bmean * SD (n=4-7)

Figure 8.1 Effect of 5-AZA-CdR and THU on colony formation by murine fibroblasts. 3T3 or 3T3-CD3-V5 cells were plated at 100 cells per well. Drug ucposurc was 50 h.

CONTROL 10% dFdC 1WMdFdC IO-'MTIIU

Figure 8.2 Effect of dFdC and THU on colony formation by m u r k fibroblasts. 3T3 or 3T3-CD3-V5 cells were plated at 100 cells per well. Drug exposure was 50 h.

Figure 8.3 Effect of ARA-C and THU on colony formation by murine fibroblasts. 3T3 or 3T3-CD3-V5 cells were plated at 100 cells per well. Drug exposure was 50 h.

The new cytosine nucleoside analogs have considerable promise in cancer therapy. dFdC exerts its cytotoxic effect primarily by inhibiting DNA synthesis [30]. This clinically active cytosine nucleoside analog has demonstrated good response rates in tumor therapy [9, 151. The experimental agent 5-AZA-CdR has the novel action of inhibiting DNA methylation which can thus result in an induction of differentiation of neoplastic cells [lo, 181 and the activation of silent tumor suppressor genes [7, 17,291. In dinicaI studies, 5-AZA-CdR showed favorable activity against several types of hematological malignancies [19, 32, 361. In a pilot study in stage IV metastatic non-small cell lung cancer. 5-AZA-CdR produced some very interesting responses, including one patient still alive > 6 years post-treatment [24].

Current conventional chemotherapy of most advanced metastatic cancers is disappointing with low response rates and short life expectancies. Both dFdC and 5-AZA-CdR have the potential to be very effective antitumor agents if their dose-limiting hematotoxicity [9, 241 could be circumvented. The degradative enzyme CR deaminase can convert by hydrolytic deamination 5-AZA-CdR, dFdC and ARA-C to pharmacologically inactive compounds [3, 5, 251. One approach to overcome the hematopoietic toxicity produced by these cytosine nucleoside analogs would be to introduce the CR deaminase gene into normal marrow precursors to render them resistant to their cytotoxic action.

Several studies utilizing different drug resistance genes have shown that chemoprotection of hematopoietic cells is feasible. Gene transfer of dihydrofolate reductase, glutathione S-transferase or MDR into murine hematopoietic progenitors has been shown to confer drug resistance to methotrexate, alkylating agents or MDR drugs, respectively 16, 14, 31, 34). A clinical trial in breast cancer patients utilizing MDR gene transfer to bestow chemoprotection has been initiated [28].

The human CR deaminase cDNA has been cloned and expressed in our laboratory and has been shown to encode a 146-amino-acid protein of 48.7 kilodaltons [13]. The small size of the CR deaminase cDNA facilitates its genetic manipulation. Cells transduced with CR deaminase showed marked ARA-C resistance in vitm and a large increase in enzyme activity

(231. In a preliminary investigation [8], we observed that transduced

fibroblast cells were also cross-resistant to other cytosine nudecjside analogs (5-AZA-CdR and dFdC). The present study confirmed and extended our initial report. We demonstrated by colony formation assays that the CR deaminase gene transfer into 3T3 cells conferred substantial drug resistance to 5-AZA-CdR and dFdC. To illustrate the cause-effect relation between the overexpression of the CR deaminase gene and the drug resistance phenotype, we utilized the CR deaminase inhibitor, THU. The inhibitory effect on CR deaminase and hence reversal of the drug resistance became increasingly evident as the concentration of THU was augmented.

Colony formation by the CR deaminase-transduced 3T3-CD3-V5 cells was not considerably inhibited by 5-AZA-CdR, whereas this analog produced a pronounced cytotoxic effect on the non-transduced 3T3 cells (Fig. 8.1). When 5-AZA-CdR was added simultaneously with THU, the inhibitory activity of this analog in the 3T3-CD3-V5 cells was restored depending on the concentration of the CR deaminase inhibitor (Table 8.1).

Treatment with dFdC practically did not affect colony formation by 3T3- CD3-V5 cells whereas the control 3T3 cells were very sensitive to this analog (Fig. 8.2). THU enhanced the inhibitory action of dFdC on colony survival by transduced cells (Table 8.1). Likewise, ARA-C did not noticeably alter colony formation by 3T3-CD3-V5 cells, but almost abolished cell survival by parental cells (Fig. 8.3). Colony formation by transduced cells was markedly reduced by ARA-C in the presence of THU (Table 8.1). These studies with THU clearly indicate that the enhanced level of CR deaminase activity in the transduced cells is primarily reponsible for drug resistance to these cytosine nucleoside analogs. It should be noted that drug resistance to cytosine nucleoside analogs can also occur by several other mechanisms [20].

To determine whether the reversal of drug resistance in the transduced cells by THU correlated with its inhibition of CR deaminase, we prepared cytosol extracts from these cells and investigated the effect of this CR deaminase inhibitor directly on the enzyme. The CR deaminase activity measured in 3T3-CD3-V5 cells was over 30-fold higher than that noted in the 3T3 cells (Table 8.2). When increasing concentrations of THU were added to the cytosol extracts from transduced cells, enzyme activity progressively fell, reaching a level equivalent to that in non-transduced cells (Table 8.2). Since in the case of dFdC and ARA-C, lo4 M THU

showed almost complete reversal of drug resistance in the CR deaminase- transduced cells and inhibited deaminase activity in these cells by =+97%. the correhtion between the effects of this enzyme inhibitor on the cells and on the enzyme is very good. THU was the first inhibitor of CR deaminase to be identified 141. Its clinical importance was shown when it was noted that THU in combination with ARA-C significantly increased the plasma lwel of AM-C in patients with solid Nmors 1121.

Neff and Blau [26] have also demonstrated, using a different retroviral vector (LCDSN), that gene transfer of CR dearninase into 3T3 cells confers ARA-C resistance. They employed a growth assay to show a 4.5-fold increased resistance to ARA-C by ID50, compared with cells transduced with a control vector. The enzyme activity measured in transduced fibroblast cells was 7-fold lower than that determined in our present study with clone 3T3-CD3-V5. These investigators also transduced the lymphoid leukemic cell line CCRF-CEM and showed some drug resistaxce to growth inhibitory effects of ARA-C and dFdC. They observed that the resistance to growth inhibition by ARA-C could be rwersed by 2.5 x 104 M THU.

Shr6der et al. [33] transfected the CR deaminase cDNA into murine fibroblast cells using a mammalian expression vector and demonstrated an augmentation in CR deaminase activity and ARA-C resistance of about 3- fold. They determined a s well that CD34+ selected human peripheral blood progenitor cells express a low lcvel of CR deaminase activity.

In conclusion, if our in vitro results that CR deaminase gene transfer can make cells resistant to different cytosine nucleoside analogs can be extended in vivo to normal human hematopoietic stem cells, it may be possible to improve tumor therapy with dFdC and 5-AZA-CdR. The protection of the hematopoietic stem cells by gene therapy will permit the use of more intensive therapy with these analogs to increase their clinical effectiveness against tumor cells without encountering the problem of severe myelosuppression. Furthermore, the CR deaminase-transduced cells can be used as a model for screening d i i r en t CR deaminase inhibitors in an intact cellular system.

Acknowledgments We thank Louise F. Momparler for her valuable technical assistance in this study.

References

1. Allay JA, Dumenco LL. Kw ON, Liu L, Gerson SL (1995) Retroviral transduction and expression of the human alkyltransferase cDNA provides nitrosourea resistance to hematopoietic cells. Blood 85: 3342

2. Bertino JR (1990) W'urning the tables" - Making normal marrow resistant to chemotherapy. J Natl Cancer Inst 82: 1234

3. Bouffard DY, Laliberti J , Momparler RL (1993) Kinetic studies on 2.2'- difluorodeoxycytidine (gemcitabine) with purified human deoxycytidine kinase and cytidine deaminase. Biochem Pharmacol45: 1857

4. Camiener GW (1968) Studies of the enzymatic deai ia t ion of aracytidine: V. Inhibition in vitro and in vivo by tetrahydrouridine and other related pyrimidine nucleosides. Biochem Pharrnacol 17: 1981

5. Chabot GG, Bouchard J, Momparler RZ (1983) Kinetics of deamination of 5-aza-2'-deoxycytidine and cytosine arabinoside by human liver cytidine deaminase and its inhibition by 3-deazauridine, thymidine or uracil arabinoside. Biochem Pharrnacol32: 1327

6. Cony CA, DeSilva AD, Holland CA, Williams DA (1990) Serial transplantation of methowexate-resistant bone marrow: protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 75: 337

7. Cbti S, Momparler RL (1997) Activation of the retinoic acid receptor l3 gene by 5-aza-2'-deoxycytidiie in human DLD-1 colon carcinoma cells. Anti-Cancer Drugs 8: 56.

8. Eliopoulos N, Bovenzi V, Momparler LF, Cournoyer D, Momparler RL (1996) Gene transfer of human cytidine deaminase cDNA into murine cells confers resistance to cytosine arabinoside, 5-aza-2'-deoxycytidine and 2',2'-diiuorodeoxycytidine. 9th NCI-EORTC symposium on new drugs in cancer therapy. Ann Oncol7 (Suppl 1): 58

9. Fossella FV, Lippman SM, Shin DM, Tarassoff P, CalayagJung M, Perez-Soler R, Lee JS, Murphy WK. Glisson B, Rivera E, Hong WK (1997) Maximum-tolerated dose defied for single-agent gemcitabiie: a phase I dose-escalation study in chemotherapy-naive patients with advanced non-small-cell lung cancer. J Clin Oncof 15: 310

10.Jones PA, Taylor SM (1980) Cellular differentiation, cytidine analogs and DNA methylation. Cell 20: 85

11.Keating MJ, McCredie KB, Bodey GP, Smith TL, Gehan EA, Freircich EJ (1982) Improved prospects for long-term survival in adults with acute myelogenous leukemia. JAMA 248: 2481

l2.Kreis W, Chan K, Budman DR. Schulrnan P, Allen S, Weiselberg L, Lichtman S, Henderson V, Freeman J, Deere M, Andreeff M, Viciguena V (1988) Effect of tetrahydrouridine on the clinical pharmacology of 1- beta-D-arabinofuranosylcytosine when both drugs are coinfused over three hours. Cancer Res 48: 1337

13.Laliberte J, Momparler RL (1994) Human cytidine deaminase: P d c a t i o n of enzyme, cloning, and expression of its complementary DNA. Cancer Res 54: 540 1

14.Etourneau S, Greenbaurn M, Cournoyer D (1996) Retrovirus-mediated gene transfer of rat glutathione S-transferase Yc confers in vim resistance to alkylating agents in human leukemia cells and in primary mouse hematopoietic progenitor cells. Hum Gene Ther 7: 831

15.Lund B, Hansen OP, Theilade K, Hanscn M, Neijt J P (1994) Phase I1 study of gemcitabiie (2',2'-difluorodeoxycytidine) in previously treated ovarian cancer patients. J Natl Cancer Inst 86: 1530

16.Markowitz D, Goff S, Bank A (1988) A safe packaging line for gene transfer: separating viral genes on two different plasmids. J Virol 62: 1120

91

17.Merlo A, Herman JG. Mao L, Lee DJ, Gabrielson E. Burger PC. Baylin SB, Sidransky D (1995) SCpG island methylation is associated with transcriptional silencing of the tumor suppressor pl6/CDKN2/MTSl in human cancers. Nature Med 1: 686

18.Momparler RL (1985) Molecular, cellular and animal pharmacology of 5-aza-2'-deoxycytidine. Pharmacol Ther 30: 287

19.Momparler RL, Rivard GE, Gyger M (1985) Clinical trial on 5-aza-2'- deoxycytidine in patients with acute leukemia. Pharmacol Ther 30: 277

20.Momparler RL, Onetto-Pothier N (1989) Drug resistance to cytosine arabinoside. In: Kessel D (ed) Resistance to antineoplastic drugs. CRC Press, Boca Raton, p 353

2l.Momparler RL, Laliberte J (1990) Induction of cytidine deaminase in HL-60 myeloid leukemic cells by 5-aza-2'-deoxycytidine. Leuk Res 14: 75 1

22.Momparler RL, Laliberte J , Eliopoulos N, Beausejour C, Coumoyer D (1996) Transfection of murine fibroblast cells with human cytidine deaminase cDNA confers resistance to cytosine arabinoside. Anti- Cancer Drugs 7: 266

23.Momparler RL, Eliopoulos N, Bovenzi V, Lktoumeau S, Greenbaum M. Cournoyer D (1996) Resistance to cytosine arabinoside by retrovirally mediated gene transfer of human cytidine deaminase into murine fibroblast and hematopoietic cells. Cancer Gene Ther 3: 331

24.Momparler RL, Bouffard DY, Momparler LF, Dionne J. BClanger K, Ayoub J (1997) Pilot phase 1-11 study on 5-aza-2'-deoxy@dine (Decitabine) in patients with metastatic lung cancer. Anti-Cancer Drugs 8: 358

25.MUer WEG, Zahn RK (1979) Metabolism of I-6-D- arabinofuranosyIuracil in mouse W l78Y cells. Cancer Res 39: 1102

26.Neff T. Blau CA (1996) Forced expression of cytidine deaminase confers resistance to cytosine arabiioside and gemcitabiie. Exp Hcrnatol 24: 1340

27.0netto N, Momparler RL, Momparler LF, Gyger M (1987) In uitro

biochemical tests to evaluate the response to therapy of acute leukemia with cytosine arabiioside or 5-aza-2'-deoxycytidine. Semin Oncol 14: 23 1

28.0'Shaughnessy JA, Cowan KH. Nienhuis AW, McDonagh KT, Sorrentino BP. Dunbar CE, Chiang Y, Wilson W. Goldspiel B, Kohler D. Cottler-Fox M, Leitman S, Gottesman M, Pastan 1, Denicoff A, Noone M. Gress R (1994) Retroviral mediated m s f e r of the human multidrug resistance gene (MDR-1) into hematopoietic stem cells during autologous transplantation after intensive chemotherapy for metastatic breast cancer. Hum Gene Ther 5: 89 1

29.0ttenon GA, Khleif SN, Chen W, Coxon AB, Kaye FJ (1995) CDKN2 gene silencing in lung cancer by DNA hypennethylation and kinetics of p16INK4 protein induction by 5-aza-2'-deoxycytidiine. Oncogene 11: 1211

30.Plunkett W, Huang P, Gandhi V (1995) Preclinical characteristics of gemcitabiie. Anti-Cancer Drugs 6(Supp16): 7

31.Podda S, Ward M, Himelstein A, Richardson C, de la Flor-Weiss E, Smith L, Gottesrnan MM, Pastan I, Bank A (1992) Transfer and expression of the human multiple drug resistance gene into live mice. Proc Natl Acad Sci USA 89: 9676

32.Rivard GE, Momparler RL, Demers J , Benoit P, Raymond R, Lin KT, Momparler LF (1981) Phase I study on 5-aza-2'deoxycytidine in children with acute leukemia. Leuk Res 5: 453

33.Shr6der JK. Kirch C, Flasshove M, Kalweit H, Seidelmann M. Hilger R, Seeber S, Schctte J (1996) Constitutive overexpression of the cytidine

dearninase gene confers resistance to cytosine arabiioside in uitro. Leukemia 10: 1919

34.Sorrentino BP. Brandt SJ, Bodine D. Gottesman M, Pastan I. Cline A, Nienhuis AW (1992) Selection of drug-resistant bone marrow cells m

uivo after retroviral transfer of human MDR1. Science 257: 99

35.Steua1-t CD. Burke PJ (1971) Cytidiie dearninase and the development of resistance to arabiiosyl cytosine. Naturr New Biol233: 109

36.Zagonel V, Lo Re G, Marotta G, Babare R, Sardeo G. Gattei V, De Angelis V, Monfardiii S, Pinto A (1993) 5-aza-2'-deoxycytidine (Decitabine) induces trilineage response in unfavourable myelodysplastic syndromes. Leukemia 7(Suppll): 30

CHAPTER9: Article 3

Retroviral transfer and long-term expression of human cytidine deaminase cDNA in hematopoietic cells following transplantation in mice

Nicoletta EliopoPlosl, Veronica Bovenbl, Oanh N.L. Lei. Louise F. Momparlerl, Mona Greenbaama, Sylvah L&toarneaw, Denis Collmoyera, and Richard L. Momparlerl 'Dtpartement de pharrnacologie, Universite de Montreal, Centre de recherche pckiiatrique, Hdpital Ste-Justine, and 2Departments of Medicine and Oncology, Montreal General Hospital, Montreal, Quebec, Canada.

Published in Gene Therapy S(l1): 1545-1551, 1998.

Running title: Cytidine deaminase gene transfer and in vivo expression

Correspondence to: Dr. Richard L. Momparler, Centre de recherche pidiatrique, Hdpital Ste-Justine, 3175 Chemin Cdte SteCatherine, Montr&l, Quebec H3T 1C5, Canada. tel: (514) 345-4691; fax: (514) 3454801

The chemotherapeutic effectiveness of cytosine nudeoside analogues used in cancer therapy is limited by their dose-dependent myelosuppression. A way to overcome this problem would be to insert the drug-resistance gene, cytidine deaminase (CD), into normal hematopietic cells. CD catalyzes the deamination and pharmacological inactivation of cytosine nudeoside analogues, such as cytosine arabiioside (ARA-C). The objective of this study was to determine if we could obtain long term persistence and expression of proviral CD in hematopietic cells following transplantation of CD-transduced bone marrow cells in mice. Murine hematopietic cells were transduced with an MFG retroviral vector containing CD cDNA and transplanted into lethally irradiited mice. The recipient mice were administered three courses of 10-15 hour i.v. infusions of ARA-C (75-1 10

mg/kg). Blood, marrow and spleen samples were obtained and analyzed for CD proviral DNA by PCR, CD activity by enzyme assay, and drug resistance to ARA-C by clonogenic assay. We detected the presence of the CD proviral DNA in most of the samples examined. Approximately one year after transplantation several mice showed increased expression of CD activity in these tissues and some mice displayed signs of ARA-C resistance. These data demonstrate that persistent in vivo expression of proviral CD can be achieved in transduced hematopoietic cells and indicate some potential of this gene for chemoprotection to improve the eficacy of cytosine nucleoside analogues in cancer therapy.

Keywords: cytidine deaminase, cytosine arabiioside, gene transfer, hematopoietic cells, drug resistance, in vivo expression

Hematopoietic toxicity produced by intensive treatment with many anticancer drugs limits their curative potential. The transfer and expression of drug resistance genes in normal blood cells may protect them from the toxic effect of chemotherapy and pennit dose escalation to increase clinical efiicacy.1

Several investigators have inserted drug resistance genes into murine bone marrow cells to evaluate the potential of chemoprotection in mice. In these studies, the transplantation of hematopoietic cells transduced with the genes for multiple drug resistance (MDR),U dihydrofolate reductase (DHFR)ts or O~alkylguanintDNA alkyltransferase (alkyItransferase)6.7 resulted in expression of these drug resistance genes and in some cases chemoprotection against the respective antineoplastic drugs. Clinical trials on chemoprotection of cancer patients with the introduction of the MDR gene into hematopoietic progenitor cells are presently under investigati0n.e-11

Our laboratory has been investigating the drug resistance gene, cytidiie deaminase (CD), which codes for an enzyme that inactivates cytosine nucleoside analogues, such as cytosine arabinoside (ARA-C), 2',2'-difluorodeoqcytidine (dFdC) and 5-aza-2'4eoxycytidine (5-AZA- CdR).12.13 Myelosuppression constitutes a major limiting factor for the utilization of high doses of these analogues in cancer therapy.1"'

In previous studies, we purified, cloned and expressed the human CD cDNA.18 We demonstrated increased resistance to ARA-C,19.m dFdC and 5-AZA-CdR2122 in murine fibroblasts following retroviral gene transfer of CD. In addition, transduction of normal murine bone marrow cells conferred in uitro chemoprotection from ARA-C.20

In the present report, we likewise used MFG-CD retroviral virions to transduce murine bone marrow cells and then transplanted these into syngeneic mice that were lethally irradiated. Following treatment with ARA-C, we detected the presence of proviral CD and increased CD activity in various hematopoietic tissues, and signs of in uitro drug resistance to ARA-C.

Materials and methods

Cell culture of mPrine fibroblasts GPcJ2-86 ecotropic retrovirus-packaging cells (A. Bank, Columbia University, New York)z were grown in Dulbecco's Modified Essential Medium (Canadian Life Technologies, Burlington, Ontario) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Wisent Technologies, S t Bruno, Quebec) and 5 p g / d gentamicin (Canadian Life Technologies). and incubated a t 37'C with 7% CO2.

Retrovid vector and virus-producing cell lines The cDNA for human CD was cloned into the pMFG retroviral vector (R. Mulligan, Harvard University, Cambridge, Mass.) to yield the pMFG-CD construct, a s we described previously.19 The plasmid pMFG-CD was transfected into GP+E-86 packaging cells. Clone GP+M6-CD3 was isolated and displayed elevated CD expression and ARA-C resistance.19 This producer cell l i e had an estimated viral titer of -4 x 10s infective particles/ml.m The pMFGLacZ vector (R. Mulligan) was similarly used to generate the clone GP+G86-LacZ.

Donor mice and transduction of bone marrow cells GP+E-86-CD3 and GWE-86-LacZ cells were sublethally irradiated with 20 Gy (cobalt source) and plated a t a density of 3 to 5 x 106 cells per 10 cm diameter tissue culture dish, 12 to 24 hours preceeding bone marrow transduction. Hematopoietic cells were harvested from the hind legs femurs and tibias of 7 to 21-week old male C3H HeJ mice (Jackson Laboratories, Bar Harbor, Maine) 48 hours after a single intraperitoneal injection of 150 mg/kg 5-fluorouracil (David Bull Laboratories, Vaudreuil, Quebec). Fresh nucleated marrow cells (3 to 5 x 106 cells) were layered onto each monolayer of irradiated viral-producing cells. Cocultures were incubated for 72 hours in alpha minimal essential medium (Gibco, Grand Island, NY) supplemented with 20% FBS, 10 m g / d bovine serum albumin (Boehringer Manheim, Laval, Quebec), 0.3 m g / d iron-saturated human transferrin (Boehringer Mannheim), 5 pg/ml gentamicin (Canadian Lie Technologies), 4 pg/ml polybrene (Sigma Chemical Company, StLouis, MO), 10% conditioned medium from the WEHI-3B murine myelomonocytic

cells= and 10% conditioned medium from the HTB-9 primary bladder carcinoma cell line.=

Transplantation and recipients Following cocultivation, the nonadherent hematopoietic cells were carrfully collected, concentrated by centrifugation, and 1 to 2 x 106 cells injected via the tail vein into irradiated female syngeneic recipient mice (Jadcson Laboratories, Bar Harbor, Maine) exposed to 10.5 Gy irradiation 17-20 hours earlier. Animals that received marrow transduced with MFG-CD were termed 'CD mice" whereas those that were transplanted with MFG Lac2 transduced cells were called 'Lac mice". The transplanted mice were provided with 2.5rngJml neomycin in the drinking water starting one week before and until 4 weeks after transpbtation.

ARA-C infusions and tissue acquisition The mice were administered three cycles of ARA-C (Upjohn, Don Mills. Ontario) administered as a single 10 to 15 hour infusion at a total dose of 75 to 110 mg/kg at intervals of 4 to 5, 6 to 8, and 8 to 10 months post- transplantation. The iv infusions were performed as described previously.~s Briefly, unanesthetized mice were placed in restraining cages (ScientSc Products, Irvine, CA) with a wood tongue blade taped to the bottom. A 25 gauge needle, attached to a butterfly and 5 cc syringe, was inserted into the lateral tail vein and immobilized to the wood blade by Transpore 3M tape. The syringe containing sterile ARA-C in 0.45% NaCl was attached to a Harvard Model 975 infusion pump (South Natick, MA) and the drug infused at a rate of 0.22 ml/hr. Food pellets were placed in the restraining qige during the infusion. Five days after every treatment, peripheral blood was obtained from the tail vein. At 11 to 13 months post-transplantation, 9 CD mice and 6 Lac mice were sacrificed to obtain the marrow and spleen. Blood was collected by cardiac puncture. These tissues were used for PCR analysis, CD enzyme activity measurements and clonogenic assays.

Clonogenic assay Clonogenic assays were carried out using bone marrow and spleen cells. The latter cells were incubated for 7 minutes in 0.83% NH4Cl to hemolyze red blood cells. The cells from the mice were plated at 10s cells/dish

(marrow) or 5 x 1& cells/ dish (spleen) in RPMI 1640 medium (Canadian Life Technologies) containing 30% WEHI-3B conditioned medium. 1pM hydrocortisone (Sigma Chemical Company) and 0.36% Sea Plaque agarose (FMC BioProducts, Rockland, Maine). Cultures were plated in triplicate 1.5 ml aliquots in 35 mm diameter Petri dishes. A R A 4 was added in graded concentrations ranging from 0 to 0.5 pM. Colonies of greater than 5 0 cells were enumerated after 2 to 3 weeks of culture a t 37'C and 7% CO2. Large colonies were isolated for PCR analysis.

DNA isolation and polymerase chain reaction (PCR) Genomic DNA was prepared from blood, marrow, and spleen cells as well a s from individual or pooled hematopoietic colonies utilizing the IDPure Genomic DNA kit (ID Labs Biotechnology, London, Ontario). Briefly, cells were lysed with a guanidie thiocyanate solution and DNA was adsorbed on silica gel and then eluted with TE buffer (10rnmol/l Tris, lmmol/l EDTA, pH 7.4).

Separate PCR reactions were performed with 3 distinct sets of primers. To detect the MFG-CD consrmct, the sense oligonucleotide 5' GGT GGA CCA TCC TCT AGA CTG 3' was used with the antisense oligonucleotide 5' AGC AGC TCC TGG ACC GTC ATG 3'. for the amplification of a 421 bp fragment. To verify the presence of a n internal standard, a 363 bp S-globin product was amplied using the primers 5' GAA GTT GGG TGC TTG GAG AC 3' (sense) and 5' GGA AGG TTG AGC AGA ATA GC 3' (antisense). To determine the origin of cells given that male mice served a s donors for female recipients, sense primer 5' CTC CTG ATG GAC AAA CTT TAC G 3' and antisense primer 5' TGA GTG CTG ATG GGT GAC GG 3' were employed for amplification of a 444 bp Y- chromosome fragment? For the PCR reaction, genomic DNA was subjected to an initial denaturation a t 94°C for 2 min, followed by amplification for 40 cycles, each including denaturation a t 94'C for 30 sec, annealing a t 57OC for 30 to 4 5 sec, and extension a t 72-C for 3 0 to 9 0 sec, and a single cycle elongation step a t 72'C for 5 min. The reaction mixture (25 or 50 pl) contained 1.25 units of Taq polymerase (Pharmacia, Baie d'Urfi, Quebec or ID Proof. ID Labs, London. Ontario). Resulting PCR products were resolved by 2% agarose gel electrophoresis, visualized by ethidium bromide staining and photographed under W light.

Enzyme assay Bone marrow. spleen and blood cells were assayed for CD activity as described pxviously.~ The cells (0.5 to 15 x 106) were washed in phosphatebuffered saline (PBS), suspended in Tris-HC1 (SmM, pH 7.4) and dithiothreiotol (5 mM), and rapidly freezethawed three times. The cytosolic extract was recovered in the supernatant following high speed centrifugation. For the enzyme assay, the cytosol was placed in a mixture containing 50 mM Tris-HC1 and 0.5 pCi 'H-cytidine (ICN Biomedicals, Irvine. California). The reaction was incubated for 30 rnin at 37'C and then stopped with cold HCl(0.001N). The mk was poured onto Whatman P-81 phosphocellulose discs and the radioactivity bound assessed by scintillation counting. Under these conditions, cytosine nucleosides b i d to the fdter, but not the deaminated uracil nucleosides. One unit of CD activity was defmed a s the quantity of enzyme that catalyzes the deamination of 1 nmole of 3H-cytidie per min a t 37°C. Total protein concentration was measured using the Bio-Rad protein assay (Bio-Rad Laboratories. Mississauga, Ontario).

Hematopoietic cells from male mice were transduced with MFG-CD or MFG-LacZ virions ex viuo and then transplanted into irradiated female syngeneic mice. Recipients were administered a n ARA-C infusion at intervals of 6 or more weeks for 3 treatments. The peripheral blood was collected a t 5 days after each ARA-C treatment At 11 to 13 months after transplantation, the mice were sacrificed and their blood, marrow and spleen analyzed for proviral integration and CD expression.

Detection of CD provims in blood at different time intervals after transplantation To establish the presence in blood of the CD transgene a t different times after bone marrow transplantation, the blood was examined by PCR (Table 9.1). For these CD mice PCR analysis was performed on DNA isolated from blood collected 5 days after each cycle of ARA-C administered a t intervals of 45 ,6-8 , and 8-10 months. The mice were ~ a ~ c e d a t 11-13 months. Results showed a detectable CD provirus in a t least 75% of the mice studied. The % of mice with CD positive blood did not diminish with time. CD proviral DNA was not detected in Lac mice (data not shown).

Detection of CD provims in different tissues at 11-13 months post- transplantation In order to obtain evidence of long term engraftment of transduced cells in various hematopoietic tissues, PCR reactions were performed to detect CD provirus and Y-chromosome. Figure 9.1 displays PCR analysis of genomic DNA from blood, bone marrow and spleen cells obtained a t 11-13 months post-transplantation. In 7 out of 9 mice, we detected the CD transgene in all three tissues. Of the other 2 mice, CD mouse #2 was positive for CD proviral DNA only in the bone marrow, whereas CD mouse #3 was positive in the spleen and blood. As expected, samples from Lac mice failed to yield a proviral CD signal; only Lac mouse #4 is included in Figure 9.1.

An additional PCR reaction was performed to evaluate the engraftment of male donor hematopoietic cells into the female recipients by the demonstration of the presence of the Y- chromosome. We detected this DNA fragment in all the three tissues of wery CD mouse examined (Figure 9.1). This result further confirms that the MFG-CD positive PCR signals

originated h m the integrated CD gene in male donor marrow cells. Of the 6 Lac mice also analyzed, a positive signal for the Y-chromosome was observed in 4 mice (data not shown). The fact that two of the 6 Lac mice analyzed did not exhibit a Y positive signal may reflect low efficiency of gene transfer, weak engraftment or shorter persistence of the MFGLacZ

' transduced cells. PCR detection of the %-globin gene was also performed to verify the quality of genomic DNA isolated h m each tissue. The B- globin was easily detected in the blood, bone marrow, and spleen of all CD and Lac mice tested.

Analysis of hcmatopoietic colonies for presence of CD pmvirrw Genomic DNA was prepared from a single colony or pools of 2-5 colonies derived from bone marrow cells harvested from mice 11-13 months post- transplantation. This DNA was used for PCR analysis for detection of CD provirus, Y-chromosome and B-globm gene (Table 9.2). In several assays of individual colonies from CD mouse #6, #8 and #9, we observed the proviral CD in 44%, 20% and 50%, respectively, of the colonies analyzed. For the pooled colonies, the positive CD signal varied from 0 to 100% for CD mouse #1 and CD mice #4 to #7. For mouse CD #6, which had the highest CD enzyme activity in marrow (Figure 9.2), 100% of pooled colocies contained the CD proviral sequences, whereas we detected CD in 44% of the single colonies. For this mouse, 60% of individual colonies derived from spleen displayed positive PCR signals for MFG-CD (data not presented).

To determine the efficiency of engraftment of male donor cells, Y- chromosome screening was performed. For CD mouse #1 and CD mice #4 to #7, a t least 75% of pooled hematopoietic colonies scored positive by PCR for the Y-chromosome fragment. For CD mouse #6 and #9, 100% and 80% of single colonies, respectively, displayed the positive Y signal. However, for CD mouse #8, even though we detected 20% MFG-CD positive colonies, we did not observe any Y positive bands in the individual colonies. This occurrence was probably due to the very small amounts of DNA isolated from a single hematopoietic colony resulting in our performing PCR a t the limit of detection which may have led to false negative results in some cases.

Exprrssio~ of human CD enzyme activity in merent tissues after transplantation CD enzyme activity was measured in various tissues to determine if the CD provirus was capable of expression at 11 to 13 months after transplantation of transduced hematopoietic cells. As depicted in Figure 9.2, bone marrow cells of several CD mice e:l-'bi:rxi increased levels of CD activity as compared with the mean value from 5 Lac mice. CD mouse #6 marrow cells displayed a striking >80-fold augmentation in enzyme activity when compared to Lac controls. Bone m a i w cells from CD mouse #I, #3. #4 and #8 showed a 2.9 to 5.4-fold higher CD activity than Lac mice. These differences of enzyme activity between Lac mice and CD mouse #1, #3. #4. #6 and #8 were significant. P 5 0.05 (Student's t test).

In Figure 9.2, analysis of the spleens of recipients of transduced marrow revealed that for several CD mice, the CD enzyme activity surpassed that of Lac mice. Notably, CD mouse #6 and #8 showed a 20- and %fold increase, respectively, in CD activity a s compared with control Lac mice. CD mouse #I, and #7 showed a 2.2 and 3.3-fold elevation in CD activity, respectively, as compared to the activity in spleens from Lac mice, P 5 0.05 (Student's t test).

Analysis of the blood cells showed >loo-fold increase in CD activity for CD mouse #6 when compared to control mice (Table 9.3). CD mouse #3 and #4 showed a 5 and 10-fold rise in CD activity respectively. Strikingly, CD mouse #6 displayed from 20 to 130-fold augmentation in CD enzyme activity in the d i i rent tissues analyzed when compared to the control Lac mice.

Clonogenic assay on marrow and spleen cells for A R A 4 resistance In uitro clonogenic assays were performed to evaluate drug resistance to ARA-C in bone marrow and spleen cells collected from mice 11-13 months after transplantation. At an ARA-C concentration of 0.5 JIM, the donogenic progenitor cells of several CD mice demonstrated a higher survival than those of Lac mice. A range of 2 to 6% survival in presence of ARA-C was observed by clonogenic assays of marrow cells from CD mouse #I, #3 and #4 a s compared to less than 1% survival for marrow cells from Lac mice (Table 9.4). Notably, the marrow cells from CD mouse #6 (which also displayed the highest CD enzyme activity) showed 29% survival in presence of ARA-C. All the colonies that were formed in the presence of

104

ARA-C from CD mouse #6 showed a positive PCR signal for MFG-CD (data not shown). For colonies generated from spleen cells, 23 and 17% survival in presence of ARA-C were noted for CD mouse #4 and #6, respectively.

Table 9.1 PCR analysis for detection of CD proviral DNA in blood cells from CD mice collected after different ARA-C treatment cycles

ARA-C CD proviral CD proviral treatment positive mice positive mice cycle (no. / total) (%I

1. 314 75

.blood samples obtained 5 days after ARA-C treatment bblood samples obtained 3 months after 3rd ARA-C treatment

Table 9.2 PCR analysis of bone marrow-derived colonies from mice transplanted with MFG-CD transduced marrow cells

Colony % PCR-oositive colonies Mouse analysis Y-ch CD &glob

- - -

CD#6 single 100 44 89

CD#8 single 0 20 100

CD#9 single 80 50 90

CD# 1 pooled 80 0 80

CD#4 pooled 100 17 100

CD#5 pooled 100 33 100

CD#6 pooled 100 100 100

CD#7 pooled 75 25 100

For single colony analysis, DNA isolation was performed on an individual colony, and for each mouse, 5-10 different colonies were analyzed by PCR.

For pooled colony amlysis, DNA isolation was performed on a pool of 2-5 colonies, and for each mouse, 3-6 different pools were analyzed by PCR.

' Table 9.3 Cytidine dcaminase (CD) activity in blood cells

Mouse Blood CD activity Relative increase (V/mg)

Lac 0.4 5 0.2 1

P ( 0.05 (Student's t test) for the differences between Lac and CD #3, #4, and #6. U, Units (mean * SE ) defined as nmoles cytidine dearninated per rnin (n = 3-4).

Table 9.4 Effect of ARA-C on clonogenic assays

Mouse % survival in 0.5uM ARA-C Marrow Spleen

Lac (mean of 6) c1 <1

n=3; S.E.< 10%; ND=not determined

Figure 9.1 Detection by PCR of Y-chromosome, CD provirus, and 13- globin in various tissues at 11-13 months after transplantation. PCR reactions were carried out with 3 different sets of primers using purified genomic DNA from blood, bone marrow, and spleen cells from 9 CD mice and 6 Lac mice at 11-13 months post-transplantation. Only Lac mouse #4 is shown. Amplification of specific 444bp. 421bp and 363bp bands indicates the presence of the donor Y- chromosome, the CD proviral DNA, and the 5-globi internal standard, respectively. H20, H20 used as template.

Figure 93 Exprnsion of human CD enzyme activity in bone marrow and spleen cells a t 11-13

months ahcr transplantation. CD enzyme activity was measured in cell extracts of bone marrow

and spleen & from CD and Lac mice a t 1 1 - 1 3 months following transplantation with

t r a d u c e d marrow ails. The Lac column rrpments the mean value from 5 Lac mice. For the

mmmw &. P 5 0.05 (Student's t t t ) for the diKermces be- Lac and CD Y1. Y3. *4. Y b

and 48. For the spleen cells, P 2 0.05 (Student's t test) for the differences between Lac and CD

Y l , 6. Y 7 and Y8. U, Units (mean i SE) defamed a s nmoles cytidine dcaminatcd per min In-3-61.

The myelosuppression produced by many antineoplastic drugs has prevented the utilization of intensive doses which may improve antitumor response. For instance, the clinical effectiveness for the treatment of lung or breast cancer with cytosine nucleoside analogues, such as ARA-Cl5 and the new promising agents, dFdClb and 5-AZA-CdRlT is limited by their hematopoietic toxicity. The efficacy of chemotherapy with these analogues would be improved if their dose-dependent myelotoxicity could be circumvented. One way to overcome this serious side effect would be to overexpress the gene for CD in normal hematopoietic progenitor cells to render them drug resistant.

Cytidik deaminase can deaminate cytosine nucleoside analogues and thus catalyze their conversion to pharmacologically less active uracil derivatives.lz.l3 We have inserted the CD cDNA into the MFG retroviral plasmid expression vector. The cDNA for human CD contains only 438 bpi* facilitating genetic manipulation. We succeeded in transfecting and, subsequently, transducing murine cells in vitro.l920 The transduced fibroblast cells demonstrated significantly increased expression of CD proviral RNA and enzyme activity, as well as drug resistance to ARA4.m In addition, in a separate study we observed drug resistance in these cells to other cytosine nucleoside analogues, dFdC and 5-AZA-CdR.21.n We also reported that clonogenic murine hematopoietic cells transduced with CD showed resistance to ARA-C.m

In this study we demonstrate that CD-transduced bone marrow cells transplanted into mice can persist for more than one year and express CD. The persistence of the MFG-CD construct in many tissue samples after transplantation indicates the successful transduction of some pluripotent bone marrow cells capable of long term engraftment The detection of CD provirus by PCR in many hematopoietic colonies more than one year post- transplantation indicates transduction and persistence of the transgene in hematopoietic progenitor cells that can survive in vivo and can form colonies in vitro.

For mouse CD #6, which showed the highest CD enzyme activity, 44% of individual colonies were found to be positive for MFG-CD provirus (Table 9.2). For this same mouse, 60% of single colonies derived from spleen displayed positive PCR signals for MFG-CD (data not presented).

The increased levels of CD enzyme activity observed in the bone marrow, spleen and blood in many mice when assayed a t 11-13 months indicates long-term expression of the proviral CD gene. The low CD enzyme activity obtained in some of the tissue samples may be due to low transduction frequency and/or provirus inactivation possibly by de novo methylation of cytosine residues in the retroviral promoter for proviral CD.u Our results are in agreement with several studies utilizing other drug resistance genes that also showed long-term expression after transplantation of transduced cells into mice. n.7

Our investigation constitutes the first in vivo CD gene transfer study reported. The data presented here using the MFG-CD construct dearly demonstrate the feasibility of obtaining in vivo long term persistence of CD and long term expression of human CD in hematopoietic cells. We plan to transplant CD transduced marrow cells into tumor bearing mice to determine if curability with cytosine nucleoside analogues can be obtained using chemoprotection as reported by Zhao et al.2' for the DHFR gene and methotrexate therapy. Future improvements in vectors to increase efficiency of transduction and expression of CD in human hemato~oietic cells will facilitate the use of this gene for chemoprotection to increase the therapeutic effectiveness of cytosine nucleoside analogues in cancer therapy.

Acknowledgments This work was supported by Grant # MT-13754 from the Medical Research Council of Canada.

1 I3

References

1. Bertino JR. TTurning the tables" - making normal m m w resistant to chemotherapy. J Natl Cancer Inst 1990; 82: 1234-1235.

2. Hanania EG et aL Resistance to tax01 chemotherapy produced in mouse marrow cells by safety-modified retroviruses containing a human MDR- 1 transcription unit Gene Ther 1995; 2: 279-284.

3. Podda S, et aL Transfer and expression of the human multiple drug resistance gene into live mice. Proc Natl Acad Sci USA 1992; 89: 9676- 9680.

4. Corcy CA, DeSilva AD, Holland CA, Williams DA. Serial transplantation of methotrexate-resistant bone marruw: protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 1990; 75: 337-343.

5. Zhao SC, Li MX, Banerjee D, Schweitzer BI. Mineishi S, Gilboa E, Bertino JR. Long-term protection of recipient mice from lethal doses of methotrexate by marrow infected with a double-copy vector retrovirus containing a mutant dihydrofolate reductase. Cancer Gene hempy 1994; 1: 27-33.

6. May JA et aL Retroviral transduction and expression of the human alkyltransferase cDNA provides nitrosocrea resistance to hematopoietic cells. Blood 1995; 85: 3342-3351.

7. Allay JA, Davis BM, Gerson SL. Human alkyltransferase-transduced murine myeloid progenitors are enriched in vivo by BCNU treatment of transplanted mice. Exp Hematol1997: 25: 1069-1076.

8. Hesdorffer C et aL Human MDR gene transfer in patients with advanced cancer. Hum Gene h e r 1994: 5: 1151-1 160.

9. Deisseroth AB, Holmes F, Hortobagyi G, Champlin R. Use of safety- modified retroviruses to introduce chemotherapy resistance sequences

into normal hematopoietic cells for chemoprotection during the therapy of breast cancer: a pilot trial. Hum Gene Ther 1996; 7: 401-416.

10.0'Shaughnessy JA et aL Retroviral mediated transfer of the human multidrug resistance gene (MDR-1) into hematopoietic stem cells during autologous transplantation after intensive chemotherapy for metastatic breast cancer. Hum Gene Ther 1994; 5: 89 1-9 1 1.

11.Hesdorffer C et aL Phase I trial of retroviral-mediated transfer of the human MDRl gene a s marrow chemoprotection in patients undergoing high-dose chemotherapy and autologous stem-cell transplantation. J Clin Onco11998; 16: 165-172.

12.Bouffard DY, Laliberte J , Momparler RL. Kinetic studies on 2 '3- difluorodeoxycytidine (gemcitabiie) with purified human deoxycytidine kinase and cytidine deaminase. Biochem Phannacol 1993; 45: 1857- 1861.

13.Chabot GG, Bouchard J , Momparler RL. Kinetics of deamination of 5- aza-2'4eoxycytidiie and cytosine arabiioside by human liver cytidine deaminase and its inhibition by 3-deazauridine, thymidine or uracil arabioside. Biochem Phannawl1983; 32: 1327-1328.

14.Rivard GE et aL Phase I study on 5-aza-2'deoxycytidiie in children with acute leukemia. Lack Res 1981; 5: 453-462.

15.Czaykowski PM, Samuels T, Oza A. A durable response to cytarab'ie in advanced breast cancer.. Clin Oncol 1997; 9: 181-183.

16.Fossella FV et aL Maximum-tolerated dose defined for singleagent gemcitabine: a phase I dostescalation study in chemotherapy-naive patients with advanced non-small-cell lung cancer. J Clin Oncol 1997; 15: 310-316.

17.Momparler RL et aL Pilot phase 1-11 study on 5-aza-2'-deoxycytidine (Decitabine) in patients with metastatic lung cancer. Anti-Cancer Drugs 1997; 8: 358-368.

18.Laliberte J , Momparler RL. Human cytidine deaminase: Purification of enzyme, cloning, and expression of its complementary DNA. CmcerRes 1994; 54: 5401-5407.

19.Momparler RL et aL Transfection of murine fibroblast cells with human cytidine deaminase cDNA confers resistance to cytosine arabiioside. Anti-Cancer Drugs 1996; 7: 266-274.

20.Momparler RL et aL Resistance to cytosine arabiioside by retrovirally- mediated gene transfer of human cytidine deaminase into murine fibroblast and hematopoietic cells. Cancer Gene Ther 1996; 3: 331438.

2l.Eliopoulos N et aL Gene transfer of human cytidine deaminase cDNA into murine cells confers resistance to cytosine arabiioside, 5-aza-2'- deoxycytidiie and Z.2'-difluorodeoxycytidine. 9th NCI-EORTC symposium on new drugs in =cer therapy. Ann Oncol 1996; 7 (Suppl 1): 58.

22.Eliopoulos N, Cournoyer D, Momparler RL. Drug resistance to Saza-2'- deoxycytidiie, 2.2'-difluorodeoxycytidine and cytosine arabiioside conferred by retroviral mediated transfer of human cytidine deaminase cDNA into murine cells. Cancer Chem Phnnnawl (in press).

23.Challita P-M, Kohn DB. Lack of expression from a retroviral vector after transduction of murine hematopoietic stem cells is associated with methylation in vivo. Proc Natl Acad Sci USA 1994; 91: 2567-2571.

24.Zhao SC, Banejee D, Mineishi S, Bertino JR. Post-transplant methotrexate administration leads to improved curability of mice bearing a mammary tumor transplanted with marrow transduced with a mutant human diiydrofolate reductase cDNA. Hum Gene Ther 1997; 8: 903-909.

25.Markowitz D. Goff S, Bank A. A safe packaging line for gene transfer: separating viral genes on two different plasrnids. J Virol 1988; 62: 1120-1 124.

26.Murphy WH, Umovitz HB, Maryanski JL., Abrams GD. Characterktion of transplantable myelomonocytic leukemia WEHI-3B in syngeneic BALBjc mice. Proc Soc Exp Biol Med 1978; 157: 556-564.

27.Hoang T, McCullough EA. Production of leukemic blast growth factor by a bladder carcinoma cell line. Blood 1985: 66: 748-751.

28.Momparler RL, Gonzales FA. Effect of intravenous infusion of 5-aza-2'- deoxycytidine on survival time of mice with L12 10 leukemia Cancer Res 1978; 38: 2673-2678.

29.Drize N et aL Long-term maintenance of hematopoiesis in irradiated mice by retrovirally transduced peripheral blood stem cells. Blood 1997; 5: 1811-1817.

3O.Momparler RL, Laliberte J. Induction of cytidine deaminase in HMO myeloid leukemic cells by 5-aza-2'-deoxycytidine. Lack Res 1990; 14: 751-754.

PART THREE

DISCUSSION

CHAPTER 10: DISCUSSION

10.1 Evaluation of merimental results

Current chemotherapy for advanced metastatic cancer is often not very effective (Wingo et al., 1995). Cytosine nudeoside analogs ARA-C, dFdC and 5-AZA-CdR show promising antitumor activity in patients with advanced lung or breast cancer (Czaykowski et al, 1997; Fossella et al., 1997; Momparler et al., 1997). Since more intensive chemotherapy with cytosine nucleoside analogs may lead to a better clinical response, their curative potential may be improved if their dose-limiting side effect of myelosuppression could be prevented. To achieve this, one approach would be to render normal bone marrow cells resistant to the cytotoxicity of cytosine nucleoside analogs by introducing into these cells the drug resistance gene, CD, which codes for an enzyme that inactivates by deamination this class of antineoplastic agents.

To determine the possible use of CD for gene transfer experiments, we first constructed the retroviral vector by insertion of the human CD cDNA into it (Annex I). This vector, caIIed pMFG-CD (Figure 10.1), was used to transfect a murine fibroblast ecotropic packaging cell line. Clones of MFG-CD virus-producing cells were isolated and demonstrated to have an increased CD enzyme activity, increased expression of the mRNA for the CD transgene, and to possess the ARA-C resistance phenotype.

Nco I Barn HI

pMFG - t PA

Nco l Barn HI

CD CDS k5-J Nco I + Barn HI digest T4 Ligase

SD ATG TGA

pMFG - CD

Figure 10.1 Molecular design of the retroviral vector for the expression of cytidiie deaminase cDNA (CD CDS) in target cells. pMFG-CD was constructed by cloning CD CDS between the Ncol and BamHl sites. Long terminal repeat (LTR); packaging sequence (y); splice donor (SD); splice acceptor site (SA).

Using the MFG-CD virions generated by one producer cell line (Figure 4.1), we transduced NIH 3T3 mouse fibroblasts (Chapter 7). Two clones of transduced cells were isolated after ARA-C selection, 3 T X D 3 - V5 (V5) and 3T3CD3-V6 (V6), and shown to overexpress the CD transgene. These clones also displayed resistance to the inhibitory action of ARA-C on cell growth, on DNA synthesis, and on colony formation. Clone V5 exhibited a higher CD enzyme activity and more resistance to AliA-C as compared to done V6. This result indicates that there is a correlation between the extent of expression of the CD transgene and drug resistance to ARA-C. huthermore, the clone V5 showed stronger expression of the spliced proviral mRNA transcript suggesting that the higher splicing correlated with increased CD enzyme zctivity and increased ARA-C resistance. Our results are in agreement with those of Krall et al. (1996) who determined in hematopoietic cells, using also the MFG vector, that enhanced amounts of spliced proviral mRNA correlated with higher transgene expression.

Our next goal was to investigate if CD gene transfer could confer drug resistance to other cytosine nucleoside analogs, such as the promising experimental drugs, 5-AZA-CdR and dFdC (Chapter 8). We analyzed by clonogenic assay the V5 clone and observed that these cells were resistant not only to ARA-C, but were also cross-resistant to 5-AZA- CdR and dFdC. In order to demonstrate that the drug resistance phenotype was due to the overexpression of the CD gene, we used THU, a competitive inhibitor of the CD enzyme. We showed that THU restored the sensitivity of the V5 clone to the inhibition of colony formation produced by cytosine nucleoside analogs. This reversal of drug resistance by THU was directly related to its concentration. Our following aim was to determine if the reversal of drug resistance in CD-transduced cells by THU correlated with its inhibition of CD enzyme activity. To accomplish this we measured the CD activity in cytosolic extracts of these cells, in the presence or absence of THU. We showed that there was a high correlation (0.87 - 0.96) between the inhibition of CD activity produced by THU and its reversal of drug resistance. These experiments using THU support that the major cause of the drug resistance to cytosine nucleoside analogs is the augmented CD activity in the CD-transduced cells. An interesting application of our CD-transduced cells would be as a model for evaluating the potency of new inhibitors of the CD enzyme activity.

Neff and Blau (1996) also showed in vitro chemoresistance to ARAC following gene transfer of CD into 3T3 cells, but using a different retroviral construct, the LCDSN vector. The CD activity that was detected using their vector was lower (7-fold) than what we obtained with the MFGCD construct in our V5 done. These researchers additionally transduced CEM human lymphoid leukemic cells and conferred resistance to the inhibition of cell growth by the cytosine nucleoside analogs ARA-C and dFdC. They used 2.5 X l(r M THU to reverse the ARA-C resistance. whereas with our transduced 3T3 cells we did not use a concentration greater than 104 M THU. Another group of investigators, Shriider et A. (1996), introduced the cDNA for CD into A9 murine fibroblast cells and showed an up to 3-fold enhancement in CD enzyme activity and an up to 3.5-fold increase in the ICso of ARA-C to inhibit growth and DNA synthesis.

In order to evaluate the potential of the CD transgene for hematopoietic chemoprotection we transduced murine marrow cells with the MFG-CD retroviral vector (Chapter 7). We obtained an 80% eficiency of gene transfer determined by PCR analysis of colonies for the presence of the CD transgene. Clonogenic assay on the CD-transduced marrow cells revealed significant resistance to ARA-C. The hematopoietic colonies from the CD-transduced cells demonstrated almost complete drug resistance to even the highest concentrations of ARAC (i.e. 104 M and l W M), a t which less than 3% of the Lac Z control cells survived. It is interesting to note that the 104 M ARA-C concentration corresponds to the plasma concentrations achieved in patients administered conventional dose ARA-C (Slevin et al., 1982) whereas the 10-5 M concentration corresponds to the plasma levels obtained during high dose therapy with ARAC (Capizzi et al., 1983).

An adcjitional goal was to determine if it was possible to obtain long- term in Yivo transgene expression following transplantation in mice of murine hematopoietic cells transduced with MFG-CD (Chapter 9). In these experiments (Figure 10.2), we transduced primary bone marrow cells from male mice with the CD transgene and then transplanted these cells into female syngeneic mice. AII mice received three cycles of ARA-C a t intervals of 6 or more weeks. At over 10 months post-transplantation. the CD proviral DNA was detected by PCR in the majority of blood, marrow and spleen samples obtained from the CD-recipient mice, demonstrating long-

term persistence of the CD proviral DNA sequences. These findings suggest that the transduction of pluripotent hematopoietic cells possessing the capacity of long-term engraftment occurred in these experiments.

In order to determine that the detection of the CD transgene in different tissues was due to its presence in the transduced donor male marrow cells, we used PCR to assay for the presence of the Y-chromosome fragment We detected the Y-chromosome fragment in the blood, manow and spleen cells of all CD recipient mice, providing further evidence that the CD positive signals were due to the CD proviral DNA integrated into the transplanted male marrow cells. We noted that although the Y- chromosome was detected in all 9 CD mice, it was not detected in 2 out of 6 Lac control mice which may have been due to low gene transfer eificiency, poor engraftment or shorter persistence of the Lac Ztransduced marrow cells. It is possible that these results may have been due to the absence of selection by ARA-C treatments for the Lac Ztransduced control cells in contrast to the CD-transduced cells.

We also conducted clonogenic assays and determined by PCR analysis that many of the hematopoietic colonies formed in uitm from CD mice were positive for the MFG-CD provirus. The presence of the CD transgene in these colonies was detected more than 10 months from transplantation, signi@ing that the marrow progenitors that had been successfully transduced were capable of surviving in mice and capable of producing in vitm colonies. In addition, we detected in some CD mice signs of in vitro drug resistance to ARA-C in the hematopoietic cells harvested 1 1-13 months after transplantation.

Assays for CD enzyme activity of hematopoietic tissues obtained from CD mice at 11-13 months following transplantation, displayed significantly higher levels of activity than the control Lac mice. These results demonstrate long-term CD transgene in uivo expression which can persist for at least 11 months. The low CD enzyme activity detected in different tissues assayed post-transplantation may have resulted from low eficiency of transduction. low donor cell engraftment, and/or inactivation of CD transgene expression. Low expression of the target gene in the MFG retroviral vector has been reported to sometimes occur in Yivo and be related to de nouo methylation of cytosine residues within the Moloney murine leukemia virus LTR promoter region (Challita and Kohn, 1994). Notably, CD mouse #6 exhibited a 20, 80, and 130- fold increase in CD

enzyme activity in spleen, marrow, and blood cells, respectively. As stated previously, the in vim ARA-C treatments may have selected for high CD- expressing hematopoietic cells which thereafter resulted in elevated CD enzyme activity in the hematopoietic tissues.

In uivo selection of hematopoietic cells containing the MDRl proviral DNA has been reported (Podda et al.. 1992; Sorrentino et al.. 1992). Selection for the MGMT transgene in mice treated with aUcylating agents has also been shown (Allay et al., 1997).

Our study in mice is the first publication on in uivo CD transgene expression. Our data, although preliminary, show promise for the use of the CD gene for chemoprotection to improve the therapeutic efficacy of cytosine nucleoside analogs in cancer treatment.

Harvest of m a r r o w cells

Transduction

virus-&oducing

I packaging cells

Transplantation 6 Ara-C t r e a t m e n t

Hematopoietic t i ssues

PCR Enzyme assays

Clonogenic a s says

Figure 10.2 Experiment for gene transfer in mice. The bone marrow cells were harvested from male mice and transduced with retroviral particles containing the cDNA for cytidine deaminase (CD). The transduced hernatopoietic cells wen transplanted into femalc syngeneic recipients and treated by i.v. infusion with ARA-C. Hematopoietic tissues were analyzed for the presence and expression of CD and in vitro drug resistance to ARA-C.

10.2 Advantages of using the CD gene for chemoprotection in cancer theram with cytosine nucleoside analo~s

Since cytosine nucleoside analogs, such as ARA-C, dFdC and 5-AZA- CdR, are S-phase spedfic agents, their major dose-limiting toxicity is myelosuppression. The doses of these drugs can be increased considerably without encountering serious non-hematopoietic toxicity. This is not possible with other anticancer drugs (e.g., MDR drugs). The advantage of intensive doses in tumor therapy is that more drug can penetrate the target tumor, especially the central core of the tumor, in order to obtain concentrations above the minimal cytotoxic range. This is essential if curative therapy is the goal. Only the hematopoietic toxicity produced by cytosine nucleoside analogs limits their chemotherapeutic potential. The use of gene therapy to insert a drug resistance gene into the normal hematopoietic cells to protect them from drug toxicity is an interesting approach that merits investigation. Importantly, chemoprotection is only required during drug treatment.

Our in vitro data indicate that it is possible to confer drug resistance to cytosine nucleoside analogs by transfer of the CD gene into murine hematopoietic cells. One advantage of using CD in gene transfer is that it is a small gene (-440 bp) which facilitates genetic manipulation in design of new vectors for this purpose. Our preliminary in uivo data indicate that it is possible to transplant CD-transduced marrow cells and obtain long- term increased expression of this gene in hematopoietic cells.

Both dFdC and 5-AZA-Cdl? show promising activity in tumor therapy, especially in non-small cell lung cancer (Fossella et al., 1997; Momparler et al., 1997) where conventional chemotherapy has little or no success in the treatment of the metastatic form of this disease. ARA-C produced an interesting response in advanced breast cancer that should be further investigated (Czaykowski et al., 1997). In the USA there are about 150.000 deaths each year due to lung cancer (Wingo et al., 1995). Novel approaches for the chemotherapy of this disease should be explored. The use of gene therapy for chemoprotection from the hematopoietic toxicity produced by cytosine nucleoside analogs is one approach that may have enormous potential in the chemotherapy of metastatic cancer which is not very responsive to conventional chemotherapy.

10.3 Approaches to kcilitate the use of Pene therapy for chemo~rotection

The success of gene therapy depends not only on the therapeutic gene, but also on the type of vector and target cell which both rnay d e c t the efficiency of gene transfer and the degree of expression. The major diff~culty to overcome is the requirement to obtain adequate levels of transgene expression in the transduced normal cells for the desired chemoprotective effect (Rafferty et al.. 1996).

10.3.1 Modification of retroviral vectors to improve gene transfer and expression

Considerable research is being conducted on retroviral vectors to make them better vehicles for gene transfer. For the Moloney leukemia virus-based vectors, expression in hematopoietic progenitor cells was shown to be hindered by LTR enhancer sequences and the presence of a transcriptional repressor at the primer biding region of the retroviral genome (Baum et al., 1995; Licht et al., 1997). Novel viral vectors were designed to surmount some of these obstacles. For example, the LTR sequences of the myeloproliferative sarcoma virus can be associated with the primer biding region sequences of the murine embryonic stem cell virus. This novel construct improved the efficiency of transfer of the MDRl gene and the level of drug resistance in transduced hematopoietic cell lines (Baum et al.. 1995).

Challita and Kohn (1994) reported that suppression of transgene transcription in hematopoietic cells frequently occurs due to gene silencing by DNA methylation a t the promoter region. A possible way to overcome this problem is to remove the CpG sequences, where methylation takes place, from the promoter region. It is important in the use of gene therapy for confening chemoprotection that the transgene not be down-modulated in hematopoietic stem cells (Rafferty et al., 1996). Design of vectors with strong tissue-specific expression of the target gene may considerably augment the chances for successful gene therapy (Licht et al.. 1997).

10.3.2 Hematopoietic cells as targets for gene therapy

The majority of the mature hematopoietic cells have a short half-We. Since the goal of the gene transfer is long-term gene expression, early hematopoietic progenitor cells must be transduced with the therapeutic gene (Dexter and Spooncer, 1987; Rafferty et al., 1996). In order to guard against the acute bone manow toxicity during treatment with antineoplastic drugs, it is possible that protection of the committed hematopietic progenitor cells may be sufficient However, chemoprotection conferred to the early stem cells has the advantage that many cycles of chemotherapy may be administered without requiring additional transplants of transduced cells due to the long-term survival of these cells. One major obstacle in gene transfer into early hematopoietic stem cells is that these cells comprise fewer than 0.01% of the bone marrow cells. This may be overcome by using fluorescence activated cell sorting (FACS) to select CD34+ hematopoietic stem cells to enrich the population of early progenitor cells to bc transduced with the drug resistance gene.

Retroviral gene transfer has the requirement that the target cells be in a state of cell division. Since most early hematopoietic progenitor cells are in a resting state, this poses a problem. In order to bypass this diculty, growth factors may be employed to activate cell division of the stem cells (Luskey et al., 1992).

10.3.3 Ways to overcome the potential problems of retroviral vectonr

As discussed earlier (Chapter 4), when retroviral gene transfer is used, there is the risk that the gene randomly integrated into the host genome may activate an oncogene or inactivate a tumor suppressor gene, consequently resulting in mutagenesis that can lead to transformation of a normal cell into a cancer cell. This is apparently a very rare event since it has not yet been observed in more than 2,000 patients in which gene therapy was performed predominantly with retroviral vectors (Richter, 1997; Sandhu et al.. 1997). Another potential problem is that the drug- resistance gene may be accidentally inserted into tumor cells that contaminate the bone marrow (Licht et al., 1997). CD inhibitors may become indispensable tools in the unfortunate event Chemotherapy with

a cytosine nucleoside analog and a CD inhibitor can be used to eradicate tumor cells that were accidentally b.ansduced with CD. Another approach that can be used to overcome this problem is to purify the hematopoietic stem cells from the marrow to remove them from the contaminating tumor cells (Licht et al., 1997; Rafferty et al., 1996).

10.4 Future studies

10.4.1 Bicistronic vectors

Most treatments for tumor therapy utilize a combination of drugs to overcome the problem of drug resistance (Rafferty et al, 1996). Efforts to confer chemoresistance to normal hematopoietic cells must take into consideration the conventional use of drug combinations. Gene therapy with bicistronic vectors containing two drug resistance genes is an approach that can be used to protect normal hematopoietic cells against the toxicity of two different anticancer drugs. To achieve &his objective, a double gene vector can be designed for example, by the insertion of the encephalomyocarditis virus internal ribosomal entry site (IRES) between the two drug resistance genes (Morgan et al., 1992). The expression cassette contains the two cDNAs and a single promoter which in combination with the IRES allows the translation of these two open reading frames from one mRNA.

Doroshow et al. (1995) demonstrated wider spectrum of drug resistance following transduction of NIH 3T3 cells with a double gene vector using separate promoters for the genes MDRl and GST. Sevenl investigators have reported in vitro chemoprotection studies with the use of two-gene vectors containing an IRES. Suzuki et al. (1997) utilized a bicistronic vector to confer protection to murine bone marrow cells against combinational chemotherapy. These inve~tigators demonstrated concurrent resistance to vincristine and nitrosourea by the expression of the MDRl and MGMT genes, respectively. The highest level of drug resistance was observed for the drug resistance gene that was placed upstream from the IRES. Galipeau et al. (1997) used a bicistronic vector to introduce the genes for MDRI and mutant DHFR into normal hematopoietic cells. They observed that the growth of transduced human lymphoblastic leukemic cells was not inhibited by ucposure to paclitaxel in

combination with trimetrexate, in contrast to control cells that showed complete suppression of growth by these agents. Similarly, in a donogenic assay. this vector conferred drug resistance in murine myeloid progenitor cells to paclitaxel and trimetrucate. Recently, Fantz et al. (1998) reported that drug resistance to fluoropyrimidine and antifolates was demonstrated in fibroblasts transfected with a retroviral vector construct coding for the thymidylate synthase and dihydrofolate reductase genes.

10.4.2 Animal studies

Animal models can be used to optimize the conditions for chemoprotection and to test its application in rodents with tumors. The fmt priority in these experiments should be to obtain very high transduction efficiency and expression of the CD transgene in murine marrow cells. Possible ways to achieve these aims are to use a new vector and a new packaging cell line that produce very high viral titers. The use of a vector containing CD-IRES-GFP would be most useful in these types of experiments. The green fluorescence protein (GFP) from the jellflsh. Aequorea victoria, would permit the nooinvasive assessment of CD gene trar.sfer efficiency since GFP expression would be easily determined by fluorescence microscopy (Chalfie et al., 1994; Persons et al., 1997).

After transplantation of these CD-transduced marrow cells into recipient mice two major objectives should be accomplished. The fmt is to show chemoprotection from intensive doses of cytosine nucleoside analogs, looking especially at the granulocyte count which when it falls too low, results in life-threatening infections in patients. The second objective would be to demonstrate in mice with tumors that chemotherapy with cytosine nucleoside analogs is significantly superior in the animals transplanted with CD-blood cells, as compared to the control mice. A very important component in these studies is that different dose-schedules of cytosine nucleoside analogs will have to be investigated to determine the optimal conditions to demonstrate both chemoprotection and increase in chemotherapeutic effectiveness.

10.4.3 Clinical studies on chemoprotection for tumor therapy

A clinical study using the CD gene for chemoprotection of cancer patients is outlined in Figure 10.3. Bone marrow or peripheral blood progenitor cells would be harvested from patients before drug treatment. Cells expressing the CD34+ antigen would be purified by an affinity column containing the immobi id antibody to this antigen. These pWed CD34+ hernatopoietic cells will be expanded ex uivo by use of hematopoietic growth factors. A vector system with very high and efficient gene transfer and expression will be used to transduce these hematopcietic cells with the CD gene. If transfer efficiency is low, in uitro selection with cytosine nucleoside analogs could be used to eliminate untransduced cells. The CD-modified blood cells would then be transplanted back into patients after intensive chemotherapy with cytosine nucleoside analogs. Chemoprotection of the hematopoietic cells from the toxic effects of cytosine nucleoside analogs (Figure 10.4) should be functional for subsequent cycles of intensive chemotherapy. If the CD- transduced hematopoietic cells show long-term survival and adequate expression of the transgene, only one transplantation would be required. It is estimated that patients with metastatic cancer would necessitate at least 5 cycles of intensive chemotherapy (Momparler et al.. 1997). In order for this type of therapy to be successful, both the optimal dose-schedule for the cytosine nucleosides analogs and the optimal conditions for expression of the CD transgene in normal blood cells should be used.

CHEMOPROTECTEON

Q j j - 0 +-

HARVEST FICOLL AFFlNlW CDJ* + CELLS COLUMN

1 GROWTH FACTORS

*- --- TRANSDUCE

CD GENE

CHEMOTHERAPY TRANSFUSE 5dZACdR

Figure 10.3 Illustration of a hypothetical clinical protocol for chemoprotection. Bone marrow cells from a patient with metastatic cancer are harvested. The CD34+ cells are purified by centrifugation and affinity chromatography and their growth stimulated with growth factors. Following ex vivo transduction with retroviral particles containing the cytidie deaminase (CD) cDNA, the cells are transfused back into the patient. The patient is then administered high-dose chemotherapy with a cytosine nucleoside analog (e.g., 5-AZA-CdR).

Theoretical objective of chemoprotection against

hematopoietic toxicity of cytosine nucleoside analogs

Time (days)

Figure 10.4 Hypothetical hernogram to illustrate chemoprotection from the hematopoietic toxicity produced by cytosine nucleoside analogs. Gene transfer of cytidine deaminase into normal hernatopoietic cells may have the potential to reduce drug- induced myelosuppression, particularly the granulocytopenia, which is the major dose-limiting toxicity of cytosine nucleoside analogs.

132

REFERENCES

Abratt RP, Bezwoda WR, Falkson G, Goedhals L. Hacking D, and Rugg TA. EEcacy and safety profde of gemcitabiine in non-small-cell lung cancer: a phase I1 study. Journal of Clinical Oncology 12(8): 1535-1540, 1994.

Aksentijevich 1, Cardarelli CO, Pastan 1, and Gottesman MM. Retroviral transfer of the human MDRl gene confers resistance to bisantrese- spec ik hematotoxicity. Clinical Cancer Research 2: 973-980, 1996.

Allay JA, Davis BM, and Gerson SL. Human alkyltransferase-transduced murine myeloid progenitors are enriched in v i m by BCNU treatment of transplanted mice. Experimental Hematology 25: 1069-1076, 1997.

Allay JA, Dumenco LL, Koc ON, Liu L, and Gerson SL. Retroviral transduction and expression of the human alkyltransferase cDNA provides nitrosourea resistance to hematopoietic cells. Blood 85(11): 3342-3351, 1995.

Alt FW, Kellems RE, Bertino JR, and Schimke RT. Selective multiplication of dihydrofolate reductase genes in methotrexate-resistant variants of cultured murine cells. Journal of Biological Chemistry 253(5): 1357- 1370,1978.

Bane j e e D, Schweitzer B1, Volkenandt M, Li MX, Waltham M, Mineishi S, Zhao SC, and Bertino JR. Transfection with a cDNA encoding a Ser31 or Ser34 mutant human dihydrofolate reductase into Chinese hamster ovary and mouse marrow progenitor cells confers methotrexate resistance. Gene 139(2): 269-274, 1994b.

Banejee D, Zhao SC, Li MX, Schweitzer B1, Mineishi S, and Bertino JR. . Gene therapy utilizing drug resistance genes: a review. Stem Cells 12(4):

378-385, 1994a.

Banejee D. Zhao SC, Tong Y. Steinhen J, Gritsman K, and Bertino JR Transfection of a nonactive site mutant murine DHFR cDNA (the tryptophan 15 mutant) into Chinese hamster ovary and mouse marrow

progenitor cells imparts MTX resistance in vitro. Cancer Gene Therapy l(3): 181-184, 1994~.

Barrand MA, Hepell-Parton AC, Wright KA, Rabbitts PH, and Twentyman PR. A 190-kilodalton protein overexpressed in non-P-glycoprotein-

' containing multidrug-resistant cells and its relationship to the MRP gene. Journal of the National Cancer Institute 86(2): 110-1 17, 1994.

Baum C, Hegewisch-Becker S, Eckert HG, Stocking C, and Ostertag W. Novel retroviral vectors for efficient expression of the multidrug resistance (mdr-1) gene in early hematopoietic cells. Journal of Vilogy 69(12): 7541-7547, 1995.

Baum C, Margison GP, Eckert HG, Fairbairn LJ, Ostertag W, and Rafferty JA. Gene transfer to augment the therapeutic index of anticancer chemotherapy. Gene Therapy 3(1): 1-3, 1996.

Bender CM, Pao MM, and Jones PA. Inhibition of DNA methylation by 5- aza-2'-deoxycytidine suppresses the growth of human tumor cell lines. Cancer Research 58(1): 95-101, 1998.

Bertino JR. Toward improved selectivity in cancer chemotherapy: the Richard and Hinda Rosenthal Foundation Award Lecture. Cancer Research 39(2 Pt 1): 293-304, 1979.

Bertino JR. "Turning the Tables" - Making normal marrow resistant to chemotherapy. Journal of the National Cancer Institute 82(15): 1234- 1235,1990.

Bertino JR, Donohue DM, Simmons B, Gabrio BW, Silber R, and Huemekens FM. The "induction" of dihydrofolic reductase activity in leukocytes and erythrocytes of patients treated with amethopterin. Journal of Clinical Investigation 42(4): 466-475, 1963.

Bertolini F, de Monte L, Corsini C, Lazzari L, Lauri E, Soligo D, Ward M, Bank A, and Malavasi F. Retrovirus-mediated transfer of the multidrug

resistance gene into human haemopoietic progenitor cells. British Journal of Haematology 88(2): 318-324, 1994.

Blau CA, Neff T, and Papayannopoulou T. The hematological effects of folate analogs: implications for using the dihydrofolate reductase gene

'

for in vivo selection. Human Gene Therapy 7: 2069-2078, 1996.

Bodensteiner DC and Doolittle GC. Adverse haematological complications of anticancer drugs. Clinical presentation, management and avoidance. Drug Safety 8(3): 213-224, 1993.

Bodie DM. Seidel NE, Gale MS, Nienhuis AW, and Orlic D. Efficient retrovirus transduction of mouse pluripotent hematopoietic stem cells m o b i i d into the peripheral blood by treatment with granulocyte colony stimulating factor and stem cell factor. Blood 84(5): 1482-1491, 1994.

Boesen J J , Nooter K, and Valeno D. Circumvention of chemotherapy- induced myelosuppression by transfer of the mdrl gene. Biotherapy 6(4): 29 1-302, 1993.

Boffetta P and Kaldor JM. Secondary malignancies following cancer chemotherapy. Acta Oncologica 33(6): 59 1-598, 1994.

Bogden JM, Eastman A, and Bresnick E. A system in mouse liver for the repair of 06-methylguanine lesions in methylated DNA. Nucleic Acids Research 9(13): 3089-3103, 1981.

Boris-Lawrie K, and Ternin HM. The retroviral vector. Replication cycle and safety considerations for retrovirus-mediated gene therapy. Annals of the New York Academy of Sciences 716: 59-71, 1994.

Bouffard DY, Lalibertk J , and Momparler RL. Kinetic studies on Z.2'- diiluorodeoxycytidine (gemcitabine) with p d e d human deoxycytidine kinase and cytidine deaminase. Biochemical Pharmacology 45(9): 1857- 1861, 1993.

Braun SE, McIvor RS, Davidson AS, Hanna M. Traycoff CM, Berebetsky DA. Gonin R, Broxmeyer HE, and Cometta K. Retrovirally mediated gene transfer of Arg 22 and Tyr 22 forms of dihydrofoIate reductase into the hematopoietic cell line K562: a comparison of inethotrexate resistance. Cancer Gene Therapy 4(1): 26-32. 1997.

Buller AL, Clapper ML, and Tew KD. Glutathione-S-transferases in nitrogen mustard-resistant and -sensitive cell lines. Molecular Pharmacology 31(6): 575-578, 1987.

Bunting KD, Lindahl R and Townsend AJ. Oxazaphosphorine-specific resistance in human MCF-7 breast carcinoma cell lines expressing transfected rat class 3 aldehyde dehydrogenase. Journal of Biological Chemistry 269(37): 23197-23203, 1994.

Bunting KD, Webb M, Giorgianni F, Galipeau J , Blakley RL. Townsend AJ, and Somentino BP. Coding region-specific destabii t ion of mRNA transcripts attenuates expression from retroviral vectors containing class 1 aldehyde dehydrogenase cDNAs. Human Gene Therapy 8(13): 1531-1543, 1997.

Burris HA, Moore MJ, Andersen J , Green MR. Rothenberg ML, Modiano MR, Cripps MC, Portenoy RK, Storniolo AM, Tarassoff P, Nelson R, Dorr FA, Stephens CD, Von Hoff DD. Improvements in survival and clinical benefit with gemcitabiie as first-lime therapy for patients with advanced pancreas cancer: a randomized trial. Journal of Clinical Oncology 15(6): 2403-2413, 1997.

Cacciamani T, Vita A, Cristalli G, Vicenzetti S, Natalini P, Ruggieri S, Arnici A, and Magni G. Purification of human cytidine deaminase: molecular and enzymatic characterization and inhibition by synthetic pyrimidine analogs. Archives of Biochemistry and Biophysics 290(2): 285-292, 1991.

Camiener GW. Studies of the enzymatic deamination of aracytidine: V. Inhibition in vitru and in viuo by tetrahydrouridine and other related

pyrimidine nucleosides. Biochemical Pharmacology 17: 198 1- 199 1, 1968.

Camiener GW and Smith CG. Studies of the enzymatic deamination of cytosine arabiioside. I. Enzyme distribution and species specificity. Biochemical Pharmacology 14(10): 1405-1416, 1965.

Capizzi RL, Yang JL., Cheng E, Bjornsson T, Sahasrabudhe D, Tan RS, and Cheng YC. Alteration of the pharmacokinetics of high-dose Ara-C by its metabolite, high Ara-U in patients with acute leukemia. Journal of Clinical Oncology l(12): 763-771, 1983.

Cannichael J and Walling J. Advanced breast cancer: investigational role of gemcitabine. European Journal of Cancer 33(Suppl 1): S27-30, 1997.

Carter CW Jr. The nucleoside deaminases for cytidine and adenosine: structure, transition state stabilkation, mechanism, and evolution. Biochimie 77: 92-98, 1995.

Casper ES, Green MR, Kelsen DP, Heelan RT, Brown TD, Flombaum CD, Trochanowski B, and Tarassoff PG. Phase I1 trial of gemcitabine (2',2'- diiuorodeoxycytidine) in patients with adenocarcinoma of the pancreas. Investigational New Drugs 12(1): 29-34, 1994.

Catimel G, Vermorken JB, Clavcl M, De Mulder P, Judson 1, Sessa C, Piccart M, Bruntsch U, Verweij J , Wanders J, Franklin H, and Kaye SB. A phase I1 study of Gemcitabiie (LY188011) in patients with advanced squamous cell carcinoma of the head and neck. Annals of Oncology 5(6): 543-547, 1994.

Chabner BA. Cytidine Analogues. In: Cancer Chemotherapy and Biotherapy, second edition, edited by Bruce A. Chabner and Dan L. Longo. Lippincott-Raven Publishers, Philadelphia 2 13-233, 1996.

Chabner BA, Johns DG, Coleman CN, Drake JC, and Evans WH. Purification and properties of cytidine dcaminase from normal and

leukemic granulocytes. The Journal of Clinical Investigation 53(3): 922- 931, 1974.

Chabot GG. Bouchard J, and Momparler RL. Kinetics of deamination of 5- aza-2'- deoxycytidine and cytosine arabiioside by human liver cytidine deaminase and its inhibition by 3-deazauridine, thymidine or uracil arabiioside. Biochemical Pharmacology 32(7): 1327-1328, 1983.

C h ~ ~ e M, Tu Y. Euskirchen G, Ward WW, and Prasher DC. Green fluorescent protein a s a marker for gene expression. Science 263(5148): 802-805. 1994.

Challita PM and Kohn DB. Lack of expression from a retroviral vector after transduction of murine hematopoietic stem cells is associated with methylation in uivo. Proceedings of the National Academy of Sciences of the United States of America 91(7): 2567-2571. 1994.

Chaudhary PM, Mechetner EB, and Roninson IB. Expression and activity of the multidrug resistance P-glycoprotein in human peripheral blood lymphocytes. Blood 80(11): 2735-2739, 1992.

Chaudhary PM, and Roninson 1B. Expression and activity of P- glycoprotein, a multidmg eftlux pump, in human hematopoietic stem cells. Cell 66(1): 85-94, 1991.

Cheng YC, Tan RS, Ruth JL, and Dutschman G. Cytotoxicity of 2'-fluoro-5- iodo-1-6-D-arabiofuranosylcytosine and its relationship to deoxycytidine deaminase. Biochemical Pharmacology 32(4): 726-729, 1983.

Chinnasamy N, Rafferty JA, Hickson I, Lashford LS, Longhurst SJ, Thatcher N, Margison GP, Dexter TM, and Fairbairn LJ. Chemoprotective gene transfer Ii: multilineage in vivo protection of haemopoiesis against the effects of a n antitumor agent by expression of a mutant human 06-alkylguanine-DNA alkyltransferase. Gene Therapy 5: 842-847, 1998.

138

Chou TC, Arlin Z. Clarkson BD, and Philips FS. Metabolism of 1-i3-D- arabiiofuranosylcytosine in human leukemic cells. Cancer Research 37: 3561-3570. 1977.

Cline MJ, Stang H, Mercola K, Morse L, Ruprecht R. Brown J , Salser W. Gene transfer in intact animals. Nature 284(5755): 422425, 1980.

Cole SP, Sparks KE, Fraser K, Loe DW, Grant CE, Wilson GM, and Declcy RG. Pharmacological characterization of multidrug resistant MRP- transfected human tumor cells. Cancer Research 54(22): 5902-5910, 1994.

Coleman CN, Stoller RG, Drake JC, and Chabner BA. Deoxycytidine kinase: properties of the enzyme from human leukemic granulocytes. Blood 46(5): 791-803, 1975.

Coles B and Ketterer B. The role of glutathione and glutathione transferases in chemical carcinogenesis. Critical Reviews in Biochemistry and Molecular Biology 25(1): 47-70, 1990.

Corey CA, DeSilva AD, Holland CA, and Williams DA. Serial transplantation of methotrexate-resistant bone marrow: protection of murine recipients from drug toxicity by progeny of transduced stem cells. Blood 75(2): 337-343, 1990.

Cormier Y, Eisenhauer E, Muldal A, Greg R, Ayoub J, Goss G, Stewart D, Tarasoff P, and Wong D. Gemcitabiie is an active new agent in previously untreated extensive small cell lung cancer (SCLC). A study of the National Cancer Institute of Canada Clinical Trials Group. Annals of Oncology 5(3): 283-285, 1994.

Costello JF, Berger MS, Huang HS, and Cavenee WK. Silencing of p16/CDKN2 expression in human gliomas by methylation and chromatin condensation. Cancer Research 56(10): 2405-2410, 1996.

Cbtk S, and Momparler RL. Activation of the retinoic acid receptor beta gene by 5-aza-2'-deoxycytidine in human DLD-1 colon carcinoma cells. Anti-Cancer Drugs 8(1): 56-61. 1997.

CBtk S, and Momparler RL. Antineoplastic action of all-trans retinoic acid '

and 5-aza-2'-deoxycytidine on human DLD-1 colon carcinoma cells. Cellular Pharmacology 2: 221-228, 1995.

Crawford J , Ozer H. Stoller R. Johnson D, Lyman G, Tabbara I, Kris M. Grous J , Picozzi V, Rausch G. Smith R, Gradishar W, Yahanda A, Vincent M, Stewart M, and Glaspy J. Reduction by granulocyte colony- stimulating factor of fever and neutropenia induced by chemotherapy in patients with small-cell lung cancer. New England Journal of Medicine 325(3):164-170, 1991.

Creusot F, Acs G, and Christrnan JK. Inhibition of DNA methyltransferase and induction of Friend erythroleukemia cell differentiation by 5- azacytidine and 5-aza-2'-deoxycytidine. Journal of Biological Chemistry 257(4): 204 1-2048, 1982.

Crone TM, Goodtzova K, Edara S, and Pegg AE. Mutations in human 0 6 -

alkylguanine-DNA allcyltransferase imparting resistance to 0 6 -

benzylguanine. Cancer Research 54(23): 6221-6227, 1994.

Crone TM and Pegg AE. A single amino acid change in human 0 6 -

alkylguanine-DNA alkyltransferase decreasing sensitivity to inactivation by 06-benzylguanine. Cancer Research 53(20): 4750-4753, 1993.

Czaykowski PM, Samuels T, and Oza A. A durable response to cytarabine in advanced breast cancer. Clinical Oncology (Royal College of Radiologists) 9(3): 181-183, 1997.

D'Hondt V, Caruso M, and Bank A. Retrovirus-mediated gene transfer of the multidrug resistance-associated protein (MRP) cDNA protects cells from chemotherapeutic agents. Human Gene Therapy 8(15): 1745-1751, 1997.

Davis BM, Reese JS, KO$ ON, Lee K. Schupp JE, and Gerson SL. Selection for G156A 06-methylguanine-DNA methylmnsferase gene-transduced hematopoietic progenitors and protection from lethality in mice treated with Ob-benzylguanine and 1,3-bis (2-chloroethy1)-l-nitrcsourea. Cancer Research 57(22): 5093-5099, 1997.

Deisseroth AB. Holmes F, Hortobagyi G, and Champk R. Use of safety- modified retroviruses to introduce chemotherapy resistance sequences into normal hematopoietic cells for chemoprotection during the therapy of breast cancer: a pilot trial. Human Gene Therapy 7(3): 401416, 1996.

Deisseroth AB, Kavanagh J , and Champiin R. Use of safety-modified retroviruses to introduce chemotherapy resistance sequences into normal hematopoietic cells for chemoprotection during the therapy of ovarian cancer: a pilot trial. Human Gene Therapy 5(12): 1507-1522, 1994.

Deisseroth AB, and Pizzomo G. The use of chemotherapy resistance in cancer treatment. The Cancer Journal from Scien=c American 60-69, 1997.

DelaFlor-Weiss E, Richardson C, Ward M, Himelstein A, Smith L, Podda S, Gottesman M, Pastan I, and Bank A. Transfer and expression of the human multidrug resistance gene in mouse erythroleukemia cells. Blood 80(12): 3106-311 1, 1992.

Devereux S. Corney C, Macdonald C, Watts M, Sullivan A, Goldstone AH, Ward M, Bank A, and Lich DC. Feasibility of multidrug resistance (MDR-1) gene transfer in patients undergoing high-dose therapy and peripheral blood stem cell transplantation for lymphoma. Gene Therapy 5(3): 403408, 1998.

Dexter TM and Spooncer E. Growth and differentiation in the hemopoietic system. Annual Review of Cell Biology 3: 423441, 1987.

Dicker AP. Volkenandt M, Schweitzcr B1, Banerjee D, and Bertino J R Identification and characterization of a mutation in the dihydrofolate reductase gene from the methofmxate-resistant Chinese hamster ovary cell line Pro3 Mtx RXI. Journal of Biological Chemistry 265(14): 8317- 8321,1990.

Dolan M, Moschel R, and Grindcy G. Depletion of mammalion 06 alkylguanine-DNA alkyltransferase activity by 06-benzylguanhe provides a means to evaluate the role of this protein in protection against carcinogenic and therapeutic aUcylating agents. Proceedings of the National Academy of Sciences USA 87: 5368-5372.1990.

Donahue RE, Kessler SW, Bodhe D, McDonagh K, Dunbar C, Goodman S, Agricola B, Byme E, Raffeld M, Moen R, Bacher J, Zsebo KM, and Nienhuis AW. Helper virus induced T cell lymphoma in nonhuman primates after retroviral mediated gene transfer. The Journal of Experimental Medicine 176(4): 1 125-1 135, 1992.

Doroshow JH, Metz MZ. Matsumoto L, Winters KA, Sakai M, Muramatsu M, and Kane SE. Transduction of NIH 3T3 cells with a retrovirus c@g both human MDRl and glutathione Stransferase - produces broad-range multidrug resistance. Cancer Research 55: 4073-4078, 1995.

Driscoll JS, Marquez VE, Plowman J , Liu PS, Kelly JA, and Barchi JJ Jr. Antitumor properties of 2(lH)-pyrimidiione riboside (zebularine) and its fluorinated analogues. Journal of Medicinal Chemistry 34(11):3280- 3284.1991.

Eckert HG. Stockschl2ider M, Just U, Hegewisch-Becker S, Grez M. Uhde A, Zander A. Ostertag W, and Baum C. High-dose multidrug resistance in primary human hematopoietic progenitor cells transduced with optimized retrovial vectors. Blood 88(9): 3407-3415, 1996.

Fantz CR, Shaw D, Moore JG, and Spencer HT. Retroviral coexpression of thymidylate synthase and dihydrofolate reductase confers

fluoropyrimidine and antifolate resistance. Biochemical and Biophysical Research Communications 243(1): 6-12. 1998.

Flasshove M, Bane jee D, Bertino JR, and Moore MA. Increased resistance to methotrexate in human hematopoietic cells after gene transfer of the Ser31 DHFR mutant Leukemia 9(Suppll): S34-37, 1995a.

Flasshove M, Banejee D, Leonard JP, Mineishi S, Li MX, Bertino JR, and Moore MA. Retroviral transduction of human CD34+ umbiical cord blood progenitor cells with a mutated diiydrofolate reductase cDNA. Human Gene Therapy 9(1): 63-71, 1998.

Flasshove M, Banejee D. Mineishi S, Li MX, Bertino JR, and Moore MA. Ex vivo expansion and selection of human CD34+ peripheral blood progenitor cells after introduction of a mutated dihydrofolate reductase cDNA via retroviral gene transfer. Blood 85(2): 566-574, 1995b.

Flintoff WF, Davidson SV, and Siminovitch L. Isolation and partial characterization of three methotrutate-resistant phenotypes from Chinese hamster ovary cells. Somatic Cell Genetics 2(3): 245-261, 1976.

Fojo AT, Ueda K, Slamon DJ, Poplack DG, Gottesman MM, and Pastan I. Expression of a multidrug-resistance gene in human tumors and tissues. Proceedings of the National Academy of Sciences of the United States of America 84(1): 265-269. 1987.

Fossella N, Lippman SM, Sh'i DM, Tarassoff P, CalayagJung M, Perez- Soler R, Lee JS, Murphy WK, Glisson B, Rivera E, and Hong WK. Maximum-tolerated dose defied for single-agent gemcitabiie: a phase I dose-escalation study in chemotherapy-naive patients with advanced non-small-cell lung cancer Journal of Clinical Oncology 15(1): 310-316, 1997.

Friedrnann T. Overcoming the obstacles to gene therapy. Scientific American June: 96-101, 1997.

Fruehauf S, Boesen JJ , Breems DA, Hoft R, Hundsdorfer P, Zeller WJ. Lowenberg B, Ploemacher RE, Haas R, and Valerio D. Retroviral transfer of the multidrug resistance-1 gene into lineage-committed and primitive hemopoietic cells. Stem Cells 13 (Suppl3): 93-99, 1995.

Furth JJ and Cohen SS. Inhibition of mammalian DNA polymerase by the 5'-triphosphate of 1-beta-d-arabmofuranosy1cytosine and the 5'- &phosphate of 9-beta-d-arabinofuranoxyladenine. Cancer Research 28(10): 2061-2067, 1968.

Galipeau J , Benaim E, Spencer HT, Blakley RL, and Sorrentino BP. A bitistronic retroviral vector for protecting hematopoietic cells against antifolates and pglycoprotein emuxed drugs. Human Gene Therapy 8: 1773-1783, 1997.

Gerson SL, Miller K, and Berger NA. 06-alkylguanine-DNA allcyltransferase activity in human myeloid cells. Journal of Clinical Investigation 76(6): 2106-21 14, 1985.

Gianni AM, Bregni M, Siena S, Villa S, Sciorelli GA, Ravagnani F, Pellegris G, and Bonadonna G. Rapid and complete hemopoietic reconstitution following combined transplantation of autologous blood and bone marrow cells. A changing role for high dose chemo-radiotherapy? Hematological Oncology 7(2): 139-148, 1989.

Gonzalgo ML, Hayashida T. Bender CM, Pao MM, Tsai YC, Gonzales FA, Nguyen HD, Nguyen TT, and Jones PA. The role of DNA methylation in expression of the p19/p16 locus in human bladder cancer cell lines. Cancer Research 58(6): 1245-1252, 1998.

Graff JR, Herman JG, Lapidus RG, Chopra H, Xu R, Jarrard DF, Isaacs WB, Pitha PM, Davidson NE, and Baylin SB. E-cadherin expression is silenced by DNA hypermethylation in human breast and prostate carcinomas. Cancer Research 55(22): 5195-5199, 1995.

Graham FL and Whitmore GF. Studies in mouse Lcells on the incorporation of 1-beta-D-arabiinofuranosylcytosine into DNA and on

inhibiton of DNA polymerase by 1-beta-D-arabiiofuranosylcytosine 5'- triphosphate. Cancer Research 30(11): 2636-2644, 1970.

Gran C. B~yum A, Johansen RF, hvhaug D, and Seeberg EC. Growth inhibition of granulocyte-macrophage colony-forming cells by human cytidine deaminase requires the catalytic function of the protein. Blood gl(l1): 4127-4135, 1998.

Grant CE, Valdimarsson G, Hipfner DR, Almquist KC, Cole SP, and Deeley RG. Overexpression of multidrug resistance-associated protein (MRP) increases resistance to natural product drugs. Cancer Research 54(2): 357-361, 1994.

Grant S. Ara-C: Cellular and molecular pharmacology. Advances in Cancer Research 72: 197-233, 1998.

Greenbaum M, Etourneau S, Assar H, Schecter RL, Batist G, and Cournoyer D. Retrovirus-mediated gene transfer of rat glutathione S- trasferase Yc confers akylating drug resistance in NIH 3T3 mouse fibroblasts. Cancer Research 54: 4442-4447, 1994.

Guchelaar HJ, Richel DJ, and van Knapen A. Clinical, toxicological and pharmacological aspects of gemcitabiie. Cancer Treatment Reviews 22: 15-31, 1996.

Hanania EG, and Deisseroth AB. Serial transplantation shows that early hernatopoietic precursor cells are transduced by MDR-1 retroviral vector in a mouse gene therapy model. Cancer Gene Therapy l(1): 21-25, 1994.

Hanania EG, Fu S, Roninson 1, Zu Z, Gottesman MM, and Deisseroth AB. Resistance to tax01 chemotherapy produced in mouse marrow cells by safety-modiiied retroviruses containing a human MDR-1 transcription unit. Gene Therapy 2(4): 279-284. 1995 b.

Hanania EG, Fu S, Zu Z, Hegewisch-Becker S, Korbling M, Hester J , Durett A, Andreeff M, Mechetner E. Holzmayer T, Roninson IB, Giles

RE. Berenson R, Heimfeld S, and Deisseroth AB. Chemotherapy resistance to tax01 in clonogenic progenitor cells following transduction of CD34 selected marrow and peripheral blood cells with a retrovirus that contains the MDR-1 chemotherapy resistance gene. Gene Therapy 2(4): 285-294, 1995.

Hanania EG, Giles RE, Kavanagh J , Ellerson D, Zu 2, Wang T, Su Y, Kudelka A, Rahman Z. Holmes F. Hortobagyi G, Claxton D, Bachier C. Thall P, Cheng S, Hester J. Ostrove JM, Bird RE, Chang A, Korbling M, Seong D, Cote R, Holzmayer T, Met-tetner E, Heimfeld S, Berenson R, Burtness B, Edwards C. Bast k, Andreeff M, Champlin R, and Deisseroth AB. Results of MDR-1 vector modification trial indicate that granulocyte/macrophage colony-forming unit cells do not contribute to posttransplant hematopoietic recovery following intensive systemic therapy. Proceedings of the National Academy of Sciences of the United States of America 93(26): 15346-15351, 1996.

Hegewisch-Becker S, Hanania EG. Fu S, Korbling M, Deisseroth AB, and Andreeff M. Transduction of MDRl into human and mouse haemopoietic progenitor cells: use of rhodamine (Rh123) to determine transduction frequency and in vivo selection. British Journal of Haematology 90(4): 876-883, 1995.

Heinemann V, Hertel LW, Grindey GB, and Plunkett W. Comparison of the cellular pharmacokinetics and toxicity of 2',2'-difluorodeoxycytidine and 1-beta-D- arabiiofuranosylcytosine. Cancer Research 48(14): 4024- 4031. 1988.

Heinemann V, Xu YZ, Chubb S, Sen A, Hertel LW, Grindey GB, and Plunkett W. Inhibition of ribonucleotide reduction in CCRF-CEM cells by 2.2'-difluorodeoxycytidine. Molecular Pharmacoiogy 38(4): 567-572. 1990.

Herman JG, Latif F. Weng Y. Lerman MI, Zbar B, Liu S, Samid D. Duan DS, Gnarra JR, Linehan WM, and Baylin SB. Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma.

Proceedings of the National Academy of Sciences of the United States of America 91(21): 9700-9704, 1994.

Hertel LW, Kroin JS, Misner JW, and Tustin JM. Synthesis of 2-deoxy-2.2- difluoro-D-ribose and 2-deoxy-2,2-diiuoro-D-ribofuranosyl nucleosides. Journal of Organic Chemistry 53: 2406-2409.1988.

Hesdorffer C, Antman 8, Bank A, Fetell M, Mears G, and Begg M. Human MDR gene transfer in patients with advanced cancer. Human Gene Therapy 5(9): 1151-1 160,1994.

Hesdorffer C, Ayello J , Ward M, Kaubisch A, Vahdat L, Balrnaceda C, Garrett T, Fetell M, Reiss R, Bank A, and Antman K. Phase I trial of retroviral-mediated transfer of the human MDRl gene a s marrow chemoprotection in patients undergoing high-dose chemotherapy and autologous stem-cell transplantation. Journal of Clinical Oncology 16(1): 165-172, 1998.

Hickson I, Fairbairn W, Chinnasamy N. Lashford LS, Thatcher N, Margison GP, Dexter TM, and Rafferty JA. Chemoprotective gene transfer I: transduction of human haemopoietic progenitors with 0 6 -

benzylguanine-resistant 06-alkylguanine-DNA alkyltransferase attenuates the toxic effects of 06-allcylating agents in vitro. Gene Therapy 5: 835-841,1998.

Ho DHW. Distribution of b a s e and deaminase of I-iS-D- arabiiofuranosylcytosine in tissues of man and mouse. Cancer Research 33: 2816-2820, 1973.

Ho DHW, Carter CJ, Brown NS, Hester J , McCredie K, Benjamin RS, Freireich EJ, and Bodey GP. Effects of tetrahydrouridine on the uptake and metabolism of l-f3-D-arabiofuranosy1cytosine in human normal and leukemic cells. Cancer Research 40(7): 2441-2446, 1980.

Hock RA and Miller AD. Retrovirus-mediated transfer and expression of drug resistance genes in human haematopoietic progenitor cells. Nature 320I6059): 275-277, 1986.

Huang P, Chubb S, Hertel LW, Grindey GB, and Plunkett W. Action of 2',2'-difluorodeoxycytidine on DNA synthesis. Cancer Research Sl(22): 61 10-6117, 1991.

Hussain A. Lewis D, Yu MF, and Melera PW. Construction of a dominant selectable marker using a novel dihydrofolate reductase. Gene 112(2): 179-188. 1992.

Huynh H, Alpert L, and Pollack M. Silencing of the mammary-derived growth inhibitor (MDGI) gene in breast neoplasms is associated with epigenetic changes. Cancer Research 56(21): 4865-4870, 1996.

Jelinek J , Fairbairn W . Dexter TM, Rafferty JA, Stocking C, Ostertag W, and Margison GP. Long-term protection of haematopoiesis against the cytotoxic effects of multiple doses of nitosourea by retrovirus-mediated expression of human 06-alkylguanine-DNA alkyltransferase. Blood 87(5): 1957-1961, 1996.

Jones PA. DNA methylation errors and cancer. Cancer Research 56(11): 2463-2467, 1996.

Jones PA, and Taylor SM. Cellular differentiation, cytidine analogs and DNA methylation. Cell 20(1): 85-93, 1980.

Kantajian H, Barlogie B, Plunkett W, Velasqun W, McLaughlin P, Riggs S, and Cabanillas F. High-dose cytosine arabiioside in non-Hodgkin's lymphoma. Journal of Clinical Oncology l(11): 689-694, 1983.

Karon M, and Shirakawa S. The locus of action of 1-beta-D- arabimofuranosylcytosine in the cell cycle. Cancer Research 29(3): 687- 696, 1969.

Kastan MB, Schlaffer E, Russo JE, Colvin OM, Civin CI, and Hilton J. Direct demonstration of elevated aldehyde dehydrogenase in human hematopoietic progenitor cells. Blood 75(10): 1947-1950, 1990.

Kaufmann M and von Minckwitz G. Gemcitabine in ovarian cancer: an overview of saiety and efficacy. European Journal of Cancer 33(Suppl 1): S31-33, 1997.

Keating MJ, McCredie KB, Bodey GP. Smith TL. Gehan E, and Freireich EJ. Improved prospects for long-term survival in adults with acute myelogenous leukemia. Journal of the American Medical Association 248(19): 2481-2486. 1982.

Kern DH, Morgan CR, and Hildebrand-Zanki SU. In vitro pharmacodynamics of 1-beta-D-arabinof~lilllosylcytosine: synergy of antitumor activity with cis-diamminedichloroplatinum (11). Cancer Research 48(1): 117-121, 1988.

Kessel D. Circumvention of resistance to anthracyclines by calcium antagonists and other membrane-perturbing agents. Cancer Surveys 5(1): 109-127, 1986.

Kim CH, Marquez VE, Mao DT, Haines DR, and McConnack JJ. Synthesis of pyrimidin-2-one nucleosides as acid-stable inhibitors of cytidine deaminase. Journal of Medicinal Chemistry 29(8): 1374-1380, 1986.

Kirch HC, Schr6der J , Hoppe H, Esche H, Seeber S, and Schutte J. Recombinant gene products of two natural variants of the human cytidine deaminase gene confer different deamination rates of cytarabine in vitro. Experimental Hematology 26(5): 421-425, 1998.

KOC ON, Allay JA. Keunmyoung L, Davis BM, Reese JS, and Gerson SL. Transfer of drug resistance genes into hematopoietic progenitors to improve chemotherapy tolerance. Seminars in Oncology 23(1): 46-65, 1996.

Krall WJ, Skelton DC, Yu XJ, Riviere I, Lehn P, Mulligan RC, and Kohn DB. Increased levels of spliced RNA account for augmented expression from the MFG retroviral vector in hematopoietic cells. Gene Therapy 3(1]: 3748, 1996.

Kreis W, Budman DR. Chan K, Allen SL, Schulman P. Lichtman S, Weiselberg L, Schuster M, Freeman J. Akerman S, Attas L, and Vindguerra V. Therapy of refractory/relapsed acute leukemia with cytosine arabinoside plus tetrahydrouridine (an inhibitor of cytidine deaminase) - a pilot study. Leukemia 5(11): 991-998, 1991.

Kreis W, Chan K. Budman DR. Schulman P. Allen S, Weiselberg L, Lichtman S, Henderson V, Freema? 9, Deere M, Andmeff M, and Vinciguerra V. Effect of tetrahydrouridine on the clinical pharmacology of 1-beta-D-arabiofuranosy1Cytosine when both drugs are coinfused over three hours. Cancer Research 48(5): 1337-1342, 1988.

Kreis W, Lesser M, Budman DR, Arlin Z, DeAngelis L. Baskind P. Feldman EJ, and Akerman S. Phenotypic analysis of 1-B-D- arabinofuranosylcytosine deamination in patients treated with high doses and correlation with response. Cancer Chemotherapy and Pharmacology 30: 126-130, 1992.

KChn K, Bertling WM, and Ernmrich F. Cloning of a functional cDNA for human cytidie deaminase (CDD) and its use as a marker of monocyte/macrophage differentiation. Biochemical and Biophysical Research Communications 190(1): 1-7, 1993.

Kwok WW, Schuening F, Stead RB, and Miller AD. Retroviral transfer of genes into canine hemopoietic progenitor cells in culture: a model for human gene therapy. Proceedings of the National Academy of Sciences of the United States of America 83(12): 45524555, 1986.

Laliberte J, Marqua VE, and Momparler RL. Potent inhibitors for the deamination of cytosine arabinoside and 5-aza-2'-deoxycytidine by human cytidine deaminase. Cancer Chemotherapy and Pharmacology 30(1): 7-1 1, 1992.

Laliberte J, and Momparler RL. Human cytidine deaminase: purification of enzyme, cloning, and expression of its complementary DNA. Cancer Research 54(20): 5401-5407, 1994.

Litourneau S, Greenbaum M, and Cournoyer D. Retrovirus-mediated gene transfer of rat glutathione S-transferase Yc confers in vitro resistance to alkylating agents in human leukemia cells and in clonogenic mouse hematopoietic progenitor cells. Human Gene Tnerapy 7: 831-840, 1996.

Lewis AD. Hickson 1D. Robson CN, Hanis AL, Hayes JD. Griffiths SA, Manson MM, Hall AE, Moss JE, and Wolf CR. Amplification and increased expression of alpha class glutathione-S-transferase-encoding genes associated with resistance to nitrogen mustards. Proceedings of the National Academy of Sciences of the United States of America 85(22): 851 1-8515, 1988.

Li MX, Banerjee D, Zhao SC, Schweitzer B1, Mineishi S, Giboa E, and Bertino JR. Development of a retroviral construct containing a human mutated dihydrofolate reductase cDNA for hematopoietic stem cell transduction. Blood 83(11): 3403-3408, 1994.

Licht T. Aksentijevich 1, Gottesman MM, and Pastan I. Efficient expression of functional human MDRl gene in murine bone marrow after retroviral transduction of purified hematopoietic stem cells. Blood 86(1): 111- 121, 1995b.

Licht T, Aran JM, Goldenberg SK, Vieira WD, Gottesman MM, and Pastan I. Cytotoxic drug selection of murine bone marrow cells following transfer of the MDRl (multidrug resistance) gene increases gene expression and chemoresistance in vivo. Blood 86(Suppll): 243, 1995a

Licht T, Herrmann F, Gottesman MM, and Pastan 1. In vivo drug-selectable genes: a new concept in gene therapy. Stem Cells 15(2):104-111, 1997.

Lie SO and Slerdahl S. High-dose cytosine arabiioside in the treatment of childhood malignancies. Seminars in Oncology 12(2 Suppl 3): 160-165, 1985.

Lin KT. Momparler RL, and Rivard GE. High-performance liquid chromatographic analysis of chemical stability of 5-aza-2'-

151

deoxycytidine. Journal of Pharmaceutical Sciences 70(11): 1228-1232, 1981.

Linkesch W. Kuhrer 1, and Wagner A. Rhu-GM-CSF after ultrahigh dose carboplatin. VP-16, cyclophosphamide with ABMT in refractory germ

'

cell cancer. Roceedings of the American Society of Clinical Oncology 9:141, 1990.

Littlefield JW. Hybridization of hamster cells with high and low folate reductase activity. Proceedings of the National Academy of Sciences of the United States of America 62(1): 88-95, 1969.

Liu PS, Marque2 VE, Driscoll JS, Fuller RW. McCormack JJ. Cyclic urea nucleosides. Cytidine deaminase activity a s a function of aglycon ring size. Journal of Medicinal Chemistry 24(6): 662-666. 1981.

Ludwig H, Fritz E, Kotzrnam H, Hocker P, Gisslinger H, and Barnas U. Exythropoietin treatment for anemia associated with multiple myeloma. New England Journal of Medicine 322(24): 1693-1699, 1990.

Ludwig H, Fritz E, Leitgeb C, Pecherstorfer M, Samonigg H, and Schuster J. Prediction of response to erythropoietin treatment in chronic anemia of cancer. Blood 84(4): 1056-1063, 1994.

Lund B, Hansen OP, Theilade K, Hansen M, and Neijt JP. Phase I1 study of gemcitabiie (T.2'-dfiuorodeoxycytid'me) in previously treated ovarian cancer patients. Journal of the National Cancer lnstitute 86(20): 1530- 1533, 1994.

Luskey BD, Rosenblatt M, Zsebo K, and Williams DA. Stem cell factor, interleukin-3, and interleukind promote retroviral-mediated gene transfer into murine hematopoietic stem cells. Blood 80(2'.. 796402, 1992.

Magni M, Shammah S, Schiro R, Mellado W, Dalla-Favera R, and Gianni AM. Induction of cyclophosphamide-resistance by aldehyde- dehydrogenase gene transfer. Blood 87(3): 1097-1 103. 1996.

Markowitz D, Goff S, and Bank A. A safe packaging line for gene transfer separating viral genes on two different plasmids. Journal of Vilogy 62: 1120-1 124, 1988a

Markowitz D, Goff S, and Bank A. Construction and use of a safe and eficient amphotropic packaging cell line. Virology 167: 400406, 1988b.

Marquez VE, Liu PS, Kelley JA. Driscoll JS, McCormack JJ. Synthesis of 1,3-diazepin-2-one nudeosides as transition-state inhibitors of cylidine deaminase. Journal of Medicinal Chemistry 23(7): 713-775, 1980.

Maze R, Carney JP, Kelly MR, Glassner BJ, W i a m s DA, and Samson L. Increasing DNA repair methyltransferase levels via bone marrow stem cell transduction rescues mice from the toxic effects of 1,3-bis(2- chloroethy1)-1-nitrosourea, a chemotherapeutic alkylating agent Proceedings of the National Academy of Sciences of the United States of America 93(1): 206-210, 1996.

Maze R, Carney JP, Moritz T, Kelley MR, Mackay W, Samson L, and W i a m s DA. Nitrosourea (Nu)-induced hematopoietic stem cell (HSC) damage and reduction of delayed myelosuppression in vivo using a 0 6 - methylguanine methyltransferase (MGMT) retroviral vector. Experimental Hematology 22: 681, 1994.

McCormack J J , Marquez VE, Liu PS, Vistical DT, and Driscoll JS. Inhibition of cytidine deaminase by 2-oxopyrimidine riboside and related compounds. Biochemical Pharmacology 29: 830-832, 1980.

McIvor RS and Simonsen CC. Isolation and characterization of a variant diiydrofolate reductase cDNA from methotrexate-resistant murine W178Y cells. Nucleic Acids Research 18(23): 7025-7032, 1990.

McLachlin JR, Eglitis MA, Ueda K, Kantoff PW, Pastan IH, Anderson WF, and Gottesman MM. Expression of a human complementary DNA for the multidrug resistance gene in murine hernatopoietic precursor cells

with the use of retroviral gene transfer. Journal of the National Cancer Institute 82(15): 1260-1263, 1990.

Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, Baylin SB. Sidransky D. SCpG island methylation is associated with transcriptional silencing of the tumor suppressor pl6/CDKN2/MTSl in human cancers. Nature Medicine l(7): 686-692. 1995.

Meropol NJ, Miller LL, Kom EL, Braitman LE, MacDemott ML, Schuchter LM. Severe myelosuppression resulting from concurrent administration of granulocyte colony-stimulating factor and cytotoxic chemotherapy. Journal of the National Cancer Institute 84(15]: 1201-1203, 1992.

Michon J. Bouffet E, Bernard J, et al. Administration of recombinant human GM-CSF after autologous bone marrow transplantation: A study of 21 stage IV neuroblastoma patients undergoing double intensification regimen. Proceedings of the American Society of Clinical Oncology 9:184, 1990.

Mickisch GH and Schroeder FH. From laboratory expertise to clinical practice: multidrug-resistance-based gene therapy becomes a d a b l e for urologists. World Journal of Urology 12: 104-11 1, 1994.

Mikita T and Beardsley GP. Functional consequences of the arabinosylcytosine structural lesion in DNA. Biochemistry 27(13): 4698- 4705. 1988.

Momparler RL. Effect of cytosine arabinoside 5'-triphosphate on mammalian DNA polymerase. Biochemical and Biophysical Research Communications 34(4): 464-471, 1969.

Momparler RL. Molecular, cellular and animal pharmacology of 5-aza-2'- deoxycytidine. Pharmacology and Therapeutics 30(3): 287-299, 1985.

Momparler RL, Bouchard J, and Samson J. Induction of differentiation and inhibition of DNA methylation in HL-60 myeloid leukemic cells by 5-AZA-2'- deoxycytidine. Leukemia Research 9(11): 1361-1366. 1985a.

Momparler RL, BouDlard DY, Momparler LF, Diome J , Bklanger K, and Ayoub J. Pilot phase 1-11 study on 5-aza-2'-deoxycytidine (Decitabiie) in patients with metastatic lung cancer. Anti- Cancer Drugs 8(4): 358- 368, 1997.

'

Momparler RL, and Fischer GA: Mammalian deoxynudeoside kinase. I. Deoxycytidine kinase: p ~ r ~ c a t i o n , properties, and kinetic studies with cytosine arabiioside. Journal of Biological Chemistry 243(16): 4298- 4304, 1968.

Momparler RL, and Laliberte J. Induction of cytidine deamhase in HL60 myeloid leukemic cells by 5-aza-2'-deoxycytidine. Leukemia Research 14(9): 751-754, 1990.

Momparler RL, Rivard GE, and Gyger M. Clinical trial on 5-aza-2'- deoxycytidine in patients with acute leukemia. Pharmacology and Therapeutics 30(3): 277-286, 1985b.

Momparler RL, Samson J , Momparler LF, and Rivard GE. Cell cycle effects and cellular pharmacology of 5-am-2'-deoxycytidine. Cancer Chemotherapy and Pharmacology 13(3): 191-194, 1984.

Moreb JS, Schweder M, Gray B, Zucali J , and Zori R. In vitm selection for K562 cells with higher retrovirally mediated copy number of aldehyde dehydrogenase class-1 and higher resistance to 4- hydroperoxycyclophosphamide. Human Gene Therapy 9(5): 61 1-619, 1998.

Moreb JS, Schweder M, Suresh A, and Zucali JR. Overexpression of the human aldehyde dehydrogenase class 1 results in increased resistance to 4-hydroperoxycyclophosphamide. Cancer Gene Therapy 3(1): 24-30, 1996.

Morgan RA, Couture L, Elroy-Stein 0 , Ragheb J, Moss B, and Anderson WF. Retroviral vectors containing putative internal ribosome entry sites: development of a polycistronic gene transfer system and applications to human gene therapy. Nucleic Acids Research 20(6): 1293-1299.1992-

Moritz T. Mackay W, Glassner BJ, Williams DA, Samson L. Retrovirus- mediated expression of a DNA repair protein in bone marrow protects hematopoietic cells from nitrosourea-induced toxicity in vitro and in uiuo. Cancer Research 55(12): 2608-2614, 1995.

Morris JA and McIvor RS. Saturation mutagenesis at dihydrofolate reductase codons 22 and 31. A variety of amino acid substitutions conferring methotrexate resistance. Biochemical Pharmacology 47(7): 1207-1220, 1994.

Mulligan RC. Gene transfer and gene therapy. Principles, Prospects, and Perspective. In: Etiology of Human Disease at the DNA Level, edited by Lidsten J and Pettersson U, Raven Press, Ltd, 143-189 , 1991.

Mulligan RC. The basic science of gene therapy. Science 260: 926-931, 1993.

Neff T and Blau CA. Forced expression of cytidine deaminase confers resistance to cytosine arabinoside and gemcitabine. Experimental Hematology 24: 1340-1346, 1996.

Noble S and Goa KL. Gemcitabiie. A review of its pharmacology and clinical potential in non-small cell lung cancer and pancreatic cancer. Drugs 54(3): 447-472, 1997.

Nygaard P and Sundstrom C. Low cytidine deaminase levels in human hematopoietic cell lines. Leukemia Research 1 l(8): 681-685, 1987.

O'Shaughnessy JA. Chemoprevention of breast cancer. JAMA 275(17): 1349-1353.1996.

O'Shaughnessy JA, Cowan KH, Nienhuis AW, McDonagh KT. Sorrcntino BP, Dunbar CE, Chiang Y, Wilson W, Goldspiel B, Kohler D, Cottler-Fox M, Leitman S, Gottesman M, Pastan 1, Denicoff A, Noone M, Gnss R. Retroviral mediated transfer of the human multidrug resistance gene (MDR-1) into hematopoietic stem cells during autologous

transplantation after intensive chemotherapy for metastatic breast cancer. Human Gene Therapy S(7): 891-91 1, 1994.

Onetto N, Momparler RL, Momparler LF, and Gyger M. In vitm biochemical tests to evaluate the response to therapy of acute leukemia with

' cytosine arabiioside or 5-aza-2'-deoxycytidine- Seminars In Oncology 14: 231-237, 1987.

Otterson GA, Khleif SN, Chen W, Coxon AB, and Kaye FJ. CDKN2 gene silencing in lung cancer by DNA hypermethylation and kinetics of p16 INK4 protein induction by 5-aza-2'-deoxycytidine. Oncogene 1 l(6): 1211-1216. 1995.

Parkinson DR. Pluda JM, Cazenave L, Ho P. Sorensen JM, Sznol M, and Christian MC. Investigational anticancer agents. In: Cancer Chemotherapy and Biotherapy, second edition, edited by Bruce A. Chabner and Dan L. Longo. Lippincott-Raven Publishers, Philadelphia, 509-528, 1996.

Pastan 1, Gottesman MM, Ueda K, Lovelace E, Rutherford AV, and Willingham MC. A retrovirus carrying an MDRl cDNA confers multidrug resistance and polarized expression of P-glycoprotein in MDCK cells. Proceedings of the National Academy of Sciences of the United States of America 85(12): 44864490, 1988.

Pate1 M, Sleep SE, Lewis WS, Spencer HT, Mareya SM, Sorrentino BP, and Blakley RL. Comparison of the protection of cells from antifolates by transduced human dihydrofolate reductase mutants. Human Gene Therapy 8(17): 2069-2077.1997.

Pegg AE. Mammalian 0s-akylguanine-DNA alkyltransferase: regulation and importance in response to akylating carcinogenic and therapeutic agents. Cancer Research SO(19): 6119-6129, 1990.

Pegg AE, Boosalis M. Samson L, Moschel RC, Byers TL, Swenn K, and Dolan ME. Mechanism of inactivation of human 06-wlguanine-DNA

157

alkyltransferase by Wbenzylguanine. Biochemistry 32(45): 11998- 12006,1993.

Persons DA, Allay JA, Allay ER, Smeyne RJ, Ashmun RA. Sorrentino BP, and Nienhuis AW. Retroviral-mediated transfer of the green fluorescent protein gene into murine hematopietic cells facilitates scoring and selection of transduced progenitors in vitm and identification of genetically modified cells in uivo. Blood 90(5): 1777-1786, 1997.

Peters WP. Rosner G, Ross M, Vredenburgh J , Meisenberg B, G i r t C, and Kurtzberg J. Comparative effects of granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) on priming peripheral blood progenitor cells for use with autologous bone marrow after high-dose chemotherapy. Blood 81(7): 1709-1719, 1993.

Pliml J and Sorm F. Synthesis of a 2-deoxy-D-ribofuranosyl-5-azacytosine. Preliminary Communication. Collection Czechoslov. Chem. Commun. 29:2576-2577, 1964.

Plagemann PG. Behrens M, and Abraham D. Metabolism and cytotoxkity of 5-azacytidine in culhlred Novikoff rat hepatoma and P388 mouse leukemia cells and their enhancement by preincubation with pyrazofurin. Cancer Research 38(8): 2458-2466, 1978b.

Plagemann PG, Man R, and Wohlhueter RM. Transport and metabolism of deoxycytidiie and 1-beta-D-arabinofuranosylcytosine into cultured Novikoff rat hepatoma cells, relationship to phosphorylation, and regulation of triphosphate synthesis. Cancer Research 38(4): 978-989, 1978a.

Plunkett W, Huang P, and Gandhi V. Preclinical characteristics of gemcitabiie. Anti-Cancer Drugs 6(Suppl6): 7-13, 1995.

Podda S, Ward M, Himelstein A, Richardson C, de la Flor-Weiss E, Smith L, Gottesman M, Pastan 1, and Bank A. Transfer and expression of the human multiple drug resistance gene into live mice. Proceedings of the

National Academy of Sciences of the United States of America 89(20): 9676-9680, 1992.

Radin AX, Zhoa XL, Woo TH, Colvin OM, and Hilton J. Structure and expression of the cytosolic aldehyde dehydrogenase gene in cyclophosphamide-resistant murine leukemia L1210 cells. Biochemical Pharmacology 42(10): 1933-1939, 1991.

Raffer: JA, Hickson I, Chi iasamy N, Lashford LS, Margison GP. Dexter TM. and Fairbairn LJ. Chemoprotection of normal tissues by transfer of dr . ' mc, genes. Cancer and Metastasis Reviews 15(3): 365-383, 1':

Raviv Y, Pollard HB, Bruggemann EP, Pastan I, Gottesman MM. Photosensitized labeling of a functional multidrug transporter in living drug-resistant tumor cells. Journal of Biological Chemistry 265(7): 3975-3980, 1990.

Razin A, and Cedar H. DNA methylation and gene expression. Microbiological Reviews 55(3): 451-458, 199 1.

Razin A, and Riggs AD. DNA methylation and gene function. Science 210(4470): 604-610, 1980.

Reese JS, Koq ON, Lee KM, Liu L, M a y JA, Phiiips WP Jr, and Gerson SL. Retroviral transduction of a mutant methylguanine DNA methyltransferase gene into human CD34 cells confers resistance to 0 6 -

benzylguanine plus 1.3-bis (2-chIoroethy1)-1-nitrosouna. Proceedings of the National Academy of Sciences of the United States of America 93(24): 14088-14093, 1996.

Richardson C and Bank A. Prese1ection of transduced murine hematopoietic stem cell populations leads to increased long-term stability and expression of the human multiple drug resistance gene. Blood 86(7): 2579-2589, 1995.

Richel DJ, Colly LP, Kluin-Nelemans JC, and Wiemze R The antileukaemic activity of 5-aza-2'-deoxycytidine (Aza-dC) in patients with relapsed and resistant leukaemia. British Journal of Cancer 64(1): 144-148, 1991.

Richter J. Gene transfer to hematopoietic cells - the clinical experience. European Journal of Haernatology 59: 67-75, 1997.

Rivard GE, Momparler RL, Demers J , Benoit P, Raymond R, Lin KT, and Momparler LF. Phase I study on 5-aza-2'-deoxycytidine in children with acute leukemia. Leukemia Research 5(6): 453-462, 1981.

Robson CN, Lewis AD, Wolf CR, Hayes JD, Hall A, Proctor SJ, Hanis AL, and Hickson lD. Reduced levels of drug-induced DNA cross-linking in nitrogen mustard-resistant Chinese hamster ovary cells expressing elevated glutathione S--sferase activity. Cancer R:search 47(22): 6022-6027. 1987.

Rockwell S and Grindey GB. Effect of 2',2'-difluorodeoxycytidine on the viability and radiosensitivity of EMT6 cells in uitro. Oncology Research 4(4-5): 151-155, 1992.

Roninson 1B. From amplification to function: the case of the MDRl gene. Mutation Research 276(3): 151-161, 1992b.

Roninson IB. The role of the MDRl (P-glycoprotein) gene in multidrug resistance in uitro and in vivo. Biochemical Pharmacology 43I1): 95-102, 1992a.

Rosenberg SA, Aebersold P, Cornetta K, Kasid A, Morgan RA, Moen R, Karson EM, Lotze MT, Yang JC, Topalian SL, Merino MJ, Culver K, Mier AD, Blaese RM, Anderson WF. Gene transfer into humans: immunotherapy of patients with advanced melanoma using tumor- infiltrating lymphocytes modified by retroviral gene transduction. New England Journal of Medicine 323: 570-578, 1990.

Roth JA, and Christian0 RJ. Gene therapy for cancer: what have we done and where are we going? Journal of the National Cancer Institute 89(1): 21-39, 1997.

Rund D, Dagan M, Dalyot-Herman N, Kimchi-Sarfaty C, Shoenlein W, Gottesman MM, and Oppenheim A. Efficient transduction of human hematopoietic cells with the human multidrug resistance gene 1 via SV40 pseudovirions. Human Gene Therapy 9(5): 649-657, 1998.

Runnebaum lB. Basics of cancer gene therapy. Anticancer Research 17(4B): 2887-2890, 1997.

Saccone S, Besati C, Andreozzi L, Della Valle G, Garattini E, and Terao M. Assignment of the human cytidine deaminase (CDA) gene to chromosome 1 band p35- p36.2. Genornics 22(3): 661-662, 1994.

Sandhu JS, Keating A, and Hozumi N. Human gene therapy. Critical Reviews in Biotechnology 17(4): 307-326, 1997.

Santi DV, Garrett CE, and Barr PJ. On the mechanism of inhibition of DNA-cytosine methyltransferases by cytosine analogs. Cell 33(1): 9-10, 1983.

Scaradavou A, Isola L, Rubinstein P, Galperin Y, Najfeld V, Berlin D, Gordon J , and Weinberg RS. A murine model for human cord blood transplantation: near-tenn fetal and neonatal peripheral blood cells can achieve long-term bone marrow engraftment in sublethally irradiated adult recipients Blood 89(3): 1089-1099, 1997.

Schecter RL, Alaoui-Jamali MA, and Batist G. Glutathione S-transferase in chemotherapy resistance and in carcinogenesis. Biochemistry and Cell Biology 70(5): 349-353, 1992.

Schimke RT, ffiufman RJ. Alt FW, and Kellems RF. Gene amplification and drug resistance in cultured murine cells. Science 202(4372): 1051- 1055, 1978.

Schuening F, Storb R, Nash R, Stead RB. Kwok WW, and Miller AD. Retroviral transfer of genes into canine hematopoietic progenitor cells. Advances in Experimental Medicine and Biology 241: 9-18, 1988.

Schwartz GN, Hakim F, Zujewski J. Szabo JM. Cepada R. Riseberg D, Warren MK, Mackall CL, Setzer A, Noone M, Cowan KH. O'Shaughnessy J, and Gress RE. Early suppressive effects of chemotherapy and cytokine treatment on committed versus primitive haemopoietic progenitors in patient bone marrow. British Journal of Haematology 92(3): 537-547, 1996.

Schwarzenberger P, Spence S, Lohrey N, Krniecik T, Longo DL. Murphy WJ, Ruscetti FW, and Keller JR. Gene transfer of multidrug resistance into a factor-dependent human hematopoietic progenitor cell line: in uivo model for genetically transferred chemoprotection Blood 87(7): 2723-273 1, 1996.

Shaffer DW, Smith LS, Burris HA, Clark GM, Eckardt JR, Fields SM, Weiss GR, Rinaldi DA, Bowen KJ, Kuhn JG, and Von Hoff DD. A randomized phase I trial of chronic oral etoposide with or without granulocyte- macrophage colony-stimulating factor in patients with advanced malignancies. Cancer Research 53(24): 5929-5933, 1993.

Shipp MA, Takvorian RC, and Canellos GP. High-dose cytosine arabinoside. Active agent in treatment of non-Hodgkin's lymphoma. American Journal of Medicine 77(5): 845-850, 1984.

Shr6der JK, Kirch C, Flasshove M, Kalweit H, Seidelmann M. Hilger R, Seeber S, and Schctte J. Constitutive overexpression of the cytidine deaminase gene confers resistance to cytosine arabinoside in vitro. Leukemia 10: 1919-1924,1996.

Simonsen CC, and Levinson AD. Isolation and expression of a n altered mouse dihydrofolate reductase cDNA. Proceedings of the National Academy of Sciences of the United States of America 80(9): 2495-2499, 1983.

Sirotnak FM, Kurita S, and Hutchison D. On the nature of a transport alteration determining resistance to amethopterin in the L1210 leukemia. Cancer Research 28(1): 75-80, 1968.

Slevin ML, Piall EM, Aherne GW, Johnston A, and Lister TA. The pharrnacokinetics of cytosine arabiioside in the plasma and cerebrospinal fluid during conventional and high-dose therapy. Medical and Pediatric Oncology lO(Supp1 1): 157-168, 1982.

Sorrentino BP, Brandt SJ, Bodine D, Gottesman M, Pastan 1, Cline A, and Nienhuis AW. Selection of drug-resistant bone marrow cells in vivo after retroviral transfer of human MDR-1. Science 257(5066): 99-103, 1992.

Spencer HT, Sleep SE, Rehg JE, Blakley RL, and Sorrentino BP. A gene transfer strategy for making bone marrow cells resistant to trimetrexate. Blood 87(6): 2579-2587, 1996.

Sreerama L and Sladek NE. Identification and characterization of a novel class 3 aldehyde dehydrogenase overexpressed in a human breast adenocarcinoma cell line exhibiting oxazaphosphorine-specific acquired resistance. Biochemical Pharmacology 45(12): 2487-2505, 1993.

Srirnatkandada S, Schweitzer B, Moroson BA, Dube S, and Bertino JR. Amplification of a polymorphic dihydrofolate reductase gene expressing an enzyme with decreased biding to methotrexate in a human colon carcinoma cell lime, HCT-8R4, resistant to this drug. Journal of Biological Chemistry 264(6): 3524-3528, 1989.

Stadler WM. Kuzel TM, Raghavan D, Levine E, Vogelzang NJ. Roth B, and DOH FA. Metastatic bladder cancer: advances in treatment. European Journal cf Cancer 33 (Suppl 1): S23-26, 1997.

Stead RB, Kwok WW, Storb R, and Miller AD. Canine model for gene therapy: inefficient gene expression in dogs reconstituted with autologous marrow infected with retroviral vectors. Blood 71(3): 742- 747, 1988.

Steuart CD, and Burke PJ. Cytidine deaminase and the development of resistance to arabiiosyl cytosine. Nature - New Biology 233(38): 109- 110. 1971.

Stone RM, Berg DT. George SL, Dodge RK, Paciucci PA, Schulmvl P. Lee EJ, Moore JO, Powell BL, and Schiffer CA. Granulocyte-macrophage colony-stimulating factor after initial chemotherapy for elderly patients with primary acute myelogenous leukemia. Cancer and Leukemia Group B. New England Journal of Medicine 332(25): 1671-1677, 1995.

Stryckmans P, DeWitte T, Bitar N, Marie JP, Suciu S, Sohbu G. Debusscher L, Bury J , Peetermans M, Andrien JM, Fiere D. Bron D, Dekker A, and Zittoun R. Cytosine arabiioside for induction, salvage. and consolidation therapy of adult acute lymphoblastic leukemia. Seminars in Oncology 14(2 Suppl 1): 67-72, 1987.

Suzuki M, Sugimoto Y, Tsukahara S, Okochi E, Gottesman MM, and Tsuruo T. Retroviral coexpression of two different types of drug resistance genes to protect normal cells from combination chemotherapy. Clinical Cancer Research 3: 947-954, 1997.

Teng YS, Anderson J E and Giblett ER. Cytidine deaminase: a new genetic polymorphism demonstrated in human granulocytes. American Journal of Human Genetics 27(4): 492-497, 1975.

Trillet-Lenoir V, Green J, Manegold C, Von Pawel J , Gatzemeier U, Lebeau B, Depierre A, Johnson P, Decoster G, Tomita D, and Ewen C. Recombinant granulocyte colony stimulating factor reduces the infectious complications of cytotoxic chemotherapy. European Journal of Cancer 29A(3): 319-324, 1993.

Tzcng E, Shears LL, Lotze MT, and Billiar TR. Vectors and Delivery Systems. Gene Therapy. Current Problems in Surgery 33(12): 989-1005, 1996.

Uckert W and Walther W. Retrovirus-mediated gene transfer in cancer therapy. Pharmacology and Therapeutics 63: 323-347, 1994.

Ueda K, CardareUi C, Gottesman MM, and Pastan I. Expression of a full- length cDNA for the human "MDR1" gene confers resistance to colchicine, doxorubicin, and vinblastine. Proceedings of the National Academy of Saences of the United States of America 84(9): 3004-3008, 1987.

Vicenzetti S, Cambi A, Balducci E, Natalini P. Volpini R, and Vita A. Human placenta cytidine deaminase: a zinc metalloprotein. Biochemistry and Molecular Biology International 42(3): 469-476. 1997.

Vicenzetti S. Cambi A. Neuhard J, Garattini E, and Vita A. Recombinant human cytidine deaminase: expression, purification, and characterization. Protein Expression and Purification 8(2): 247-253. 1996.

Vinh DB, and McIvor RS. Selective expression of methotrexate-resistant dihydrofolate reductase (DHFR) activity in mice transduced with DHFR retrovirus and administered methotrexate. Journal of Pharmacology and Experimental Therapeutics 267(2): 989-996,1993.

Vita A, Cacciamani T, Natalini P, Ruggieri S, and Magni G. Cytidine deaminase from human spleen. Advances in Experimental Medicine and Biology 253B: 71-77. 1989.

Vita A, Vicenzetti S, Amici A, Ferretti E, and Magni G. Cytidine deaminase: a rapid method of purification and some properties of the enzyme from human placenta. Advances in Experimental Medicine and Biology 309B: 235-238, 1991.

Von Hoff DD and the San Antonio Drug Development Team. Activity of gemcitabine in a human tumor cloning assay as a basis for clinical trials with gemcitabiie. Investigational New Drugs 14(3): 265-270. 1996.

Wang G. Weiss C, Sheng P, and Bresnick E. Retrovirus-mediated transfer of the human 06-methylguanine-DNA methyltransferase gene into a

murine hematopoietic stem cell line and resistance to the toxic effects of certain aUrylating agents. Biochemical Pharmacology 51(9): 1221-1228, 1996.

Ward M. Pioli P, Ayello J. Reiss R. U n i G, Richardson C. Hesdorfier C, and '

Bank A. Retroviral transfer and expression of the human multiple drug resistance (MDR) gene in peripheral blood progenitor cells. Clinical Cancer Research 2: 873-876, 1996.

Ward M, Richardson C, Pioli P. Smith L, Podda S, Goff S, Hesdorfier C, and Bank A. Transfer and expression of the human multiple drug resistance gene in human CD34+ cells. Blood 84(5): 1408-1414. 1994.

Weichselbaum RR and Kufe D. Gene therapy of cancer. The Lancet 349(SupplII): 10-12, 1997.

Wentworth DF and Wolfenden R. On the interaction of 3,4,5,6- tetrahydrouridiie with human liver cytidiie deaminase. Biochemistry 14(23): 5099-5105, 1975.

Werkheiser WC. Specific biding of 4-amino folic acid analogues by folic acid reductase. Journal of Biological Chemistry 236(3): 888-893, 1961.

Whartenby KA. Abraham GN, Calabresi PA, Abboud CN, Calabresi P, Marmgi A, and Freeman SM. Gene-modified cells for the treatment of cancer. Pharmacology and Therapeutics 66(1): 175-190, 1995.

Williams DA, Hsieh K, DeSiva A, and Mulligan RC. Protection of bone marrow transplant recipients from lethal doses of methotrexate by the generation of methotrexate-resistant bone marrow. Journal of Experimental Medecine 166(1): 2 10-2 18, 1987.

Wingo PA, Tong T, and Bolden S. Cancer Statistics, 1995. CA- A Cancer Journal for Clinicians 45(l): 8-30, 1995

Zagonel V, Lo Re G, Marotta G. Babare R, Sardeo G. Gattei V, De Angelis V, Monfardini S, and Pinto A. 5-Aza-2'-deoxycytidine (Decitab'ie)

induces trilineage response in unfavourable myelodysplastic syndromes. Leukemia f(Supp1 1): 30-35. 1993.

Zaman GJ, Flens MJ. van Leusden MR. De Haas M, Mulder HS, Lankelma J, Pinedo HM, Scheper RJ, Baas F, Broxterman H, and Borst P. The human multidrug resistance-associated protein MRP is a plasma membrane drug-efflux pump. Promdings of the National Academy of Sciences of the United States of America 91(19): 8822-8826, 1994.

Zhao SC, Banejee D. Mineishi S, and BeAno J R Post-transplant methotrexate administration leads to improved curabiity of mice bearing a mammary tumor transplanted with marrow transduced with a mutant human dihydrofolate reductase cDNA. Human Gene Therapy 8: 903-909, 1997.

Zhao SC, Li MX, Banejee D, Schweitzer B1, Mineishi S, Gilboa E, and Bertino JR. Long-term protection of recipient mice from lethal doses of methotrexate by marrow infected with a double-copy vector retrovirus containing a mutant dihydrofolate reductase. Cancer Gene Therapy l(1): 27-33, 1994.

Article: Ransfection of marine fibroblast cells with haman cytidine deaminase cDNA confers resistance to cytosine arabinoside. Momparler RL, Laliberte J, Eliopoulos N. Beausejour C, and Cournoyer C. Anti-Cancer Drugs 7: 266-274, 1996.

make the n o m l bone marrow cells resistant to the toxic effens of this drug. Our labontory Ins recently cloned and expressed the cDNA for human CR deaminase." The objective of this study a= to determine if tnnsfection of the CR d a m i n a x cDSA into mammalian cells would make them resistant to An-C. Our results show that tnnsfected cells hJd 3 increased expression of CR deaminase and showed signs of drug resistance to An<.

Materials a n d methods

Vector design

The expression vector containing the human CR deaminase cDNA n x s constructed as follows. The plasmid pBluescript KSII containing the complete cDS.4 sequence for human CR deaminase" u2s used as the temphte for PCR amplifying a 465 bp DSA fngment of the CR deaminase protein coding sequence. This PCR ulls performed in a hlJ Research thermocycler using the following oligonucleotide primers:

Kc01

5'-TAC CAC CAT CGC CCA GAA GCG T-3' (Pl) BarnHI * 5'-TCG GCA CCA TCC CGC TGT CAC Ta(P2)

The oligonucleotide P1 contained a tic01 recognition site 31 the 5' end and the first five codons of the human CR deaminase gene in the sense orientation. Oligonucleotide P2 contained 3 BamHI recognition site at the 5' end and the last five codons of the human CR deaminase gene in the antisense orientation. This PCR was performed with 2.5 units ID- Proof (Taq) DNA polymerase (ID Labs Biotechnol- o p . London. Ontario. Canada) using the buffer sup- plied by the manufacture plus 1.5 mM MgCh and 0.1 m\l of each of the deoxynucleotide tripho- sphates in a reaction volume of 50 pl. The template (25 ng plasmid DNA) was denatured for 2 min at 94'C followed by 60°C for 1 min and 72'C for 5 min for one cycle. and amplified for 15 cycles. each cycle consisting of denaturation for 30 s at 94'C. annealing for 30 s a t 60'C and extension for 2.5 min at 72'C with a terminal 5 min extension at 72'C. The 465 bp DNA fngment ms precipitated with ethanol.

The plasmid expression vectors pMFG-PA and pMFG-Lac2 used in this study were obrained from R Mulligan (Whitehead Institute. Cambridge. MA). In these vectors. expression of the insened

CR deaminase cDXA confers Am-C resistance

sequence is promoted by the Moloney murine leukemia virus long terminal repeat (LTR). The presence of m t u n l splice donor and acceptor sites. n o m l l y used to g e n m t e the subgenomic enumn- script of the vim, appears to improve the expression of cloned ins en^.".'^ To maintain the proper positioning of the cloned insen. the insen n x s cloned bemeen the Kc01 and BamHl sites of the vector. This am accompliihrd by inuoducing an Xco1 site at the stan codon of the protein coding sequences and a BanlHI site shonly after the stop codon.

Both the pMFG-[PA plasmid and the 465 bp PCR amplified CR deamirux fngment upere digested with Kcol and BamHI, the appropriate fragments were sepanted by 2% aprose gcl elemphoresis. purified with QIAquick spin column (QIAGES, Chatswonh. tA) and lipred svith T4 DSA lipse. Competent Bcherichia coli were tnnsformed with the consuuct and individual colonies of tnnsfor- mans were screened for insenion by PCR and resuiction e n q m e digests. Large-scale prepantions of plasmids were produced by swndard methods and the plasmids purified with QIAprep plasmid kit (QIAGEU). The resulting plasmid with the CR deaminase gene was named phlFG-CD. The term- iml 3' region of hIFG and initial 5' coding region of the gene were sequenced by the dideoxynucleotide chain termination method using a Phamcia Auto- matic DNA Sequencer with fluoro-dATP to verify the desired sequence at the initial ATG codon region of the CR deaminase.

Cell culture techn~ques

Cells were grown in DMEM medium (Canadian Life Technologies, Burlington. Ontario. Canada) supplemented with 10% heat-inactivated fetal calf serum (Wiient Technologies, St Bruno. Quebec. Canada) and 5 pg/ml genwmycin (Canadian Life Technolo- gies) and incubated at 37°C and 7% C02. GP + E86 murine ecotropic packaging cells, which are derived from NIH 3 T j mouse fibroblasts, were obtained from Dr A Bank (Columbia Lhiversity)?' The GP +E86 cells were cc-mnsfec- red with the p u r s e d plasmid DNAs pMFG-CD (or pYFG-LacZ) and pSV2-neo in 3 10: 1 molar ntio. using the w n d a r d calcium phosphate precipitation method and a calcium phosphate tnnsfection kit (Pharmacia. Baie d'Urfi.. Quebec. Canada).'" At 72 h after mnsfeaion. G418 (Geneticin. Promega. Madison. WI) at 400 pg/ml u3s added to the medium and the cells were selected in this medium for

RL dfompaler et al.

14 &ys. Clones of cells resisant to G418 were isolated by ring cloning or by dilution. Thereafter the cells were maintained in medium without G418.

Clonogenic 3553ys w x e performed as follows. hliquors of l ml of 1'20 cells/ml were plated in u,ells of a 12-well Costar dish: 18-20 h later Ara-C a% added at the indicated concentntions and the incubation continued for an additional 12-14 days. The well were then suined with 0.5% methylene blue in 50% methanol and colonies of greater than 10' cells were counted.

For the hiistochemical suin for fl-galactosidse the monolayer cells were nzshed with phosphate buffered uline (PBS) and futed at 4°C in 2% pan-

. - fomuldehyde/O.Wo gluunldehyde in phosphate buffer. The fixed cells were suined at 37'C with 1 mg/ml X-Gal (Life Technologies) solution con- mining KFdCN)b in phosphate buffer until 3 blue color develops.

DSA synthesis as.uys were performed as follows. Cells were diluted to 10'/ml and 1 ml aliquors were placed in wells of a 12-well Cosur dish. After incubation overnight. Am-C n m added at the indicated concentrations with 0.2 pCi [3~lthymidine (20 Ci/mmol) and the incubation continued for an additional 6 h. The amount of ndioactivicy incorporated into DSA was determined after trypsiniza-

@ tion as described p r e v i o ~ s l y . ~ ~

Enzyme assay

In vitro assay for CR deamimse was performed using a modification of a previously described method." Monolayer cells (5 x 10') were uypsi- nized, centrifuged and washed once in PBS. recm- trifuged again and resuspended in 200 pl of 5 m\l Tris-CI. pH 7.4. and 5 m\l dithiothreitol. The cell suspension was then subjected to three cycles of npid freezing and tha-xing. The mixture was centrifuged at nuximum speed in a microfuge at 5°C for I5 min. The molaricy of the s u p e m u n t (cytosol) u3s then increased to 50 m\l Tris-CI. pH 7.4. Dif- ferent dilutions of the cytosol were used in 30 min incubation at 37'C to measure the conversion of ['~lcytidine to [%xid ine as described previously."

Northern and RNA dot blot analysis

Total RVA was isolated from cells by a modified method of Chomczynski and sacchi" using the Ultnspec-11 W A Isolation kit (ID Labs Biotechnol-

om). Briefly. the cells were Iysed with 3 gunidine solution. the RSA precipiuted with isopropanol and purified uith 3 min . The RVA n z s eluted from the r a i n with TE buffer (10 mM Tris. pH 8: 1 m\l EDTA) and stored at -70°C.

For dot blot amlysis. RVAuudi1u:ed in a mixture of 20 x SSC (I x SSC is 0.15 hl Sac1 and 15 mM Sa citnte at pH 7.0) and 37% formaldehyde. The u m - ples were heated at 65'C for 5 min and spotted on Hybond-S nylon membnne (Amersham. Oakville. Onurio. Cam&) in a Hoeffer slot blot appantus. For Xonhern blot analpis RSA samples in loading buffer were heated to 6YC for 5 min and loaded onto 1.5% agarose-1.1% fomuldehyde gel. Electro- phoresis n a s performed in a 0.035 M sodium bonte buffer-1.1% formldehyde. After LIT photography of the gel. the R S A u w blotted onto S p n n - S nylon membnne with 20 x SSC using a TurboBlotter device (Schleicher Rr Schuell. Keme. SH). Before hybridimtion. the nylon membrane were w~5ht.d with 20 x SSC. baked at80"C and cross-linked using a Bioslink W linker (0.3 ~/cm').

Forsynthesis of the probe. the complete cDSA for human CR deaminase u ~ s labeled by the nndom prime method of Feinberg and vogelstein2' using the kit from Boehringer hhnnhcim (Dowal. Que- bec. Canada) and [Z-"P]~CTP from ICS (Missis- uuga. Ontario. Canada). The membnnes were prehybridized in 6 x SSC. 2 x Dmhardt's solution and 1% SDS at 65°C for 2.5 h. and hybridized overnight at 65'C in 10 ml of prehybridization solution conuining 2 x 10' c.p.m. of the DSA probe. The membnne n n s then washed with 2 x SSC and 0.1% SDS at room temperature. and then with 0.2 x SSC: 0.1% SDS at 60'C for I5 min. The blot was exposed to Kodak X-Omt film with intensifying screens at -70% for 24-48 h.

PCR analysis and Southern blotting

A PCR a s u y was used to verify the presence of the MFG-CD construct in tnnsfected cells. The oligonucleotides:

5'-GGT GGA CCA TCC T C T AGA CTG-3' (P3) 5'-AGC AGC T C C TGG ACC GTC ATG3' (P4)

were used as primers with genornic DSA in the PCR to amplify a specific 421 b p fngrnent as predicted by the DSA sequence of the phlFG-CD construct. The sense oligonucleotide P3 a m about 270 bp doa.nstream from the splice acceptor region of AlFG and 2 b p upstream from the s u n of the enu coding region. The antisense oligonucleotide P4 was from

CR deaminar~ cDXA confers AraC resistance

positions 378-398 of the CR deaqimse coding Results mion . Genomic DSA a x s isolated from the

9 G P + ~ cells n.ith In ~ i o r b DSA liit (ID ~ a b s Biotechnolom) by cell lvsis with puanidine thiocva- CR deaminase expression

mte. DSA adso&on on si l iu gel and elution a k h TE buffer. This PCR rills performed as described above with about 1 ng genomic DSA denatured at 9YC for 2 min and amplified for 25 cycles, each q c l r consisting of demtuntion for 1 min at 94°C. annealing for 1 min at 56°C and enension for 1 min at 7 X with a terminal 5 min extension at 7L'C. The reaction mixture n z s sepanted on 2% aprose elec- trophortAs.

For Southern blot analysis, genomic DSA n z s Lolared by cell lysis and digestion with proteinase K fo!:owed by extnction with phenol~hloroform- isoamyl alcohol and precipiwtion with ethanol. The DSA nxs digested xvith iVcol and BantHI. s e p ~ n t e d by electrophorrsis on a 1% agarose and tnnsferred to a nylon membnne. The ~Vcol-BamH1 fragments of MFG-CD containingthe open reading fnme of CR deaminax were ndiolabeled by nndom prime reaction and hybridized with the rnembnne as described above. In order to determine coov num-

CP + E86 cells were co-tnnsfected with p>IFG-CD and pSV2-neo or with p>lFC-Lac2 and pSW-neo. Sevenl clones that showed resistance to C.118 were b la red and tested for CR deaminase or p'-galactosidase expression. The pXlFC-CD mnsfected clones that showed the highat CR deaminase expression (CP + E86-CD; and GP iE86-CD4) were selected for funher investigation. A histochemical stain showed efficient expression o i p-galactosi&se in 3 G418 resistant clone (GPfE86-Lad) isolated from p>lFG-kc2 tnnsfected celL.(&w not shown).

Enzymatic assays for CR dearninase activity were performed on the non-tnnsfected CP + E86 cells. GP - E86-LacZ. GP + E86-CD3 m d GP + E86-CD4 cells (Table 1). CR deaminase activit). for the GP - E86 and C P + E86-Lac2 cells mas very low. of the order of 1-4 unitsfmg. The same enzyme activity was 1086 units/mg for the GP +E86-CD3 cells and 227 unirsfmg for the GP + E86-CD4 cells.

. " genomic DSA from CP+E8G cells followed by digeation with NcoI and BamHl. Table 1. CR dearninase activity in d~flerent cell lines

Cell line CR dearninase activity (uniwmg)'

Western blot analysis GP -E86 4.0 + 1 .oD GP A E86-LacZ

Electrophoreses of 5 pg protein of cytosols from GP + ~ 8 ~ ~ 0 3 different cells lines were run on 15% ~olsacn.la- GP&E86-C04 mide-SDS gels with molecular weight m a r h k i n d the tnnsfened el~crrophoretiuIly to a

CR deaminase activity was measured in cell exhans ol dlnerent cell lines as described under 'Methods:

nitrocellulose membnne (Schleicher Sr Schucll). 'Une 01 anivirv is defined as nmoles CR deaminaled mr mtn. After treatment of the membnne with blocking solution. nbbit CR deaminase antiserum (1 : 2000 dilution) was incubated with the membnne. The ECL kit from Amenham was used for antigen detection as described by the manufacturer. Briefly, after washing. the membnne a x s incubated with goat anti-nbbit IgG-horseradish peroxidase complex (1 :ZOO0 dilution). The rnembnne was then nxshed, incubated with substnte and exposed to film for 10 min for luminescent detection of the antigen. For prepantion of antibody to CR deaminw, the 14 amino acid sequence at the C-terminus a x selected for immunogenicit). as described by Jameson and ~rolf." This peptide was linked to keyhole limpet hernocyanin (KW) and injected s.c. with Frrund's adjuvant into rabbits. The an tben was collected at different intervals after immunization.

The mRVA expression of CR deaminase of the different clones was also evaluated using the CR cDSA as the probe. In a RSA dot blot analysis, it w s not pas.+' , detect CR deaminase expression in 5 pg c KYA in either the CP + E86 or GPf E86-Lac2 cells (Figure 1). In contnst, it was possible to detect CR deaminase expression mRVA in as little a s 0.5 pg of total RVA in both the GP fE86-CD3 and GP + E86CD4 cells. The exp-es- sion of mRVA was greater in the CP + E86-CD3 cells as compared to the GPfE86-CD4 cells. Similar results were obtained in a Xonhern blot of these different clones (Figure 2). It n x s nor possible to detect CR deaminase mRNA in cells that b d not h e n tnnsfected with the pMFG-CD consrruit.

hY. .~fomparler ct al.

no RNA

Rgum 1. Dot blot analysis of RNA. Different amounts of total RNA Imm the indicated cell lines were bloned onto a neutral nylon membrane and hybridized with *P-labeled CR deaminase cDNA pmbe as described under Methods.

2.4 k b i A .

1.4 kb-

flgum 2 Northern blot hybriduatlon ol 5 pg total RNA from each ol the mdlcated cell Ilnes. The membrane was hybndued w~th *P-labeled CR deaminase cDNA probe as descnbed under Methods.

whereas the mRUA expression for lhi gene was greater in the GP fEs6-CD3 cells as compared to the GP + E86-CD4 cells. Of note. two transcripts of approximately 2800 and 1900 bp were detected in the GP+E86-CD3 cells. Using the hlFG vector. Dwarki ef a~'%lso detected two uanscripts for human factor v l I l in 3T3 cells which differ in size by about 0.9 kb.

flgum 3. Western blot analysis to deten thu expression of human CR deaminase. Initially. 5 pg of protein from the indicated cell lines was separated by polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane. Alter incubation with rabbit CR deaminase anti. serum. the presence of CR deaminase was detected by chemiluminescence as described under Methods.

The expression of the CR deamimse protein in the different clones was also evaluated by Western blotting. Antibodies to the !.uman CR deaminue were generated in rabbits by injection of 3 C-terminal peptide fragment coupled to KLH protein. In Figure 3 the Western blot showed that it was possible to detect the presence of the CR deaminase protein in both the GP + E86-CD3 and GP + E86- CD4 cells, bur not in the GP + E86 and GP +E86- Lac2 cells. The estimated molecular weight of CR

CR deamirtase cD.U cot@ A r a C resistance

daminase uxs about 16 kDa which is the u m e as the same concenmtions of Am-C produced less the purified protein from human placenta and from than 10% inhibition. Am-C at .\I produced the CR deaminase pGFX bactefid expression 20.3% inhibition of DSA synthesis for the vector.'" Again, the expression of CR d e a m i m e GP+ESGCD4 cells indicating that this clone was was greater in the GP+ESG-CD3 cells than the rzore sencitive to thh nucleoside amlog t h n the GP i E86-CD4 cells. GP i E86-CD3 cells.

PCR and Southern blot analysis Table 2 Inhibition of DNA synthesis by Ara-C in d~tferent cell lmes

In order to verify the presence of the tnnsteaed Cell line Inhibition (%) of DNA synthesis at Ara-C hlFG-CD constmct in cells, a PCR was performed concentratton using purified genomic DXA from the cells as the 10" M lo-' M 10'' M remplatc. 3 sense primer downstream from the hlFG splice acceptor region and m antisense primer from the coding region of CR deaminase. A DSA band of the predicted size of 421 bp was amplified from both the GP + E86-CD3 and GP + E86-CD4 cells. but not in the GP+E86-hcZ cells (Flgure 4). In a second n~ethod to confirm the presence of MFG-CD in the tnnsfected cells, purified genomic DSA was digested with Ncol and BanrHl and analyzed by Southern hybridization with CR deaminase cDSA probe (Figure 5. upper panel). This analysis revealed a DSA band of approximxely 0.45 lib which is similar to the predicted size of the coding region of CR deaminase cDSA. The intensir). of the DSA band \vas greater in the GP + E86-CD3 cells as compared to the GPtE86-CD4 cells, suggesting that the former cells contained more copies of the gene. An estimate of the copy number n'as performed by loading variable amounts of CR deaminase cDSA corresponding to a specific number of vector copies per cell to a constant amount of genomic DSA from non-mnsfected cells before South- ern blot analysis (Figure 5, lower panel). Densitometric analysis suggests that GP + E86-CD4 cells conuin one copy whereas the GP + E86-CD3 cells contain three or four copies of the tnnsfected CR Jeaminase gene.

Effect of Am-C on DNA synthesis and colony formation

The effects of An-C on DSA synthesis as determined by the incorpontion of ndioactive thymidine into DNA in both the CR deaminase tnnsfected and non-tnnsfected cells are shoum in Table 2. An-C at concentntions of lo-'. lo-' and lo-$ &I produced 14.8.66.2 and 91.7% inhibition of DSA synthesis, respectively. for the GP + E66 cells. A similar level of inhibition of DSA synthesis with these concentntions of hn-C was obsemed with the GP + E86-Lac2 cells. For the GP + E66-CD3 cells

Cells were exposed lo the indbcaled concentralions 01 Ara-C and ['H).lhymdme lor 6 hr and DNA synthesis determmd as descrtbed under Methods .Mean =SD. n=3.

A clonogenic 3553). was used to evaluate the cytotoxic action of An-C on the non-tnnsfected and CR deaminase-tnnsfected cells (Table 3). An-C at concentntions of lo-' or hl completely suppressed colony formation for both the GP+E86 and GPiE66-hcZ cells. At lo-' h1 An-C. colony for- marlon for the GP+ES6-CD3 cells and the GPiE86-CD4 cells was greater than 90°/o of the control value without drug. Colony formation at

hl An-C was reduced to about 73 and 24% of the control value for the GP + E86-CD3 cells and the GP + E86-CD4 cells, respectively.

Table 3. Etfecl of Ara-C on colony Ionnation of cells trans- leded with CR deaminase cDNA

Cell line Experi- No. of colonies at Ara-C concentration men! no.

0 :O'eM TO-'M ~ o - ~ M lo- 'M

Cells *ore expose6 to the lndcatoa ConcentraLonS 01 Ara.C lor 12-14 aays and the number 01 colonies lormed determned as dexnbed under Methods.

CR draminuse cDXA confen A r a t reskfance

late with the e n q m e activity and RY.4 expression as discussed above. Of note. the m.0 different vector tranxriprs detected by Sonhern blot amlysis in the clone GP + E86-CD3 most cerminly represent the spliced and unspliced tnnscriprs resulting from the inclusion of tht hloloney virus splice donor and acceptor sigmls in the pMFG construct. The approximtely 0.9 kb difference in size bemeen t h n e nvo tranxriprs is also in accord with this inter- prention, as this n ~ ? ; also the difference in size of the two mRSA species o b x w e d when the XlFG vector was used to express the h u m n factor \'Ill.'"

In order to detect the presence of the LIFG-CD construct in the tnnsfected cells. we purified the DSA from these cells to use as the template in a PCR reaction using a sen.e primer downstream from the 3IFG rplice acceptor region and an antisense primer from the coding region of CR deaminax cDSA. The predicted 411 bp DSA was amplified from the GP + E86-CD3 and GP + ES6-CD1 cells, but not from the GP+E86-Lac2 cells (Figure 4).

Southern blot analysis was also performed on these cells with purified genomic DSA digested with Xcol and BamHI and probed with CR d a m i - nase cDSA. A DSA band of approximately 0.45 kb was detected in both the GP+E86-CD3 and GP + E86-CD4 cells, but not in the GP + E86 cells (Figure 5, upper panel). The intensity of this band. which is similar to the size of the coding region of CR deaminase cDSA, was greater for the GP + E86- CD3 cells thnn the GP + ES6-CD4 cells, indicating that there were more copies of the LIFG-CD construct integnted into the genomic DSA of the GP + E86-CD3 cells. A more precise estimate of the copy number for CR deaminase cDSA in these cells is in agreement with our analpis (Figure 5, lower panel). There appears to b e only one copy of CR deaminase cDSA in the GP+E86-CD4 cells. whereas the GP iE86-CD3 cells appear to contain three to four copies of the tnnsfected gene.

We have also determined that the increased expression of CR deaminase in the tnnsfected cells conferred drug resistance to An-C as evaluated by inhibition of DSA synthesis and a colony formation. An-C is a very potent inhibitor of DS.4 synthesis due to its incorporation into DNA. whew it produces a chain termination-like effect." An-C at

31 produced 91.5% inhibition of DShsynthesis for the parennl GP +E86 cells. as compared with only 7.6 and 20.3% inhibition for the GP + E86-CD3 and GP + E86-CD4 cells. respectively (Table 2). In a clonogenic assay, we detected no colonies for the GP - E86 and GP + E86- Lac2 cells in the presence of hl An-C, whereas the number of colonies

for the GP + E86-CD3 and GP + E86-CD1 cells a= similar to that of the untreated controls (Table 3). Even 31 a concentration of lo-' hl An-C 3 greater number of colonies appexed for the GP + E8G CD3 cells as compared with the GP + E86-CD4 cells. This correlates with the gra te r expression of CR d a m i - mse in the former cell line. It is interesting to note that I1 An-C is in the range ofthe p lasm levels obtained with conventional dose Am-C whereas

JI An-C is in the nnge obnined with high dose . * r a ~ . ~ ~ . = -

The GP + E86-CD clones showed increased expression of CR deaminasr and drug resistance to An-C for more than 4 months in culture without the presence of a selecting agent indicating t h t 3 snble expression of the drug resiswnce phenotype. The GP&E86 cell h e is an ecotropic packaging cell. Our & n suggest that the pIlFG-CD expression vector was integnted into the genomic DSA of the GP i E86 cells and efficiently expressed. Some invesigators have reponed the reduction or loss of expression of integnted gene with tirne.2n.s We did not obseme the silencing of expression of the integrated CR deaminase cDSA in the clones of GP + E86-CD cells. Our preliminary &n indicate that the GP + ES6-CD clones can by retrovinl-mediated tnn5ft.r confer Ara-C resistance to murine hemtopoietic cells, suggesting that the human CR deaminase gene may have the potential for gene thenpy.' Another impomnt application of this gene m y be as a positive selectable marker to enrich the expression of a second gene of interest from a common vector.

Acknowledgments

We thank Mona Greenbaum for technical advice and Louise F blomparler for technical assisnncr.

References

1. Bcnmo JR. 'Turning the tables'-mxkmp; normal mrrorv rcssunt to chemothcnpy. J.lhr1 Cancrr lnsr 1990 82 1?3i-5.

2. Corcy CA. DcStlva AD. Holland CA. W~llmm DA. Send tnmphnutton of mcthomute-resbunt bonc marrow protection of murmc rcc~picnu from drug toxicq by progcny of mnsduccd rtcm cells. BImd 1990; 75. 337- 43.

3. Bancqcr D. Schwcinzcr BI. Volkcnandt .\I, ef 51. Tnns- fearon w!th a cDSA cncodtng a Scdl o: Scr3.l muunt human dihydrofolatc rrducusc into Chincx hamster o\.ar). and moux marrow progcnnor cclls confers mcth- oueutc mbmncc. Cone 1994. 139. 269-&1.

human complcmcnuq DSA for rhc multtdrup msisuncc gene in munne hmtopoic t ic prccunar cell with the uw of rctrovtnl scnc tran4cr. I .Val1 C7ncer Ins1 1990. 82: 1260-3.

5. S o m t m o BP. Bnndt SJ. Badmc D. el d Sclca~on of d r u p m u a n t b n c rmrmw cell\ rn moaftcr Mrovml rnn.&r of human .lIDRl. S n m c e 1992: 257. 99-103

6. Pod& S. Uird XI. Himclztem A. er aL T n m f n and rxprcuion of the human multiple drug rcsruncc Erne into I s c mxcc Pmc.VadAcadScz CS.4 1992: 89- 967680.

cluuth~one S-mn$fcnw Yc confcn nlkvhtmr d ~ c . . . rstrwncc tn SlH 3T3 mowc fibmblsb. Cancer fin- 1991: % 4442-7.

8. Gmlener GU: Smnh CG. Studtcr of the emvrmr- dca- mtnarxon of q tosmc znbmoadr . I Ennmc dranautnon and spccm ,pccBcrty. Biushm Phnnnacoll965. 14. 140i-16

9 Xlullcr WEG. Zahn Rti .\lcubolmm of 1-8.o-anbinofur- morluncrl In mouw LUITRY cell. CaftcerR"s 1979: 39 ,,"7-7 ..v- . .

10 Ho DHW FWI E. Clxntcal phamucolop\. of I-8-D- nr~binofur~no\lqtosine Clm Pharmacol 1972. 12 94-51.

11. Swum CD. Burke PJ Cyrxdmc dcxninau: and the dmcl. opmrnt of rcrrmncc to a n b i n o q l c y o w c . .Varurc.Vm Btol 1971: 233. 109-10

12. Oncrto S. hlomp~rler RL. .\lomparlcr LF. Gyger XI In cmo bimhcmml test, to c\-aluate the rrqmnw to therapy of acute leukemu wi:h qzormc anbmovdc or 5- ara-2'-droqqt1dmc. Scmrn Oncol 1987: 14. 231-7.

13 Bodcy GP. Frcscrch EJ. .\Ionto RW. Hcu.lc11 JS. C}tosxne anhmostde (SSC4Wi8) thcnpy for acute leukemia in adult. Canccr Chrmorher Rep 1969: 53. 59-66

1.1. b l ~ k n i - J. Llompxlcr RL. Human cytidmc dmmin.wz purificmmn of cnzymrr. clonmp, and cxprcrrion of m cDSA. Cancer Res 1994: 54: 5401-7.

15. Jaffcc EX!. Dnnoff G. Cohcn LK. er a 1 Htph efficiency tnnsfcr mto pnmaq tumor cxplanb w h o u t cell sclcchon. Cancer Rrr 1993. 53: 2221-36

16 Dwrkl VJ. Belloni P. Sqar. T, el al. Gene rhcr~py for h r m o ~ h t l u A. ~ r o d u a l o n of thcnrwutx Ic\.cl\ of human fino;\.lll 8n &in micc. Pmc .6ar l~cadSct L'SA 1995: 92. 1023-7.

I-. hlnrkowtz D. Goff S. Bank A A u f c ~ ~ c k a c m c lmc for . . . gene tr~nrfcr. sepuarmy \ t n l p e n s on IWO d~ffercnr p h r m A . J lrml 19W. 62 1120-24.

18 Xldler AD. .\Idler DG. G r c u p . Lynch C\I. Cw of rrtro- r i n l vccton for pcne tnmfcr and m p n x t o n .llrrhalr Enqmol 1993: 217: 581-99.

19. Xlomparln RL bltbcni- J. lnduamn of q n d m c dcaml- wu in HL-60 mycloid lcukcmtc cclh by 5- azadcoxyqt~dmc. h h r a R"s 1590. 14 7514.

20. Chomqmkr P. Sxchi S . Srnplc.wcp mcfhul of RXA r w l o o n by actd puanidtntum rhcq;rrurc-phenol- chloroform c s n n t o n . Anal Biochzn 19H7: 162. 1 5 6 9

21. Fcinbcrp AP. Vopclrtcm BA. A t t ihnquc for n d n h h c l - 1"s DSA rnrrinmn cndunucl~=u f n p c n t r to hlgh specific nntvty. Anal Kxxhrm 1983; 132 6-13.

22. J a m ~ w n Bh. Wolf H. Thc mtrpcnc mder a nmel alp* nthm for prrdmmg znrtgcna determ~mnt\ CAKIOS 19W: 4 181-6.

23. Ho DHW. G n e r CJ. Broun SS. cr a1 Efftits of tetrA- hydroundmc on the "puke and m e u b o l ~ m of 1.8-1,- a n b m o f u n n o ~ l q t m m e m human normal and l ruktm!~ cell* Cancer Res 1980: 40 244146.

21 Krcih W, Chm K. Budnun DR. rr a 1 Effcn of rrtrahy- droundmc o n Ihc c l m r ~ l phlmroloy?. of I-8-D-anhs- nofur~no4qtosme u hen both d r u p arc comfuvd over three houn. C a ~ c r r R ~ J 19U. 48. 1337-42

25. i\(ompnrlcr RL. Kmettc and template t u d m utth l-8-rr a n b m o f u n n o ~ l q ~ o \ l n e 5'-trtpho>phxe 2nd rmmmalur. d:oxynbnuclcrc actd plymen-. .SlolPharmacol19-2. 8 362-70.

26. Ruztum XI. R I V ~ C. Pr~islrr HD Pharm~coktn~t~c p a n - m n m o f I-8-D-anbmofur~norlqtoo~nc (An-C) and thrtr rclmon*h!p to intncrlluhr mcubol~\m of Am-C. toxrtty and re\pon\e of punentr w t h ncutc nonlgmphrr q t i c Icukcmn trca~cd w t h convcnr~onal 2nd h ~ g h do-. An-C. Scmrn Oncol 1947. 14 (suppl 1). 1414.

27. G p p u i RL. Yanp J-L. Chcnp E, ef a1 Altcnrmn of Ihc phamucokiner~cs of hlyh d o ~ An.C hy ILT mcmbol!lc. hlph An.C in pauentr utth cute Icukcmra. J Clln Oncol 1983: 1: i63-71.

28. H w x r J. Lcvm AS. Dtckron K. Unique plttcrn of p l n t muutions amtnp after gene tnnsfcr into mammalun cellr. EllBO J 1987: 6: 63-7.

29. Karluon S. Bodtnc DhI. P e w L. Papqannopoulou T. Scxh;lus AW. Exprcwon of the humm beu plonm penr fo!lowmp m r o r m l mcdntcd tnnrfcr znto mdtlptcntll l hcrmtopoictic progenlton of micc Pmc Xall Acad Scl L5A 19x8: 85 6062-6.

Documents

€¦ · · 2004-11-29demonstrated the ARA-C resistance phenotype by clonogenic assay. ... Long-term expression of the CD transgene by enzyme activity assay was ... 4.2.2 Advantages