intratumoral activation and enhanced chemotherapeutic ... · therapy strategy1 ling chen2 and david...

[CANCER RESEARCH 55, 581-589. February 1. ITO]

Intratumoral Activation and Enhanced Chemotherapeutic Effect ofOxazaphosphorines following Cytochrome P-450 Gene Transfer:

Development of a Combined Chemotherapy/Cancer GeneTherapy Strategy1

Ling Chen2 and David J. Waxman3

Division of Cell and Molecular Biology, Departmenl of Biology, Boston University, Boston, Massachusetts 02215

ABSTRACT

Cyclophosphamide and its isomer ifosfamide are cell cycle-nonspecificalkylating agents that undergo bioactivation catalyzed by liver cyto-chrome P-450 enzymes. The therapeutic efficacy of these oxazaphospho-

rine anticancer drugs is limited by host toxicity resulting from the systemic distribution of activated drug metabolites formed in the liver. Sincetumor cells ordinarily do not have the capacity to activate oxazaphospho-

rines, we examined whether introduction into tumor cells of a cDNAencoding CYP2B1, a major catalyst of oxazaphosphorine activation, sensitizes the cells to the cytotoxic effects of Cyclophosphamide and ifosfamide. Here we show that 9L gliosarcoma cells stably transfected with acDNA encoding rat CYP2B1 are highly sensitive to Cyclophosphamide andifosfamide cytotoxicity as compared to parental 9L cells or 9L cellstransfected with an EscherÃ¬chiacoli ÃŸ-galactosidase gene. The CYP2B1enzyme inhibitor metyrapone protects the CYP2B1-expressing 9L cellsfrom oxazaphosphorine cytotoxicity, demonstrating that the chemosensi-

tivity of these cells is a direct consequence of intracellular prodrug activation. Moreover, CYP2B1-expressing 9L cells potentiate the cytotoxiceffects of Cyclophosphamide and ifosfamide toward cocultured CYP2B1-negative 9L tumor cells. This "bystander effect" does not require cell-cell

contact, and therefore may have the therapeutic advantage of distributingcytotoxic drug metabolites to a wide area within a solid tumor mass. Invivo experiments using Fischer 344 rats implanted s.c. with CYP2B1-

expressing 9L tumor cells demonstrated that intratumoral expression ofthe CYP2B1 gene provides a substantial therapeutic advantage over thatprovided by liver cytochrome P-450-dependent drug activation alone;

Cyclophosphamide treatment resulted in complete growth inhibition ofCYP2Bl-positive tumors, whereas only a modest growth delay effect wasobtained with CYP2B1 -negative tumors. These studies establish thatdrug-activating CYP genes may be useful for the development of novel

combined chemotherapy/gene therapy strategies for cancer treatmentutilizing established cancer chemotherapeutic agents.

INTRODUCTION

Conventional chemotherapy aims to kill malignant cells withoutmajor toxicity to normal host cells and tissues. Although some notablesuccesses in the treatment of specific tumor types (e.g., childhoodleukemias) have been achieved using this approach, more limitedsuccess has been obtained in the treatment of solid tumors. Thisfailure is primarily due to the low therapeutic index of many antican-

cer drugs, as well as to the intrinsic or acquired drug resistance thatoften characterizes tumor cells in the clinic. To circumvent thesedifficulties, novel approaches based on the application of the principles of gene therapy to cancer therapeutics have gained wide attention(1, 2). One therapeutically attractive strategy is based on the direct

Received 8/3/94; accepted 11/22/94.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported in part by NIH Grant CA49248 (D. J. W.).- Present address: Division of Cancer Pharmacology, Dana-Farber Cancer Institute, 44

Binney Street, Boston, MA 02115.3 To whom requests for reprints should be addressed, at Department of Biology,

Boston University, 5 Cummington Street, Boston, MA 02215.

transfer to the tumor of a drug susceptibility gene, or "suicide gene,"

which encodes an enzyme that can activate a prodrug intratumorally,thereby rendering the tumor cells sensitive to chemicals which areotherwise noncytotoxic. One widely studied example is the HSV-TK4

gene, the protein product of which activates the antiviral drug andnucleoside analogue ganciclovir (3, 4). Tumor cells transduced withthe HSV-TK gene using retroviral or adenoviral vectors are killed by

ganciclovir, while normal host cells, which do not express HSV-TK,are unaffected (5-7). A second example is the bacterial cytosine

deaminase gene, which confers chemosensitivity to the relativelynontoxic 5-fluorouracil precursor 5-fluorocytosine (8-10). Althoughboth //SV-rA7ganciclovir and cytosine deaminase/5-fluorocytosine

may potentially be useful for cancer gene therapy, ganciclovir wasoriginally introduced into the clinic for treatment of herpes viralinfection (11-13), while 5-fluorocytosine is an antifungal drug (14).

There are, therefore, few detailed biochemical or pharmacologicalstudies on the application of these drugs in cancer treatment.

The Oxazaphosphorines, Cyclophosphamide and ifosfamide (Fig. 1),are therapeutically inactive cancer chemotherapeutic prodrugs thatmust first be activated by liver CYP metabolism to achieve therapeutic effects in cancer patients. Conceivably, CYP gene transfer incombination with Cyclophosphamide or ifosfamide treatment couldcorrespond to a useful therapeutic strategy to increase the activity andimprove the therapeutic index of this class of chemotherapeuticagents. CYP gene transfer could provide an opportunity to minimizethe systemic toxicity associated with conventional oxazaphosphorinetreatment, which derives from drug activation in the liver by aCYP-catalyzed 4-hydroxylation reaction (Fig. 1), followed by release

into the circulation of activated drug metabolites. Although theseactivated metabolites can effect tumor regression, their formation isaccompanied by inevitable cytotoxic side effects toward critical hosttissues such as bone marrow and kidney. Since tumor cells do notnormally express significant levels of CYP enzymes, introductionof these enzymes into tumor cells by gene transfer may sensitizethe cells to Oxazaphosphorines as a consequence of the resultantdirect, intratumoral prodrug activation. This could provide a basisfor improved anticancer effects using oxazaphosphorine-based

chemotherapy.Recently, the specific liver cytochrome P-450 enzyme catalysts of

Cyclophosphamide and ifosfamide activation were identified in bothrat (15, 16) and human (17) liver. Three specific rat liver cytochromeP-450 enzymes, CYP2B15 (phÃ©nobarbitalinducible), CYP2C6 (con-

stitutively expressed), and CYP2C11 (constitutively expressed andpresent only in adult male rats), are major catalysts of Cyclophosphamide activation in adult rat liver (15), while a rat CYP3A enzyme

4 The abbreviations used are: HSV-TK, herpes simplex virus Ihymidine kinase; CYP,

cytochrome P-450; CYP2B1, CYP form 2B1; 9L-Z, 9L cells that stably express E. coliÃŸ-galactosidase; 9L-ZP, 9L cells that stably express E. coli ÃŸ-galactosidase and ratCYP2B1; 9L-P3, 9L-P13 and L450-2, 9L cells that stably express rat CYP2B1; X-Gal,5-bromo-4-chloro-3-indoyl-ÃŸ-M-galactoside.

5 Individual cytochrome P-450 enzymes are designated according to the systematic

gene nomenclature for CYPs (40).

581

Research. on September 4, 2020. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

CYTOCHROME P-450-BASED GENE THERAPY

4-hydroxy

Cyclophosphamide :

R, â€¢Rj Â«â€”CH2CH2CI

RS â€¢H

Ifosptiamide :

RI â€¢Rj â€¢â€”CH2CH2CI

R2 â€¢H

Fig. 1. Pathways of cytochrome P-450-catalyzed Cyclophosphamide and ifosfamide

metabolism.

additionally contributes to the activation of ifosfamide (16). Corresponding human enzymes CYP2B6 and CYP3A4 activate cyclophos-

phamide and ifosfamide in human liver (17). Of these enzymes,CYP2B1 is the most active (15). Studies carried out in stably trans-

fected cultured cell lines (17, 18) and in transgenic Drosophila (19)have demonstrated that the expression of CYP2B genes, which are notpresent in these cells, can confer sensitivity to Cyclophosphamide.However, it is not known whether tumor cells can be similarlysensitized to oxazaphosphorine cytotoxicity. Moreover, the extent towhich this drug sensitivity can be transferred to adjacent, non-CYP-

expressing tumor cells via a bystander effect is unclear. Finally, thetherapeutic benefits of this strategy in vivo are uncertain given thevery high level of endogenous CYP expression in liver, coupled withthe high activity of several liver CYP enzymes with oxazaphosphorinesubstrates (15-17). The present paper evaluates these important ques

tions through a series of in vitro and in vivo studies based on ex vivogene transfer using the rat 9L gliosarcoma model. Together, thesestudies establish the feasibility of this approach and suggest thatsubstantial therapeutic benefits may be achieved by using Cyclophosphamide and CYP2B1 in a novel combined chemotherapy/cancer genetherapy strategy.

MATERIALS AND METHODS

Chemicals. Cyclophosphamide and ifosfamide were obtained from theDrug Synthesis and Chemistry Branch of the National Cancer Institute (Be-thesda, MD). 4-Hydroperoxy Cyclophosphamide was obtained from Nova

Pharmaceutical Corporation (Baltimore, MD). Metyrapone was purchasedfrom Aldrich Chemical Co. (Milwaukee, WI).

Stable Transfection of 9L Cells. Rat 9L gliosarcoma cells (20) weregrown in a-MEM (GIBCO-BRL) containing 10% fetal bovine serum, 10

units/ml penicillin, and 10 mg/ml streptomycin. Cells were maintained in ahumidified atmosphere of 5% CO2/95% air. 9L cells were cotransfected witha rat CYP2B1 expression plasmid (pMT2-CYP2Bl; kindly provided by Drs.

Milton Adesnik and Allison Reiss, New York University Medical School) andplasmid pCMV-ÃŸgal.Neo (a generous gift from Dr. H. Li, Dana-Farber CancerInstitute) in a molar ratio of 10:1 using Lipofectin (GIBCO-BRL) according tothe instructions of the manufacturer. The plasmid pCMV-ÃŸgal.Neo contains a

neomycin phosphotransferase gene, which confers resistance to G418, and alsothe lacZ (/3-galactosidase) gene of Escherichia coli, which serves as a control

and provides a convenient cell marker. Stable transfectants were cloned underselection in 1 mg/ml G418 (GIBCO-BRL). Cell lines resistant to G418 werecloned, propagated, and evaluated. L450-2, another CYP2Bl-expressing 9Lgliosarcoma-derived cell line, was prepared by similar methods, and was

kindly provided by Dr. Antonio Chiocca (Massachusetts General Hospital).To test for drug sensitivity, 1 X IO5 cells were plated in 30-mm tissue

culture plates (Falcon 3046) in duplicate. Drugs were added 18â€”24h after

seeding. Cells were allowed to grow for times ranging up to 5 days after drugtreatment, and the final cell number was then determined. Cells were rinsedwith PBS or HBSS, dispersed using trypsin-EDTA (GIBCO-BRL), and then

counted with a hemocytometer. Results are expressed as a growth ratio, i.e., the

number of cells in plates containing a drug as a percentage of the corresponding drug-free controls (mean Â±range of duplicate determinations).

Coculture Experiments. Parental 9L cells were plated in the bottom wellsof culture plates (Falcon 3502), and CYP2B1-negative or CYP2B1-positivecells were plated into 25-mm cell culture inserts (0.45 /j.m pore size; Falcon3090). Culture media were removed 18-24 hr later by aspiration. Culture

medium without drug (1.0 ml) was added to the bottom well, and 1.0 mlmedium containing drug was added to the upper cell culture insert. Cellnumbers were determined 5 days later, as described above.

Tumor Growth Delay Studies. 9L cells were grown s.c. as solid tumors infemale Fischer 344 rats using procedures approved by the Boston UniversityInstitutional Animal Care and Use Committee. Adult female Fischer 344 rats(120-150 g) were inoculated at 2 x IO6 cells/s.c. site; CYP2B1-negative cells

(parental 9L or 9L-Z cells) were injected on one thigh and CYP2B1-positivecell (9L-ZP or L450-2 cells) on the other thigh. This strategy was used to

control for any potential effects that s.c. growth of the 9L tumor might have onliver CYP-dependent Cyclophosphamide activation activity. Drug treatment

was initiated 7 days after tumor implantation. Rats were randomized anddivided into two groups. One group was treated with Cyclophosphamide at 100mg/kg body weight given as a single i.p. injection. Another group was givenan injection of saline as control. Tumor size was monitored by caliper measurement at times ranging up to 7-8 weeks, at which point the animals were

sacrificed.Western Blot and CYP2B1 Activity Analysis. Microsomal proteins pre

pared from cultured 9L cells by differential centrifugation (21) were electro-

phoresed through 10% SDS/polyacrylamide gels (20 /ng protein/lane), transferred to nitrocellulose, and then probed with polyclonal rabbit anti-CYP2Blantibodies (22, 23). Phenobarbital-induced rat liver microsomes (1 jag) wereused as a positive control. CYP2Bl-dependent enzyme activity was measuredby monitoring 7-ethoxycoumarin O-deethylation (24) and testosterone16ÃŸ-hydroxylation (23) in isolated 9L microsomal fractions.

RESULTS

Stable Expression of CYP2B1 Gene in 9L Gliosarcoma Cells.9L cells were cotransfected with an expression plasmid encoding ratCYP2B1 and a ÃŸ-galactosidase expression plasmid containing a neo

mycin resistance gene in a 10:1 molar ratio. Cell lines resistant toG418 were selected and cloned. Western blot analysis of isolated 9Lcell microsomes using a rabbit polyclonal antibody specific to CYP2Bshowed a single protein band of approximately 52 kilodaltons corresponding to the molecular mass of purified CYP2B1 in samplesprepared from the clonal cell line designated 9L-ZP. No CYP2B1protein was detected in parental 9L cells or in 9L-Z cells, which wereshown to express ÃŸ-galactosidase (X-Gal staining) but not CYP2B1(Fig. 2; data not shown). The clonal cell lines 9L, 9L-Z, and 9L-ZPwere used for further studies. Analysis of CYP2B1-dependent enzymeactivities (see "Materials and Methods") verified that the CYP2B1

transformant 9L-ZP expresses CYP2B1 protein in an enzymaticallyactive form and at a level corresponding to ~l-2% that of phenobar-bital-induced adult male rat liver, while CYP2B1 activity (testosterone 16ÃŸ-hydroxylation) was not detectable in parental 9L cells or in9L-Z cells (data not shown). A similar level of CYP2B1 expressionwas obtained in several independent 9L/CYP2Bl/ÃŸ-galactosidasetransformants (9L-ZP1, 9L-ZP2, 9L-ZP3, and 9L-ZP6), as well as inseveral transformants that do not express ÃŸ-galactosidase (clonal celllines designated 9L-P3, 9L-P13, and L450-2) (data not shown).

Effects of Oxazaphosphorines on Cultured 9L and 9L/ZP Cells.We first tested whether 9L cells which express CYP2B1 are sensitiveto the cytotoxic effects of Cyclophosphamide and ifosfamide.CYP2B1-positive (9L-ZP and L450-2) and CYP2Bl-negative (parental 9L and 9L-Z) cells were cultured with various concentrations of

Cyclophosphamide or ifosfamide. The number of viable cells present5 days after drug treatment was then determined. As shown in Fig. 3A,Cyclophosphamide inhibited the growth of CYP2Bl-positive cells in aconcentration-dependent manner (IC50 = ~70 P.M). Growth of

582



CYTOCHROME P-45Ãœ-BASED GENE THERAPY

,2B2

V2B1

1 2 3 5 6Fig. 2. Western blol analysis of CYP2B1 in parental 9L cells and in 9L cells that stably

express CYP2B1. Microsomal proteins prepared from cultured cells (20 fAgprotein/lane)were electrophoresed on 10% SDS/polyacrylamide gels, transferred to nitrocellulose, andprobed with polyclonal rabbit anti-CYP2Bl antibodies as described in "Materials andMethods." 9L-ZP1 (Lane 2) corresponds to a second clone derived from the same

selection as 9L-ZP, expresses CYP2B1, and exhibits an oxazaphosphorine sensitivity verysimilar to that of 9L-ZP (data not shown). Phenobarbital-induced rat liver microsomes (0.5

or 1 ftg; Lanes 5 and 6, respectively) were used as a standard for CYP2B1 (2B1, lowerband of doublets in Lanes 5 and 6).

CYP2B1 -positive cells was also inhibited by ifosfamide, but thisrequired a somewhat higher drug concentration (IC50 = â€”145/MM)

(Fig. 3B). These findings are consistent with our earlier observationthat CYP2B1 activates ifosfamide with a 3-4-fold lower catalytic

efficiency (Vmax/ATm)than cyclophosphamide (16). In contrast, parental 9L cells and 9L-Z cells manifested no adverse effects when grown

in the presence of millimolar concentrations of cyclophosphamide orifosfamide. In control experiments, we established that CYP2B1-positive and CYP2B1-negative cells are both inherently sensitive to

activated cyclophosphamide, which we presented to the cells in theform of 4-hydroperoxycyclophosphamide (Fig. 3C).

Effects of P-450 Enzyme Inhibition on Oxazaphosphorine Sensitivity of CYP2Bl-positive 9L Cells. To verify that the expressionof CYP2B1 per se is responsible for the chemosensitivity of theCYP2B1-positive cells to cyclophosphamide and ifophamide, we useda CYP2B1-selective enzyme inhibitor, metyrapone (24), to inhibit

cellular CYP2B1 activity. In the presence of 10 JU.Mmetyrapone, thecytotoxic effects of cyclophosphamide and ifosfamide toward 9L-ZP

cells were nearly eliminated (Fig. 4). By contrast, metyrapone did notblock the cytotoxic effect of the chemically activated derivative,

_,â€” 9L _.â€ž_aâ€”9L_Z

-oâ€”- 9L-ZP â€”uâ€” L450-2

100

1-

A. CPA

=0=

0.5 1.5

O.l-

0.01 0.02

Drug Concentration (mM)Fig. 3. Cytotoxicity of oxazaphosphorines toward CYP2Bl-negative cells (parental 9L

and 9L-Z) and CYP2B1-positive cells (9L-ZP and L450-2). Cells (1 X IO5) plated in

duplicate in 30-mm tissue culture plates were treated with the indicated concentrations ofcyclophosphamide (CPA ), ifosfamide (IFA ), or 4-hydroperoxy cyclophosphamide (4HC)(A-C, respectively). Surviving cells were counted 5 days after beginning drug treatment asdescribed in "Materials and Methods." The effect of drugs on cell survival was expressed

as growth ratio (values shown in units of percentage), i.e., cell number in plates containingdrug as a percentage of the corresponding drug-free controls [mean Â±range (hars) forduplicate determinations]. Final cell number (X 10 4) in drug-free controls = 140 Â±8

(9L), 135 Â±7 (9L-Z), 140 Â±10 (9L-ZP), and 130 Â±6 (L450-2) for each of the indicated

cell lines.

583




100-50-T1OQOZT*|2E0.Ã¼TÃ•iCLCL

Ã¼T

*1ECMÂ£T1|Â£"r

i2p:

E2Ã¼lT,,,,iÃ‡Coi'2E^,n.

Ã¼TT1CL24CLÃ¼ElB[1]12"E^^u.cCTA|Q.2Â£L-ZL-ZPI

15" Â°~

E2isO

"*I

Fig. 4. CYP2B1 enzyme inhibitor metyrapone (MTP) blocks the cytotoxic effects ofcyclophosphamide (CPA) and ifosfamide (/FA) but not 4-hydroperoxy cyclophosphamide(4HC) on CYP2B1-expressing cells. 9L-Z and 9L-ZP cells (1 X 10^) were treated witheither 1 HIMcyclophosphamide, 2 mM ifosfamide, or 10 UM 4-hydroperoxy cyclophosphamide in the absence or presence of 10 /AMmetyrapone, as indicated. Controls receivedno drug treatment. Cell numbers were determined 5 days after beginning drug treatment.Data [mean Â±range (bars) for duplicate determinations] are expressed as growth ratio(percentage) relative to drug-free controls.

4-hydroperoxy cyclophosphamide (Fig. 4), a finding that is consistentwith metyrapone protection via inhibition of CYP2B1-catalyzed ox-

azaphosphorine activation. Therefore, the chemosensitivity ofCYP2Bl-expressing cells to cyclophosphamide and ifosfamide is

dependent on the presence of a functional CYP2B1 enzyme withinthese cells.

Analysis of Bystander Cytotoxicity Effect. We next examinedwhether CYP2Bl-negative 9L cells can be rendered susceptible tocyclophosphamide cytotoxicity when cocultured with CYP2B1-expressing tumor cells. Parental 9L cells and CYP2Bl-positive 9L-ZP

cells were used for these experiments since they have similar doublingtimes in culture. Equal numbers of 9L and 9L-ZP cells were mixed,

and the mixed culture was then treated with cyclophosphamide. Weanticipated that if cyclophosphamide cytotoxicity was restricted to theCYP2B1-positive cells, then the total cell number would be decreasedby approximately 50% compared to drug-free controls, as predicted

on the basis of the selective, but near complete (>90%), cytotoxicityof cyclophosphamide toward 9L-ZP cells, which comprise one-half ofthe mixed cell population. On the other hand, if the CYP2B1-positivecells chemosensitize the adjacent CYP2Bl-negative cells, then both

cell types should be eliminated following treatment of the coculturewith cyclophosphamide. As shown in Fig. 5, more than 80% of thetotal cell population was eradicated when the mixed culture wastreated with cyclophosphamide. Moreover, the CYP2B1 enzyme inhibitor, metyrapone, could largely abrogate this effect. The cells in themixed culture showed a similar pattern of sensitivity to ifosfamide,albeit at a somewhat higher drug concentration. In contrast, there wasno killing of either cell population when 9L-Z cells were mixed with

parental 9L cells (data not shown). These studies demonstrate thatCYP2Bl-positive cells confer a bystander killing effect on adjacentCYP2Bl-negative cells by a mechanism that involves CYP2B1

enzyme activity.

We subsequently monitored the effect of cyclophosphamide treatment on both CYP2B1-negative and CYP2B1-positive cells withinthe mixed cell population. Cells marked with the lacZ gene (ÃŸ-

galactosidase), which can be identified as blue cells after staining thecultures with the ÃŸ-galactosidase substrate X-Gal, were used to dis

tinguish the two types of cells in culture. Equal numbers of parental9L cells were mixed with /acZ-marked CYP2Bl-positive cells (9L-

ZP). Following cyclophosphamide treatment, cells were fixed andstained with X-Gal to reveal the cytotoxicity of cyclophosphamide tothe two-cell populations. As illustrated in Fig. 6, cyclophosphamidedramatically inhibited growth of the CYP2B1-positive cells (blue-

stained cells). A substantial, albeit somewhat lower inhibition of thegrowth of the CYP2B1-negative cells (unstained cells) was observed.

The few remaining cells showed marked morphological abnormalitiesand may no longer be viable. The CYP2B1 inhibitor metyraponeprotected both cell types from cyclophosphamide killing; however,microscopic evaluation revealed morphological distortions in some ofthe CYP2B1-positive cells but not in the CYP2B1-negative cells.

These findings indicate that 9L cells that express CYP2B1 are moresusceptible to cyclophosphamide cytotoxicity as a consequence ofprodrug activation that occurs within the tumor cell, but that substantial cytotoxicity toward adjacent CYP2Bl-negative cells also occurs.

We next assessed whether this bystander killing of CYP2B1-negative cells by the adjacent CYP2B1-positive cells requires directcell-cell contact by analogy to the case of HSV-TK-positive and//SV-7"/L-negative tumor cells and ganciclovir treatment (6, 25-27).

100-

(0cc

1o

50-

0D_I-

Q

iSÃ¼

Fig. 5. Cytotoxicity of oxazaphosphorines toward mixed cultures of 9L and 9L-ZPcells. Equal numbers of parental 9L cells were mixed with 9L-ZP cells (total initial cellnumber = 1 X 105/30-mm tissue culture dish). Cells were untreated or were treated with

either 1 HIM cyclophosphamide (CPA) or 2 mM ifosfamide (Â¡FA) in the absence orpresence of 10 (Â¿Mmetyrapone (MTP), as indicated. Cell numbers were determined 5 daysafter beginning drug treatment. Data [mean Â±range (bars) of duplicates] are expressed asgrowth ratio (percentage) relative to drug-free controls. In control experiments, cyclo

phosphamide (1 mM) and ifosfamide (2 HIM)exhibited no significant cytotoxicity towardmixed cultures of 9L and 9L-Z cells (data not shown).

584




A. 9L/lacZ/2Bl + 9L (1:1) B. + CPA (ImM, 5 days) C. + CFA, +MTP

- / -s Triâ€¢_*"j"^ - Aâ€¢-'ii -v^f>.. v r*

Sea

/'*#

r-AT-Ã¬i?Â»:â€¢Â»-'S.'.j=r~>Â».l

-â€¢â€¢i*:^riÂ¿f1 v Ã'RAPÃ•^S^Â¡SÃ•Â« / "* ''â€¢-

Fig. 6. Histochemical analysis of/acZ-marked CYP2B l-positive cells and unmarked CYP2B l-negative cells in a mixed cell population. Equal numbers of parental 9L cells (IX IO5)were mixed with /acZ-marked CYP2Bl-expressing 9L-ZP cells ("9L/lacZ/2Bl") (A) and the effect of cyclophosphamide (CPA) (B) or cyclophosphamide with metyrapone (MTP) (C)

on the surviving cells was assayed as in Fig. 5. Shown are cells fixed in 0.5% glutaraldehyde 5 days after beginning drug treatment and then stained with X-Gal for 4 h to visualizethe 9L-ZP cells (dark staining).

For these experiments, parental 9L cells were seeded in the bottomchamber of Falcon coculture inserts, and either CYP2B1-positive(9L-ZP) or CYP2B1-negative (9L-Z) cells were placed in the top

chamber of the coculture inserts. The two cell populations are physically separated in this coculture system but share the same culturemedium. Cyclophosphamide treatment for 5 days killed not only theCYP2B1-positive 9L cells in the top chamber (data not shown) but

also the parental 9L cells cultured in the bottom chamber (Fig. 7A).The killing of both cell populations could be effectively blocked bythe CYP2B1 inhibitor metyrapone. In contrast, there was no killing ofeither cell population when 9L-Z cells were cocultured with parental

9L cells. To assess whether the bystander killing of coculturedCYP2B1-negative cells is dependent on the number of coculturedCYP2B1-positive cells, a variable number of 9L-ZP cells (rangingfrom IO4 to IO6 cells) was placed in Falcon culture inserts andcocultured with IO5 parental 9L cells. Fig. IB demonstrates that the

cytotoxicity of cyclophosphamide toward the 9L cells in the bottomculture chamber (abscissa) is directly correlated with the initial number of 9L-ZP cells in the top chamber (ordinate). Thus, in the case of

CYP2Bl/cyclophosphamide, the bystander killing effect is at leastpartly due to the transfer to the non-CYP-expressing cells of solublecytotoxic metabolite(s) formed via cytochrome P-450-catalyzed drug

metabolism. This bystander effect is therefore distinct from that of the//SV-rAyganciclovir system, where intimate cell-cell contact is nec

essary for bystander cytotoxicity to occur (26, 27).Effects of CYP2B1 Expression on Cyclophosphamide Sensitiv

ity of 9L Tumors in Vivo. The studies described above establish that9L gliosarcoma cells that are stably transfected to express CYP2B1become highly sensitive to cyclophosphamide and ifosfamide cytotoxicity. We next used these cells as an ex vivo gene transfer model toevaluate in vivo the feasibility of using the CYP2Bl/oxazaphophorinesystem for cancer gene therapy. An in vivo tumor growth delay studywas carried out to compare the cyclophosphamide sensitivity ofCYP2B1-negative 9L tumors to that of CYP2Bl-expressing 9L tu

mors. CYP2B1-negative cells (9L and 9L-Z) and CYP2B1-expressingcells (9L-ZP and L450-2) were grown s.c. as solid tumors in female

Fischer 344 rats. Female rats were used for these experiments sinceliver microsomal cyclophosphamide activation in this animal model iscatalyzed primarily by a single P-450 enzyme (CYP2C6) (15) and at

a rate (15, 16) that is comparable to that observed in human livermicrosomes (17). Cyclophosphamide treatment of CYP2B1-express

ing tumors led to complete inhibition of tumor growth (Table 1; Fig.8). The CYP2Bl-negative 9L tumors showed some growth delay

following cyclophosphamide treatment, but this antitumor effect wasshort-term, with aggressive tumor growth eventually returning. The

temporary growth delay of the parental 9L tumors results from theactivation of cyclophosphamide by cytochrome P-450 present in the

liver, which in the case of adult female rats is catalyzed primarily bycytochrome P-450 form 2C6 (15). These in vivo tumor model studies

establish that intratumoral expression of the CYP2B1 gene and theassociated intratumoral prodrug activation can render solid tumorshighly susceptible to oxazaphosphorine treatment in vivo.

DISCUSSION

Rat 9L gliosarcoma cells were used as a model to assess the utilityof CYP gene transfer as a paradigm for chemosensitization of tumorsby introduction of genes for drug-metabolizing enzymes that activate

known, established cancer chemotherapeutic agents. 9L cells, originated from a rat brain tumor (20), can be grown in culture or can beimplanted either s.c. or intracranially in Fischer 344 rats. 9L cellsexpress cytochrome P-450 reducÃase, which transfers electrons required for all microsomal CYP-dependent enzyme reactions, but

contain little or no endogenous CYP enzyme activity, making themwell suited as a recipient cell line for experiments involving cytochrome P-450 gene transfer. The primary goals of the present studies

were: (a) to evaluate whether expression of CYP2B1 in this gliosarcoma line sensitizes the cells to oxazaphosphorines; (b) to establish

585




Ofio

_^ 100-

0)Ãœ_lO)D>

â€¢E50-

co

m 9L-Z (Top)

9L-ZP (Top)

O3

OO

OiteQo

Q_Ãœ

CLÃœ

Q_O

o

o>

150-

100-

50-

0.5 1 2.5 59L-ZP : 9L Ratio

Fig. 7. Soluble factors are involved in the bystander cytotoxicity of cyclophosphamidetoward CYP2B1-negative cells. A, parental 9L cells (1 X IO5) were plated in the bottom

well of 30-mm culture plates. Upper chambers of Falcon culture inserts were seeded with9L-Z or 9L-ZP cells (1 X IO6) as indicated. The two cell populations were thus separated

by a 0.45-fim pore size membrane which prevents direct contact between the two cell

populations. Cells were treated with 1 mM cyclophosphamide (CPA) with or without 10liM metyrapone (MTP), or received no drug treatment as control. Cell numbers remainingin the bottom chamber were determined 5 days after beginning treatment. B, experimentalsystem is the same as in A, except that the initial number of 9L-ZP cells in the top chamber(Ordinale) was varied from IO4 to IO6, i.e., a ratio ranging from 0.1 to 10 relative to the

initial number of 9L cells in the bottom chamber. Abscissa, final number of 9L cells in thebottom chamber 5 days after growth in the presence of 1 ITIMcyclophosphamide. Data[mean Â±range (bars) of duplicates] are presented as growth ratio (percentage) relative todrug-free controls.

whether adjacent, non-CYP-containing cells become drug sensitivevia a bystander effect; and (c) to ascertain whether this chemosensi-

tization in vitro translates into a therapeutic advantage in vivo in thecontext of an intact liver system, which has the capacity to catalyze

oxazaphosphorine activation at an overall rate that greatly exceedsthat of the tumor itself. Our findings establish that transfer of oxaza-phosphorine-activating CYP2BÃŒgene into 9L tumor cells does indeed

render these cells preferentially susceptible to cyclophosphamide andifosfamide, both in vitro and in vivo, and therefore, it may be usefulfor application to cancer therapy.

In vitro and in vivo studies of the HSV-TK/ganciclovir system haveindicated that //SV-7~/i-transduced cells treated with ganciclovir exerta bystander killing of non-//SV-rA'-transduced cells which they con

tact (6, 25-27). The precise mechanistic basis for the bystander killing

effect remains unclear, but it appears to involve transfer of activatedganciclovir metabolites or other toxic substances through cell-cell

contact. This bystander effect can be of great therapeutic significancebecause it indicates that eradication of the tumor can, in principle, beachieved even if only a subset of a tumor cell population is effectivelytransduced with the drug sensitivity gene. Consequently, we carriedout experiments to model whether CYP gene transfer is associatedwith a bystander effect, i.e., whether CYP-expressing cells can sen

sitize adjacent tumor cells to cyclophosphamide. We observed thatCYP2B1-positive cells do confer a bystander killing of CYP2B1-

negative cells by a mechanism that requires enzymatically activeCYP2B1. This bystander killing effect involves, at least in part,intercellular transfer of soluble cytotoxic metabolite(s), as indicatedby the chemosensitivity conferred by 9L-ZP cells to parental 9L cells,

even when contact between the two cell populations is prevented.Conceivably, the bystander killing that we observed could additionally involve cell-cell contact mechanisms as well. 4-Hydroxycyclo-

phosphamide formed by CYP2B1 (Fig. 1) is believed to be readilydiffusible across cell membranes (28), and it is likely that the releaseof this primary metabolite, or perhaps its cytotoxic decompositionproducts phosphoramide mustard and acrolein, contribute to the lethaleffect of cyclophosphamide on neighboring CYP2B1-negative cells.

Other mechanisms such as the transfer of apoptotic signals from dying9L-ZP cells to 9L cells could also play a role. Further investigations

will be required to elucidate in detail the mechanisms of cell death inthis ex vivo model of CYP gene transfer.

The lack of a requirement for cell-cell contact to achieve CYP2B1/

cyclophosphamide bystander cytotoxicity may represent an importanttherapeutic advantage of cytochrome P-450 gene therapy over the//SV-TO/ganciclovir system by providing for more extensive distri

bution of activated drug within a tumor mass. In addition, unlike//SV-rKVganciclovir, which produces activated metabolites the cyto

toxicity of which is limited to cells in the DNA synthesis (S phase) ofthe cell cycle, the CYP2B1/oxazaphosphorine system generates metabolites that are effective in killing tumor cells in a cell cycle-independent manner. Thus, the toxicity to tumor cells of phosphor-amide mustard-derived interstrand DNA cross-links becomesmanifest at whichever point the tumor cells begin to replicate, result-

Table 1 Effect of cyclophosphamide on CYP2Bl-negati\>e and CYP2B1-positive 9Ltumors grown s.c. in Fischer 344 rats

Complete tumor growth inhibition"

Tumor Saline Cyclophosphamide

9L9L-Z9L-ZPL450-20/110/90/90/110/110/98/911/11

" Rats were injected with 2 X 10* CYP2B1-negative tumor cells (9L or 9L-Z) or

CYP2B1-positive tumor cells (9L-ZP or L450-2). Cyclophosphamide was administered as

a single i.p. injection at 100 mg/kg body weight 7 days after tumor implantation. Thecompleteness of tumor growth inhibition in the cyclophosphamide-treated 9L-ZP andL450-2 tumors was assessed by palpation or by anatomical examination 7-8 weeks after

cyclophosphamide treatment. Results were combined from three independent experiments. Representative tumor growth curves for an experiment involving 9L-Z and 9L-ZPtumors are shown in Fig. 8.

586




1000-

CM

E

coCD 500

O

IO 20 30 40 50

Days After Tumor ImplantationFig. 8. Growth inhibitory effects of cyclophosphamide toward 9L-Z and 9L-ZP tumors

grown in rivo. Female Fisher 344 rats (n = 5) were each inoculated with 9L-Z cells andwith 9L-ZP cells by s.c. injection into the ouler thighs (2 X Hf 9L-Z cells on the rightthigh and 2 X 10* 9L-ZP cells on the left thigh of each rat). Seven days after tumor

implantation, each rat received a single i.p. injection of cyclophosphamide (100 mg/kgbody weight) or saline as control. Tumor areas were measured until the rats weresacrificed. Data shown are mean Â±SEM (bars) for n = 5 tumors/group.

ing in a higher fractional tumor cell kill. Another potential advantageof cytochrome P-450-based cancer gene therapy is the possibility of

augmenting a local antitumor effect of a drug, via CYP gene transfer,in combination with the selective inhibition of the liver CYP enzymesinvolved in systemic prodrug activation. Since the major liver CYPcatalyst(s) of oxazaphosphorine activation, primarily CYP2C6 in ourfemale rat model (15), are biochemically distinct from CYP2B1, itshould be possible to test the effects of liver CYP inhibition by usingappropriate CYP isoform-selective inhibitors, such as 21,21-dichlo-roprogesterone and 3,5-dicarbethoxy-2,6-dimethy 1-4-ethy1-1,4-dihy-

dropyridine (29, 30). If selective inhibition of the corresponding majorhuman liver P-450 catalysts of oxazaphosphorine activation (17) can

be achieved, this could provide a significant therapeutic advantage byminimizing the host tissue toxicity that results from the liver CYP-

mediated systemic exposure to activated metabolites which invariablyoccurs during conventional chemotherapy.

An important finding of the present study is the demonstration thatCYP2Bl/cyclophosphamide is highly effective with respect to itschemotherapeutic potential in vivo, thus, it is a good candidate forfurther preclinical development as a target for cancer gene therapy.Intratumoral CYP2B1 expression conferred a striking enhancement ofantitumor cytotoxic effect (Fig. 8; Table 1) that was without anyapparent increase in host toxicity. This increased therapeutic effectcould not be fully anticipated for several reasons. First, endogenousCYP enzymes active in cyclophosphamide metabolism are alreadyexpressed at significant levels in liver tissue (15, 16). Moreover, at thetime of cyclophosphamide treatment, 7 days after tumor implantation,the 9L tumors in our studies were just palpable and thus were small insize compared to liver. In addition, the specific content of CYPprotein in the CYP2Bl-transfected tumor cells was low compared to

liver (see Fig. 2). Thus, the major fraction of activated cyclophosphamide in circulation in the 9L-ZP and the L450-2 tumor-bearing rats

is undoubtedly liver derived rather than tumor derived. Nevertheless,

complete tumor regression following cyclophosphamide treatmentwas observed in both of these 9L/CYP2B1 tumors but not in theCYP2B1-negative 9L and 9L-Z tumors. Apparently, there is a verysubstantial "proximity effect" in the case of intratumoral cyclophos

phamide activation. This may indicate that the primary metabolite4-hydroxycyclophosphamide/aldophosphamide (Fig. 1) or perhaps itsplasma protein-stabilized sulfhydro adduci (31) has limited access to

the tumor vasculature or a lower degree of cell membrane permeability when formed in the liver than would be anticipated on the basis ofearlier studies (28, 32). Alternatively, given the short intrinsic half-lifeof 4-hydroxycyclophosphamide [r,/s = 5.2 and 3.3 min in rat and

human plasma, respectively; (32)], significant decomposition of thisprimary metabolite may occur before it reaches the tumor from its siteof generation in the liver. Thus, the effective intratumoral concentration of alkylating metabolites may be substantially higher in the caseof CYP2B1-expressing 9L tumors as compared to non-CYP2Bl-

expressing 9L tumors, despite the much higher inherent metaboliccapacity of liver in these rats.

An additional possibility is that the substantially enhanced cytotox-icity of cyclophosphamide toward CYP2B1-expressing 9L tumors

results from sensitization of the tumor cells to phosphoramide mustard, the primary DNA-alkylating metabolite, by the protein-alkylat-ing metabolite acrolein. Acrolein, derived from 4-hydroxycyclophos-

phamide/aldophosphamide by chemical decomposition, is formed inequimolar amounts with phosphoramide mustard (Fig. 1), and is animportant contributing factor to cyclophosphamide-associated cardio-

toxicity (33) and to the endocrine toxicities that cyclophosphamidecan have, as indicated by its importance for the depletion of serumtestosterone levels and the modulation of liver CYP and steroidmetabolizing enzyme profiles following cyclophosphamide treatment(34). Although acrolein derived from liver cyclophosphamide activation does not mediate the antitumor activity of the parent drug (28), itis conceivable that in our CYP gene transfer/intratumoral cyclophosphamide activation model, acrolein formed locally potentiates thecytotoxic effects of phosphoramide mustard, perhaps by a glutathionedepletion mechanism (compare Refs. 33 and 35). Further studies willbe required to evaluate these possibilities.

In a recent study, the expression of foreign genes within 9L cellswas shown to have a potential negative influence on tumor growth invivo, even in the case of 9L tumors expressing control genes, such asneomycin phosphotransferase (36). While the mechanism underlyingthis effect is still unclear, one possibility is that the protein productsof these foreign genes may act as strong immunogens, thereby decreasing the growth of genetically modified 9L tumors, even in theabsence of cytotoxic drug treatment (36). In our studies, however,genetically modified 9L cells (CYP2B1 or ÃŸ-galactosidase) only

regressed in the case of the CYP2BÃŒgene, and then again, onlyfollowing cyclophosphamide treatment (Fig. 8). In addition, sinceCYP2B1 is an endogenous rat liver protein, the expression with 9Ltumor cells of CYP2B1 seems less likely than a foreign gene toprovide a significant immunogenic stimulus. In this context, the use ofmammalian genes for cancer gene therapy, i.e., rat CYP2B1 for modelstudies using rat tumors or its human counterpart, CYP2B6 [whichalso activates cyclophosphamide (17)] for human tumors, has thepotential advantage of avoiding complications due to possible immu-nological responses associated with the expression of HSV-TK or

other foreign genes.Taken together, the present studies establish a model system for

further investigation of the therapeutic utility of transferring oxaza-phosphorine-activating cytochrome P-450 genes into tumor cells. The

killing of tumor cells by oxazaphosporines is shown to proceed in anefficient manner even if only a subset of a tumor cell population isefficiently transfected with the drug-activating cytochrome P-450

587




gene. A substantial improvement in the therapeutic activity of cyclophosphamide or ifosfamide may thus be anticipated when these drugsare combined with CYP2B1 gene transfer, even if the efficiency ofCYP gene transfer that can be achieved using viral vectors or othergene transfer approaches is less than 100% with respect to genetransduction into tumor cells. This approach also has the potential forbeing superior to alternative pharmacological strategies, e.g., in-tratheal injection of 4-hydroperoxycyclophopsphamide in the case of

brain tumors (37, 38), as a consequence of the sustained, intracellulargeneration of activated oxazaphosphorine metabolites that occurswhen CYP2B1 is expressed within tumor cells. In addition, thechemotherapy/gene therapy concepts and strategies developed in thisstudy may potentially be extended to other established cancer chemo-therapeutic agents (39) and other cytochrome P-450 genes (40) using

any one of a number of approaches to achieve selective, targeted genedelivery to tumor cells. Recombinant retrovirus (6, 25) and recombinant adenovirus (41) have been used as gene transfer vehicles todeliver the HSV-TK gene into brain tumors in situ and to sensitize

brain tumor cells to the prodrug ganciclovir. Recent studies havesuggested that the CYP2Bl/oxazaphosphorine system may also beapplicable to the chemotherapeutic treatment of brain tumors usingsimilar approaches, i.e., local injection of retrovirus carrying theCYP2BI gene, followed by treatment with cyclophosphamide (42). Inthe more general case of systemic tumors, selective delivery of theCYP2BI gene to tumor cells might best be achieved using tumortissue-specific DNA-regulatory sequences to direct the expression ofthe CYP prodrug-activating enzyme within tumor cells. DNA-regulatory sequences from several tumor-expressed genes, including a-fe-

toprotein (hepatoma) (43), tyrosinase (melanoma) (44, 45), the oncogene ERBB2 (pancreatic cancer) (46), carcinoembryonic antigen (lungcancer) (47), and DF3/MUC1 (breast cancer) (48) have been shown todirect selective expression of a prodrug activating enzyme in each ofthe indicated tumor cell types. These or other DNA elements mightalso prove useful in achieving targeted gene therapy in the case ofCYP2Bl/cyclophosphamide or CYP2Bl/ifosfamide. The goal of future studies will be to optimize drug efficacy by increasing thespecificity and selectivity of anticancer oxazaphosphorines, whileminimizing the dose-limiting systemic toxicities traditionally

associated with the use of these agents.

ACKNOWLEDGMENTS

We wish to thank Drs. Milton Adesnik and Allison Reiss, New YorkUniversity Medical School, for providing the expression plasmid pMT2-

CYP2B1 and Dr. Antonio Chiocca, Massachusetts General Hospital, for providing L450-2 cells. Initial experiments described in this paper were carriedout at Dana-Farber Cancer Institute, Boston, Massachusetts.

REFERENCES

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

29

30.1. Israel, M. A. Molecular approaches to cancer therapy. Adv. Cancer. Res., 61: 57-85,

1993.2. Anderson, W. F. Editorial: Gene therapy for cancer. Hum. Gene Ther., 5.- 1-2, 1994.

3. Moolten, F. L., and Wells, J. M. Curability of tumors bearing herpes thymidine kinase 31.genes transferred by retroviral vectors. J. Nail. Cancer Inst., 82: 297-300, 1990.

4. Borrelli, E., Heyman, R., Hsi, M., and Evans, R. M. Targeting of an inducible toxic 32.phenotype in animal cells. Proc. Nati. Acad. Sci. USA, 85: 7572-7576, 1988.

5. Ezzeddine, Z. D., Martuza, R. L., Platika, D., Short, M. P., Malick, A., Choi, B., and 33Breakefield, X. O. Selective killing of glioma cells in culture and in vivo by retrovirustransfer of the herpes simplex virus thymidine kinase gene. New Biol., 3: 608-614,

1991.6. Ram, Z., Culver, K. W., Walbridge, S., Blaese, R. M., and Oldfield, E. H. In situ 34

retroviral-mediated gene transfer for the treatment of brain tumors in rats. CancerRes., S3: 83-88, 1993.

7. Moolten, F. Tumor chemosensitivity conferred by inserted herpes thymidinekinase genes: paradigm for a prospective cancer control strategy. Cancer Res., 46: 355276-5281, 1986.

8. Mullen, C. A., Kilstrup, M. and Blaese, R. M. Transfer of the bacterial gene for

588

cytosine deaminase to mammalian cells confers lethal sensitivity to 5-fluorocytosine:a negative selection system. Proc. Nati. Acad. Sci. USA, 89: 33-37, 1992.

Huber, B. E., Austin, E. A., Good, S. S., Knick, V. C., Tibbels, S., and Richards, C. A.In vivo antitumor activity of 5-fluorocytosine on human colorectal carcinoma cellsgenetically modified to express cytosine deaminase. Cancer Res., 53: 4619-4626,

1993.Mullen, C. A., Coale, M. M., Lowe, R., and Blaese, R. M. Tumors expressing thecytosine deaminase suicide gene can be eliminated in vivo with 5-fluorocytosine andinduce protective immunity to wild type tumor. Cancer Res., 54: 1503-1506, 1994.Field, A., Davies, M., and Dewitt, C. 9-[2-Hydroxy-l-(hydroxymethyl)ethoxy]meth-

yl]guanine: a selective inhibitor of herpes group virus replication. Proc. Nati. Acad.Sci. USA, 80: 4139-4143, 1983.

Smith, K. O., Galloway, K. S., and Kennell, W. L. A new nucleoside analog,9-|2-hydroxy-l-(hydroxymethyl)ethoxy]methyl]guanine, highly active in vitroagainst herpes simplex virus types 1 and 2. Antimicrob. Agents Chemother., 22:55-61, 1982.Smee, D., Martin, J., and Verheyden, J. Anti-herpesvirus activity of the acyclicnucleoside 9-(l,3-dihydroxy-2-proposymethyl)guanine. Antimicrob. Agents Che-mother., 23: 676-682, 1980.Bennett, J. E. Goodman and Oilman's The Pharmacological Basis of Therapeutics.

A. G. Oilman, T. Rail, A. S. Nies, and P. Taylor (eds),. Vol. 8. pp. 1165-1181. New

York: Pergamon Press, Inc., 1990.Clarke, L., and Waxman, D. J. Oxidative metabolism of cyclophosphamide: identification of the hepatic monooxygenase catalysts of drug activation. Cancer Res., 49:2344-2350, 1989.Weber, G. F., and Waxman, D. J. Activation of the anti-cancer drug ifosphamide byrat liver microsomal P450 enzymes. Biochem. Pharmacol., 45: 1685-1694, 1993.

Chang, T. K. H., Weber, G. F., Crespi, C. L., and Waxman, D. J. Differentialactivation of cyclophosphamide and ifosphamide by cytochromes P450 2B and 3A inhuman liver microsomes. Cancer Res., 53: 5629-5637, 1993.

Doehmer, J., Seidel, A., Oesch, F., and Glatt, H. R. Genetically engineered V79Chinese hamster cells metabolically activate the cytostatic drugs cyclophosphamideand ifosfamide. Environ. Health Perspect., 88: 63-65, 1990.

Jowett, T., and Wajidi, M. F. Mammalian genes expressed in drosophila: a transgenicmodel for the study of mechanisms of chemical mutagenesis and metabolism. EMBO,10: 1075-1081, 1991.

Barker, M., Hoshino, T., Gurcay, O., Wilson, C. B., Nielsen, S. L., Downie, R., andEliason, J. Development of an animal brain tumor model and its response to therapywith l,3-bis(2-chloroethyl)-l-nitrosourea. Cancer Res., 33: 976-986, 1973.

Waxman, D. J., Lapenson, D. P., Morrissey, J. J., Park, S. S., Gelboin, H. V.,Doehmer, J., and Oesch, F. Androgen hydroxylation catalyzed by a cell line (SDÃŒ)that stably expresses rat hepatic cytochrome P-450 PB-4 (UBI). Biochem. J., 260:81-85, 1989.Waxman, D. J., and Walsh, C. Phenobarbital-induced rat liver cytochrome P450.

Purification and characterization of two closely related isozymic forms. J. Biol.Chem., 257: 10446-10457, 1982.Waxman, D. J. Rat hepatic P450IIA and P450IIC subfamily expression using catalytic, immunochemical, and molecular probes. Methods Enzymol., 206: 249-267,

1991.Waxman, D. J., and Walsh, C. Cylochrome P-450 isozyme 1 from phenobarbital-

induced rat liver: purification, characterization, and interactions with metyrapone andcytochrome b5. Biochemistry, 22: 4846-4855, 1983.

Culver, K. W., Ram, Z., Wallbridge, S., Ishii, H., Oldfield, E. H., and Blaese, R. M.In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors [see comments). Science (Washington DC), 256: 1550-1552,

1992.Freeman, S. M., Abboud, C. N., Whartenby, K. A., Packman, C. H., Koeplin, D. S.,Moolten, F. L., and Abraham, G. N. The "bystander effect": tumor regression when a

fraction of the tumor mass is genetically modified. Cancer Res., 53: 5274-5283, 1993.

Bi, W. L., Parysek, L. M., Warnick, R., and Stambrook, P. J. In vitro evidence thatmetabolic cooperation is responsible for the bystander effect observed with HSV tkretroviral gene therapy. Hum. Gene Ther., 4: 725-731, 1993.Sladek, N. E. Metabolism of oxazaphosphorines. Pharmacol. Ther., 37: 301-355,

1988.Halpert, J., Jaw, J. Y., Cornfield, L. J., Balfour, C., and Mash, E. A. SelectiveÂ¡nactivationof rat liver cytochromes P450 by 21-chlorinated steroids. Drug Metab.Dispos., 17: 26-31, 1989.Correla, M. A., Yao, K., Wrighton, S. A., Waxman, D. J., and Rettie, A. E.Differential apoprotein loss of rat liver cytochromes P450 after their inactivation by3,5-dicarbethoxy-2,6-dimethyl-4-ethyl-l,4-dihydropyridine: a case for distinctproteolytic mechanisms? Arch. Biochem. Biophys., 294: 493-503, 1992.Hohorst, H. J., Bielicki, L., and Voelcker, G. The enzymatic basis of cyclophosphamide specificity. Adv. Enzyme Regul., 25: 99-122, 1986.Hong, P. S., Srigritsanapol, A., and Chan, K. K. Pharmacokinetics of 4-hydroxycy-clophosphamide and metabolites in Ihe rat. Drug Metab. Dispos., 19: 1-7, 1991.

Friedman, H. S., Colvin, O. M., Aisaka, K., Popp, J., Bossen, E. H., Reimer, K. A.,Powell, J. B., Hilton, J., Gross, S. S., Levi, R., Bigner, D. D., and Griffith, O. W.Glutathione protects cardiac and skeletal muscle from cyclophosphamide-inducedtoxicity. Cancer Res., 50: 2455-2462, 1990.

Chang, T. K., and Waxman, D. J. Cyclophosphamide modulates rat hepatic cytochrome P450 2C11 and steroid 5 a-reductase activity and messenger RNA levels

through the combined action of acrolein and phosphoramide mustard. Cancer Res.,53: 2490-2497, 1993.

Gurtoo, H. L., Hipkens, J. H., and Sharma, S. D. Role of glutathione in themetabolism-dependent toxicity and chemotherapy of cyclophosphamide. Cancer Res.,41: 3584-3591, 1981.




36. Tapscott, S. J., Miller, A. D., Olson, J. M., Berger, M. S., Groudine, M., and Spence,A. M. Gene therapy of rat 9L gliosarcoma tumors by transduction with selectablegenes does not require drug selection. Proc. Nati. Acad. Sci. USA, 91: 8185-8189,

1994.37. Phillips, P. C., Than, T. T., Cork, L. C., Hilton, J., Carson, B. S., Colvin, O. M., and

Grochow, L. B. Intrathecal 4-hydroperoxycyclophosphamide: neurotoxicity, cÃ©rÃ©bro-

spinal fluid pharmacokinetics, and antitumor activity in a rabbit model of VX2leptomeningeal carcinomatosis. Cancer Res., 52.- 6168-6174, 1992.

38. Friedman, H. S., and Oakes, W. J. New therapeutic options in the management ofchildhood brain tumors. Oncology, 6: 27-36, 1992.

39. LeBlanc, G. A., and Waxman, D. J. Interaction of anticancer drugs with hepaticmonooxygenase enzymes. Drug Metab. Rev., 20: 395-439, 1989.

40. Nelson, D. R., Kamataki, T., Waxman, D. J., Guengerich, F. P., Estabrook, R. W.,Feyereisen, R., Gonzalez, F. J., Coon, M. J., Gunsalus, I. C, Gotoh, O., Okuda, K.,and Nebert, D. W. The P450 superfamily: update on new sequences, gene mapping,accession numbers, early trivial names of enzymes, and nomenclature. DNA CellBiol., 12: 1-51, 1993.

41. Chen, S. H., Shine, H. D., Goodman, J. C., Grossman, R. G., and Woo, S. L. Genetherapy for brain tumors: regression of experimental gliomas by adenovirus-mediatedgene transfer in vivo. Proc. Nati. Acad. Sci. USA, 91: 3054-3057, 1994.

42. Wei, M. X., Tamiya, T., Chase, M., Boviatsis, E. J., Chang, T. K. H., Kowall, N. W.,Hochberg, F. H., Waxman, D. J., Breakefield, X. O., and Chiocca, E. A. Experimental

tumor therapy in mice using the cyclophosphamide-activating cytochrome P450 2B1gene. Hum. Gene Trier., 5: 969-978, 1994.

43. Huber, B. E., Richards, C. A., and Krenitsky, T. A. Retroviral-mediated gene therapy

for the treatment of hepatocellular carcinoma: an innovative approach for cancertherapy. Proc. Nati. Acad. Sci. USA, 88: 8039-8043, 1991.

44. Vile, R. G., and Hart. I. R. In vitro and in vivo targeting of gene expression tomelanoma cells. Cancer Res., 53: 962-967, 1993.

45. Vile, R. G., and Hart, I. R. Use of tissue-specific expression of the herpes simplex

virus thymidine kinase gene to inhibit growth of established murine melanomasfollowing direct intratumoral injection of DNA. Cancer Res., 53: 3860-3864, 1993.

46. Harris, J. D., Gutierrez, A. A., Hurst, H. C., Sikora, K., and Lemoine, N. R. Genetherapy for cancer using tumor-specific prodrug activation. Gene Therapy, 1:170-175, 1994.

47. Osaki, T., Tanio, Y., Tachibana, I., Hosoe, S., Kumagai, T., Kawase, I., Oikawa, S.,and Kishimoto, T. Gene therapy for carcinoembryonic antigen-producing human lungcancer cells by cell type-specific expression of herpes simplex virus thymidine kinasegene. Cancer Res., 54: 5258-5261, 1994.

48. ManÃ³me, Y., Abe, M., Hagen, M. F., Fine, H. A., and Kufe, D. W. Enhancersequences of the DF3 gene regulate expression of the herpes simplex virus thymidinekinase gene and confer sensitivity of human breast cancer cells to ganciclovir. CancerRes., 54: 5408-5413, 1994.

589



1995;55:581-589. Cancer Res Ling Chen and David J. Waxman Gene Therapy StrategyTransfer: Development of a Combined Chemotherapy/Cancerof Oxazaphosphorines following Cytochrome P-450 Gene Intratumoral Activation and Enhanced Chemotherapeutic Effect

Updated version

http://cancerres.aacrjournals.org/content/55/3/581

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/55/3/581To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



intratumoral activation and enhanced chemotherapeutic ... · therapy strategy1 ling chen2 and david...

Documents