[CANCER RESEARCH 56, 1331-1340. March 15. 1996] Sensitization of Human Breast Cancer Cells to Cyclophosphamide and Ifosfamide by Transfer of a Liver Cytochrome P450 Gene1
Ling Chen, David J. Waxman,2 Dongshu Chen, and Donald W. Kufe
Division of Cancer Pliarrnacology, Dana-Farber Cancer Institute, Han'artl Medical School. Boston. Massachusetts 02115 [L. C., D. C., D. W. K.]. and Division of Cell and Molecular Biology, Department of Biology, Boston University, Boston, Massachusetts 02215 [L. C., D. J. W.j
ABSTRACT based on the introduction of genes such as HSV-rk3' or bacterial CD, which render mammalian cells sensitive to the otherwise nontoxic The cancer chemotherapeutic agent Cyclophosphamide (CPA) and its antiviral agent and nucleoside analogue ganciclovir (3, 4) or to the isomer ifosfamide (IFA) are alkylating agent prodrugs that require me antifungal drug 5-fluorocytosine (5-7), respectively. This general tabolism by liver cytochrome P450 (P450) enzymes for antitumor activity. The therapeutic effectiveness of these oxazaphosphorines is limited by the strategy may be extendable to conventional anticancer drugs, in par hematopoietic, renal, and cardiac toxicity that accompanies the systemic ticular those nontoxic prodrugs that are subject to enzymatic activa distribution of liver-derived activated drug metabolites. Transfer of a liver tion by endogenous drug-metabolizing enzymes (8). Transfer of genes cytochrome P450 gene, CYP2BI, into human breast MCF-7 cancer cells is encoding such prodrug-activating enzymes directly into tumor cells presently shown to greatly sensitize these cells to oxazaphosphorine tox may conceivably augment the endogenous activity associated with icity as a consequence of the acquired capacity for intratumoral CPA and extratumoral (e.g., liver) drug metabolism, thereby providing for a IFA activation. Thus, CPA and IFA were highly cytotoxic to MCF-7 cells more effective anticancer effect (9). following stable transfection of CYP2BJ but exhibited no toxicity to pa The oxazaphosphorine CPA and its isomer IFA are cell cycle- rental tumor cells or to a ß-galactosidase-expressing MCF-7 transfectant. independent alkylating agents that have a broad spectrum of activity This cytotoxicity could be appreciably blocked by the CYP2B1 inhibitor metyrapone. Cell cycle analysis revealed that CPA arrested the CYP2B1- against a variety of neoplasms, including breast cancer (10). CPA and expressing cells, but not CYP2B1-negative cells, at G2-M phase. A strong IFA are both devoid of alkylating activity and must first be activated bystander cytotoxicity effect that does not require direct cell-cell contact in the liver by cytochrome P450 (CYP) enzymes to manifest their was mediated by CYP2Bl-expressing MCF-7 cells on non-CYP2Bl cells. latent cytotoxic potential. The therapeutic efficacy of these drugs is Intratumoral CYP2B1 expression conferred a distinct therapeutic advan largely dependent on the patient's liver enzyme function with respect tage when treating MCF-7 tumors grown in nude mice with CPA, as to prodrug activation and on the tolerance of host tissues to the revealed by a 15-20-fold greater in vivo cytotoxicity, determined by tumor systemic distribution of activated metabolites formed in the liver. excision/colony formation assay, and by the substantially enhanced anti- However, administration of agents that induce liver P450 enzymes, tumor activity, monitored by tumor growth delay, for CYP2B1-expressing MCF-7 tumors as compared to CYP2B1-negative control tumors. These such as phénobarbitalor prednisone (11). accelerates the rate but not the net extent of CPA activation and appears to have little impact on enhanced therapeutic effects were obtained without any apparent increase drug efficacy (12-14). Recently, the specific P450 enzymes that in host toxicity. To evaluate the extent to which a CPA/P4SO gene therapy strategy may be generally applicable to other tumor cell types, a replica catalyze the activation of CPA and IFA have been identified in both tion-defective recombinant adenovirus carrying the CYP2B1 gene driven rat and human liver. The P450 enzymes designated CYP2B1, by the cytomegalovirus (CMV) promoter Ad.CMV-2Bl was constructed CYP2C6, CYP2C11, and CYP3A1 contribute to the activation of and used to infect a panel of human tumor cell lines. Ad.CMV-2Bl CPA and IFA in adult rat liver (15, 16), whereas the corresponding infection rendered each of the cell lines highly sensitive to CPA and IFA human gene products CYP2B6 and CYP3A4, in addition to several cytotoxicity, with substantial chemosensitization seen at multiplicities of CYP2C enzymes, have been identified as the catalysts of oxazaphos infection as low as 10. The CPA/P450 prodrug activation system may thus phorine activation in human liver (17, 18). CPA and IFA are hydroxy- serve as a useful paradigm for further development of novel cancer gene lated by these P450s at C-4 to yield 4-hydroxy-CPA and 4-hydroxy- therapy strategies that utilize drug susceptibility genes to significantly potentiate the antitumor activity of conventional cancer chemotherapeutic IFA, respectively, which exist in equilibrium with the corresponding agents. ring-opened aldophosphamides. These primary metabolites then un dergo spontaneous ß-eliminationto yield acrolein and an electrophilic mustard (phosphoramide mustard or ifosphoramide mustard), which is INTRODUCTION the therapeutically active. DNA-alkylating metabolite (10). In addition to their widespread use in conventional cancer chemo Gene therapy using drug susceptibility genes is emerging as a therapy, oxazaphosphorines can be used in high dose regimens to treat promising strategy for cancer chemotherapy. This approach to cancer metastatic breast cancer when combined with autologous bone mar treatment is based on the premise that a significant therapeutic ad row transplantation (19-21). In this setting as well, the therapeutic vantage can be gained by direct transfer to tumor cells of a drug susceptibility gene, or "suicide gene." which encodes an enzyme that efficacy of these drugs is limited by host toxicity as a result of the systemic distribution of activated drug metabolites that have signifi can catalyze the intratumoral activation of an anticancer prodrug ( 1, cant cytotoxic side effects, including cardiac and renal toxicity. In 2). Widely studied model systems that have shown some promise are addition, IFA can have significant neurotoxic side effects, which are mediated by its /V-dealkylation metabolites (22, 23). Since tumor cells Received 9/21/95; accepted 1/17/96. typically do not express detectable levels of CPA- or IFA-activating 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 with CYP enzymes, we have considered the possibility that transduction of 18 U.S.C. Section 1734 solely to indicate this fact. tumor cells with an oxazaphosphorine-activating P450 gene may 1This investigation was supported in part by NIH Grant CA49248 (to D. J. W.) and by Army Grant #DAMD 17-94-J-4394. The content of the information does not necessarily reflect the position or the policy of the government, and no official endorsement should 1The abbreviations used are: HSV-tk, herpes simplex virus thymidine kinase; CD, be inferred. cytosine deaminase; CPA, Cyclophosphamide; IFA, ifosfamide; P450 or CYP. cytochrome •¿ Towhom requests for reprints should be addressed, at Department of Biology. P450: CYP2B1. CYP form 2B1; XTT, 2,3-bis[2-methoxy-4-nitro-5-sultbphenyl|-2//- Boston University, 5 Cummington Street. Boston, MA 02215. Phone: (617) 353-7401; tetrazolium-5-carboxanilide inner salt: CMV, cytomegalovirus; 4HC, 4-hydroperoxy- Fax: (617) 353-7404; E-mail: [email protected]. CPA; ßGal,ß-galactosidase. 1331
Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. ['450 BASCO CASfKR (iliNh TllliKAI'Y sensitize the tumor cells to CPA and IFA by a mechanism that Tumor Growth Delay Measurements. Female homo/ygous (nu+/nu + ) involves direct, intracellular prodrug activation. Consistent with this nude athymic Swiss mice (Taconic, Germanlown. NY), 20-30 g, were used. A possibility, stable transduction of the gene encoding the oxazaphos- single s.c. 17/3-estradiol pellet (1.7 mg. 60-day release pellet; Innovative phorine-activating P450 form CYP2B1 (15) into rat 9L gliosarcoma Research. Toledo. OH) was implanted I day prior to s.c. injection of MCF-7, MCF-7-Z. or MCF-7-P tumor cells in the exponential growth phase (2 X IO7 (24) and C6 glioma cells (25) renders these rodent brain tumor cell lines highly sensitive to CPA and IFA. However, the extent to which cells/0.2 ml/injection site). Each mouse received two tumor implantations, a CYP2B1-negative tumor (parental MCF-7 or MCF-7-Z) on one flank and a CYP2B1 gene transfer confers a therapeutic advantage when treating CYP2B1-expressing tumor (MCF-7-P2 or MCF-7-P9) on the other. Drug tumors of non-central nervous system origin and the potential appli treatment was initiated 4 to 5 weeks later. CPA was given by i.p. injection at cability of this strategy to human cancer cells in an in vivo situation is 150 mg/kg body weight per day for two consecutive days. Tumor size was uncertain. In the present report, we have evaluated the therapeutic determined by external caliper measurement, and host toxicity was monitored impact of transferring the CYP2B1 gene into human MCF-7 breast by body weight measurement. cancer cells when combined with oxazaphosphorine treatment in vitro Tumor Excision Assay. CPA treatment was initiated 6 weeks after tumor and m vivo. In addition, we have constructed a replication-defective implantation, at which time the tumors were approximately 50-KX) mrrr in recombinant adenovirus carrying the CYP2BI gene that facilitates size. The mice were then treated by i.p. injection of CPA at a dose of 150 efficient transduction of tumor cells with CYP2B1. Breast cancer and mg/kg body weight, followed by a second drug injection 20 h later. At 4 h after other tumor cell types transfected with CYP2BI are shown to acquire the second CPA injection, the mice were sacrificed and soaked briefly in 75% high sensitivity to CPA. without a detectable increase in host toxicity. ethanol. The tumors were excised, suspended in DMEM cell culture medium (GIBCO). and minced under sterile conditions. The tumor tissue was then incubated for 15 min at 37°Cwith shaking in a solution of 500 units/ml of MATERIALS AND METHODS collagenase (Sigma) containing 0.2 mg/ml of DNusc (Sigma). The samples were filtered through a Cell Strainer (Fisher Scientific), washed twice with Chemicals. CPA was purchased from Sigma Chemical Co. (St. Louis. DMEM. and then suspended in DMEM supplemented with 10% FCS. The MO). IFA was obtained from Bristol-Myers Squibb (Princeton, NJ). 4-Hy- single-cell suspensions were counted and plated at densities of 10', IO4, and droperoxy-CPA was obtained from Nova Pharmaceutical Corporation IO5 cells/well of a 6-well plate, in duplicate, for a determination of cell (Baltimore, MD). Metyrapone was purchased from Aldrich Chemical Co. viability by a colony-forming assay. Cells were grown for 10-12 days, and (Milwaukee, WI). colonies (>50 cells) were then stained with crystal violet and counted. Results Cell Culture. Human tumor cell lines were obtained from American Type are expressed as the surviving cell fraction ±SEM for the drug-treated groups Culture Collection (Rockville. MD). Cells were grown as monolayers in the compared to untreated controls. The untreated tumor cell suspension had culture media recommended by American Type Culture Collection, supple a plating efficiency (colony-forming activity) of 5 to 9% in separate mented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 experiments. units/ml penicillin, and 100 ng/ml streptomycin. Cells were maintained in a Recombinant Adenovirus. A recomhinant adenovirus carrying the 5% CO,/95% air humidified atmosphere. CYP2B1 coding sequence was prepared by homologous recombination in the Stable Transfection of MCF-7 Cells. MCF-7 cells were cotransfected human embryonic kidney cell line 293 (27). A shuttle plasmid. pCMV-2Bl. with a plasmid containing a rat CYP2B1 cDNA (pMT2-CYP2Bl; kindly was first constructed by excising the HSV-lk gene from the adenovirus shuttle provided by Dr. Milton Adesnik. New York University, New York. NY) and plasmid pCMV-HSV-tk (28) and then cloning in its place the CYP2B1 coding pSV2Neo at a molar ratio of 20:1 using Lipofectin (GIBCO-BRL) according to the manufacturer's instructions. pSV2Neo contains a neomycin phospho- sequence, excised from pMT2-CYP2Bl. The resulting pCMV-2Bl shuttle plasmid contains the CYP2B1 gene under the control of the human CMV transferase gene, which confers resistance to G418. pCMV-ßGal.Neo. which immediate-early promoter and enhancer, followed by SV40 polyadenylation contains the neomycin phosphotransferase gene and the Escherichia coli IticZ signals, and flanked by type 5 adenovirus map units().()-1.3 and 9.3-17.3. This gene (24), was used to transfect MCF-7 cells to establish a cell line that stably shuttle plasmid was transfected by calcium phosphate precipitation into 293 expresses ßGal.Stable transfectants were cloned under selection in 1 mg/mi cells together with plasmid pJM17, which contains an El region insertion at G418 (GIBCO-BRL). Cell lines resistant to G418 were cloned, propagated, 3.7 map units of the adenoviral genome (Ret. 29; pJM17 kindly provided by and evaluated. Cytotoxkity Assay. To test for drug sensitivity. 4 X K)4tumor cells were Dr. Frank Graham. McMaster University, Hamilton. Ontario. Canada). Re comhinant adenovirus was isolated from a single plaque and expanded in 293 plated in each well of a 6-well tissue culture plate (Falcon 3046) in duplicate. cells. The resulting recomhinant adenovirus, Ad.CMV-2BI, was verified by Drugs were added approximately 20 h after seeding. Cells were allowed to PCR and by restriction enzyme digestion. The virus was confirmed to be grow for 6 to 7 days after drug treatment, and the final viable cell number was replication-defective by the absence of a cytopathic effect in a HeLa cell assay. then determined by trypan blue exclusion and cell counting. Cells were rinsed Ad.CMV-ßGal is a structurally similar replication-deficient recombinunt ad with PBS. dispersed using trypsin-EDTA (GIBCO-BRL), and then counted enovirus (a generous gift of Dr. R. Crystal. New York Hospital. New York. with a hemocytometer. In some cases, cell number was assessed using a colorimetrie cell proliferation assay (XTT assay) that measures the mitochron- NY), which contains a ßGalreporter gene. Large scale production of recom binunt adenovirus was performed by growth in 293 cells, followed by purifi drial dehydrogenase activity of viable cells (26). Flow Cytometric Analysis. Following drug treatment under conditions cation by double cesium gradient ultracentrifugation as described (27). The indicated in the text, cells were trypsinized, centrifuged. resuspended in PBS. tilers of purified adenovirus were determined in a spectrophotometer at 260 nm and then gently vortexed while 80% ethanol was added to a final concentration and by plaque assays. of 40%. The fixed cells were treated with RNase A (10 jug/ml) at 37°Cfor 30 Adenovirus-mediated Gene Transduction in Tumor Cells. Cells (0.5- 1 X K)'1)were plated on 30-mm diameter tissue culture plates. Recomhinant min and then resuspended in a solution of 50 ng/ml propidium iodide. Cell adenovirus, Ad.CMV-2B 1 or Ad.CMV-ßGal. was added to the cells at various cycle analysis was performed using a FACScan instrument (Becton Dickin son). Results are presented as the number of cells versus the amount of DNA multiplicities of infection. The transduced cells were replated and treated with as indicated by fluorescence intensity. CPA. Six to 7 days after drug treatment, surviving cells remaining attached lo Coculture Experiments. Parental MCF-7 cells were plated in the bottom the culture plate were counted, and the results were compared to non-drug- wells (30 mm in diameter) of Falcon 6-well coculture plates. CYP2B1- treated control plates. expressing cells (MCF-7-P) or CYP2B1-negative cells (MCF-7-Z) were plated Western Blot Analysis. Cell lysates prepared from cell lines (stably trans in 25-mm cell culture inserts (0.45 /urn pore size: Falcon 3090). Culture fected MCF-7 cells, as well as adenovirus-infected cells) were electrophoresed medium was removed 18-24 h later by aspiration. Culture medium without through 10% SDS-polyacrylamide gels (10-30 fig/well), transferred to nitro drug (1.0 ml) was added to the bottom well, and 1.0 ml medium containing cellulose, and then probed with polyclonal rabbit anti-CYP2Bl antibodies (30, drug at a 2X concentration was added to the upper cell culture insert. Cell 31 ). Phenobarbital-induced rat liver microsomes ( 1 jug) were used as a positive numbers were determined 6 to 7 days later. control tbrCYP2BI protein. 1332
Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY
Immunohistochemistry. MCF-7 cells cultured on 30-mm plates were in fected with Ad.CMV-2Bl at a multiplicity of infection of 50. After 48 h, the A.150-100-50-n-LCPAy1iIiÕiiUfi11IsI1y.fa•P*.MMCF-7D monoluyers were fixed with Histochoice tissue fixature MB (Amresco. Solon. MCF-7-ZH OH I. The monoluyers were then treated with 5% goat serum, washed with PBS, and reacted with anti-CYP2Bl antibody at room temperature for 1 h. P2a After washing in PBS. the bound primary antibody was detected by an immunoperoxidase procedure using the Vectastain ABC kit (Vector Labora PSE2 tories Inc.. Burlingame. CA) and diaminobenzidine tetrahydrochloride as chro- P90 mogen. P26 RESULTS Expression of CYP2B1 Gene in Human Breast Cancer MCF-7 100 250 500 1000 2000 Cells. To establish human breast cancer MCF-7 cell lines that stably IFAà iJSjafi!«t express the CYP2BI gene. MCF-7 cells were cotransfected with plusmids encoding rat CYP2BI and a neomycin-resistant gene. Cells 150-100-50-n.B. resistant to G418 were then selected and clonal lines isolated. Western -& blot analysis of cell ly.sates using anti-CYP2BI antibody revealed a NN Similar results were obtained with the CYP 2BI-expressing clonal r*1*r^* r"** cell lines MCF-7-P5 and -P8 (data not shown). Additional experi li. ul li. O ^ÃŒ"^ enzyme inhibitor (Ref. 32; Fig. 3). In control experiments, metyrapone alone had no effect on cell viability (data not shown). Moreover, metyrapone did not block the •¿ 4—CYP2B2 cytotoxic effect of the chemically activated CPA derivative. 4HC •¿ *~CYP2B1 (Fig. 3). This metyrapone protection effect demonstrates that the presence of a catalytically active CYP2B1 enzyme is required for the oxazaphosphorine chemosensiti vity of CYP2BI -expressing cells to be manifest. Cell Cycle Perturbation by CPA in the CYP2B1 -expressing Cells but not in CYP2Bl-negative Cells. To elucidate the effect of 1 23 4 5 6 7 8 9 10 11 12 13 1415 Fig. I. Immunodetection ol'CYP2Bl in parental MCF-7 cells and in MCF-7 cells that intracellular CPA activation on the cell cycle progression of breast cancer cells, we examined the cytokinetic consequences of CPA stably express CYP2B1. Microsoma] proteins prepared from the indicated cultured cell lines (20 ^ig protein/lane) were analyzed on a Western blot probed with pol\clonal rabbit treatment in CYP2BI-expressing cells, MCF-7-P2 and MCF-7-P9, anti-CYP2B 1 antibodies. Phenobarbital-induced rat liver microsomes ( 1 jig) were used as and in CYP2B1 -negative MCF-7-Z and parental MCF-7 cell controls. a standard for CYP2B1 (lower hand of doublet in Lane 15). Clonal MCF-7-derived cell lines P2. P3. P5. P8. P9. and P26 express CYP2BI protein, whereas parental MCF-7 and Cells were harvested and stained with propidium iodine, and the lacZ-transfecled MCF-7 cells (Z and Z2| do not. distribution of cells at each stage of the cell cycle was then determined 1333 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY the G,-M phase 24 h after CPA treatment. This effect was even more •¿ NoDrug striking 60 and 84 h after CPA treatment (Fig. 4ß).Further evidence ~150-ÕS5ioo-E^= S CPA for a block at the S-phase and the GrM phase was provided by the S CPA + MTP progressive decrease in the number of 0,,-G, cells at 60 and 84 h posttreatment. These findings suggest that the intracellular activation \/ 0 IFA \/ t¡Ii of CPA allows the tumor cells continuing egress out of G,,-G, into the \f\/ H IFA + MTP \»/ S phase but prevents cells from traversing from G-.-M into G()-G|, a j^ ^/ Å’H 4HC process that ultimately leads to cell death. SO-tOO 's ^1IIII / CYP2Bl-expressing Cells Mediate a Bystander Killing Effect. s ID 4HC + MTP SS- We next examined whether CYP2BI-expressing human breast cancer wS,-fe5^5X5iil/ n-LLTTtÃ\' cells mediate a CPA-dependent bystander killing of coculturcd MCF-7 MCF-7-Z MCF-7-P2 MCF-7-P9 CYP2B1 -negative cells. MCF-7-Z cells and MCF-7-P2 and -P9 cells Fig. 3. Meiyrapone (MTP) selectively blocks the cylotoxic effects of CPA and IFA on were used for these experiments since they have similar doubling CYP2BI-expressing cells. MCF-7. MCF-7-Z. MCF-7-P2. and MCF-7-P9 cells (4 X IO4) times in culture (data not shown). MCF-7-Z cells and CYP2B1 were treated with either I m\i CPA, 1 m.MIFA, or 25 JAM4HC in the ahsence or presence of 50 /ÃŒMmetyrapone,as indicated. Corresponding controls received no drug treatment. transfectants were mixed at ratios of 1:1 and 4:1 and then were treated Data (mean ±range for duplicate samples) are expressed as final cell number determined with CPA in cell culture. As shown in Fig. 5A, a significant fraction 6 days after drug treatment. of the total cell population was eradicated when the mixed culture was exposed to CPA. even when the ratio of CYP2B1-negative to by flow cytometry. As is seen in the histograms presented in Fig. 4A, CYP2B1-positive cells was 4:1. Moreover, the CYP2B1 enzyme the cell cycle profile of the CYP2B1 -negative cells was unaffected by inhibitor, metyrapone. could substantially block this effect (data not CPA treatment. By contrast, for CYP2B1-expressing cells, there was shown). Thus, CYP2B1 -positive cells confer a P450 metabolism- a substantial accumulation of S-phase cells, as well as an increase in dependent bystander killing of adjacent CYP2BI-negative cells. A 36h Culture 72h Culture No CPA + CPA (24h) No CPA + CPA (60h) 210 210 140 170 MCF-7 G2-M 220 300 180 190 MCF-7-Z n ...... P2(+CPA) ^) 50- 50- ---A--- P9(-CPA) Ãœ ---A--- P9I+CPA) •¿ E25H 25- -V— MCF-7-Z(-CPA) V MCF-7-Z(+CPA) 0 20 40 60 80 0 20 40 60 80 Hours Post-CPA Treatment 1334 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY Iut i af muorili Expression of CYP2B1 Substantially Enhances 150- A. Mixed culture D -CPAD the Antitumor Effect of CPA in Vivo. In view of the positive results obtained in the in vitro experiments described above, we carried out +CPA¿Iil*N an in vivo tumor growth delay study to compare the CPA sensitivity of CYP2Bl-negative MCF-7 tumors to that of CYP2B1-expressing 10°- MCF-7 tumors grown s.c. as solid tumors in female nude mice. Since x^ the growth of MCF-7 breast cancer is estrogen dependent, a 60-day ß-estradiolrelease pellet was implanted prior to tumor inoculation. co J3 CYP2B1-expressing cells formed solid tumors in nude mice that E exhibited growth rates comparable to the CYP2Bl-negative tumors 50- growing in the same animals (Fig. 6A, upper panel). Moreover, all "05 four tumors responded to CPA given 35 days after tumor implanta O tion, as demonstrated by the cessation of tumor growth from day 35 to day 42 (Fig. 6A, lower panel). However, the CYP2B1 -negative tumors were only partially responsive, because they resumed their _ 1iiii.«I:p h-ii-^ initial growth rate after a delay of 7-10 days. This partial response is most likely due to the activation of CPA by cytochrome P450 en ü CM CM O) 05 a. CL Q_ CL zymes present within the livers of these nude mice. By contrast, substantial tumor regression was observed in the case of the CYP2B1 - positive tumors during a 4-week period following CPA treatment. This enhanced cytotoxicity of CPA toward MCF-7 tumors express "oE75"0)EjCCo-culture D -CPA• ing CYP2B1 was achieved without a detectable increase in host toxicity. This is indicated by comparing the CPA-induced body +CPAI1+ weight loss over a 1-2-week period after drug treatment in nude mice containing CYP2B1-negative tumors versus in mice that additionally host a CYP2B1-positive tumor (Table 1). A similar degree of modest weight loss (up to 12-15% decrease) was observed in both cases. We 50-ÃœE0~ also observed no difference in the antitumor effect of CPA toward CYP2B1-negative tumors grown alone as compared to CYP2B1- negative tumors grown in nude mice that also contained a CYP2B1- 25-CQC~~ positive tumor at a second s.c. site (data not shown). This indicates that the contribution of the CYP2B1-positive tumor to the formation of circulating, cytotoxic drug metabolites is small when compared to that of the host liver. U-1OB.n0 To ascertain whether the enhanced growth delay effect exhibited by MCT1I1F-7-Z +iA»JT,32 + P9 the CYP2B1-expressing tumors reflects enhanced localized drug cy Fig. 5. C YP2B l -expressing cells mediate a bystander cytotoxic effect toward totoxicity during the period immediately following drug treatment, we CYP2BI -negative cells in the presence of CPA. A, cytotoxicity of CPA toward mixed used a tumor excision assay to quantitate, by in vitro colony formation cultures of parental MCF-7 and CYP2B1-expressing MCF-7 cells. MCF-7 cells and MCF-7-Z. -P2. or -P9 cells were mixed in the ratios indicated (total initial cell num activity, antitumor activity induced in vivo over a 24 h period follow ber = 4 X 104/30-mm tissue culture dish). Cells were untreated or were treated with 1 HIM ing CPA treatment. CYP2B1 -expressing tumor cells and CYP2B1 - CPA. Cell numbers (mean ± range of duplicates) were determined 6-7 days after beginning drug treatment. B. parental MCF-7 cells (2 x IO4) were plated in the bottom negative tumor cells were grown in nude mice and then were treated well of 30-mm culture plates. Upper chambers of Falcon culture inserts were seeded with with CPA. The tumors were excised from the mice, dispersed to give the CYP2B1-expressing cells. MCF-7-P2 or MCF-7-P9. or with CYP2B1-negative MCF- a single-cell suspension, and then plated on culture dishes. The 7-Z cells (2 X IO5), as indicated along the X-axis. The two cell populations were thus number of surviving tumor cells was then determined by a colony separated by a 0.45 /urn pore size membrane that prevents direct contact between the two cell populations. Cells were treated with 1 mM CPA or received no drug treatment as formation assay. As shown in two separate experiments presented in control. Cell numbers in the bottom chamber were determined 6 days after drug addition Fig. 65, CPA induced a 15-20-fold greater killing of the CYP2B1- (mean ±range of duplicates). expressing tumor cells as compared to the parental MCF-7 or MCF- 7-Z tumor cells. Together, the tumor growth delay and the tumor To assess whether direct cell-cell contact is required for this by excision studies establish that the intratumoral activation of CPA by stander killing effect, parental MCF-7 cells were seeded in the bottom CYP2B1 renders human breast cancer cells highly susceptible to chamber of a Falcon coculture plate, and either CYP2B1 -positive cells oxazaphosphorine cytotoxicity in vivo without an increase in host (MCF-7-P2 or -P9) or CYP2B1-negative cells (MCF-7-Z) were toxicity. placed in the upper chamber of the coculture inserts. In this manner, Adenovirus-mediated Transduction of the CYP2B1 Gene to the two cell populations are physically separated but share the same Breast Cancer Cells Confers Chemosensitivity to CPA. Recombi culture medium. As shown in Fig. 5B, CPA treatment killed not only nant adenovirus has been widely used as a highly efficient vector to the CYP2Bl-positive cells in the upper Falcon coculture inserts but deliver therapeutic genes to a variety of cell types, both in vitro and also the parental MCF-7 cells cultured in the bottom wells. The killing in vivo (33). To assess whether transduction of CYP2B1 into other of both cell populations was significantly reduced by the CYP2B1 tumor cell lines can sensitize them to CPA, we constructed a repli inhibitor metyrapone (data not shown). In control experiments, there cation-defective recombinant adenovirus that carries the CYP2BI was no killing of either cell population when MCF-7-Z cells were gene. This was accomplished by homologous recombination of a cocultured with parental MCF-7 cells. Thus, cytochrome P450-cata- pCMV-2Bl shuttle plasmid with the adenovirus 5 genome plasmid pJM17 in 293 cells (see "Materials and Methods"). The resultant lyzed CPA activation leads to the formation of soluble cytotoxic metabolite(s) that can mediate a significant bystander killing effect. recombinant adenovirus, Ad.CMV-2Bl, contains an insertion of the 1335 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY CYP2BÃŒgene coding sequence in the adenovirus El region under the Table I Influence iif CYP2BI-expressing tumor an CPA-induced hoily weight /<>.\s control of the CMV promoter. Adenovirus infection of MCF-7 cells, Nude mice carrying CYP2B1-positive or CYP2B1-negative MCF-7 tumors were followed by Western blot analysis using anti-CYP2B 1 antibody, dem treated with two single injections of CPA 4-5 weeks after tumor implantation, as described in Fig. 6/1. The presence of a CYP2B1-positive tumor is shown to have no onstrated the expression of CYP2B1 protein in the cultured tumor significant impact on CPA-induced host toxicity associated with body weight loss. Body cells 48 h after infection with Ad.CMV-2Bl, but not after infection weights generally stabili/ed within 2 weeks of drug treatment and then began to increase. Data shown are mean ±SD for the indicated number of mice. The percentage of weight values relative to / = 0 time point are shown in italics. weight (g) of mice hearing Body weight (g) of mice hearing o tumors-CPA29.3CYP2BI -positive tumors CYP2BI -negative Weeks after CFAtreatment0I2Mice/groupBody CPA30.3 ±0.610030.3 ±1.4I(X)26.5 ±2.2im27.3 ±0.9103.430.8 ±1.387.fi25.9 ±2.487.627.3 ±0.8105.1II+ ±1.085.517+CPA31.2 ±1.587.fi5 E with the ßGalconstruct Ad.CMV-ßGal (data not shown). Moreover, co DISCUSSION Previously, rat gliosarcoma 9L and C6 cells were used to investi gate the impact of cytochrome P450 gene transfer on the responsive ness of rodent tumors to CPA (24, 25). In the present study, human breast cancer MCF-7 cells were used to test whether the paradigm of cytochrome P450 gene transfer/oxazaphosphorine chcmoscnsitization can also be applied to human cancers, as well as to tumors of 100 200 300 non-central nervous system origin. Our primary goals were: (a) to CPA (mg/kg) establish whether the expression of a cytochrome P450 gene, Fig. 6. Evaluation of growth-inhihitory and cytotoxic effects of CPA toward CYP2BI- CYP2B1, sensitizes human breast cancer cells to the prodrugs CPA expressing and CYP2BI-negative tumors grown in a nude mouse model. A, growth delay and IFA; (b) to determine the impact of intracellular CPA activation experiment. Female athymic nude mice were inoculated with either of two CYP2BI- on cell cycle progression; (c) to evaluate the therapeutic potential of expressing cells, MCF-7-P2 and MCF-7-P9, or with the CYP2BI-negative tumor cells MCF-7-Z and parental MCF-7 cells, by s.c. injection of 2 X IO7cells/0.2 ml in the flanks intratumoral CPA activation in vivo using human breast cancer cells of the animals as described in "Materials and Methods." About 5 weeks after tumor grown as solid tumors in a nude mouse model; and (cl) to develop implantation, each mouse received CPA ( 150 mg/kg body weight, daily for 2 days. i.p.). recombinant adenovirus that can efficiently deliver the CYP2BI gene Tumor sizes were measured for 2 months, at which time the animals were sacrificed to avoid complications relating to expiration of the 17ß-estradiolpellets (see "Materials and into human tumor cells for additional preclinical studies of cyto Methods"). Results are presented as the means of four to six individual animals; hurs. SE. chrome P450-based cancer gene therapy. fl. tumor excision assay shown for two independent experiments. Mice bearing parental We found human breast cancer MCF-7 cells to be well suited as a MCF-7 (D). MCF-7-Z (O), MCF-7-P2 (•).and MCF-7-P9 (•)tumors were treated with CPA at 150 mg/kg body weight two times (see "Materials and Methods") or saline as recipient cell line for cytochrome P450 gene transfer studies, since control. Tumors were excised 4 h after the second dose, and single-cell suspensions were these cells contain little or no endogenous CYP enzyme activity but prepared lor assay of colony formation activity measured IO to 12 days later. Results are presented as the survival fractions as compared to control groups (means of n = 4 express significant amounts of NADPH cytochrome P450 reducÃase mice/group for each experiment: burs. SE). (34), which transfers electrons required for all microsomal cyto- 1336 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASF.D CANCER GENE THERAPY B Fig. 7. Immunohistochemical staining of Ad.CMV-2B I-infected and uninfected MCF-7 breast cancer cells with polyclonal anti-CYP2Bi antibody. A, shown are Ad.CMV-2Bl- infected MCF-7 cells exhibiling strong positive cellular CYP2B1 immunostaining in about 25% of cells (large arrows) and more moderate immunostaining in 45-50% of the cells (small arrowhead) 48 h after infection at a multiplicity of 50. X200. B. uninfected MCF-7 cells exhibited no CYP2B1-positive staining. X200. chrome P450-dependent enzyme reactions. Transfer of the oxazaphos- with these findings, Hoechst dye 33528 staining revealed several phorine-activating CYP2B1 gene rendered MCF-7 tumor cells highly morphological characteristics of apoptosis, including nuclear conden susceptible to oxazaphosphorine cytotoxicity, thus confirming and sation, cell shrinkage, and the presence of apoptotic bodies following extending our previous findings in a rat glioma cell model (24). The CPA treatment of CYP2B1 -expressing but not parental MCF-7 cells fact that CYP2B! gene transfer also confers a striking chemosensiti- (data not shown). zation of MCF-7 tumors to CPA treatment in vivo, in the context of an Rats implanted with 9L gliomas that express CYP2B1 exhibit a intact mouse liver which contains a high level of CPA-activating striking growth inhibition following CPA treatment that far exceeds cytochrome P450 (35), lends strong support to our earlier conclusion the modest tumor growth delay effect obtained with non-CYP2Bl- that there is a very substantial proximity effect in the case of intratu- expressing 9L tumors (24). However, a recent report has indicated that moral CPA activation in CYP2B1-expressing tumor cells. This effect 9L tumor cells engineered to express foreign genes can be potentially is not achieved when CPA is activated within the liver during con immunogenic in a syngeneic host (43), raising the question of whether ventional CPA therapy, and this undoubtedly contributes to the strik host immune response, rather than intratumoral drug activation per se, ing positive impact that intratumoral CYP2BI gene transfer has on the is the key factor contributing to the CPA-induced growth inhibition responsiveness of MCF-7 tumors to systemic CPA treatment. that we had observed. To address this question, we used in the present The establishment of stable, CYP2BI-expressing MCF-7 cell lines study an immune-incompetent nude mouse model to evaluate the enabled us to investigate the cytokinetic effects of CPA following its responsiveness of CYP2B1-expressing tumor cells to CPA treatment direct activation within breast cancer target cells, thus extending in vivo. Using this model, we were able to demonstrate a striking earlier cell cycle studies carried out with mafosfamide. a chemically therapeutic benefit of intratumoral CYP expression in the form of a activated CPA derivative (36). Direct, intratumoral CPA activation substantial tumor growth delay effect following CPA treatment; no led to a delayed S-phase progression and G,-M arrest that likely is a such effect was obtained with MCF-7 tumors engineered to express consequence of the DNA cross-links and single-strand breaks that are ßGal.Moreover, using an independent experimental approach, CPA induced by P450-activated CPA (37-39). Conceivably. P450-acti- induced a 15-20-fold greater cytotoxic response in CYP2B1-express vated CPA may induce apoptosis/programmed cell death subsequent ing tumors compared to control MCF-7 tumors when measured by a to the initial DNA alkylation and cell cycle disruption events. Other tumor excision assay 24 h after initiation of drug treatment. Thus, anticancer drugs, including etoposide (40) and cisplatin (41), induce using two distinct assay systems, we demonstrate that expression of G2-phase cell cycle arrest, which in turn leads to activation of an cytochrome P450 gene 2BI by human tumor cells does indeed trans apoptotic cascade. Indeed, the chemically activated CPA derivatives late into a marked therapeutic advantage in vivo. mafosfamide and 4HC cause DNA fragmentation and chromatin Despite the striking enhancement in CPA-dependent antitumor condensation, two characteristics of apoptosis (36, 42). In agreement activity, no additional host toxicity was associated with the presence 1337 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY No Infection 100 Ad.CMV-ßgal (MOI = 50) Ad.CMV-2B1(MOI = 10) --*- Ad.CMV-2B1(MOI = 50) SO- •¿ -T- Ad.CMV-2B1(MOI = 100) 100 £ o ü 100 50- O 50- 0.5 CPA (mM) Fig. 8. Adenoviru.s-medialed transfer oÃ*CYl'2Hlgene to breast tumor cells sensiti/es cells to CPA and IFA. A, human MCF-7 breast cancer cells. li. human ZR-75-1 breast cancer cells. Cells were infected with Ad.CMV-2BI at multiplicities of infection (MOI) ranging from K)-KM), as indicated. Uninfected cells or cells infected with Ad.CMV-ßGal were used as control. About 48 h later, cells were replated n\ 4 x IO4cells/well on 6-well plates. CPA was then added at concentrations ranging from 0 to I m\i. After 7 days incubation, viable cells were counted; hurs. SF. (', human tumor cells MDA-MB231 (a poorly differentiated breast cancer cell line). T98G (a gliohlastoma cell line), and DIJI45 (a prostate cancer cell line) were infected with Ad.CM V-2BI at multiplicities of infection of 50 and treated with I HIMCPA 48 h after infection. The number of surviving cells was determined using an XTT colorimetrie assay. The effects of CPA on cell survival are expressed as growth ralio (%). i.e.. cell number (or XTT activity) in piales containing CPA as a percentage of the corresponding drug-free controls: hars, SE. of CYP2BI within the MCF-7 tumors, as indicated by the similar able by combining P450 gene transfer/CPA therapy with the selective degree of CPA-induced weight loss over a I-2 week period following inhibition of liver P450-catalyzed drug activation. Further preclinical drug treatment in nude mice bearing CYP2BI-negative tumors versus model studies will be required to evaluate the therapeutic potential of mice that additionally contain a CYP2BI-positive tumor. Further this combined modulator approach. evidence that the effects of intratumoral CYP2BI are primarily local The application to human cancer treatment of a gene therapy ized, rather than systemic, is provided by our observation that the strategy using drug susceptibility genes, such as cytochrome P450. modest growth-inhibitory effect of CPA on non-P450-containing imposes two essential requirements: that gene transfer to the tumor be MCF-7 tumors is not enhanced by the presence in the same animals of selective and that it be efficient. Retroviruses have been used to a CYP2BI-positive MCF-7 tumor at a second distal s.c. site. These transduce the HSV-tk gene into tumor cells, both in vitro and in vivo findings are consistent with the fact that the CYP2BI-positive tumor (3, 4, 44, 45). Retroviral vector systems are limited, however, by the is small and contains a very low specific content of P4SO protein as difficulty in producing high viral titers for efficient gene transduction, compared to that of the host liver. Accordingly, intratumoral transfer by the restriction that only actively dividing tumor cells become of the CYP2B1 gene under conditions that are sufficient to enhance infected, and by the potential risks of insertional mutagenesis as a the localized antitumor effect of CPA is unlikely to significantly consequence of the random integration of retrovirus. In this regard, increase host toxicity associated with an increased formation of sys recombinant adenovirus presents a promising alternative vector sys temic, cytotoxic drug metabolites. This suggests that the antitumor tem. Adenovirus can infect a broad spectrum of dividing and nondi- effect of CPA, in the case of CYP2BI-expressing tumors, may largely viding cells and has little or no significant potential for viral integra be independent of liver drug activation, in which case a decrease in tion and insertional mutagenesis (33, 46). The El gene-disrupted host toxicity without compromising antitumor effect may be achiev adenovirus. used in the present study, is replication defective but can 1338 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY be prepared to a high titer in 293 cells, which provide the Eia and Elb bination of the cell cycle nonspecific CPA/P450 system with the gene products required for complementation of the replication defect S-phase-specific ganciclovir/HSV-tk or 5-fluorocytosine/CD system in trans (47, 48). Adenovirus-based in vivo gene therapy is effective may be one possible way to take advantage of the distinct actions of for in vivo trunsduction of the HSV-tk gene for treatment of gliomas these alternative drug/suicide gene combinations in a manner that and head and neck tumors in rodent models (49, 50). Moreover, as could lead to an additive or perhaps a synergistic tumor killing effect, shown in the present study, the replication-defective recombinant as well as decrease the potential for tumor cells to develop resistance adenovirus Ad.CMV-2Bl can sensitize a variety of cancer cell types to these cancer gene therapy regimens. to CPA, including breast, glioma, and prostate cancer cells, suggesting that adenovirus may be widely useful as an efficient and therapeuti- REFERENCES cally effective vector for transfer of drug-activating cytochrome P450 1. Moolten. F. L. Drug sensitivity {"suicide") genes for selective cancer chemotherapy. genes. Recombinant adenovirus constructed with other oxazaphos- Cancer Gene Ther., /.•279-287. 1994. phorine-activating P450 genes, such as human CYP2B6 (17), may be 2. Culver. K. W., and Blaese. R. M. Gene therapy for cancer. Trends Genet., 10: useful in testing whether they provide any added pharmacological or 174-178. 1994. 3. Moolten. F. L.. and Wells. J. M. Curability of tumors hearing herpes thymidine kinase therapeutic benefit beyond that of CYP2B1. These recombinant ade genes transferred by retroviral vectors. J. Nati. Cancer Inst.. #2: 297-300, 1990. novirus may also serve as useful tools for in vivo studies to explore 4. Ezzeddine. Z. D.. Martuza. R. L.. Platika, D., Short, M. P.. Malick, A.. Choi. B.. and approaches to control the tumor cell targeting specificity that will be Breakefield. X. O. Selective killing of glioma cells in culture and in vnw by retrovirus transfer of the herpes simplex virus thymidine kinase gene. New Biol.. 3: 608-614.1991. required for clinical applications of CPA/P450-based cancer gene 5. Mullen. C. A.. Kilstrup. M.. and Blaese. R. M. Transfer of the bacterial gene for therapy. In one potentially useful approach to gene targeting speci cytosine deaminase to mammalian cells confers lethal sensitivity to 5-fluorocytosine: ficity, tumor tissue-specific DNA regulatory sequences could be used a negative selection system. Proc. Nati. Acad. Sci. USA, 89: 33-37, 1992. 6. Huber. B. E.. Austin. E. A., Good. S. S.. Knick. V. C.. Tibbels. S., and Richards. C. A. to direct the selective expression of cytochrome P450 enzymes in In vim antitumor activity of 5-fluorocytosine on human coloréela!carcinoma cells tumor cells. For example, the oncogene ERBB2 (51, 52) and the genetically modified to express cytosine deaminase. Cancer Res., 53: 4619-4626. mucin-like glycoprotein DF3/MUC1 (53) are both overexpressed in 1993. 7. Mullen. C. A.. Coale. M. M.. Lowe. R.. and Blaese, R. M. Tumors expressing the human breast carcinomas under the influence of well-defined DNA cytosine deaminase suicide gene can be eliminated in \'i\'t> with 5-fluorocytosine and regulatory elements (54-57). Construction of recombinant adenovi- induce protective immunity to wild type tumor. Cancer Res.. 54: 1503-1506. 1994. 8. LeBlanc. G. A., and Waxman. D. J. Interaction of anticancer drugs with hepatic ruses containing the CYP2BI gene under the control of these or other monooxygenase enzymes. Drug Metab. Rev.. 20: 395-439, 1989. tumor tissue-specific enhancer/promoter elements may be a useful 9. Chen. L.. and Waxman. D. J. Metabolic activation of anticancer oxazaphosphorines approach in this regard. by cytochrome P450s: development of a mode! for cancer gene therapy. In: M. R. Waterman and M. Hildebrand (eds.). Assessment of the Use of Single Cytochrome The present studies indicate that the CPA/cytochrome P450 system P450 Enzymes in Drug Research. Vol. 13. pp. 57-80. Berlin: Springer-Verlag. 1994. has several distinct advantages as compared to other prodrug/enzyme Sladek. N. E. Metabolism of oxazaphosphorines. Pharmacol Ther.. 37: 301-355. activation systems, such as ganciclovir/HSV-tk and 5-fluorocytosine/ 1988. Pichard. L.. Fahre. I.. Daujat. M.. Domergue, J., Joyeux, H., and Maure!. P. Effect of CD: (a) the P450-based approach uses a class of widely used alky- corticosteroids on the expression of cytochromes P450 and on cyclosporin A oxidase lating agent prodrugs whose chemotherapeutic potential, mechanism activity in primary cultures of human hepatocytes. Mol. Pharmacol.. 41: 1047-1055. 1992. of action, and spectrum of host toxicity are well established. By 12. Sladek, N. E. Therapeutic efficacy of cyclophosphamide as a function of its metab contrast, although ganciclovir/HSV-tk and 5-fluorocytosine/CD have olism. Cancer Res., 32: 535-542, 1972. shown promise in model systems, the ultimate effectiveness of these 13. Faber, O. K.. Mouridsen. H. T.. and Skovsted. L. The biotransformation of cyclo phosphamide in man: influence of prednisone. Acta Pharmacol. Toxicol., 35: drugs for treatment of human cancer patients remains unproven; (b) 195-200. 1974. the HSV-tk and CD genes are derived from nonmammalian sources 14. Moore. M. J. Clinical pharmacokinetics of cyclophosphamide. Clin. Pharmacokinet.. 20: 194-208. 1991. and could potentially elicit a host immune response that interferes 15. Clarke, L.. and Waxman. D. J. Oxidative metabolism of cyclophosphamide: identi with prodrug activation. This problem can be avoided by use of an fication of the hepatic monooxygenase catalysts of drug activation. Cancer Res., 49: endogenous liver cytochrome P450 enzyme, such as rat CYP2B1 in a 2344-2350. 1989. 16. Weber. G. F.. and Waxman. D. J. Activation of the anti-cancer drug ifosfamide by rat rat model (24) or the human counterpart CYP2B6 (which contributes liver microsomal P450 enzymes. Biochem. Pharmaco!., 45: 1685-1694. 1993. to the activation of CPA in human liver; Ret. 17) in human gene 17. Chang. T. K. H.. Weber. G. F., Crespi, C. L., and Waxman, D. J. Differential therapy trials; (<•)incontrast to ganciclovir/HSV-tk (58, 59), CPA/ activation of cyclophosphamide and ifosfamide by cytochromes P450 2B and 3A in human liver microsomes. Cancer Res., 53: 5629-5637, 1993. P450 exhibits a bystander cytotoxic effect that does not require direct 18. Walker. D.. Flinois, J. P., Monkman, S. C.. Beloc. C, Boddy. A. V.. Cholerton, S., cell-cell contact and is mediated by diffusible substances, most likely Daly, A. K.. Lind. M. J.. Pearson. A. D. J.. Beaune, P. H., and [die. J. R. Identification of the major human hepatic cytochrome P450 involved in activation and A'-dechlo- the activated metabolites of CPA. Since this bystander effect is roethylation of ifosfamide. Biochem. Pharmacol.. 47: 1157-1163. 1994. manifest even when as little as 20% of a tumor cell population 19. Pelers. W. P.. Ross. M.. Vredenburgh. J. J.. Meisenberg, B.. Marks. L. B.. Winer, E.. expresses CYP2BI (Fig. 5), a significant therapeutic response may be Kurtzberg. J.. Bast. R. C.. Jr.. Jones, R.. and Stipali. E. High-dose chemotherapy and autologous bone marrow support as consolidation after standard-dose adjuvant ther anticipated, even if in vivo gene transfer to tumor cells is less than apy for high-risk primary breast cancer. J. Clin. Oncol.. //: 1132-1143. 1993. 100% complete, owing to the extensive bystander tumor killing; and 20. Ayash. L. J.. Wright. J. E.. Tretyakov. O.. Gonin. R.. Elias. A.. Wheeler. C., Eder, (d) a fourth distinct advantage of the CPA/P450 system as compared J. P.. Rosowsky. A.. Animan. K., and Frei. E., III. Cyclophosphamide pharmaco kinetics: correlation with cardiac toxicity and tumor response. J. Clin. Oncol., 10: to ganciclovir/HSV-tk and 5-fluorocytosine/CD relates to the cell 995-1000, 1992. cycle dependence of each of these drugs. Ganciclovir and 5-fluoro 21. Chen. T. L.. Passos-Coelho, J. L, Noe. D. A.. Kennedy. M. J.. Black. K. C., Colvin. O. M.. and Grochow, L. B. Nonlinear pharmacokinetics of cyclophosphamide in cytosine are metabolized to activated nucleoside analogues that pref patients with melastatic breast cancer receiving high-dose chemotherapy followed by erentially kill cells in the S-phase of the cell cycle by inhibiting DNA autologous bone marrow transplantation. Cancer Res., 55: 810-816, 1995. polymerase, thereby interfering with DNA synthesis (60). By contrast, 22. Goren. M. P.. Wright, R. K.. Pratt. C. B., and Pell. F. E. Dechloroethylation of ifosfamide and neuroloxicity. Lancet, 2: 1219-1220, 1986. the DNA-alkylating metabolites of CPA and IFA kill tumor cells, in 23. Thigpen. T. Ifosphamide-induced central nervous system toxicity. Gynecol. Oncol.. a cell cycle-independent manner, by inducing DNA cross-links whose 42: 191-192, 1991. cytotoxic potential becomes manifest at whichever point the cells 24. Chen, L.. and Waxman. D. J. Intratumoral activation and enhanced chemotherapeutic effect of oxazaphosphorines following cytochrome P450 gene transfer: development of a begin to replicate. Tumor cells that have entered the quiescent (G0) combined chemotherapy/cancer gene therapy strategy. Cancer Res., 55: 581-589. 1995. state are thus more susceptible to the cytotoxic effects of CPA/P450 25. Wei, M. X.. Tamiya. T., Chase. M.. Boviatsis. E. J.. Chang. T. K. H.. Kowall, N. W., than to those of ganciclovir/HSV-tk or 5-fluorocytosine/CD, resulting Hochberg. F. H.. Waxman, D. J., Breakefield, X. O., and Chiocca, E. A. Experimental tumor therapy in mice using the cyclophosphamide-activaling cytochrome P450 2B1 in a higher fractional tumor killing in the case of CPA/P450. Com gene. Hum. Gene Ther., 5: 969-978, 1994. 1339 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. P450-BASED CANCER GENE THERAPY 26. Scudiere, D. A.. Shoemaker. R. H., Paull. K. D., Monks. A.. Tierney, S.. Nofziger. 44. Caruso. M., Pañis.Y., Gagandeep, S., Houssin, D., Salzmann. J. L., and Klat/mann. T. H.. Currens, M. J., Seniff, D., and Boyd. M. R. Evaluation of a soluble letrazolium/ D. Regression of established macroscopic liver métastasesafter in tint transduction of formazan assay for cell growth and drug sensitivity in culture using human and other a suicide gene. Proc. Nati. Acad. Sci. USA, 90: 7024-7028, 1993. tumor cell lines. Cancer Res.. 48: 4827-4833. 1988. 45. Culver. K. W.. Ram. Z.. Wallbridge. S., Ishii. H.. Oldfield, E. H., and Blaese, R. M. 27. Graham, F. L.. and Prevec. L. Manipulation of adenovirus vectors. In: E. J. Murray In vivo gene transfer with retroviral vector-producer cells for treatment of experi (ed.) Methods in Molecular Biology. Gene Transfer and Expression Protocols. Vol. 7. mental brain tumors. Science (Washington DC). 256: 1550-1552, 1992. pp. 109-127. Clifton. NJ: The Human Press. Inc.. 1991. 46. Kozarsky, K. F.. and Wilson. J. M. Gene therapy: adenovirus vectors. Curr. Opin. 28. Chen. L.. Chen. D., Manóme. Y.. Dong. Y.. Fine. H. A., and Kufe. D. W. Breast Genet. Dev.. 3: 499-503, 1993. cancer selective gene expression and therapy mediated by recombinant adenoviruses 47. Bett, A. J.. Haddara. W.. Prevec, L.. and Graham. F. L. An efficient and flexible containing the DF3/MUCI promoter. J. Clin. Invest.. 96: 2775-2782, 1995. system for construction of adenovirus vectors with insertions or deletions in early 29. McGrory. W. J.. Bautista, D. S., and Graham, F. L. A simple technique for the rescue regions 1 and 3. Proc. Nati. Acad. Sci. USA. 91: 8802-8806. 1994. of early region I mutations into infectious human adenovirus type 5. Virology. 163: 48. Graham. F. L.. Smiley. J., Rüssel.W.C.. and Nairn. R. Characteristics of a human cell 614-617. 1988. line transformed by DNA from human adenovirus type 5. J. Gen. Virol.. 36: 59-72. 30. Waxman, D. J. Rat hepatic cytochrome P-450 isoenzyme 2c. Identification as a male- 1977. specific, developmentally induced steroid 16 a-hydroxylase and comparison to a female- 49. Chen, S. H., Shine, H. D., Goodman. J. C.. Grossman. R. G.. and Woo. S. L. Gene specific cytochrome P-450 isoenzyme. J. Biol. Chem.. 259: 15481-15490. 1984. therapy for brain tumors: regression of experimental glioma* by adenovirus-mediated 31. Waxman. D. J., and Walsh. C. Phenobarbital-induced rat liver cytochrome P450. gene transfer in vim. Proc. Nati. Acad. Sci. USA. 91: 3054-3057. 1994. Purification and characterization of two closely related isozymic forms. J. Biol. 50. O'Malley. B. W. J., Chen, S. H.. Schwartz, M. R., and Woo. S. L. C. Adenovirus- Chem., 257: 10446-10457, 1982. mediated gene therapy for human head and neck squamous cell cancer in a nude 32. Waxman, D. J., and Walsh. C. Cytochrome P-450 isozyme 1 from phenoharbital- mouse model. Cancer Res.. 55: 1080-1085. 1995. induced rat liver: purification, characterization, and interactions with metyrapone and 51. Harris. J. D., Gutierrez, A. A.. Hurst. H. C., Sikora, K.. and Lemoine, N. R. Gene cytochrome b5. Biochemistry. 22: 4846-4855. 1983. therapy for cancer using tumour-specific prodrug activation. Gene Ther.. /: 170-175. 33. Ali, M.. Lemoine, N. R., and Ring, C. J. A. The use of DNA viruses as vectors for gene therapy. Gene Ther., /: 367-384. 1994. 1994. 34. Chen. G.. and Waxman. D. J. Identification of glutathione S-transferase as a deter 52. Slamon. D. J.. Clark, G. M.. Wong. S. G., Levin. W. J.. Ullrich. A., and McGuire. minant of 4-hydroperoxycyclophosphamide resistance in human breas! cancer cells. W. L. Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science (Washington DC), 235: 177-182. 1987. Biochem. Pharmacol., 49: 1691-1701. 1995. 53. Kufe. D. G.. Inghirami. M., Abe. D.. Hayes. D.. and Justi-Wheeler, H. Differential 35. Cox, P. J., Farmer, P. B., Jarman. M.. Kinas, R. W.. and Stec. W. J. Stereoselectivity reactivity of a novel monoclonal antibody (DF3) with human malignant versus benign in the metabolism of the enantiomers of cyclophosphamide in mice. rats, and rabbits. breast tumors. Hybridoma, 3: 223-232. 1984. Drug Metab. Dispos. Biol. Fate Chem., 6: 617-622, 1978. 54. Hollywood, D. P., and Hurst, H. C. A novel transcription factor. OB2-1, is required 36. Bullock, G.. Tang, C.. Tourkina, E., Ibrado, A. M., Lutzky. J., Huang. Y.. Mahoney. M. E., and Bhalla. K. Effect of combined treatment with interleukin-3 and interleu- for overexpression of the proto-oncogene c-erhB-2 in mammary tumour lines. EMBO kin-6 on 4-hydroperoxycyclo-phosphamide-induced programmed cell death or apop- J., 12: 2369-2375, 1993. tosis in human myeloid leukemia cells. Exp. Hematol., 21: 1640-1647. 1993. 55. Kovarik, A.. Peat, N., Wilson, D., Gendler, S. J., and Taylor-Papadimitriou, J. 37. Hengstler, J. G., Fuchs, J., and Oesch, F. DNA strand breaks and DNA cross-links in Analysis of the tissue-specific promoter of the MVC1 gene. J. Biol. Chem., 26 1340 Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research. Sensitization of Human Breast Cancer Cells to Cyclophosphamide and Ifosfamide by Transfer of a Liver Cytochrome P450 Gene Ling Chen, David J. Waxman, Dongshu Chen, et al. Cancer Res 1996;56:1331-1340. Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/56/6/1331 E-mail alerts Sign up to receive free email-alerts related to this article or journal. Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected]. Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/56/6/1331. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site. Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1996 American Association for Cancer Research.