Effects of hypoxia and limited diffusion in tumor cell microenvironment on bystander effect of P450 prodrug therapy.

Cytochrome P450 (CYP) enzyme 2B1 metabolizes the anticancer prodrug cyclophosphamide (CPA) to 4-hydroxy-CPA, which decomposes to the cytotoxic metabolites acrolein and phosphoramide mustard. We have evaluated the bystander cytotoxicity of CPA in combination with CYP2B1 gene-directed enzyme prodrug therapy using a cell culture-based agarose overlay technique. This method mimics the tumor microenvironment by limiting the diffusion of metabolites and by reducing the oxygen concentration to levels similar to those found in solid tumors. Under these conditions, the CYP activity of CYP2B1-expressing tumor cells was decreased by 80% compared to standard aerobic conditions. Despite this decrease in metabolic activity, a potent bystander effect was observed, resulting in up to 90% killing by CPA of a tumor cell population comprised of only approximately 20% CYP-expressing tumor cells. Similarly, transient transfection of a small fraction ( approximately 14%) of a human hepatoma Huh7 cell population with a CYP2B1 expression plasmid followed by short-term treatment with CPA (5 h) led to an eradication of 95% of the cells. No such bystander effect was observed without the agarose overlay. These findings suggest that the agarose overlay technique is very useful as an in vitro test system for investigation of the bystander effect of CYP/CPA and other enzyme/prodrug combinations under conditions that mimic the hypoxic conditions present in solid tumors in vivo.

Combination of the bioreductive drug tirapazamine with the chemotherapeutic prodrug cyclophosphamide for P450/P450-reductase-based cancer gene therapy.

Tirapazamine (TPZ) is a bioreductive drug that exhibits greatly enhanced cytotoxicity in hypoxic tumor cells, which are frequently radiation-resistant and chemoresistant. TPZ exhibits particularly good activity when combined with alkylating agents such as cyclophosphamide (CPA). The present study examines the potential of combining TPZ with CPA in a cytochrome P450-based prodrug activation gene therapy strategy. Recombinant retroviruses were used to transduce 9L gliosarcoma cells with the genes encoding P450 2B6 and NADPH-P450 reductase. Intratumoral coexpression of P450 2B6 with P450 reductase sensitized 9L tumor cells to CPA equally well under normoxic (19.6% O2) and hypoxic (1% O2) conditions. The P450 2B6/P450 reductase combination also sensitized 9L tumor cells to TPZ under both culture conditions. Interestingly, bystander cytotoxic effects were observed for both CPA and TPZ under hypoxia. Furthermore, TPZ exerted a striking growth-inhibitory effect on CPA-treated 9L/2B6/P450 reductase cells under both normoxia and hypoxia, which suggests the utility of this drug combination for P450-based gene therapy. To evaluate this possibility, 9L tumor cells were transduced in culture with P450 2B6 and P450 reductase and grown as solid tumors in severe combined immune deficient mice in vivo. Although these tumors showed little response to TPZ treatment alone, tumor growth was significantly delayed, by up to approximately four doubling times, when TPZ was combined with CPA. Some toxicity from the drug combination was apparent, however, as indicated by body weight profiles. These findings suggest the potential benefit of incorporating TPZ, and perhaps other bioreductive drugs, into a P450/P450 reductase-based gene therapy strategy for cancer treatment.

Cyclophosphamide modulates rat hepatic cytochrome P450 2C11 and steroid 5 alpha-reductase activity and messenger RNA levels through the combined action of acrolein and phosphoramide mustard.

Cyclophosphamide treatment of adult male rats leads to sustained decreases in several liver microsomal cytochrome P450 (CYP) activities, including CYP 2C11-catalyzed cyclophosphamide activation, via a process that is associated with a feminization of the overall pattern of liver enzyme expression (G. A. LeBlanc and D. J. Waxman, Cancer Res., 50:5720-5726, 1990). The present study compares the effects of cyclophosphamide and its isomeric analogue ifosphamide on the gender-dependent expression of hepatic CYP 2C11 and steroid 5 alpha-reductase in adult male rats and also examines the role of the cyclophosphamide metabolites acrolein and phosphoramide mustard in feminizing the expression of these liver enzymes. Ifosphamide (a) suppressed the male-specific CYP 2C11 mRNA and CYP 2C11-catalyzed liver microsomal testosterone 2 alpha-hydroxylation and cyclophosphamide and ifosphamide 4-hydroxylation and (b) elevated the female-dominant liver enzyme steroid 5 alpha-reductase and its mRNA 7-9 days after drug treatment, both occurring in a manner similar to that of cyclophosphamide, but requiring a 50% higher dose (180 mg/kg, single i.p. injection) to achieve these effects. This pattern of response could not be achieved by treatment of rats with acrolein or with cyclophosphamide analogues that decompose to acrolein without formation of phosphoramide mustard. In contrast, phosphoramide mustard treatment (100 mg/kg) did modulate microsomal CYP 2C11 and steroid 5 alpha-reductase activities. Treatment with a lower dose (50 mg/kg) of phosphoramide mustard or with the acrolein precursor 4-hydroperoxydechlorocyclophosphamide (200 mg/kg) alone did not affect liver enzyme expression, whereas the combination of these agents produced an overall pattern of response that was similar to that conferred by cyclophosphamide. These studies establish that ifosphamide is less potent than cyclophosphamide in modulating the pattern of cytochrome P450 and steroid 5 alpha-reductase expression and that phosphoramide mustard is responsible for the modulation of liver enzyme expression by cyclophosphamide, with acrolein potentiating the modulating activity of the mustard.

Intratumoral activation and enhanced chemotherapeutic effect of oxazaphosphorines following cytochrome P-450 gene transfer: development of a combined chemotherapy/cancer gene therapy strategy.

Cyclophosphamide and its isomer ifosfamide are cell cycle-nonspecific alkylating agents that undergo bioactivation catalyzed by liver cytochrome P-450 enzymes. The therapeutic efficacy of these oxazaphosphorine anticancer drugs is limited by host toxicity resulting from the systemic distribution of activated drug metabolites formed in the liver. Since tumor cells ordinarily do not have the capacity to activate oxazaphosphorines, we examined whether introduction into tumor cells of a cDNA encoding 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 a cDNA encoding rat CYP2B1 are highly sensitive to cyclophosphamide and ifosfamide cytotoxicity as compared to parental 9L cells or 9L cells transfected with an Escherichia coli beta-galactosidase gene. The CYP2B1 enzyme inhibitor metyrapone protects the CYP2B1-expressing 9L cells from oxazaphosphorine cytotoxicity, demonstrating that the chemosensitivity of these cells is a direct consequence of intracellular prodrug activation. Moreover, CYP2B1-expressing 9L cells potentiate the cytotoxic effects 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 distributing cytotoxic drug metabolites to a wide area within a solid tumor mass. In vivo experiments using Fischer 344 rats implanted s.c. with CYP2B1-expressing 9L tumor cells demonstrated that intratumoral expression of the CYP2B1 gene provides a substantial therapeutic advantage over that provided by liver cytochrome P-450-dependent drug activation alone; cyclophosphamide treatment resulted in complete growth inhibition of CYP2B1-positive tumors, whereas only a modest growth delay effect was obtained with CYP2B1-negative tumors. These studies establish that drug-activating CYP genes may be useful for the development of novel combined chemotherapy/gene therapy strategies for cancer treatment utilizing established cancer chemotherapeutic agents.

N,N',N''-triethylenethiophosphoramide (thio-TEPA) oxygenation by constitutive hepatic P450 enzymes and modulation of drug metabolism and clearance in vivo by P450-inducing agents.

The cancer chemotherapeutic drug N,N',N''-triethylenephosphoramide (thio-TEPA) is oxidatively desulfurated to yield the active metabolite N,N',N''-triethylenephosphoramide (TEPA) in a reaction catalyzed by the phenobarbital-inducible rat liver P450 enzyme IIB1. In the current study, the role of constitutively expressed P450 enzymes in thio-TEPA metabolism was studied using purified P450s, isolated liver microsomes, and intact rats. Metabolism of thio-TEPA (100 microM) to TEPA by uninduced adult female and male rat liver microsomes proceeded at initial rates of 0.10 and 0.28 nmol TEPA formed/min/mg microsomal protein, respectively. Although these rates are low compared to those catalyzed by phenobarbital-induced liver microsomes (3.5 nmol TEPA/min/mg), they are sufficient to contribute to the systemic metabolism of this drug. Thio-TEPA metabolism catalyzed by uninduced female liver microsomes was approximately 70% inhibitable by antibodies selectively reactive with P450 IIC6. For the uninduced male liver microsomes, which exhibit a severalfold higher rate of thio-TEPA metabolism, enzyme activity was only 15-20% inhibitable by these antibodies but was 80-85% inhibited by an anti-P450 IIC6 monoclonal antibody cross-reactive with P450 IIC11, which is expressed only in the males. Consistent with these observations, purified P450s IIC11 and IIC6 both oxidized thio-TEPA in reconstituted systems (turnover, 1.1 and 0.3 min-1 P450-1, respectively, at 100 microM substrate), while several other constitutive hepatic P450s exhibited significantly lower or undetectable activities (turnover, less than or equal to 0.15 min-1 P450-1). Metabolism of thio-TEPA by purified P450 IIC11 was associated with a time-dependent inactivation of the cytochrome analogous to that previously shown to accompany thio-TEPA metabolism catalyzed by P450 IIB1. Depletion of hepatic P450 IIC11 by cisplatin treatment of adult male rats led to a 70% reduction of TEPA formation catalyzed by the isolated liver microsomes, suggesting that cisplatin may influence thio-TEPA pharmacokinetics when these two drugs are given in combination. The extent to which hepatic P450s contribute to thio-TEPA metabolism and clearance in vivo was assessed by monitoring thio-TEPA and TEPA pharmacokinetics in rats that exhibit widely differing rates of microsomal thio-TEPA metabolism, i.e., uninduced female and male rats, and male rats treated with the P450 IIB1 inducers clofibrate and phenobarbital. In accord with the microsomal activities, conversion of thio-TEPA to TEPA was less extensive and thio-TEPA elimination slower in female than in male rats.(ABSTRACT TRUNCATED AT 400 WORDS)

Mechanisms of cyclophosphamide action on hepatic P-450 expression.

Cyclophosphamide was administered to adult male rats (130 mg/kg, single i.p. injection) and its effects on the P-450 enzymes that contribute to the activation of this drug in rat liver were then assessed. P-450-mediated cyclophosphamide 4-hydroxylase activity in isolated rat liver microsomes decreased by approximately 70% over a 9-day period following drug treatment. This decrease was due to the loss of cytochrome P-450 form 2c (IIC11), a major contributor to cyclophosphamide 4-hydroxylation in untreated male rat liver, while the other major hepatic cyclophosphamide 4-hydroxylase, P-450 PB-1 (IIC6), was largely unaffected. The loss of P-450 2c activity did not result from a decrease in P-450 reductase or from direct inactivation of the P-450 protein by cyclophosphamide or its metabolites, but rather was due to a reduction in hepatic P-450 2c protein and mRNA levels. Hepatic P-450 2a (IIIA2) and P-450 RLM2 (IIA2) were also suppressed by cyclophosphamide treatment. Serum testosterone, which contributes to the expression of P-450s 2c, 2a, and RLM2, was severely depleted in the cyclophosphamide-treated rats; however, this loss was not the direct cause of the decreases in these hepatic P-450s, since the decreases were not reversed upon restoration of normal testosterone levels by human chorionic gonadotropin stimulation of testicular androgen production. In contrast to the suppression of these testosterone-dependent P-450s, P-450 3 (IIA1), P-450j (IIE1), and the P-450-independent microsomal enzyme steroid 5 alpha-reductase were each elevated in rat liver following cyclophosphamide administration. In contrast to P-450 3 and steroid 5 alpha-reductase, however, the elevation of P-450j protein was transient and was not accompanied by an increase in P-450j-associated hepatic microsomal aniline hydroxylase activity. In vitro experiments revealed that P-450j was severalfold more susceptible to inactivation by the cyclophosphamide metabolite acrolein as compared with P-450 3. These observations suggest that P-450j protein is induced by cyclophosphamide treatment but that the protein is inactivated by the cyclophosphamide metabolite acrolein. These findings establish that cyclophosphamide treatment can modulate hepatic P-450 activities through multiple mechanisms and in a manner that may alter P-450 metabolism of cyclophosphamide and perhaps other anticancer drugs that undergo bioactivation in the liver.

Biotransformation of N,N',N''-triethylenethiophosphoramide: oxidative desulfuration to yield N,N',N''-triethylenephosphoramide associated with suicide inactivation of a phenobarbital-inducible hepatic P-450 monooxygenase.

Oxidative metabolism of the polyfunctional alkylating agent N,N',N''-triethylenethiophosphoramide (thio-TEPA) was studied in isolated rat liver microsomes and purified, reconstituted cytochrome P-450 (P-450) enzyme systems in order to elucidate the pathways of drug oxidation and to identify the possible contributions of individual P-450 enzymes to the bioactivation of this chemotherapeutic agent. Rat liver microsomes were found to catalyze conversion of thio-TEPA to its oxo metabolite, N,N',N''-triethylenephosphoramide (TEPA), in a P-450-dependent reaction that was markedly stimulated by prior in vivo treatment with drug inducers of hepatic P-450 subfamily IIB (phenobarbital), but not by pretreatment with inducers of P-450 subfamilies IA (beta-naphthoflavone) or IIE (isoniazid). Thio-TEPA depletion and TEPA formation catalyzed by phenobarbital-induced liver microsomes were both inhibited by greater than 90% by antibodies selectively reactive with P-450 PB-4 (gene product IIB1), the major phenobarbital-inducible rat liver microsomal P-450 form, but not by antibodies inhibitory toward 7 other rat hepatic P-450s. Oxidation of thio-TEPA to TEPA was also catalyzed by purified P-450 PB-4 (Km (app) 19 microM; Vmax (app) = 11 mol thio-TEPA metabolized/min/mol P-450 PB-4) following reconstitution of the cytochrome with NADPH P-450 reductase in a lipid environment. Metabolism of thio-TEPA by P-450 PB-4 was associated with a suicide inactivation of the cytochrome characterized by kinactivation = 0.096 min-1, KI = 24 microM, and a partition ratio of 136 +/- 28 (SD) mol thio-TEPA metabolized/mol P-450 inactivated. The thio-TEPA metabolite TEPA, however, did not inactivate the cytochrome, nor was it subject to further detectable metabolism. In microsomal incubations, metabolism of thio-TEPA led to the inactivation of P-450 PB-4 (steroid 16 beta-hydroxylase) as well as P-450 IIIA-related enzymes (steroid 6 beta-hydroxylase) and the P-450-independent enzyme steroid 17 beta-hydroxysteroid:NADP+ 17-oxidoreductase, as demonstrated by use of the P-450 form-selective steroidal substrate androst-4-ene-3,17-dione. In contrast, little or no inactivation of microsomal P-450 IIA-related enzymes (steroid 7 alpha-hydroxylase) or microsomal NADPH P-450 reductase was observed.(ABSTRACT TRUNCATED AT 400 WORDS)

Oxidative metabolism of cyclophosphamide: identification of the hepatic monooxygenase catalysts of drug activation.

Cytochrome P-450-catalyzed activation of cyclophosphamide to alkylating metabolites was studied in isolated rat liver microsomes and purified, reconstituted P-450 enzyme systems in order to identify the major enzymatic catalysts of drug activation in both uninduced and drug-induced liver tissue. P-450 form PB-4 (P-450 gene IIB1) activated cyclophosphamide with high efficiency [Vmax (app) = 18.2 nmol metabolite/min/nmol P-450; Km (app) = 0.16 mM] via the formation of 4-hydroxycyclophosphamide, which was quantitatively trapped as a bisulfite adduct then characterized following its conversion to cyano derivatives. Antibodies to P-450 PB-4 inhibited cyclophosphamide activation catalyzed by phenobarbital-induced adult male rat liver microsomes (specific activity, 5.4 nmol metabolite/min/mg liver microsomes) in a selective and near quantitative (greater than 80%) fashion; little or no inhibition was obtained using antibodies inhibitory towards six other rat hepatic P-450 forms. Cyclophosphamide activation catalyzed by uninduced adult male rat liver microsomes (specific activity, 0.68 nmol/min/mg), although not inhibited by anti-P-450 PB-4 antibodies, was partially inhibited (approximately 60%) by antibodies to P-450 PB-1 (gene IIC6) and more completely inhibited (greater than 95%) by antibodies reactive with both P-450 PB-1 and P-450 2c (gene IIC11). Consistent with these observations, P-450 PB-1 and P-450 2c both activated cyclophosphamide at moderate rates in reconstituted systems (turnover, 1.6-2.7 nmol metabolite/min/nmol P-450), while seven other purified hepatic P-450 forms exhibited significantly lower activities (turnover less than or equal to 0.5 nmol metabolite/min/nmol P-450). Further studies revealed that the changes in liver microsomal cyclophosphamide activation rates with age and sex and in response to in vivo administration of cisplatin primarily reflect changes in the levels of P-450 forms PB-1 and 2c. These studies establish that P-450 forms PB-1, 2c, and PB-4 are the major catalysts of cyclophosphamide activation in rat hepatic tissue and that the modulation of microsomal cyclophosphamide activation with development and in response to drug exposure largely reflects alterations in the levels of these three hepatic P-450 enzymes.

Frequent, moderate-dose cyclophosphamide administration improves the efficacy of cytochrome P-450/cytochrome P-450 reductase-based cancer gene therapy.

Transduction of tumor cells with a cyclophosphamide (CPA)-activating cytochrome P-450 (P450) gene provides the capacity for localized prodrug activation and greatly sensitizes solid tumors to CPA treatment in vivo. The therapeutic impact of this P450-based cancer gene therapy strategy can be substantially enhanced by cotransduction of P450 reductase, a rate-limiting component of P450-dependent intratumoral CPA activation. The present study examined the possibility of further improving P450/P450 reductase-based gene therapy by using a novel schedule of CPA administration, involving repeated CPA injection every 6 days and previously shown to have an antiangiogenic component. 9L gliosarcoma cells transduced with the CPA-activating enzyme couple P450 2B6/P450 reductase and grown s.c. in immunodeficient severe combined immunodeficient (scid) mice were repeatedly challenged with 140 mg/kg CPA every 6 days. Full tumor regression leading to eradication of six of eight tumors was observed when the tumor size at the time of initial drug treatment was approximately 400 mm(3) (approximately 1.5% of body weight). Little or no overt toxicity of the repeated CPA treatment regimen was observed. The same CPA schedule was much less effective in inducing regression of 9L tumors that were not transduced with P450/P450 reductase. Repeated CPA treatment of mice bearing large, late-stage P450/P450 reductase-transduced tumors (approximately 9-16% of body weight) resulted in major (> or =95%) regression in 15 of 16 tumors, with tumor eradication observed in 2 cases. Although CPA resistance was found to emerge in the population of P450/P450 reductase-transduced tumors, this resistance primarily involved a loss of expression of the transduced P450 and/or P450 reductase gene, rather than development of intrinsic cellular resistance to the activated form of CPA. These findings demonstrate that repeated CPA treatment on a 6 day schedule can be highly effective when combined with P450/P450 reductase gene therapy and suggest that repeated transduction of tumors with prodrug-activation genes may be necessary to achieve tumor eradication and a sustained therapeutic response.

Potentiation of cytochrome P450/cyclophosphamide-based cancer gene therapy by coexpression of the P450 reductase gene.

Intratumoral expression of cytochrome P450 2B1 sensitizes tumor cells to the cytotoxic action of the alkylating agent prodrug cyclophosphamide (CPA) and provides a novel strategy for cancer gene therapy that may enhance the selectivity and the effectiveness of this class of antitumor drugs [L. Chen and D. J. Waxman, Cancer Res., 55: 581-589, 1995]. P450-catalyzed drug metabolism is obligatorily dependent on electron input from the flavoenzyme NADPH-P450 reductase (RED), which is widely expressed in many cell types, including tumor cells. Here, we investigate the potential utility of combining RED gene transfer with CPA-based P450 gene therapy. Rat 9L gliosarcoma cells stably expressing either basal or elevated (up to 10-fold increase) levels of RED, in the presence or absence of P450 2B1, were selected and characterized. RED overexpression substantially increased the sensitivity of these cells to CPA, but only when combined with P450 2B1 expression. An enhanced cytotoxic response was also obtained when recombinant adenovirus encoding P450 2B1 was used to deliver the P450 gene to RED-overexpressing tumor cells. CPA cytotoxicity was substantially decreased by the RED inhibitor diphenyleneiodonium chloride or by the P450 inhibitor metyrapone, providing evidence of its dependence on the catalytic contributions of both protein components of the P450 metabolic pathway. Conditioned media from P450 2B1-expressing and RED-overexpressing tumor cells treated with CPA exhibited increased formation of the primary 4-hydroxy metabolite and greater cell contact-independent bystander cytotoxic potential compared to tumor cells containing P450 2B1 and basal levels of RED. Evaluation of the impact of P450/RED combination gene therapy using a s.c. solid tumor model/tumor excision assay revealed a dramatic 50-100-fold increase in tumor cell kill in vivo over that provided by liver drug activation alone. These findings establish the importance of endogenous RED levels as a determinant of the sensitivity of tumor cells to CPA/P450 gene therapy and demonstrate the striking therapeutic effectiveness of an anticancer prodrug activation strategy based on the combination of a cytochrome P450 gene with the gene encoding RED.

Retroviral transfer of human cytochrome P450 genes for oxazaphosphorine-based cancer gene therapy.

Cyclophosphamide (CPA) and ifosfamide (IFA) are widely used anticancer prodrugs that are bioactivated in the liver by specific cytochrome P450 enzymes (CYPs). The therapeutic activity of these antitumor agents can be compromised by a low therapeutic index that is, in part, due to the systemic distribution of activated drug metabolites. Here, recombinant retroviruses were used to deliver six different CPA- or IFA-metabolizing human CYP genes to 9L gliosarcoma cells: 2B6, 2C8, 2C9, 2C18 (Met385 and Thr385 alleles), 2C19, and 3A4. Intratumoral cytochrome P450 expression conferred substantial sensitivity to CPA cytotoxicity, with the most dramatic effects seen with CYP2B6. Strong CPA chemosensitivity was also seen following transduction of CYP2C18-Met, despite a very low level of CYP protein expression (>60-fold lower than that of 2B6). In contrast to CPA, the cytotoxicity of IFA was greatest toward tumor cells transduced with CYP3A4, followed by CYPs 2B6 and 2C18-Met. A substantial further increase in chemosensitivity was achieved upon transduction of 2B6 or 2C18-Met-expressing tumor cells with P450 reductase, which provided for more efficient intratumoral prodrug activation and cytotoxicity at lower drug concentrations. With 2B6- plus P450 reductase-transduced tumor cells, CPA but not IFA conferred a strong cell contact-independent bystander cytotoxic effect on non-P450-expressing 9L cells. CPA treatment of tumors that were transduced with 2B6 or 2C18-Met together with P450 reductase and were grown s.c. in immunodeficient mice resulted in a large enhancement of the liver P450-dependent antitumor effect seen with control 9L tumors, with no apparent increase in host toxicity (growth delay of >25-50 days in P450-expressing tumors versus approximately 5-6 days without P450). CYP2B6 plus P450 reductase and CYP2C18-Met plus P450 reductase thus appear to be excellent gene combinations for use with CPA in P450/prodrug activation-based cancer gene therapy.

Differential activation of cyclophosphamide and ifosphamide by cytochromes P-450 2B and 3A in human liver microsomes.

The present study identifies the specific human cytochrome P-450 (CYP) enzymes involved in hydroxylation leading to activation of the anticancer drug cyclophosphamide and its isomeric analogue, ifosphamide. Substantial interindividual variation (4-9-fold) was observed in the hydroxylation of these oxazaphosphorines by a panel of 12 human liver microsomes, and a significant correlation was obtained between these 2 activities (r = 0.85, P < 0.001). Enzyme kinetic analyses revealed that human liver microsomal cyclophosphamide 4-hydroxylation and ifosphamide 4-hydroxylation are best described by a 2-component Michaelis-Menten model composed of both low Km and high Km P-450 4-hydroxylases. To ascertain whether one or more human P-450 enzymes are catalytically competent in activating these oxazaphosphorines, microsomal fractions prepared from a panel of human B-lymphoblastoid cell lines stably transformed with individual P-450 complementary DNAs were assayed in vitro for oxazaphosphorine activation. Expressed CYP2A6, -2B6, -2C8, -2C9, and -3A4 were catalytically competent in hydroxylating cyclophosphamide and ifosphamide. Whereas CYP2C8 and CYP2C9 have the characteristics of low Km oxazaphosphorine 4-hydroxylases, CYP2A6, -2B6, and -3A4 are high Km forms. In contrast, CYP1A1, -1A2, -2D6, and -2E1 did not produce detectable activities. Furthermore, growth of cultured CYP2A6- and CYP2B6-expressing B-lymphoblastoid cells, but not of CYP-negative control cells, was inhibited by cyclophosphamide and ifosphamide as a consequence of prodrug activation to cytotoxic metabolites. Experiments with P-450 form-selective chemical inhibitors and inhibitory anti-P-450 antibodies were then performed to determine the contributions of individual P-450s to the activation of these drugs in human liver microsomes. Orphenadrine (a CYP2B6 inhibitor) and anti-CYP2B IgG inhibited microsomal cyclophosphamide hydroxylation to a greater extent than ifosphamide hydroxylation, consistent with the 8-fold higher activity of complementary DNA-expressed CYP2B6 with cyclophosphamide. In contrast, troleandomycin, a selective inhibitor of CYP3A3 and -3A4, and anti-CYP3A IgG substantially inhibited microsomal ifosphamide hydroxylation but had little or no effect on microsomal cyclophosphamide hydroxylation. By contrast, the CYP2D6-selective inhibitor quinidine did not affect either microsomal activity, while anti-CYP2A antibodies had only a modest inhibitory effect. Overall, the present study establishes that liver microsomal CYP2B and CYP3A preferentially catalyze cyclophosphamide and ifosphamide 4-hydroxylation, respectively, suggesting that liver P-450-inducing agents targeted at these enzymes might be used in cancer patients to enhance drug activation and therapeutic efficacy.

Gene-specific oligonucleotide probes for alpha, mu, pi, and microsomal rat glutathione S-transferases: analysis of liver transferase expression and its modulation by hepatic enzyme inducers and platinum anticancer drugs.

Glutathione S-transferases (GSTs) play an important role in the detoxification of diverse electrophilic chemicals, including anticancer drugs. Gene-specific oligonucleotide probes were developed to monitor the expression of individual GST mRNAs in livers of adult male rats treated with drugs and other chemical modulators of GST expression. Northern blot analysis of total liver RNA using probes specific for individual GSTs belonging to classes alpha (GSTs Ya1, Ya2, Yc), mu (GSTs Yb1, Yb2, Yb3), pi (GST Yp), and GSTms demonstrated the expression in liver of all but Yp mRNA. Kidney GST expression was at least as high as that in liver for GSTs Ya1, Yc, and Yp, while it was substantially lower but still detectable for GSTs Ya2, Yb2, and GSTms. Several of the liver GST class alpha mRNAs, in particular Ya2, were inducible by pretreatment of rats with phenobarbital or isosafrole. In contrast, dexamethasone preferentially induced Yb1, Yb2, and Ya2, while two other inducers of liver drug metabolism, isoniazid and clofibrate, were less effective with respect to GST induction. GSTms mRNA was induced to a small extent or not at all by the agents tested. Treatment of adult male rats with the anticancer drug cisplatin increased liver expression of GST Yc mRNA and suppressed Ya1 mRNA levels with little or no major effect on several other GST mRNAs. Western blot analysis of liver cytosols prepared from the cisplatin-treated rats revealed corresponding changes in GST Yc and Ya protein levels. Comparable changes in liver GST Ya1 and Yc expression were effected by the cisplatin analogue iproplatin but not by carboplatin or transplatin. This pattern of response to these platinum drugs is comparable to that seen with respect to platinum drug-induced gonadal toxicity and modulation of liver cytochrome P450 expression, suggesting a common mechanistic basis for these diverse effects of platinum anticancer drugs on hepatic enzymes of drug metabolism. Together, these studies demonstrate the utility of oligonucleotide probes for phenotyping liver tissue for the expression of GST enzymes that can contribute to anticancer drug metabolism and resistance. They also raise the possibility of drug-drug interactions involving cisplatin and alkylating agent anticancer drugs that can be metabolized in liver by alpha-class GSTs.

Platinum anticancer drugs modulate P-450 mRNA levels and differentially alter hepatic drug and steroid hormone metabolism in male and female rats.

Treatment of male rats with the anticancer drug cisplatin leads to feminization of the profile of cytochrome P-450 and other microsomal enzymes involved in steroid hormone and drug metabolism (G.A. LeBlanc, and D.J. Waxman, J. Biol. Chem., 263: 15732-15739, 1988). The present study uses the rat model to evaluate the differential effects of cisplatin treatment on liver microsomal enzymes between genders, and also examines whether the modulation of enzyme activities by cisplatin and its analogues involves changes in P-450 gene expression. While cisplatin treatment of male rats caused a severalfold increase in female-predominant hepatic enzymes, including testosterone 5 alpha-reductase and testosterone 7 alpha-hydroxylase (P-450 form 2A1), it partially decreased the expression of these enzymes in females. The reduced expression of these estrogen-dependent enzymes in females may derive from the loss of circulating estradiol that was shown to occur in response to cisplatin treatment. Analysis of mRNA levels of individual P-450 forms revealed that the effects of cisplatin on P-450-catalyzed steroid hydroxylase activities in both male and female rats are primarily operative through the drug's effects on P-450 mRNA expression. P-450-dependent cyclophosphamide activation was significantly compromised in male rats after cisplatin administration; however, this activity was not altered in cisplatin-treated females. This sex-dependent effect of cisplatin was due to its suppression of P-450 form 2C11, a male-specific P-450 that is a major contributor to microsomal cyclophosphamide bioactivation in male rat liver. The clinically active cisplatin analogue iproplatin elicited effects very similar to those of cisplatin, while carboplatin and transplatin did not have significant effects on hepatic P-450 expression. Together, these findings demonstrate that the response of rat liver to cisplatin-induced changes in hepatic P-450 enzyme profiles and cyclophosphamide bioactivation capacity differs between the sexes, and in addition, these effects can be minimized by use of carboplatin in place of cisplatin.

Enhanced cyclophosphamide and ifosfamide activation in primary human hepatocyte cultures: response to cytochrome P-450 inducers and autoinduction by oxazaphosphorines.

The anticancer oxazaphosphorine prodrugs cyclophosphamide and ifosfamide are activated in human liver by a 4-hydroxylation reaction catalyzed by multiple cytochrome P450 (CYP) enzymes. In the present study, we used a cultured human hepatocyte model to identify possible inducers of the CYP-catalyzed activation of these two anticancer prodrugs. Treatment of primary cultures of human hepatocytes with phenobarbital, dexamethasone, or rifampin elevated hepatocyte microsomal oxazaphosphorine 4-hydroxylation by up to 200-400% of control for both drug substrates. These inductions were associated with corresponding increases in immunoreactive CYP2B6, CYP2C8, CYP2C9, and CYP3A4, all previously shown to catalyze oxazaphosphorine activation. Rifampin (1 microM, 96-h exposure) was a particularly potent inducer of ifosfamide and cyclophosphamide 4-hydroxylation, as well as of CYP3A protein levels and CYP3A-dependent testosterone 6beta-hydroxylation. CYP3A4, CYP2C8, and CYP2C9 protein levels were also increased by exposure of the hepatocytes to cyclophosphamide or ifosfamide (50 microM), which thereby enhanced their own rates of 4-hydroxylation in the cultured hepatocytes. In one human hepatocyte culture that contained the polymorphically expressed CYP3A5 in addition to the more widely expressed CYP3A4, only CYP3A4 was induced by cyclophosphamide, ifosfamide, and rifampin. These studies: (a) demonstrate an underlying metabolic basis for the clinically important oxazaphosphorine autoinduction pharmacokinetics seen with these drugs in cancer patients; and (b) identify rifampin and other CYP inducers as potentially useful for increasing the rates of cyclophosphamide 4-hydroxylation and ifosfamide 4-hydroxylation in human liver in a manner that could favorably impact the clinical pharmacokinetics of these anticancer prodrugs.

Sensitization of human breast cancer cells to cyclophosphamide and ifosfamide by transfer of a liver cytochrome P450 gene.

The cancer chemotherapeutic agent cyclophosphamide (CPA) and its isomer ifosfamide (IFA) are alkylating agent prodrugs that require metabolism by liver cytochrome P450 (P450) enzymes for antitumor activity. The therapeutic effectiveness of these oxazaphosphorines is limited by the hematopoietic, renal, and cardiac toxicity that accompanies the systemic distribution of liver-derived activated drug metabolites. Transfer of a liver cytochrome P450 gene, CYP2B1, into human breast MCF-7 cancer cells is presently shown to greatly sensitize these cells to oxazaphosphorine toxicity as a consequence of the acquired capacity for intratumoral CPA and IFA activation. Thus, CPA and IFA were highly cytotoxic to MCF-7 cells following stable transfection of CYP2B1 but exhibited no toxicity to parental tumor cells or to a beta-galactosidase-expressing MCF-7 transfectant. This cytotoxicity could be appreciably blocked by the CYP2B1 inhibitor metyrapone. Cell cycle analysis revealed that CPA arrested the CYP2B1-expressing cells, but not CYP2B1-negative cells, at G(2)-M phase. A strong bystander cytotoxicity effect that does not require direct cell-cell contact was mediated by CYP2B1-expressing MCF-7 cells on non-CYP2B1 cells. Intratumoral CYP2B1 expression conferred a distinct therapeutic advantage when treating MCF-7 tumors grown in nude mice with CPA, as revealed by a 15-20-fold greater in vivo cytotoxicity, determined by tumor excision/colony formation assay, and by the substantially enhanced antitumor activity, monitored by tumor growth delay, for CYP2B1-e xpressing MCF-7 tumors as compared to CYP2B1-negative control tumors. These enhanced therapeutic effects were obtained without any apparent increase in host toxicity. To evaluate the extent to which a CPA/P450 gene therapy strategy may be generally applicable to other tumor cell types, a replication-defective recombinant adenovirus carrying the CYP2B1 gene driven by the cytomegalovirus (CMV) promotor ad.CMV-2B1 was constructed and used to infect a panel of human tumor cell lines. Ad.CMV-2B1 infection rendered each of the cell lines highly sensitive to CPA and IFA cytotoxicity, with substantial chemosensitization seen at multiplicities of infection as low as 10. The CPA/P450 prodrug activation system may thus serve as a useful paradigm for further development of novel cancer gene therapy strategies that utilize drug susceptibility genes to significantly potentiate the antitumor activity of conventional cancer chemotherapeutic agents.

Induction of microsomal and peroxisomal enzymes by dehydroepiandrosterone and its reduced metabolite in rats.

Dehydroepiandrosterone (DHEA) given to rodents in pharmacological doses induces several hepatic enzymes including cytochromes P4504A, NADPH:P450 oxidoreductase, palmitoyl coenzyme A oxidase, and other enzymes associated with the peroxisomal beta-oxidation pathway, leading to peroxisome proliferation and development of hepatocellular carcinoma in rodents. Comparison of the inductive potency of DHEA and other intermediates of the steroid biosynthetic path demonstrated that only DHEA, 5-ene-androstene-3 beta,17 beta-diol (ADIOL), and to a lesser extent, 17 alpha-hydroxypregnenolone, a precursor of DHEA, induce cytochromes P4504A protein and other enzymes associated with the peroxisome proliferative response in vivo. ADIOL exerted its inductive response at a somewhat lower dosage than DHEA, whereas ADIOL and DHEA both induced the microsomal enzymes (P4504A and its oxidoreductase) at somewhat lower dosages than those required to induce peroxisomal enzymes. Northern analysis demonstrated increases in the mRNAs encoding the cytochromes P4504A (> 20-fold) and NADPH:P450 oxidoreductase (> 10-fold) in the livers of DHEA- and ADIOL-treated rats. Run-on transcription analysis demonstrated that DHEA induces CYP4A gene expression 11-fold at the level of transcription initiation. Comparison of the responsiveness of individual rat CYP4A genes (4A1, 4A2, and 4A3) to DHEA and ADIOL in immature versus mature male rats revealed 2-3-fold higher levels of induced CYP4A1 and 4A3 mRNAs in immature rat livers. In contrast, hepatic CYP4A2 mRNA was induced to 6-10-fold higher levels in mature rats. No basal or significant inducible expression of mRNA for CYP4A1 and 4A3 was noted in rat kidney. Significant basal levels of kidney CYP4A2 mRNA were observed only in mature animals, where they were inducible by ADIOL and DHEA to a 3-5-fold greater extent than in the kidneys of immature rats. These studies demonstrate developmental differences in the responsiveness of CYP4A mRNA levels to DHEA and ADIOL in rat kidney and liver. Moreover, the striking inducibility of CYP4A protein and mRNAs, together with the increased rates of synthesis of nascent CYP4A mRNA transcripts in hepatic nuclei from DHEA-treated rats, establish that DHEA increases the expression of these microsomal enzymes at the transcriptional level.

Evidence for enzymatic activation and oxygen involvement in cytotoxicity and antitumor activity of N,N',N''-triethylenethiophosphoramide.

The cytotoxicity of N,N',N''-triethylenethiophosphoramide (thiotepa) was studied in vitro in the MCF-7 human breast carcinoma cell line and in vivo using the EMT6 mouse mammary tumor model, under various conditions of oxygenation and in the presence and absence of Aroclor 1254-induced liver preparations. The cytotoxicity of thiotepa toward exponentially growing MCF-7 cells was markedly dependent on the presence of oxygen during the period of drug exposure, with 3 log greater cell kill at 500 microM thiotepa being observed when the cells were normally oxygenated compared with hypoxic cells. Incubation of thiotepa with an Aroclor 1254-induced rat liver S-9 homogenate, in the presence of a NADPH-regenerating system, resulted in an 8-fold increase in cytotoxicity towards the MCF-7 cells over a wide range of drug concentrations. Thiotepa was shown to be metabolized under these conditions in a NADPH- and O2-dependent reaction that was catalyzed by one or more microsomal cytochrome P-450 enzymes that were present in the S-9 fraction. The thiotepa metabolite triethylene phosphoramide, which hydrolyzes significantly faster than thiotepa, was significantly less cytotoxic toward the MCF-7 cells than was thiotepa itself, suggesting that it is unlikely to be the S-9 metabolite responsible for the observed increase in drug cytotoxicity. Moreover, triethylene phosphoramide cytotoxicity was only partially O2 dependent and was largely unaffected by incubation in the presence of the S-9 preparation, indicating a mechanism of action distinct from that of thiotepa. Tumor cell survival experiments with the EMT6 mouse mammary carcinoma system revealed that a 3.6-fold increase in thiotepa cytotoxicity was obtained by prior administration of the liver inducer Aroclor 1254 to the tumor-bearing animals, 5 days before drug treatment. Finally, the therapeutic effectiveness of thiotepa was significantly enhanced (3- to 5.8-fold increase in tumor growth delay) when an increase in oxygenation was achieved, by carbogen breathing, in animals given the perfluorochemical emulsion Fluosol-DA. These findings establish that the cytotoxic effects of thiotepa are oxygen dependent and may involve, at least in part, metabolic processes catalyzed by cytochrome P-450 enzymes.

Dehydroepiandrosterone sulfate and beta-cell function: enhanced glucose-induced insulin secretion and altered gene expression in rodent pancreatic beta-cells.

Administration of dehydroepiandrosterone (DHEA), or its sulfated form (DHEAS), controls hyperglycemia in diabetic rodents without directly altering insulin sensitivity. We show that DHEAS enhanced glucose-stimulated insulin secretion when administered in vivo to rats or in vitro to beta-cell lines, without changing cellular insulin content. Insulin secretion increased from 3 days of steroid exposure in vitro, suggesting that DHEAS did not directly activate the secretory processes. DHEAS selectively increased the beta-cell mRNA expression of acyl CoA synthetase-2 and peroxisomal acyl CoA oxidase in a time-dependent manner. Although DHEAS is a peroxisomal proliferator, it did not alter the mRNA expression of peroxisomal proliferator-activated receptor (PPAR) alpha or beta, or enhance the activity of transfected PPAR alpha, beta, or gamma in vitro. Thus, DHEAS directly affected the beta-cell to enhance glucose-stimulated insulin secretion and increased the mRNA expression of specific beta-cell mitochondrial and peroxisomal lipid metabolic enzymes. This effect of DHEAS on insulin secretion may contribute to the amelioration of hyperglycemia seen in various rodent models of diabetes.

Hepatic P450 expression in hypothyroid rats: differential responsiveness of male-specific P450 forms 2a (IIIA2), 2c (IIC11), and RLM2 (IIA2) to thyroid hormone.

Studies carried out in hypophysectomized adult rats have demonstrated that both thyroid hormone and GH can suppress hepatic expression of the steroid 6 beta-hydroxylase P450 2a (IIIA2). The present study further characterizes the influence of thyroid hormone on the expression of P450 2a and two other male-specific hepatic P450s, a steroid 2 alpha/16 alpha-hydroxylase, designated P450 2c (IIC11), and a steroid 15 alpha-hydroxylase, designated P450 RLM2 (IIA2). These studies were carried out in rats rendered hypothyroid by treatment with methimazole, which allows for the nonsurgical depletion of circulating T4, and in hypophysectomized rats. Hypothyroidism led to an increase in hepatic P450 2a (IIIA2) protein and mRNA in both male and female rats that was fully reversed by T4 replacement. In contrast, hypothyroidism decreased by 70-80% the expression of P450 2c (IIC11) activity and mRNA, but did not significantly alter the expression of P450 RLM2 (IIA2). The decrease in P450 2c (IIC11) was not reversed by T4 replacement, suggesting that it is a consequence of the loss of plasma GH pulses that occurs secondary to hypothyroidism. In agreement with these findings, T4 given to hypophysectomized rats partially suppressed the expression of P450 2a (IIIA2) mRNA, but not P450 2c (IIC11) or P450 RLM2 (IIA2) mRNA. A more complete suppression of P450 2a (IIIA2) mRNA as well as P450 2c (IIC11) mRNA was achieved when the hypophysectomized rats were treated with T3 at a supraphysiological, receptor-saturating dose. Although GH administered to intact male rats by continuous infusion fully suppressed all three male-specific P450 proteins and their mRNAs, the same treatment given to hypothyroid rats was only partially suppressive in the case of P450 2a (IIIA2) and P450 RLM2 (IIA2), unless combined with T4. In the case of P450 2c (IIC11), substantial suppression of the residual P450 present in hypothyroid rats was achieved by treatment with GH alone, despite persistent thyroid hormone deficiency. These studies demonstrate that while thyroid hormone is a negative regulator of P450 2a (IIIA2) expression and is required for the full suppression of that P450 and P450 RLM2 (IIA2) by the continuous plasma GH profiles associated with adult female rats, the suppression of P450 2c (IIC11) by continuous plasma GH is largely independent of the presence of thyroid hormone.