Reductive activation of doxorubicin by xanthine dehydrogenase from EMT6 mouse mammary carcinoma tumors

Chemico-Biological Interactions 104 (1997) 87 – 101

Reductive activation of doxorubicin by xanthine dehydrogenase from EMT6 mouse mammary carcinoma tumors Steven B. Yee 1, Chris A. Pritsos * Department of Nutrition and Biochemistry Graduate Program, Uni6ersity of Ne6ada, Reno, NV 89557, USA Received 3 September 1996; received in revised form 27 February 1997; accepted 5 March 1997

Abstract The role of enzymes in the reductive activation of various chemotherapeutic agents is an area of considerable interest in studies to better understand the selective toxicities of these agents. Xanthine dehydrogenase (XDH) is an enzyme capable of reductive activation of chemotherapeutic agents. Previously, this enzyme has not been extensively studied because of difficulties in its isolation. We recently isolated this enzyme from EMT6 mouse mammary carcinoma cells and showed that this enzyme is capable of activating mitomycin C. In this study, we examined whether XDH could activate the clinically important antineoplastic agent, doxorubicin. Drug activation was determined under aerobic and hypoxic conditions and at various pHs in order to simulate the different environments found in solid tumors. The results of these studies show that XDH reacts with doxorubicin via a two-electron reduction. This reduction is different from the modified and more extensively studied form

Abbre6iations: DOX, doxorubicin; DDOXONE, 7-deoxydoxorubicin aglycone; DOXONE, doxorubicin aglycone; HPLC, high performance liquid chromatography; ROS, reactive oxygen species; XDH, xanthine dehydrogenase; XO, xanthine oxidase. * Corresponding author. Present address: Center for Environmental Sciences and Engineering, Mail Stop 199, University of Nevada, Reno, NV 89557-0132, USA. Tel.: + 1 702 3275096; fax: + 1 702 3275051. 1 Present address: Department of Pharmacology and Toxicology, Michigan State University, East Lansing, MI 48824-1317, USA. 0009-2797/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved. PII S 0 0 0 9 - 2 7 9 7 ( 9 7 ) 0 0 0 1 9 - 7

88

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

of the enzyme, xanthine oxidase (XO), which reacts with doxorubicin via a one-electron reduction. Under hypoxic conditions, the formation of large quantities of 7-deoxydoxorubicin aglycone, a deactivation product of doxorubicin metabolism, may serve to moderate doxorubicin’s antineoplastic activity. Under aerobic conditions, however, XDH activation led to a greater rate of formation of oxygen radicals than XO thereby possibly potentiating doxorubicin’s cytotoxicity to aerobic tumor cells. Kinetic constants were determined for doxorubicin activation by XDH. © 1997 Elsevier Science Ireland Ltd. Keywords: Anthracyclines; Bioreductive activation; Doxorubicin; Quinone; Xanthine dehydrogenase; Xanthine oxidase

1. Introduction Doxorubicin (adriamycin) is an anthracycline quinone antineoplastic antibiotic that has been shown to have a wide spectrum of clinical activity against a variety of solid tumors [1 – 3]. NADPH:cytochrome P-450 reductase [3–6], xanthine oxidase [3,4,6], ferridoxin reductase [3,4,7], nitrate reductase [3,6], lipoamide dehydrogenase [6], NADH-dependent cytochrome c and flavin reductase [6,8], and NADH:cytochrome b5 reductase [9] catalyze the reductive activation of doxorubicin. Under aerobic conditions, the reduction of doxorubicin leads to the generation of reactive oxygen species (ROS). Conversely, the reduction of doxorubicin, under hypoxic conditions, results in the formation of a doxorubicin quinone methide that may act either as a nucleophile or an electrophile [4,9,10]. A considerable amount of research has been conducted on the activation of quinone antineoplastic agents, like doxorubicin and mitomycin C, by xanthine oxidase (XO; EC 1.1.3.22). Very little research, however, exists on the activation of quinone antineoplastic agents by xanthine dehydrogenase (XDH; EC 1.1.1.204), the enzymatic precursor to XO, largely due to the difficulty and poor yields involved in the isolation of XDH [11]. With the development of a rapid and high yield purification procedure for XDH, research in the role that XDH plays as a reductive enzyme for bioreductive quinone antineoplastic agents has increased [12]. Recently, Gustafson and Pritsos [13] investigated the activation of mitomycin C by XDH. XDH was determined to be a better generator of mitomycin C-alkylating metabolites, but a relatively poor generator of mitomycin C-induced oxygen radicals, than was XO. These observations suggested that the metabolic reduction of mitomycin C by the two xanthine converting enzymes differed. XO appeared to reduce mitomycin C through a one electron reduction, while XDH primarily reduced mitomycin C through a two electron reduction. Whether this pattern of metabolic reduction by XDH, in comparison to XO, is consistent for other quinone antineoplastic agents has not yet been determined. Presently, no research has been conducted on the reductive activation of the quinone antineoplastic agent doxorubicin by XDH.

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

89

XO and XDH are complex molybdenum containing flavoproteins that are involved in purine catabolism in mammals. These enzymes are similar in size, structure, and appear to represent alternate forms of the same gene product [14]. Both enzymes not only catalyze the hydroxylation of hypoxanthine to xanthine, but also the subsequent hydroxylation of xanthine to uric acid [11]. The distinction between XO and XDH is determined by the preference of each enzyme for a different electron acceptor. XO utilizes molecular oxygen as an electron acceptor, with negligible reactivity toward NAD + , to reduce molecular oxygen to a superoxide anion radical. Conversely, XDH prefers to utilize NAD + as an electron acceptor to form NADH [11,14,15]. Hence, XO is reoxidized under physiological conditions by two one electron reductions of molecular oxygen to generate two superoxide anion radicals, which go on to form hydrogen peroxide; while XDH is reoxidized under physiological conditions by the two electron reduction of NAD + to produce NADH. XDH isolated from mammalian sources, can be reversibly transformed to XO by sulfhydryl (i.e., oxidative) modification or irreversibly transformed to XO by limited proteolysis [12–17]. Within cells, XO and XDH activities appear to be localized in the cytoplasm [11,15]. The relative activities of these two enzymes in tissues, however, are tissue specific. In most tissues, XDH is the predominant form of the two xanthine converting enzymes [11,13,15,18,19]. In the present investigation, XDH mediated metabolism of doxorubicin was studied using partially purified XDH isolated from EMT6 mouse mammary carcinoma cells grown in athymic BALB/c mice. EMT6 tumor tissue was used for the enzyme preparation because EMT6 cells contain no measurable XO activity [13,20]. The potential for reductive activation of doxorubicin by XDH, in comparison to XO, was investigated by measuring doxorubicin-induced oxygen consumption. Oxygen consumption studies were conducted at two different pHs: pH 7.4 to reflect cellular physiological pH and pH 6.0 to simulate the reduced or acidic pH of the hypoxic region of solid tumor cells. To further understand the interactions of doxorubicin with XDH, the kinetic constants (i.e., Km, Vmax, and kcat) for the reduction of doxorubicin by XDH were determined at the above two pH’s. Finally, following reductive activation by either XO or XDH, the presence of doxorubicin metabolites were detected for, under aerobic and hypoxic conditions at both pH 7.4 and 6.0. Along with the oxygen consumption results, the results from the detection of doxorubicin metabolites assisted in the determination of an electron reduction process for XDH activation of doxorubicin.

2. Materials and methods

2.1. Materials Doxorubicin hydrochloride was obtained from Sigma (St. Louis, MO). Doxorubicinol, doxorubicin aglycone, doxorubicinol aglycone, and 7-deoxydoxorubicinol aglycone were kindly provided by Pharmacia (Columbus, OH). Other reagents were obtained from the following sources: acetonitrile, ammonium formate, ammonium

90

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

sulfate, anhydrous dipotassium hydrogen phosphate (dibasic), anhydrous potassium dihydrogen phosphate (monobasic), cupric sulfate, disodium ethylenediamine tetraacetate (EDTA), ethanol (denatured, unsaturated), 2-mercaptoethanol, methanol, sucrose, and sodium potassium tartate from Fisher (Fair Lawn, NJ); Folin’s and Ciocalteu’s phenol reagent, lactose, nicotinamide adenine dinucleotide (NAD + ), phenylmethylsulfonyl fluoride (PMSF), reactive red-120, reduced form of-nicotinamide adenine dinucleotide (NADH), Sephadex G-200, xanthine (sodium salt), and xanthine oxidase (grade I, from buttermilk) from Sigma (St. Louis, MO); bovine serum albumin standard (fraction V) from Pierce (Rockford, IL); and NADPH (tetrasodium salt) from Calbiochem (La Jolla, CA). All reagents were of reagent grade purity or better. All solvents were of HPLC grade purity or higher.

2.2. Enzyme preparation XDH was purified (75 fold purification) from EMT6 tumors grown in athymic BALB/c mice according to Gustafson and Pritsos [13], as modified from Suleiman and Stevens’ [12] XDH isolation from liver tissue. In addition to the use of Sephadex G-200 columns and a reactive blue 2-Sepharose column for the purification of XDH, a reactive red-120 column was employed to enhance the purification of this enzyme. Protein concentration was determined by the Lowry protein assay method [21] on a Beckman DU-64 spectrophotometer (Beckman Instruments, Palo Alto, CA) utilizing bovine serum albumin for protein standards.

2.3. Xanthine dehydrogenase acti6ity assay XDH activity was measured by monitoring the formation of uric acid at 293 nm [16]. All enzyme assays were determined using a Beckman DU-64 spectrophotometer at 25°C. The assay mixture used contained 0.1 M potassium phosphate buffer (pH 7.8), 0.2 mM xanthine, 0.4 mM NAD + , and enzyme sample. One unit of XDH activity is defined as 1 mol of uric acid formed per min.

2.4. Oxygen consumption Oxygen consumption experiments were carried out using a YSI Model 5300 Biological Oxygen Monitor (YSI, Yellow Springs, OH) equipped with a Clark electrode. The experiments were performed in 0.1 M potassium phosphate buffer at either pH 7.4 or 6.0, with 1 mM NADH, 10 mU of XO or XDH, and doxorubicin at various concentrations.

2.5. Reduction of doxorubicin as measured by NADH oxidation NADH oxidation was based on the method used by Hodnick and Sartorelli [9], whereby the velocity of doxorubicin reduction was expressed as the disappearance or oxidation of NADH. The assay mixture consisted of 0.1 M potassium phosphate buffer at pH 7.4 or pH 6.0, 0.2 mM NADH, 10 mU XDH, and doxorubicin at

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

91

various concentrations. The disappearance of NADH was monitored at 340 nm ( = 6.18 ×103 mol − 1/cm) using a Beckman DU-64 spectrophotometer at 25°C. The Michaelis-Menten kinetic parameters of Km, Vmax, and kcat were then calculated. In order to calculate kcat an approximate XDH molecular weight of 300 000 Da was used [14].

2.6. High performance liquid chromatography analysis of doxorubicin metabolites High performance liquid chromatography (HPLC) analysis of doxorubicin metabolites followed the method reported by Bolanowska, et al. [22] as modified by Gewirtz and Yanovich [23]. The HPLC system consisted of a Thermo Separation Products Spectrasystem P2000 primary gradient pump (TSP, San Jose, CA) and a Thermo Separation Products Spectrasystem AS1000 fixed loop autosampler connected to a Hewlett-Packard 1046A Programmable Fluorescence Detector (Hewlett-Packard, Wilmington, DE). HPLC operation and integration was performed by Thermo Separation Products Spectrasystem PC1000 (version. 3.0.1) software. The HPLC was fitted with an Adsorbosphere phenyl guard column (Alltech, Deerfield, IL) and an Adsorbosphere phenyl HPLC column (250 × 4.6 mm). The mobile phase of 28% acetonitrile in 0.1 M ammonium formate at pH 4.0 was always filtered and vacuum degassed before use. Detection of peaks were made at an excitation wavelength of 470 nm and an emission wavelength of 585 nm. Assay mixtures consisted of 0.1 M potassium phosphate buffer either at pH 7.4 or 6.0, 1 mM NADH, 50 M doxorubicin, and 10 mU of either XO or XDH. Incubation of the reaction mixture was carried out for 30 min with shaking at 37°C. An equal volume of 100% acetonitrile was then added to the reaction mixture to facilitate chromatography. The reaction mixture was mixed and centrifuged at 2000 rpm for 3 min at 4°C in a IEC Centra GP8R refrigerated centrifuge (IEC; Needham Heights, MA) to remove any precipitate. 100 mL of the sample was then injected onto the column. Hypoxic conditions were obtained by bubbling the reaction mixture with nitrogen gas for 5 min and then maintaining the reaction mixture, in a sealed vial, under nitrogen gas throughout the remainder of the incubation. Two 25 gauge needles placed in a rubber stopper of the reaction vial were used for the influx and efflux of the nitrogen gas. Quantitation of doxorubicin and doxorubicin aglycone was accomplished through external standards. All standards were prepared in absolute methanol to a concentration of 1 mM. All dilutions were made with 50% methanol. 7-Deoxydoxorubicin aglycone was synthesized according to Gewirtz and Yanovich [23] to be used as a qualitative standard in the above HPLC method. 7-Deoxydoxorubicin aglycone was quantified via relative molar fluorescence. 7Deoxydoxorubicin aglycone has previously been determined to have twice the relative molar fluorescence of doxorubicin [24]. The sensitivity of fluorescence detection for the various doxorubicin metabolites ranged from 2–6 pmol.

92

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

2.7. Statistical analysis Statistical analysis was performed by employing a two-tailed Student’s t-test. The level of significance was attributed to PB 0.05.

3. Results

3.1. Oxygen Consumption as an indication of reducti6e acti6ation potential In all oxygen consumption experiments, oxygen consumption (mean9 S.D.) was reported in nmol of O2 consumed/min per U of enzyme activity, with the enzyme being either XO or XDH at equal enzyme activities. Enzymatic activities for these two enzymes are expressed as mol of uric acid formed per min. Oxygen consumption studies using either 10 mU of XO or XDH showed a dose-dependent increase in oxygen consumption with respect to doxorubicin concentration (50, 100, 250 and 500 mM; Fig. 1A and Fig. 1B), reaching saturation at higher doxorubicin concentrations. Additionally, doxorubicin-induced oxygen consumption generated through doxorubicin activation appeared to be pH sensitive. Doxorubicin activation by XDH exhibited significantly (PB 0.01) greater oxygen consumption at pH 6.0 than at pH 7.4. Oxygen consumption generated by doxorubicin activation by XO slightly decreased as the pH decreased from 7.4–6.0. This decrease in oxygen consumption was not significant at lower doxorubicin concentrations of 50 and 100 mM, but appeared to be significant (PB 0.05) at the higher doxorubicin concentrations of 250 and 500 mM. An increase in oxygen consumption as a result of doxorubicin activation by XDH at pH 7.4 compared to doxorubicin activation by XO (Fig. 1A) was not significant at the lowest and highest doxorubicin concentrations, but proved to be significant (P B0.05) at the 100 and 250 mM concentrations of doxorubicin. However, at pH 6.0 a nearly 8-fold difference in oxygen consumption was observed for doxorubicin activation by XDH in comparison to doxorubicin activation by XO (Fig. 1B). This increase in oxygen consumption from doxorubicin activation by XDH was highly significant (P B0.001) for all concentrations of doxorubicin investigated.

3.2. Enzyme kinetics for the acti6ation of doxorubicin by xanthine dehydrogenase Partially purified XDH catalyzed the reduction of doxorubicin as observed by the oxidation of NADH at 340 nm. The reduction process exhibited saturation kinetics at both pH 7.4 and 6.0 (Fig. 2A). The Km and Vmax were determined from a Lineweaver-Burk plot at the respective pH (Fig. 2B). The kinetic constants, which includes Km, Vmax, and kcat, for this enzymatic reduction are tabulated in Table 1 for both pHs. The kinetic constants determined from these studies were all pH dependent. The kinetic constants are greater at pH 6.0 than at pH 7.4.

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

93

3.3. HPLC determination of doxorubicin metabolites HPLC was used to identify and quantitate doxorubicin metabolites formed through the activation of doxorubicin by 10 mU XO and XDH at pH 7.4 and pH 6.0 under aerobic and hypoxic conditions. External standards for doxorubicin and doxorubicin aglycone were linear over a range of 0.1–250 M with an averaged

Fig. 1. Doxorubicin-induced oxygen consumption upon activation by either 10 mU XO or XDH at both pH 7.4 and 6.0. Assay mixture is described in Section 2. Doxorubicin was injected into the reaction chamber to a final concentration of 50, 100, 250 and 500 mM. Values represent the mean S.D. of at least ten independent determinations. Curves were plotted using the equation for a rectangular hyperbola. (A) doxorubicin-induced oxygen consumption at pH 7.4. The r2 for XO was 0.99, while the r2 for XDH was 0.98. (B) doxorubicin-induced oxygen consumption at pH 6.0. The r2 for XO was 0.86, while the r2 for XDH was 0.96.

94

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

Fig. 2. Kinetic analysis of doxorubicin activation by XDH as measured by NADH oxidation under aerobic conditions at both pH 7.4 and pH 6.0. Values represent the mean 9 S.D. of at least three independent determinations. S.D. smaller than the symbols are not shown. (A) 10 mU XDH catalyzed oxidation of NADH as stimulated by 1, 5, 10, 50, 100 mM of doxorubicin at pH 7.4 and 6.0. Curves were plotted using the equation for a rectangular hyperbola. The r2 at pH 7.4 was determined to be 0.97, while the r2 at pH 6.0 was determined to be 0.99. (B) Kinetic analysis using a Lineweaver-Burk plot for doxorubicin activation by 10 mU XDH at pH 7.4 and pH 6.0. The linearity of each line was assessed by least squares linear regression. The r2 at pH 7.4 was determined to be 0.97 and the r2 at pH 6.0 was determined to be 0.99.

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

95

r2 \0.95. A qualitative standard for 7-deoxydoxorubicin aglycone was generated. Quantitative determinations for 7-deoxydoxorubicin aglycone were made through relative molar fluorescence with doxorubicin [24]. Doxorubicin and doxorubicin metabolites were not detected when either XO or XDH was incubated, in the absence of doxorubicin, at pH 7.4 and pH 6.0 aerobically or hypoxically. Table 2 summarizes the quantitative results for doxorubicin and doxorubicin metabolites from doxorubicin activation by 10 mU XO or XDH at pH 7.4 and 6.0 under aerobic and hypoxic conditions. Under aerobic conditions, no doxorubicin metabolite formation was observed following XO activation of doxorubicin at pH 7.4 and 6.0. Following activation of doxorubicin by XDH, the doxorubicin metabolite of doxorubicin aglycone was observed at pH 7.4, but no doxorubicin metabolites were observed at pH 6.0. No doxorubicin metabolite formation was observed for the activation of doxorubicin by XO at either pH 7.4 or 6.0 under hypoxic conditions. However, 7-deoxydoxorubicin aglycone was observed under hypoxic conditions following doxorubicin activation by XDH at both pH 7.4 and 6.0. More of the 7-deoxydoxorubicin aglycone metabolite was observed at pH 7.4 than at 6.0.

4. Discussion In this study, we observed a dose dependent increase in oxygen consumption for increasing doxorubicin concentrations upon reductive activation by either XO or XDH at both pH 7.4 and 6.0 (Fig. 1). The data indicated that, in general, at pH 7.4, the reductive activation of doxorubicin by XDH appeared to generate slightly more oxygen radicals than the reductive activation of doxorubicin by XO. However, when the pH was dropped to 6.0, a very significant difference (PB 0.001) was observed between the oxygen consumption for doxorubicin as catalyzed by XO and XDH activation. XDH activation of doxorubicin catalyzed approximately an eight fold increase in oxygen radical generation over XO activation. The generation of Table 1 Kinetic constants for the reduction of doxorubicin by 10 mU xanthine dehydrogenase at pH 7.4 and 6.0

Kma Vmaxb kcatc

pH 7.4

pH 6.0

4.5 330 1.6×10-2

37.0 920 4.4×10−2

Kinetic constants were determined from NADH oxidation as detailed in Section 2. One Unit of XDH activity is expressed as 1 mol of uric acid formed per min. a Expressed in mM doxorubicin. b Expressed in nmol of NADH oxidized/min per U XDH, nmoles of NADH oxidized is equivalent to nmoles of doxorubicin reduced. c Expressed in s−1, kcat was determined by using an approximate XDH molecular weight of 300 000 Da.

96

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

Table 2 Percentage of doxorubicin metabolite formation from activation of doxorubicin by xanthine dehydrogenase and xanthine oxidase at pH 7.4 and 6.0 under aerobic and hypoxic conditions Aerobic

Hypoxic

pH 6.0

pH 7.4

pH 6.0

pH 7.4

XDH Doxorubicin Doxorubicin aglycone 7-deoxydoxorubicin aglycone

100 9 19 0 0

79 9 11 21 9 1 0

94 91 0 69 1

30 91 0 70 9 2

XO Doxorubicin Doxorubicin aglycone 7-deoxydoxorubicin aglycone

100 97 0 0

100 914 0 0

100 91 0 0

100 9 4 0 0

Doxorubicin metabolite formation was measured by HPLC analysis as described in Section 2. Each value represents the mean9 S.D. for at least four independent determinations. a All values are expressed as a percentage of the total metabolite determined by HPLC for each series of experiments. Metabolites were determined (recovered) within the range of 64 – 76% of the theoretical yield.

oxygen radicals from XDH activation of doxorubicin, also, increased significantly (P B 0.01) as the pH was lowered from 7.4 to 6.0 by an average of about three fold. This increase correlates with the kinetic data presented in Table 1 which shows that the Vmax and kcat increased by approximately three fold as the pH was lowered from 7.4 to 6.0. Oxygen radical generation from the activation of doxorubicin by XO was significantly effected only at the highest doxorubicin concentrations investigated, when the pH was lowered from 7.4 to 6.0. At doxorubicin concentrations of 250 and 500 mM, oxygen radical production decreased significantly (PB 0.05) by over two fold as the pH was lowered from 7.4 to 6.0. However, at both pH 7.4 and 6.0 (Fig. 1), more oxygen consumption was observed from XDH activation of doxorubicin than XO activation of doxorubicin. XDH is known to reduce mitomycin C and xanthine by two electrons. A two electron reduction of doxorubicin by XDH results in the production of a doxorubicin hydroquinone, which can produce more oxygen radicals than a one electron reduction of doxorubicin. Doxorubicin hydroquinone is not stable in the presence of molecular oxygen [25,26]. Two possibilities exist for the generation of ROS from doxorubicin hydroquinone. First, a comproportionation reaction could occur between the doxorubicin hydroquinone and doxorubicin to produce two doxorubicin semiquinone radicals. These doxorubicin semiquinone radicals could subsequently react with molecular oxygen to generate two superoxide anion radicals and two doxorubicin molecules (i.e. redox cycle). Second, doxorubicin hydroquinone could become oxidized in the presence of molecular oxygen by one electron to produce a doxorubicin semiquinone radical and a superoxide anion radical. The doxorubicin semiquinone radical would then react with molecular oxygen to produce a doxorubicin molecule and a second superoxide anion radical [27,28]. A two electron

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

97

reduction of doxorubicin by XDH would explain the observed increase in oxygen consumption for XDH activation of doxorubicin, in comparison to the one electron reduction of doxorubicin by XO, observed at both pH 7.4 and 6.0. This greater production of superoxide anion radicals from a two electron reduction of doxorubicin by XDH would lead to greater ROS formation. Furthermore, the increase in oxygen consumption for doxorubicin activation by XDH, in comparison to doxorubicin activation by XO, as the pH was lowered from 7.4 – 6.0 suggests that the pH optimum for XDH activity for the reduction of doxorubicin may occur at a more acidic pH. These findings suggest that XDH would more effectively metabolize doxorubicin in the hypoxic region of solid tumors due to the more acidic pH (as compared to physiological pH) environment within the hypoxic region of solid tumors [13,29]. The slight decrease in XO activated doxorubicin as the pH was lowered from 7.4 to 6.0 suggests that the optimal pH for XO activity for the reduction of doxorubicin may occur at a more physiological pH. Therefore, the bioreductive potential of the xanthine converting enzymes for doxorubicin, in any specific environment, is dependent upon the form of the enzyme, oxidase or dehydrogenase. The reduction of doxorubicin was studied kinetically by observing the oxidation of NADH. XDH catalyzed reduction of doxorubicin appeared to obey MichaelisMenten kinetics and was thus saturable at both pH 7.4 and 6.0. The Km and the Vmax were determined for both pHs from the double reciprocal Lineweaver-Burk plot (Fig. 2 and Table 1). The Km for doxorubicin activation by XDH differs by about eight fold from pH 7.4–6.0, suggests a higher affinity for the substrate at pH 7.4. The Vmax and kcat for doxorubicin activation by XDH also differed between pH 7.4 and pH 6.0 by nearly three fold, with the greater maximal velocity and enzyme turnover at pH 6.0. The increase in oxygen consumption by three fold for doxorubicin activation by xanthine dehydrogenase as the pH was lowered from pH 7.4 – 6.0 may thus be explained by the subsequent increase in the catalytic velocity for doxorubicin activation at pH 6.0, despite the lower Km for doxorubicin binding with XDH. This enzyme kinetic data is consistent with the hypothesis that the pH optimum for enzyme activity for doxorubicin as catalyzed by XDH appears to occur at the more acidic pH of the tumor rather than at physiological pH. The Km for doxorubicin activation by XDH was found to be 66 fold less at pH 7.4 and about 4 fold less at pH 6.0 than the Km for mitomycin C activation by XDH. Furthermore, the Vmax for doxorubicin activation by XDH was considerably higher than the Vmax for mitomycin C activation. The Vmax for doxorubicin activation was 10 fold greater at pH 7.4 and 16 fold greater at pH 6.0 than for mitomycin C activation [30]. A general comparison of the kinetic constants from mitomycin C activation by XDH and the kinetic constants from doxorubicin activation by XDH, therefore, suggests that doxorubicin is a better substrate for XDH than is mitomycin C. The Km for doxorubicin activation by XDH at pH 7.4 and 6.0 is two fold greater and 27 fold greater, respectively, than the Km observed for doxorubicin activation by NADH:cytochrome b5 reductase at pH 6.6 [9]. The Km for doxorubicin activation by XDH is also 200 fold less than (at pH 7.4) and 21 fold less than (at

98

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

pH 6.0) the Km for doxorubicin activation by NADPH:cytochrome P450 reductase at pH 8.0 [31]. Although these kinetic constants were determined at different pHs, a relative comparison of Km values for doxorubicin activation from reductive enzymes can be made. These Km values suggest that doxorubicin may be a better substrate for NADH:cytochrome b5 reductase than XDH, while doxorubicin is a better substrate for XDH than for NADPH:cytochrome P-450 reductase. The kcat for doxorubicin activation by NADH:cytochrome b5 reductase was approximately 0.02 s − 1 at pH 6.6 [9]. This catalytic velocity is relatively equivalent to the kcat for doxorubicin activation by XDH at pH 7.4 and half the catalytic velocity for doxorubicin activation by XDH at pH 6.0. This kcat value for doxorubicin activation by XDH suggests that at saturating concentrations of substrate (doxorubicin) and at acidic tumor pH the rate velocity for doxorubicin activation by XDH is faster (doubled) than the rate velocity for doxorubicin activation by NADH:cytochrome b5 reductase. These data suggest that at pH 7.4 for high doxorubicin concentration, XDH activated doxorubicin at a similar rate as NADH:cytochrome b5 reductase. However, at more acidic pHs, XDH has a higher rate of activation than NADH:cytochrome b5 reductase. At substrate concentrations below saturation for these enzymes, NADH:cytochrome b5 reductase appears to be a better activator of doxorubicin than XDH. The analysis of metabolite formation (Table 2) from doxorubicin activation by XO and XDH at pH 7.4 and pH 6.0 both aerobically and hypoxically further characterizes the reductive activation of doxorubicin by XDH. Chromatographic peaks from doxorubicin upon activation by XO and XDH, at pH 7.4 and pH 6.0 under aerobic and hypoxic conditions, were compared to chromatographic peaks obtained from the various doxorubicin metabolite standards. Doxorubicin activation by XO at pH 7.4 and pH 6.0 under aerobic conditions generated no detectable doxorubicin metabolites. These results are consistent with data that show that under aerobic conditions XO activates doxorubicin exclusively through a one electron reduction [32,33]. However, when an excess amount of XO (200 mU) was used in the reaction mixture a very small amount of 7-deoxydoxorubicin aglycone was detected (data not shown). Therefore, this suggests that a very small amounts of 7-deoxydoxorubicin aglycone is generated. Theoretically, 7- deoxydoxorubicin aglycone can be generated from one electron reductions of doxorubicin by either reductive glycoside cleavage of a doxorubicin semiquinone radical or by interactions of two doxorubicin semiquinone radicals to produce a doxorubicin hydroquinone, leading to the production of 7-deoxydoxorubicin aglycone [27]. This suggests that XO activates doxorubicin by a one electron reduction. Also, doxorubicin activation by XDH at pH 6.0 under aerobic conditions showed no detectable doxorubicin metabolites. However, at pH 7.4 under aerobic conditions, XDH activated doxorubicin generated the doxorubicin aglycone metabolite (Table 2). The formation of doxorubicin aglycone was proposed to occur via the interaction of hydroxyl radicals with radical precursors of the 7-deoxydoxorubicin aglycone metabolite [34]. This appears to be consist with our data. Aerobically we only see the formation of the doxorubicin aglycone metabolite whereas under hypoxic conditions we only see the formation of the 7-deoxydoxorubicin aglycone. Hypoxi-

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

99

cally there would be no redox cycling and therefore no hydroxyl radical generation for reactions with 7-deoxydoxorubicin aglycone to form doxorubicin aglycone. Under aerobic conditions, not only is doxorubicin being metabolized to 7-deoxydoxorubicin aglycone but it is also redox cycling, allowing for the formation of hydroxyl radicals for subsequent reactions with 7-deoxydoxorubicin aglycone. This reaction under aerobic conditions generates doxorubicin aglycone at the expense of 7-deoxydoxorubicin aglycone, which is consistent with our pH 7.4 data seen in Table 2. At pH 6.0, we do not observe doxorubicin aglycone under aerobic conditions due to the very low yield of 7-deoxydoxorubicin aglycone, as can be seen in the pH 6.0 hypoxic data in Table 2. Under hypoxic conditions, doxorubicin activation by XO produced no detectable metabolites either at pH 7.4 or pH 6.0. Similar to the aerobic data for XO, a small undetectable amount of 7-deoxydoxorubicine aglycone may be formed under these conditions. XDH activation of doxorubicin under hypoxic conditions, however, generated the 7-deoxydoxorubicin aglycone metabolite at both pH 7.4 and 6.0 (Table 2), with the greatest formation 7-deoxydoxorubicin aglycone occurring at pH 7.4. This generation of 7-deoxydoxorubicin aglycone, under hypoxic conditions, suggests that doxorubicin has been reduced by XDH via a two electron reduction of doxorubicin. A two electron reduction of doxorubicin results in the formation of a doxorubicin hydroquinone. Under aerobic conditions, the doxorubicin hydroquinone is unstable due to the compounds’ sensitivity to oxidation by molecular oxygen. Conversely, under hypoxic conditions, doxorubicin hydroquinone becomes stabilized long enough to allow for electron rearrangement resulting in the loss of the glycoside group to form the doxorubicin quinone methide intermediate. This doxorubicin quinone methide can act as a nucleophile, which can react with DNA. However, if no suitable reactant is present, the doxorubicin quinone methide will react with a proton (H + ) to generate 7-deoxydoxorubicin aglycone [26]. The 7-deoxydoxorubicin aglycone metabolite has considerably less antineoplastic activity than doxorubicin and is considered to be the inactivation product of doxorubicin [25,26]. In conclusion, XDH appears to reduce doxorubicin by two electrons as opposed to the one electron reduction of doxorubicin by XO. Greater oxygen consumption from doxorubicin activation by XDH was observed than from doxorubicin activation by XO. This greater oxygen consumption could be explained by a two electron reduction of doxorubicin to produce a doxorubicin hydroquinone, which is known to be unstable under aerobic conditions, reacting with molecular oxygen to form oxygen radicals. Under hypoxic conditions, doxorubicin activation by XDH leads to the production of 7-deoxydoxorubicin aglycone, a clear two electron reduction metabolite of doxorubicin. Furthermore, the formation of a doxorubicin quinone methide from the two electron reduction of doxorubicin by XDH, under hypoxic conditions, is known to be considerably less reactive in comparison to other antineoplastic quinone methides. As a consequence of the reactivity of this quinone methide, doxorubicin quinone methide may not have the ability to alkylate DNA to a major extent. The formation of 7-deoxydoxorubicin aglycone from doxorubicin quinone methide is thought to be a pathway for drug inactivation. The 7-deoxydox-

100

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

orubicin aglycone metabolite is known to have considerably less antineoplastic activity than doxorubicin [25,26]. Therefore, under hypoxic condition, XDH may serve to inactivate doxorubicin antineoplastic activity by reducing doxorubicin to the 7-deoxydoxorubicin aglycone metabolite. However, under aerobic conditions, doxorubicin activation by XDH generates more oxygen radicals than doxorubicin activation by XO. Thus, XDH in comparison to XO, under aerobic conditions, may enhance the toxicity of doxorubicin by way of greater ROS generation.

Acknowledgements This work was supported by Public Health Service Grant CA-43660 from the National Cancer Institute and the Nevada Agricultural Experiment Station.

References [1] S.K. Carter, Adriamycin, a review. J. Natl. Cancer Inst., 55 (1975) 1265 – 1274. [2] R.T. Dorr and W.L. Fritz, in: Cancer Chemotherapy Handbook. Elsevier, New York, 1980, pp. 388–401 and 628–631. [3] G. Powis, Metabolism and reactions of quinoid anticancer agents. Phamacol. Ther., 35 (1987) 57–162. [4] H. Kappus, Overview of enzyme systems involved in bioreduction of drugs and redox cycling. Biochem. Pharmacol., 35 (1986) 1 – 6. [5] T. Komiyama, T. Kikuchi and Y. Sugiura, Generation of hydroxyl radical by anticancer quinone drugs, carbazilquinone, mitomycin C, aclacinomycin A, and Adriamycin, in the presence of NADPH-cytochrome P–450 reductase. Biochem. Pharmacol., 31 (1982) 3651 – 3656. [6] S. Pan, L. Pedersen and N.R. Bachur, Comparative flavoprotein catalysis of anthracycline antibiotic reductive cleavage and oxygen consumption. Mol. Pharmacol., 19 (1981) 184 – 186. [7] K. Ramakrishnan and J. Fisher, 7-deoxydaunomycin quinone methide reactivity with thiol nucleophiles. J. Med. Chem., 29 (1986) 1215 – 1221. [8] K. Ramakrishnan and J. Fisher, Nucleophilic trapping of 7,11-dideoxyanthracycline quinone methide. J. Am. Chem. Soc., 105 (1983) 7187 – 7188. [9] W.F. Hodnick and A.C. Sartorelli, The pH-dependent reduction of adriamycin catalyzed by NADH:cytochrome b5 reductase. Cancer Letters, 84 (1994) 149 – 154. [10] J. Cummings, L. Anderson, N. Willmott and J.F. Smyth, The molecular pharmacology of doxorubicin in vivo. Eur. J. Cancer, 27 (1991) 532 – 535. [11] C.A. Pritsos and G.L. Lustafson, Xanthine dehydrogenase and its role in cancer chemotherapy. Oncol. Res., 6 (1994) 477–481. [12] S.A. Suleiman and J.B. Stevens, Purification of xanthine dehydrogenase from rat liver: a rapid procedure with high enzyme yields. Arch. Biochem. Biophys., 258 (1987) 219 – 227. [13] D.L. Gustafson and C.A. Pritsos, Bioactivation of mitomycin C by xanthine dehydrogenase from EMT6 mouse mammary carcinoma tumors. J. Natl. Cancer Inst., 84 (1992) 1180 – 1185. [14] R. Hille and T. Nishino, Xanthine oxidase and xanthine dehydrogenase. FASEB J., 9 (1995) 995–1003. [15] D.A. Parks and N.D. Granger, Xanthine oxidase: biochemistry, distribution and physiology. Acta Physiol. Scand., 548 (1986) 87– 99. [16] F. Stirpe and E. Della Corte, The regulation of rat liver xanthine oxidase. Conversion in vitro of the enzyme activity from dehydrogenase (type D) to oxidase (type O). J. Biol. Chem., 244 (1969) 3855–3863.

S.B. Yee, C.A. Pritsos / Chemico-Biological Interactions 104 (1997) 87–101

101

[17] E. Della Corta and F. Stirpe, The regulation of rat liver xanthine oxidase. Involvement of thiol groups in the conversion of the enzyme activity from dehydrogenase (type D) into oxidase (type O) and purification of the enzyme. Biochem. J., 126 (1972) 739 – 745. [18] M.G. Batelli, E. Della Corte and F. Stirpe, Xanthine oxidase type D (dehydrogenase) in the intestine and organs of the rat. Biochem. J., 126 (1972) 747 – 749. [19] R.F. Anderson, K.B. Patel, K. Reghebi and S.A. Hill, Conversion of xanthine dehydrogenase to xanthine oxidase as a possible marker for hypoxia in tumors and normal tissues. Br. J. Cancer, 60 (1989) 193–197. [20] D.L. Gustafson and C.A. Pritsos, Oxygen radical generation and alkylating ability of mitomycin C bioactivated by xanthine dehydrogenase. Proc. West. Pharmacol. Soc., 35 (1992) 147 – 151. [21] O.H. Lowry, N.J. Rosebrough, A.L. Farr and R.J. Randall, Protein measurement with the folin phenol reagent. J. Biol. Chem., 193 (1951) 266 – 275. [22] W. Bolanowska, T. Gessner and H.A. Preisler, Simplified method for determination of daunorubicin, adriamycin, and their chief fluorescent metabolites in human plasma by high-pressure liquid chromatography. Cancer Chemo. and Pharmacol., 10 (1983) 187 – 191. [23] D.A. Gewirtz and S. Yanovich, Metabolism of adriamycin in hepatocytes isolated from the rat and rabbit. Biochem. Pharmacol., 36 (1987) 1793 – 1798. [24] J. Cummings, J.F.B. Stuart and K.C. Calman, Determination of adriamycin, adriamycinol and their 7-deoxyaglycones in human serum by high-performance liquid chromatography. J. Chrom. Biomed. Appl., 311 (1984) 125– 134. [25] C.E. Meyers, Jr. and B.A. Chabner, Anthracyclines, in: B.A. Chabner and J.M. Collins (Eds.), Cancer Chemotherapy: Principles and Practice, J.B. Lippincott, Philadelphia, 1990 pp. 356 – 381. [26] B.R.J. Abdella and J. Fisher, A chemical perspective on the anthracycline antitumor antibiotics. Env. Health Persp. 64 (1985) 3 – 18. [27] G. Powis, Metabolism and reactions of quinoid anticancer agents. In: G. Powis (Ed.), Anticancer Drugs: Reactive Metabolism and Drug Interactions, Permagon Press, Oxford, 1994 pp. 273 – 284. [28] P. Workman, Enzyme-directed bioreductive drug development revisited: a commentary on recent progress and future prospects with emphasis on quinone anticancer agents and quinone metabolizing enzymes, particularly DT-diaphorase. Oncol. Res., 6 (1994) 461 – 475. [29] I.F. Tannock and D. Rotin, Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res., 49 (1989) 4373–4384. [30] D.L. Gustafson and C.A. Pritsos, Kinetics and mechanism of mitomycin C bioactivation by xanthine dehydrogenase under aerobic and hypoxic conditions. Cancer Res., 53 (1993) 5470 – 5474. [31] N.R. Bachur, S.L. Gordon, M.V. Gee and H. Kon, NADPH cytochrome P-450 reductase activation of quinone anticancer agents to free radicals. Proc. Natl. Acad. Sci. USA, 76 (1979) 954–957. [32] S-S. Pan and N.R. Bachur, Xanthine oxidase catalyzed reductive cleavage of anthracyline antibiotics and free radical formation. Mol. Pharmacol., 17 (1980) 95 – 99. [33] J.H. Doroshow, Effects of anthracycline antibiotics on oxygen radical formation in rat heart. Cancer Res., 43 (1983) 460–472. [34] H.S. Schwartz and B. Paul, Biotransformation of daunorubicin aglycones by rat liver microsomes. Cancer Res. 44 (1984) 2480–2484.

.