Detoxification of carcinogenic aromatic and heterocyclic amines by enzymatic reduction of the N-hydroxy derivative

Cancer Letters 143 (1999) 167±171

Detoxi®cation of carcinogenic aromatic and heterocyclic amines by enzymatic reduction of the N-hydroxy derivative Roberta S. King a,1, Candee H. Teitel a, Joseph G. Shaddock b, Daniel A. Casciano b, Fred F. Kadlubar a,* a

Division of Molecular Epidemiology, National Center for Toxicological Research, Jefferson, AR 72079-9501, USA b Division of Genetic Toxicology, National Center for Toxicological Research, Jefferson, AR 72079-9501, USA Received 24 November 1998; received in revised form 27 January 1999; accepted 1 February 1999

Abstract The metabolic activation pathways associated with carcinogenic aromatic and heterocyclic amines have long been known to involve N-oxidation, catalyzed primarily by cytochrome P4501A2, and subsequent O-esteri®cation, often catalyzed by acetyltransferases (NATs) and sulfotransferases (SULTs). We have found a new enzymatic mechanism of carcinogen detoxi®cation: a microsomal NADH-dependent reductase that rapidly converts the N-hydroxy arylamine back to the parent amine. The following N-OH-arylamines and N-OH-heterocyclic amines were rapidly reduced by both human and rat liver microsomes: NOH-4-aminoazobenzene, N-OH-4-aminobiphenyl (N-OH-ABP), N-OH-aniline, N-OH-2-naphthylamine, N-OH-2-amino¯uorene, N-OH-4,4 0 -methylenebis(2-chloroaniline) (N-OH-MOCA), N-OH-1-naphthyamine, N-OH-2-amino-1-methyl-6phenylimidazo[4,5-b]pyridine (N-OH-PhIP), N-OH-2-amino-a -carboline (N-OH-Aa C), N-OH-2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (N-OH-MeIQx), and N-OH-2-amino-3-methylimidazo[4,5-f]quinoline (N-OH-IQ). In addition, primary rat hepatocytes and human HepG2 cells ef®ciently reduced N-OH-PhIP to PhIP. This previously unrecognized detoxi®cation pathway may limit the bioavailability of carcinogenic N-OH heterocyclic and aromatic amines for further activation, DNA adduct formation, and carcinogenesis. q 1999 Published by Elsevier Science Ireland Ltd. Keywords: Heterocyclic amine; Aromatic amine; Detoxi®cation; N-hydroxy; PhIP; Aa C; IQ; MeIQx

1. Introduction An N-hydroxy (N-OH) amine reductase activity * Corresponding author. Tel: 11-870-543-7204; fax: 11-870543-7773. E-mail address: [email protected] (F.F. Kadlubar) 1 Present address: Department of Biomedical Sciences, College of Pharmacy, University of Rhode Island, Kingston, RI 02281, USA Abbreviations: Aa C, 2-amino-a -carboline; ABP, 4-aminobiphenyl; IQ, 2-amino-3-methylimidazo[4,5-f]quinoline; MeIQx, 2amino-3,8-dimethylimidazo[4,5-f]quinoxoline; 3-MC, 3-methylcholanthrene; N-OH, N-hydroxy; N-OH-MBA, N-hydroxy-Nmethylbenzylamine; MOCA, 4,4 0 -methylenebis(2-chloroaniline); PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine

was ®rst reported in pig liver microsomes in 1973, and this activity was not altered by the presence of oxygen, carbon monoxide, EDTA, cyanide, nor sodium azide [1]. It required the presence of NADH or NADPH, although NADH was the preferred cofactor. This enzyme system had a pH optimum of 6.3 and its puri®cation showed that it consisted of cytochrome b5, cytochrome b5 reductase, and a third partially characterized protein [2]. More recently, in human liver microsomes, this enzyme system has been studied for its ability to catalyze the reduction of N-hydroxy-sulfamethoxazole [3] and of several Nhydroxy-aminoguanidines (amidoximes) that are used

0304-3835/99/$ - see front matter q 1999 Published by Elsevier Science Ireland Ltd. PII: S 0304-383 5(99)00119-6

168

R.S. King et al. / Cancer Letters 143 (1999) 167±171

as therapeutic drugs [4]. We now report characterization of this NADH-dependent and oxygen-insensitive N-OH amine reductase activity in rat and human liver microsomes, and in intact rat hepatocytes and human HepG2 cells. This investigation initially arose as a consequence of in vitro studies with the heterocyclic amine carcinogen, PhIP. Primary hepatocytes from Aroclortreated (CYP1A-induced) rats produced N-oxidized PhIP metabolites. However, hepatocytes from untreated rats produced no N-oxidized metabolites when incubated with PhIP even though these hepatocytes contained signi®cant CYP1A activity. In addition, incubations of N-OH-PhIP with primary hepatocytes from both untreated and CYP1Ainduced F344 rats quickly and completely converted the N-OH-PhIP to the parent amine. We have now investigated the mechanism of this reduction, and have found it to be catalyzed by the cytochrome b5-dependent pathway ®rst described some 25 years ago [1]. 2. Materials and methods 2.1. N-Hydroxy-amines Purity of each N-OH-amine available in the authors' laboratory was assessed by measuring Fe 31 reducingequivalents and by HPLC analysis. Compounds were dissolved in argon-saturated dimethylsulfoxide/ ethanol (4/1) at 48C to a concentration of 5 mM. Solutions of N-OH-amines were stored in liquid nitrogen to prevent decomposition. The following compounds were chosen for study: N-OH-Aa C, N-OH-ABP, NOH-IQ, N-OH-MeIQx, N-OH-MOCA, N-OH-PhIP, N-OH-4-aminoazobenzene, N-OH-2-amino¯uorene, N-OH-aniline, N-OH-N-methylbenzylamine (N-OHMBA), N-OH-1-naphthylamine, and N-OH-2naphthylamine. 2.2. Human and rat liver microsomal preparations Human tissues were obtained from the International Institute for the Advancement of Medicine (Exton, PA). Microsomal fractions were prepared as described previously [5], and protein concentrations were determined by biuret assay. Rat liver microsomal prepara-

tions were purchased from In Vitro Technologies (Baltimore, MD). N-OH-Reductase activity was determined in human and rat liver microsomes by measuring NADH- and enzyme-dependent loss of N-OH-amine utilizing a colorimetric assay [1]. Microsomal assay conditions were optimized for pH, incubation time, and concentration of NADH, microsomal protein, and the N-OHMBA substrate. The optimal conditions for N-OHMBA and human liver microsomes, which were used for all other microsomal assays, were: 0.1 mM N-OH-amine, 1 mM NADH, 0.3 mg/ml microsomal protein, 0.1 M potassium phosphate (pH 6.3), and 10 min incubation at 378C. 2.3. Rat liver primary hepatocytes Cells were isolated from adult male F344 rats and viability counts were performed as previously reported [6]. The total number of cells required were centrifuged (5 min at 50 £ g) and resuspended in serum-free Williams' Medium E (WE, Gibco/BRL) supplemented with 1% bovine serum albumin (BSA) to a ®nal cell concentration of 5 £ 106 viable cells/ml. Cell suspensions were utilized immediately for incubation with 3H-N-OH-PhIP as described below. 2.4. Human HepG2 cells Cells were grown to 60±70% con¯uency in WE supplemented with 10% fetal bovine serum. Cells were removed from the ¯asks by washing three times with 1 £ phosphate-buffered saline, trypsinized with 0.05% trypsin, and centrifuged (10 min at 225 £ g) to obtain a cell pellet. Cells were resuspended in WE and cell counts were taken with a hemocytometer. The total number of cells required were subjected to centrifugation (10 min at 225 £ g) and resuspended in serum-free WE supplemented with 2% BSA for a ®nal cell concentration of 10 £ 106 cells/ml. For some experiments, cells were treated with 3-methylcholanthrene (3-MC) after growing to 60±70% con¯uency. 3-MC was added to cultures in fresh medium to a ®nal concentration of 10 mM with 1% DMSO. After a 24-h incubation (5% CO2 at 378C), cells were removed from ¯asks, counted, and resuspended as described for cells without 3-MC treatment.

R.S. King et al. / Cancer Letters 143 (1999) 167±171 Table 1 Reduction of N-OH-arylamines by human liver microsomes from two individuals with high N-OH-reductase activity a Rate of reduction (nmol/min per mg total protein)

N-OH-IQ N-OH-MeIQx N-OH-Aa C N-OH-PhIP N-OH-1-naphthylamine N-OH-MOCA N-OH-2-amino¯uorene N-OH-2-naphthylamine N-OH-aniline N-OH-aminobiphenyl N-OH-N-methylbenzylamine N-OH-4-aminoazobenzene a b

Individual #10

Individual #16

±b ± 1.52 ^ 0.43 1.56 ^ 0.13 1.97 ^ 1.04 2.89 ^ 0.76 3.06 ^ 0.44 3.50 ^ 1.50 3.58 ^ 1.95 3.61 ^ 0.92 5.00 ^ 0.37 5.39 ^ 1.11

0.29 0.32 ±b ± ± ± ± ± ± ± 3.69 ^ 0.09 ±

Rates were measured as described in Section 2. Not measured.

2.5. Intact cell assays Reactions were started by adding 3H-N-OH-PhIP (100 mCi/mmol; ®nal concentration, 50 mM) to the intact cells suspended in serum-free WE (as described above) in a shaking water bath at 378C. At the indicated intervals after addition of N-OH-PhIP, 200 ml was transferred to a 1.5-ml cryotube and fast frozen in liquid nitrogen. Samples were stored at 2808C until analysis by HPLC. In some reactions, cells were

169

treated with N-OH-MBA (0.1, 0.5, or 1 mM) just prior to adding N-OH-PhIP. 2.6. HPLC analysis Frozen samples from the intact cell incubations were thawed and centrifuged to pellet the cell membrane and protein. A 30-ml aliquot of the supernatant was analyzed by HPLC on a Supelcosil C18 column (4:6 £ 200 mm, 5 mm) using a photo-diode array detector and a ¯ow scintillation counter. The elution was conducted at 1.0 ml/min with a linear gradient over 20 min from 45% to 55% methanol in 20 mM diethylamine-acetate (pH 6.4). Under these conditions, N-OH-PhIP eluted at 14.1 min, and PhIP eluted at 15.1 min. 3. Results When this N-OH-reductase activity was ®rst studied in pig liver microsomes, only N-OH-alkylamines were reported to be substrates [1]. Our current studies show that many carcinogenic and non-carcinogenic N-hydroxy heterocyclic and aromatic amines are substrates for this microsomal NADH-dependent N-OH-reductase in rat and human liver microsomes. Every N-OH-aryl or heterocyclic amine available in the authors' laboratory was checked for purity by HPLC, and 12 compounds were chosen for study. Table 1 shows that all 12 N-OH-amines were reduced by the human microsomal NADH-dependent N-OH-

Fig. 1. Reduction of N-OH-PhIP to PhIP in intact rat primary hepatocytes and intact human HepG2 cells: effect of inhibition and induction. A: Rat primary hepatocytes, inhibition of reduction in presence of N-OH-MBA. B: Human HepG2 cells, inhibition of reduction in the presence of N-OH-MBA. C: Human HepG2 cells, effect of CYP1A induction on overall rate of reduction.

170

R.S. King et al. / Cancer Letters 143 (1999) 167±171

Table 2 Comparison of N-OH-amine reduction with 4-ABP-N-hydroxylation in human liver microsomes a Human liver microsomes no.

Reduction of N-OHABP (nmol/min per mg), n ˆ 3

Reduction of N-OHMBA (nmol/min per mg), n ˆ 6

N-Hydroxylation of ABP b (nmol/min per mg), n ˆ 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

0.92 ^ 0.48 0.89 ^ 0.16 0.79 ^ 0.72 0.53 ^ 0.42 0.36 ^ 0.39 0.47 ^ 0.14 0.63 ^ 0.07 0.24 ^ 0.08 0.30 ^ 0.25 1.67 ^ 0.58 ±c ± ± ± ± ±

±c ± ± 3.63 ^ 0.05 2.40 ^ 0.29 ± ± 3.53 ^ 0.85 ± 5.00 ^ 0.37 3.09 ^ 0.07 3.31 ^ 0.08 2.85 ^ 0.19 2.49 ^ 0.05 4.78 ^ 2.29 3.69 ^ 0.09

0.072 0.040 0.075 0.023 0.054 0.090 0.184 0.056 0.110 0.030 0.153 0.097 0 0.290 0.092 0.124

a

Rates were measured as described in Section 2. Unpublished results, George J. Hammons, Division of Molecular Epidemiology, NCTR. c Not measured. b

reductase, including N-OH-heterocyclic amines, NOH-aromatic amines, and N-OH-alkylamines. The N-OH-reductase activity with all 12 N-OH-amines was comparable in rat liver microsomes (data not shown). The enzyme-dependent reduction of N-OH-ABP was subsequently measured in liver microsomal samples from ten individuals and compared to the enzyme-dependent N-hydroxylation of ABP in the same samples (Table 2). In each individual, the rate of N-OH reduction was greater than the rate of Nhydroxylation (2.7- to 55-fold); and there was no signi®cant correlation between oxidation and reduction (r ˆ 0:29). These rates of N-hydroxylation of ABP compare favorably to those observed in other studies with human liver microsomes [5]. In addition, there was signi®cant interindividual variation in both N-OH-ABP reduction and ABP N-hydroxylation. Reduction of N-OH-ABP varied nearly 7-fold over the ten individual samples, while ABP N-hydroxylation varied about 13-fold. We also studied the reduction of N-OH-PhIP in primary hepatocytes from untreated F344 rats to con®rm our previous results and to study further the

mechanism of the reduction. The reduction of N-OHPhIP to PhIP in primary hepatocytes from untreated F344 rats was very rapid (5.1 nmol/min per 10 7 cells) and complete reduction was observed only 30 min after treating the cells with N-OH-PhIP (Fig. 1A). To provide evidence that the hepatocellular reduction of N-OH-PhIP was catalyzed by the same N-OHreductase enzyme system characterized in human and rat liver microsomes, N-OH-MBA was added to the incubation mixture. In the presence of this competitive substrate, the rate of reduction of N-OH-PhIP was slowed but not completely prevented (2.1 nmol/ min per 10 7 cells) (Fig. 1A), indicating that the reduction of N-OH-PhIP and N-OH-MBA was catalyzed by the same enzyme system. We also examined this N-OH-reductase activity in human HepG2 cells. In these intact cell incubations, the rate of N-OH-PhIP reduction (0.15 nmol/min per 10 7 cells) was appreciably slower than in rat hepatocytes (Fig. 1B). This slower rate of reduction was expected, as the rate of metabolism in HepG2 cells is generally much slower than in rat hepatocytes. Addition of the competitive substrate, N-OH-MBA, decreased the rate of reduction of N-OH-PhIP to

R.S. King et al. / Cancer Letters 143 (1999) 167±171 7

0.10 nmol/min per 10 cells, again indicating that the reduction of N-OH-PhIP and N-OH-MBA was catalyzed by the same enzyme system. Our previous experiments with rat hepatocytes showed that induction of CYP1A activity affected the overall oxidation-reduction balance. N-Hydroxylation of PhIP was observed only in CYP1A-induced hepatocytes; no N-oxidized metabolites (N-OH-PhIP or N-OH-PhIP-glucuronides) were observed in hepatocytes from untreated animals. We conducted similar experiments with HepG2 cells treated with 3-MC to induce CYP1A activity. Fig. 1C shows that induction of CYP1A decreased the overall rate of N-OH-PhIP reduction. Addition of the competitive substrate, NOH-MBA, decreased the rate of reduction even further (data not shown). 4. Discussion The enzymatic N-hydroxylation of aromatic and heterocyclic amines has been studied for many years; however, this is the ®rst study to report enzymatic reduction of the N-hydroxy derivative of carcinogenic aromatic and heterocyclic amines back to the parent amine. We found that each N-OH amine examined (n ˆ 12) was reduced to the amine in an enzymedependent and NADH-dependent manner with both human and rat liver microsomes. Reduction of NOH-PhIP to PhIP was also observed in intact rat liver hepatocytes and human HepG2 cells. Comparison of the rates of ABP N-oxidation and N-OH-ABP reduction in human liver microsomes showed that the rate of reduction was 3±55-fold higher than the rate of Noxidation. However, it should be emphasized that comparisons of the rates of oxidation vs. reduction in vitro may not accurately re¯ect the relative rates in vivo. Nevertheless, in intact rat hepatocytes, overall N-oxidation of PhIP was only observed in CYP1Ainduced cells, even though both induced and uninduced hepatocytes quickly reduced N-OH-PhIP to PhIP. In intact HepG2 cells, CYP1A induction decreased the overall rate of conversion of N-OHPhIP to PhIP as compared to uninduced cells. Future studies will be necessary to quantify the contribution of

171

this pathway to the in vivo situation, including determination of oxidation-reduction cycling in vivo, and further characterization of the enzyme system involved. As indicated by the 16 individual human liver microsomes studied, this N-OH-amine reduction exhibits considerable inter-individual variation. Furthermore, this variation in N-OH amine reduction does not correlate with the widely observed inter-individual variation in CYP-dependent N-hydroxylation of carcinogenic aromatic and heterocyclic amines [5,7]. Thus, this pathway seems to add a new level of complexity to the study of the role of aromatic and heterocyclic amines in the etiology of human carcinogenesis.

References [1] F.F. Kadlubar, E.M. McKee, D.M. Ziegler, Reduced pyridine nucleotide-dependent N-hydroxy amine oxidase and reductase activities of hepatic microsomes, Arch. Biochem. Biophys. 156 (1973) 46±57. [2] F.F. Kadlubar, D.M. Ziegler, Properties of a NADH-dependent N-hydroxy amine reductase isolated from pig liver microsomes, Arch. Biochem. Biophys. 162 (1974) 83±92. [3] A.E. Cribb, S.P. Spielberg, G.P. Grif®n, N4-Hydroxylation of sulfamethoxazole by cytochrome P450 of the cytochrome P4502C subfamily and reduction of sulfamethoxazole hydroxylamine in human and rat hepatic microsomes, Drug Metab. Dispos. 23 (1995) 406±414. [4] B. Clement, R. Lomb, W. MoÈller, Isolation and characterization of the protein components of the liver microsomal O2-insensitive NADH-benzamidoxime reductase, J. Biol. Chem. 272 (1997) 19615±19620. [5] M.A. Butler, M. Iwasaki, F.P. Guengerich, F.F. Kadlubar, Human cytochrome P-450PA (P4501A2), the phenacetin Odeethylase, is primarily responsible for the hepatic 3-demethylation of caffeine and N-oxidation of carcinogenic arylamines, Proc. Natl. Acad. Sci. USA 80 (1989) 7696±7700. [6] J.W. Oldham, D.A. Casciano, J.A. Farr, The isolation and primary culture of viable, nonproliferating rat hepatocytes, Tissue Culture Manual 5 (1979) 1047±1050. [7] R.J. Turesky, A. Constable, J. Richoz, N. Varga, J. Markovic, M.V. Martin, F.P. Guengerich, Metabolic activation of 2amino-3,8-dimethyl[4,5-f]quinoxaline and 2-amino-1-methyl6-phenylimidazo[4,5-b]pyridine by rat and human microsomes and by puri®ed rat P4501A2 and recombinant human P4501A2, Chem. Res. Toxicol. 11 (1998) 925±936.