Abstract
Kinetic parameters were determined for the hydroxylation ofN′-nitrosonornicotine (NNN),N-nitrosobenzylmethylamine (NBzMA), coumarin, and ethoxycoumarin catalyzed by rat nasal mucosa microsomes. NNN is a tobacco-specific nitrosamine that, in rats, causes tumors in the nasal cavity and esophagus, whereas NBzMA induces tumors in rat esophagus. Both nitrosamines require α-hydroxylation to exert their carcinogenic effects. NNN, NBzMA, coumarin, and ethoxycoumarin were all extensively hydroxylated by rat nasal mucosa microsomes. TheKM values for the hydroxylation of each substrate were low, ranging between 2 and 5 μM. 2′- and 5′-Hydroxylation of NNN were catalyzed to a similar extent. NBzMA was metabolized predominantly to benzaldehyde, the product of debenzylation, or methylene hydroxylation. 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), NNN, and NBzMA were inhibitors of coumarin and ethoxycoumarin hydroxylation. NNN hydroxylation by nasal mucosa microsomes was inhibited by coumarin, ethoxycoumarin, NNK, and NBzMA but not byN-nitrosodimethylamine. 8-Methoxypsoralen, a potent inhibitor of P450 2A6- and 2a5-dependent coumarin hydroxylation in human and mouse liver microsomes, also significantly inhibited NNN activation. The results of this study suggest that the four substrates examined are hydroxylated by closely related P450 enzymes in rat nasal mucosa and that a coumarin hydroxylase metabolizes both NNN and NBzMA.
N′-Nitrosonornicotine (NNN)1 and NBzMA are potent carcinogens in animal species and are believed to be carcinogenic to humans (Castonguay et al., 1984; Hecht and Hoffman, 1989; Lijinsky, 1987; Mirvish, 1995). NNN is a tobacco-specific, nicotine-derived carcinogen formed during tobacco smoking or processing (Hecht and Hoffman, 1989). In rats it induces tumors in the nasal cavity and esophagus (Hecht and Hoffman, 1989;Castonguay et al., 1984). NBzMA is a potent esophageal carcinogen in rats (Mirvish, 1995). The carcinogenicity of these compounds depends on their activation by cytochrome P450-mediated hydroxylation of the carbons α- to the nitroso group. α-Hydroxylation of NNN forms 2′ and 5′-hydroxy NNN (Hecht and Hoffman, 1989). The former decomposes to a reactive diazohydroxide, which can either pyridyloxobutylate DNA or react with water to form keto alcohol (Hecht and Hoffman, 1989; Castonguay et al., 1984). Hydroxylation of the 5′-carbon produces a related diazohydroxide, which can potentially form a DNA adduct or react with water to form lactol (Hecht and Hoffman, 1989). NBzMA can be hydroxylated at either the methyl (N-demethylation) or the methylene carbon (N-debenzylation).N-Demethylation generates formaldehyde and a DNA benzylating agent, and N-debenzylation produces a DNA methylating agent and benzaldehyde (Labuc and Archer, 1982; Peterson, 1997). The latter pathway is believed to be critical for NBzMA tumorigenesis.
One striking characteristic of nitrosamines is their tissue-specific tumor induction. This tissue specificity may be explained in part by tissue-specific activation of these compounds (Hecht and Hoffman, 1989). Whereas a significant amount of work has been carried out on the metabolism of these compounds, little is known about which P450s activate these carcinogens, particularly in extrahepatic tissues. NNN is a nasal carcinogen in the rat and is efficiently metabolized by rat nasal tissue in culture (Castonguay et al., 1984). We recently demonstrated that human P450 2A6 efficiently metabolized this nitrosamine (Patten et al., 1997), and preliminary results in our laboratory suggest NBzMA may be an even better substrate for P450 2A6. We and others have suggested that the metabolism of nitrosamines by rat esophagus and nasal tissue is qualitatively similar (Castonguay et al., 1984; Koenigsmann et al., 1988; Murphy et al., 1990), although total metabolism by nasal tissue is greater. This is most likely due to the significantly greater level of P450 enzymes in the nasal mucosa. We believe that understanding nitrosamine metabolism in rat nasal mucosa will increase our knowledge of nitrosamine activation in the rat esophagus, the tissue most sensitive to nitrosamine tumorigenesis. Therefore, in the experiments presented here, we have characterized α-hydroxylation of NNN and NBzMA in nasal mucosa microsomes and investigated a role for a nasal P450 2A enzyme in the metabolism of these nitrosamines.
Materials and Methods
Chemicals.
[5-3H]NNN (3.9 Ci/mmol), NNN, and NNK were purchased from Chemsyn Science Laboratories (Lenexa, KS). NNN metabolite standards (keto alcohol, hydroxy acid, lactol, NNN-N oxide, and keto acid) were a gift from Stephen Hecht (University of Minnesota Cancer Center, Minneapolis, MN). NBzMA was obtained from the NCI Chemical Carcinogen Repository (Midwest Research Institute, Kansas City, MO). [3H-benzyl]NBzMA (1.8 Ci/mmol) was synthesized as described elsewhere (Peterson, 1997). [5-3H]NNN and [3H-benzyl]NBzMA were purified prior to use by HPLC. All other chemicals were purchased from Sigma and Aldrich.
Microsome Preparation and Enzyme Assays.
Male Fischer-344 rats, weighing 200 to 300 g (Charles River Breeding Laboratories, Kingston, NY) were maintained on NIH-07 diet prior to sacrifice by CO2 affixation and decapitation. Microsomes were prepared from the nasal mucosa (both respiratory and olfactory epithelium) as previously described (Hadley and Dahl, 1982) and then resuspended in 50 mM potassium phosphate (pH 7.4) containing 0.3 M sucrose. The metabolism of NNN and NBzMA by these microsomes was determined over a range of concentrations (0.5–80 μM). Microsomes (4 μg) were incubated with either 0.3 μCi [5-3H]NNN or 0.6 μCi [3H-benzyl]NBzMA, an NADPH generating system (0.4 mM NADP+, 100 mM glucose 6-phosphate, and 0.4 units/ml glucose-6-phosphate dehydrogenase) in 0.2 ml of 80 mM sodium phosphate buffer (pH 7.4). For the analysis of NBzMA metabolism, semicarbazide (20 mM, pH 7.0) was included to trap benzaldehyde. The reaction mixtures were terminated by the addition of 20 μl of 0.3 M zinc sulfate and 20 μl of 0.3 M barium hydroxide and then centrifuged, filtered, and coinjected with the appropriate metabolite standards onto a reversed phase HPLC system with radioflow detection. The metabolites were separated on a 4.6 × 30-mm Waters C18-μBondapak column (Millipore, Milford, MA). NNN metabolites were separated using a linear gradient from 100% solvent A (25 mM ammonium acetate, pH 4.5) to 75% solvent A and 25% solvent B (100% methanol) in 60 min (Carmella and Hecht, 1985). NBzMA metabolites were separated with an isocratic system; the mobile phase was 15% solvent A (adjusted to pH 4.0) and 85% solvent B (Pattenet al., 1997). Coumarin and ethoxycoumarin hydroxylation was assayed by previously described methods (Yun et al., 1991).
Results and Discussion
We report here the ability of rat nasal mucosa to metabolize two carcinogenic nitrosamines and two well known P450 substrates, 7-ethoxycoumarin and coumarin. Rat nasal mucosa contains significant levels of cytochrome P450 and metabolizes a number of xenobiotics (Reed, 1993). 7-Ethoxycoumarin metabolism by rat nasal microsomes was previously reported, but no kinetic parameters were determined (Reedet al., 1986). The kinetic parameters for both 7-ethoxycoumarin O-deethylation and coumarin 7-hydroxylation are presented in table 1. TheKM values for these two reactions were similar, 1.5 and 1.8 μM, respectively. TheVmax for 7-ethoxycoumarinO-deethylation was roughly 12-fold greater than coumarin hydroxylation (table 1). 7-Ethoxycoumarin O-deethylation is catalyzed by several unrelated P450 enzymes in human and rat liver microsomes; these include 2E1, 2B, and 1A (Yamazaki et al., 1996). In contrast, the 7-hydroxylation of coumarin is mainly catalyzed by a number of highly related P450 2A enzymes, which include 2A6 (human liver), 2a5 (mouse liver), and 2A10 (rabbit nasal mucosa) (Dinget al., 1994; Raunio et al., 1988). The detection of coumarin 7-hydroxylation activity in rat nasal mucosal microsomes suggests the presence of a related P450 2A enzyme in this tissue.Bereziat and co-workers (1995) previously reported evidence that a P450 2a5-related enzyme is expressed in rat nasal mucosa.
Rat nasal mucosal microsomes catalyzed both 2′- and 5′-hydroxylation of NNN. The products of these two reactions in microsomal incubations, keto alcohol and lactol, were quantified by radioflow HPLC (Pattenet al., 1997). A low KM value, between 2 and 3 μM, was obtained for both NNN α-hydroxylation pathways (table 1). The low KM value obtained for NNN hydroxylation supports the hypothesis that the high sensitivity of the nasal cavity to tumor induction by NNN is due to efficient target tissue activation of this carcinogen. Previous studies demonstrated that the product of 2′-hydroxylation is more mutagenic than the product of 5′-hydroxylation (Hecht and Lin, 1986). Therefore, 2′-hydroxylation may be more important in tumor induction. TheVmax values for the two pathways were similar (2′ to 5′ ratio of 0.77, table 1).
The KM values obtained for both methylene and methyl hydroxylation of NBzMA were also low, ranging between 3 and 4 μM (table 1). The rate for methylene hydroxylation was calculated as the sum of benzaldehyde and benzoic acid formation. The latter assumes benzoic acid is generated by the oxidation of benzaldehyde. This is discussed below. Benzoic acid accounted for 2 to 25% of the total metabolites, depending on the concentration of NBzMA used (data not shown). The rate of formation of benzaldehyde was 5-fold greater than either that for benzyl alcohol or benzoic acid. Benzyl alcohol is most likely derived from the reaction of phenylmethanediazohydroxide (the product of methylene hydroxylation) with water. Benzyl alcohol may also form by the reduction of benzaldehyde. Although we have not excluded this possibility, it is unlikely given the formation of benzyl alcohol in the presence of semicarbazide. Also, in the absence of semicarbazide, there was no increase in the formation of benzyl alcohol. Formaldehyde formation, the other product of methyl hydroxylation, was not measured in this study. Semicarbazide, included in the analysis to trap benzaldehyde, did not seem to inhibit NBzMA activation as the total metabolism of NBzMA was unaffected by its omission. In the absence of semicarbazide, the rate of benzoic acid formation doubled, whereas the rate of benzaldehyde formation decreased by half. This parallel relationship of benzoic acid formation to benzaldehyde formation suggests that some, if not all, of the benzoic acid originates from benzaldehyde oxidation. Aldehyde dehydrogenases, which could mediate this reaction, have been reported in rat nasal mucosa (Yun et al., 1991).
The data reported here demonstrate that rat nasal mucosal microsomes contain high NBzMA debenzylation activity, i.e. efficiently generate a DNA methylating agent. This result is consistent with whole-body radioautography studies in Sprague Dawley rats treated with radioactive NBzMA containing 14C in the methyl or benzyl moiety. In that study the nasal cavity, lung, and esophagus tissues were the most extensively labeled by [14C-methyl]NBzMA (Kraft and Tannenbaum, 1980). Nasal mucosa was also extensively labeled by [14C-benzyl]NBzMA, suggesting thatN-demethylation is also carried out by this tissue; a result that agrees with the data presented here. Nasal cavity and lung, however, are not considered target organs of NBzMA tumorigenicity (Lijinsky, 1987). A lack of correlation between DNA methylation and tumor formation in lung and nasal cavity of Fischer rats has also been observed with N-nitrosomethylalkylamines (Koenigsmannet al., 1988). In these cases, it has been proposed that the animals die from esophageal tumors, due to impaired food intake, before tumors in other organs can develop; a similar argument may be evoked for NBzMA tumorigenesis.
To investigate the role of a common enzyme in the metabolism of the compounds in table 1, we carried out a series of inhibitor studies. Many nitrosamines, including NNN and NBzMA, were tested for their ability to inhibit the hydroxylation of coumarin and ethoxycoumarin (table 2). NNN and NBzMA, as well as NNK, were potent inhibitors of both reactions. Coumarin 7-hydroxylation was inhibited 73–97%, and 7-ethoxycoumarin O-deethylation was inhibited 80–97% by a 60-μM concentration of these three nitrosamines. NDEA and NPIP moderately inhibited both reactions (36–62%). NDMA and NPYR did not inhibit either coumarin 7-hydroxylation or 7-ethoxycoumarin dealkylation. The inhibition of coumarin hydroxylation by these nitrosamines, with the exception of NBzMA, roughly parallels their nasal cavity tumorigenicity. That is, the nasal carcinogens NNN, NNK, NPIP, and NDEA (Hecht and Hoffman, 1989; Kraft and Tannenbaum, 1980) are inhibitors of both coumarin and 7-ethoxycoumarin metabolism. But NDMA and NPYR, liver carcinogens that do not induce nasal tumors (Lijinsky, 1987), do not inhibit the nasal metabolism of either compound.
An attempt was made to determine if the NNN inhibition of coumarin-7-hydroxylation was competitive. The results of these experiments were confusing, most likely due to the mixture of P450s present in the nasal mucosa. The effect of NNN on coumarin metabolism was dependent on the concentration of coumarin (data not shown), suggesting there may be two coumarin hydroxylases in rat nasal mucosa. This would not be surprising, as this is the case in rabbit nasal mucosa (Maenpaa et al., 1994). Further data supporting the role of a coumarin hydroxylase or a P450 2A in the metabolism of NNN include its inhibition by coumarin and 8-MP. 10 μM Coumarin inhibited both 2′- and 5′-hydroxylation of NNN greater than 70%; higher concentrations (100 and 200 μM) produced almost 100% inhibition (table 3 and fig.1). The IC50 value for coumarin-dependent inhibition of both 5′- and 2′-NNN hydroxylation was 6 μM. 8-MP, a selective inhibitor of coumarin 7-hydroxylation in humans (P450 2A6) and mouse (P450 2a5) (Peng et al., 1993), was also a potent inhibitor of NNN α-hydroxylation in nasal mucosa microsomes. The IC50 value for 8-MP inhibition was 0.25 μM (data not shown).
NNN metabolism was also completely inhibited by NBzMA and 7-ethoxycoumarin. No inhibition was observed with an equal concentration of NDMA or NPYR (table 3). Likewise, 1 μM NBzMA hydroxylation was significantly inhibited by NNN (73%) and coumarin (96%) at concentrations of 200 μM (results not shown). The inhibition studies with NNN, NBzMA, coumarin, and 7-ethoxycoumarin suggest that the P450s that metabolize these four compounds overlap and that a coumarin hydroxylase contributes to the metabolism of both NNN and NBzMA. NDEA, NPIP, and NNK, all of which induce tumors in the rat nasal cavity (Lijinsky, 1987), may also be substrates for this P450. In this study, we do not identify the rat nasal mucosal P450 enzyme(s) that hydroxylates these nitrosamines. But it seems likely that a rat nasal mucosa enzyme highly related to P450s of the 2A family, which includes P450s 2A6, 2a5, and 2A10/11, is capable of catalyzing these reactions.
P450s 2A6, 2a5, and 2A10/11 all catalyze the 7-hydroxylation of coumarin and share 83% amino acid sequence homology (Maenpaa et al., 1994). Rabbit nasal P450 NMa, which is a mixture of P450 2A10 and 2A11 (Maenpaa et al., 1994), has lowKM values for methyl and methylene α-hydroxylation of NNK, 9 μM and 15 μM, respectively (Honget al., 1992). Rat nasal mucosal microsomes also contain a P450 enzyme with a low KM (10 μM) for both NNK α-hydroxylation pathways (Smith et al., 1992), and it is likely that the rat NNK hydroxylase is orthologous to rabbit P450 2A10/11. Human hepatic P450 2A6 has a lowKM of about 5 μM for the 7-hydroxylation of coumarin (Yamano et al., 1990). Recently, we reported that P450 2A6 is an efficient catalyst of NNN 5′-hydroxylation. TheKM of this reaction was quite low, 2.1 μM (Patten et al., 1997). We hypothesize that the high coumarin and NNN hydroxylation activity we have detected in rat nasal mucosa microsomes is catalyzed by a P450 2A enzyme(s) closely related to 2A6 and 2A10/11. It was reported recently that P450 2A3, a rat ortholog of P450 2A6, is present in rat nasal mucosa (Su et al., 1996). P450 2A3 has been expressed in a baculovirus/insect cell system (Liu et al., 1996), and preliminary results in our laboratory suggest P450 2A3 is an efficient catalyst of both NNN and NBzMA hydroxylation.
Footnotes
-
Send reprint requests to: Sharon E. Murphy, University of Minnesota Cancer Center, 420 Delaware Street SE, Minneapolis, MN 55455.
- Abbreviations used are::
- NNN
- N′-nitrosonornicotine
- NBzMA
- N-nitrosobenzylmethylamine
- NDMA
- N-nitrosodimethylamine
- NDEA
- N-nitrosodiethylamine
- NPYR
- N-nitrosopyrrolidine
- NPIP
- N-nitrosopiperidine
- NNK
- 4-(methylnitrosamino)-1(3-pyridyl)-1-butanone
- keto alcohol
- 4-hydroxy(3-pyridyl)-1-butanone
- lactol
- 2-hydroxy-5(3-pyridyl)-tetrahydrofuran
- 8-MP
- 8-methoxypsoralen
- Received January 10, 1997.
- Accepted October 9, 1997.
- The American Society for Pharmacology and Experimental Therapeutics