Abstract
N′-nitrosonornicotine (NNN) induces tumors in the rat nasal cavity and esophagus and is believed to be a causative agent for esophageal cancer in tobacco users. To exert its carcinogenic potential, NNN must be metabolically activated by α-hydroxylation at either the 2′- or 5′-carbon. We previously reported that the human cytochrome P450 (P450), 2A6, efficiently and specifically catalyzed NNN 5′-hydroxylation. P450 2A3, which is expressed in the rat nasal cavity and to a small extent in the esophagus, is closely related to P450 2A6. P450 2A3, like 2A6, is a good catalyst of NNN α-hydroxylation (Km 7 μM; Vmax17 nmol/min/nmol). However, in contrast to P450 2A6, 2A3 catalyzed both 5′- and 2′-hydroxylation of NNN. The ratio of 2′- to 5′-hydroxylation was 1:3. These data, both with P450 2A6 and 2A3, were obtained using racemic NNN. P450 2A3 catalyzed metabolism of (S)-NNN occurred exclusively at the 5′-position. The predominant pathway of (R)-NNN metabolism was 2′-hydroxylation, and occurred to a 3-fold greater extent than did 5′-hydroxylation. These data are in contrast to those obtained from a recent study of (R)- and (S)-NNN metabolism by cultured rat esophagus. In that study, (S)-NNN was metabolized predominantly by 2′-hydroxylation and (R)-NNN equally by 2′- and 5′-hydroxylation. Taken together, these data provide strong evidence that P450 2A3 is not the rat esophageal P450 that catalyzes the metabolic activation of NNN. P450 2A3 may be an important catalyst of NNN activation in rat nasal mucosa.
N′-Nitrosonornicotine (NNN)1 is an esophageal and nasal carcinogen in the rat and is believed to be a causative agent for esophageal and oral cancer in tobacco users (Hecht, 1998). This tobacco-specific nitrosamine requires metabolic activation to exert its carcinogenic potential. Activation is believed to occur through P450-catalyzed α-hydroxylation in the target tissue. It has been proposed by us and others that this tissue-specific activation contributes to the tissue-specific induction of tumors by nitrosamines (Bartsch et al., 1977; Hodgson et al., 1980; Murphy and Spina, 1994). However, what P450s catalyze the α-hydroxylation of NNN in either the esophagus or nasal mucosa have yet to be determined. Previously, we reported that the human P450 2A6, a coumarin 7-hydroxylase, is an efficient and specific catalyst of NNN α-hydroxylation at the 5′-carbon (Patten et al., 1997). In the study presented here, we characterized the metabolism of NNN by a rat coumarin 7-hydroxylase, P450 2A3, which is 85% homologous to P450 2A6 (Kimura et al., 1989).
α-Hydroxylation of NNN may occur at either the 2′- or 5′-carbon [Fig. 1; (Hecht, 1998)]. Both hydroxy NNN products formed are unstable. They decompose to the corresponding diazohydroxides, which then react with H2O to produce HPB and lactol respectively (Fig. 1). The pyridyloxobutyl diazohydroxide formed by 2′-hydroxylation of NNN can alkylate DNA, generating adducts that release HPB upon hydrolysis (Hecht, 1998). DNA adduct formation by the 5′-hydroxylation pathway has not been detected to date. The 2′-hydroxylation pathway is the predominant one in both the rat esophagus and nasal cavity (Hecht et al., 1982, 1998; Brittebo et al., 1983; Murphy et al., 1990). Preferential α-hydroxylation at the 2′-position and the detection of HPB releasing DNA adducts in these tissues, which are susceptible to NNN tumorigenesis, has led to the view that 2′-hydroxylation is the more important pathway of NNN activation (Murphy et al., 1990; Trushin et al., 1994; Hecht, 1998).
α-Hydroxylation of NNN.
Rat esophageal microsomes catalyze both the 2′- and 5′-hydroxylation of NNN (Murphy and Spina, 1994). This activity is NADPH-dependent and inhibited by carbon monoxide, consistent with the role of a P450 enzyme as the catalyst of these reactions. The total α-hydroxylation of NNN has an apparent Km of 49 μM and the ratio of 2′- to 5′-hydroxylation is 3:1 for all concentrations of NNN (Murphy and Spina, 1994). There are few data on what P450s are present in the esophagus. P450 1A1 and P450 17 mRNA have been detected by RT-PCR (Traber et al., 1992; Valle et al., 1995). P450 1A1 protein was detected by Western blot analysis, whereas P450 2E1 and 2B1 were not (Ahn et al., 1996). Using RT-PCR, we recently detected low levels of P450 2A3 expression in the esophagus (Gopalakrishnan et al., 1999). The mRNA level was 1/60th of that detected in the lung, the tissue from which the cDNA for P450 2A3 was originally isolated (Kimura et al., 1989). We did detect small amounts of P450 2A protein by Western blot analysis with an antibody to the related mouse P450 2A5 (Gopalakrishnan et al., 1999).
Rat nasal mucosa microsomes also catalyze both the 2′- and 5′-hydroxylation of NNN (Patten et al., 1998). TheKm values for these reactions are 2.5 and 2.9 μM, respectively, and the ratio of 2′- to 5′-hydroxylation is 1:3. The metabolism of NNN by both pathways is inhibited by coumarin, IC50 6 μM (Patten et al., 1998). Rat nasal tissue microsomes catalyze the 7-hydroxylation of coumarin, and P450 2A3 is one of the major P450s present in this tissue (Liu et al., 1996;Su et al., 1996; Patten et al., 1998). Therefore, one candidate for the P450 catalyst of NNN hydroxylation is P450 2A3. Further support for the role of P450 2A3 in the catalysis of NNN α-hydroxylation in rat nasal microsomes is the observed inhibition of this reaction by 8-methoxypsoralen, IC50 0.25 μM (Patten et al., 1998). 8-Methoxypsoralen is an inhibitor of both the mouse and human coumarin 7-hydroxylases, P450s 2A6 and 2A5, respectively (Maenpaa et al., 1994; Koenigs et al., 1997).
NNN has a chiral center at the 2′-carbon, and (S)-NNN accounts for between 63 and 95% of the NNN present in tobacco products (Carmella et al., 2000). However, until recently all NNN metabolism studies, including those discussed above, were carried out using racemic NNN. The metabolism of the two enantiomers of NNN by cultured rat esophagus was studied, and it was reported that (S)-NNN is metabolized predominantly by 2′-hydroxylation, whereas (R)-NNN is metabolized equally well by both 2′- and 5′-hydroxylation (McIntee and Hecht, 2000). We report here on the relative efficiency and selectivity of (S)- and (R)-NNN metabolism by baculovirus-expressed P450 2A3.
Materials and Methods
Caution.
NNN is a carcinogen and mutagen and therefore should be handled with extreme care, using appropriate protective clothing and ventilation at all times.
Chemicals.
Racemic [5-3H]NNN (purity > 97%, 27 Ci/mmol), was obtained from Moravek Biochemicals (Brea, CA.). [5-3H]NNN was purified by reverse-phase HPLC before use whenever needed. [5-3H]-(R)-NNN (29 Ci/mmol) and [5-3H]-(S)-NNN (28 Ci/mmol) were synthesized as described previously (McIntee and Hecht, 2000). NNN, HPB, and lactol were gifts from Stephen Hecht (University of Minnesota Cancer Center).
NNN Metabolism by P450 2A3.
P450 2A3 was expressed in baculovirus Sf9 insect system, and cell membrane prepared as previously described (Liu et al., 1996). P450 2A3-catalyzed metabolism of [5-3H]NNN was analyzed in an analogous fashion to P450 2A3-catalyzed metabolism ofN-nitrosobenzylmethylamine as was previously described (von Weymarn et al., 1999). The concentration of NNN ranged from 0.2 to 400 μM (0.1–3.0 Ci/mmol). Metabolite standards were added to the terminated reactions, which were frozen at −20°C until analyzed by reverse-phase HPLC with radioflow detection. Two HPLC systems were used, which have been described previously: an ammonium acetate pH 7/methanol, system I; and an ammonium acetate pH 4.6/methanol, system II (Carmella and Hecht, 1985; Patten et al., 1997). The products of the reactions were determined to be linear with time and protein concentration. The kinetic parameters for NNN metabolism were determined in duplicate. To confirm the identity of the products of P450 2A3-catalyzed NNN metabolism, metabolites were analyzed on HPLC system I, then collected and reanalyzed on system II. Unlabeled metabolites were added as internal standards.Km and Vmaxvalues were determined using EZ-FIT5 kinetics program from Perrella Scientific (Amherst, NH). This program uses a nonlinear regression method to fit the curves and the Runs test of residuals to determine statistically if experimental data are randomly distributed among the curve with 95% confidence (Perrella, 1988).
Results and Discussion
The products of P450 2A3-catalyzed metabolism of racemic [5-3H]NNN were analyzed by radioflow HPLC. Using HPLC system I (pH 7.0), only one radioactive metabolite peak was detected (data not shown). This peak coeluted with HPB and lactol, which do not separate in this system. When the reaction products were analyzed on a second HPLC system, two radioactive peaks that coeluted with lactol and HPB were detected (Fig.2A). Lactol accounted for 75% of the metabolites, and the ratio of 2′-hydroxylation to 5′-hydroxylation was 1:3. When (S)-NNN was the substrate, the only metabolite detected was lactol, the product of 5′-hydroxylation. In contrast, HPB and lactol were both products of P450 2A3-catalyzed (R)-NNN metabolism (Fig. 2C). HPB was the predominant product. The ratio of 2′- to 5′-hydroxylation was 2.5:1.
Radioflow HPLC analysis of the products of P450 2A3-catalyzed metabolism of racemic NNN (A), (S)-NNN (B), and (R)-NNN (C).
HPLC system II was used. P450 2A3 (2 pmol) was incubated with 1 μM [5-3H]NNN (3.4 Ci/mmol), or (S)- or (R)-[5-3H]NNN (3.0 Ci/mmol) for 10 min. The reaction with racemic NNN was analyzed directly on HPLC system II (panel A). The reactions with either (S)- or (R)-NNN were first analyzed on HPLC system I; the metabolite peak containing both HPB and lactol was collected and reanalyzed on HPLC system II (panels B and C).
Kinetic parameters for P450 2A3-catalyzed metabolism of racemic, (R)-, and (S)-NNN were determined and are presented in Table 1. TheKm for total α-hydroxylation (2′- and 5′-hydroxylation) of racemic NNN was 13 μM and theVmax was 17.8 nmol/min/nmol of P450. Both (R)- and (S)-NNN were efficiently metabolized by P450 2A3 with Km values between 5 and 19 μM. P450 2A3 catalyzed the 5′-hydroxylation of (S)-NNN somewhat more efficiently than (R)-NNN,Vmax/Km values were 1.8 and 0.49, respectively. TheVmax/Km value for the 2′-hydroxylation of (R)-NNN was 1.1.
Kinetic parameters of P450 2A3-catalyzed metabolism of racemic, (R)-, and (S)-NNN
The relative affinity of P450 2A3 for (R)- and (S)-NNN is quite similar. However the product distribution is strikingly different. An explanation for this is that (S)-NNN can only assume one orientation within the active site resulting in the oxidation of the 5-carbon, whereas (R)-NNN can assume two orientations allowing both 2′- and 5′-hydroxylation to occur. These data on the selectivity of oxidation of (R)- and (S)-NNN should be useful in modeling the active site of P450 2A3.
An earlier study (McIntee and Hecht, 2000) determined the relative rates of 2′- and 5′-hydroxylation of the two NNN enantiomers by cultured rat esophagus. In that study, which used 1 μM NNN, the (S)-enantiomer was metabolized by 2′-hydroxylation at a rate 6 to 8 times that of 5′-hydroxylation. (R)-NNN was metabolized equally well by both pathways. These data, particularly those for (S)-NNN are inconsistent with P450 2A3 contributing significantly to NNN metabolism in the esophagus. Previously, we reported that rat esophageal microsomes do not catalyze the 7-hydroxylation of coumarin, yet P450 2A3 efficiently catalyzes this reaction. Therefore, it appears that the level of P450 2A3 in the esophagus is insufficient to contribute significantly to NNN metabolism in this tissue. We have reached a similar conclusion for the metabolism of the potent esophageal carcinogen,N-nitrosomethylbenzylamine (von Weymarn et al., 1999).
Although P450 2A3 does not appear to be responsible for nitrosamine activation in the esophagus, it may contribute in the nasal cavity. The level of P450 2A3 mRNA in the rat nasal mucosa is more than 1000 times greater than that in the esophagus (Gopalakrishnan et al., 1999). P450 2A3 is a major P450 in rat nasal mucosa (Su et al., 1996). The metabolism of the NNN enantiomers has not been studied in the nasal mucosa; racemic NNN is metabolized by both 2′- and 5′-hydroxylation. Cultured rat nasal mucosa preferentially metabolized NNN by 2′-hydroxylation, and the ratio of 2′- to 5′-hydroxylation was between 2 and 3 (Brittebo et al., 1983). Microsomes from the nasal mucosa catalyzed 2′- and 5′-hydroxylation equally with aKm of less than 3 μM. P450 2A3 preferentially catalyzes 2′-hydroxylation of racemic NNN (Fig. 1A). Therefore, while P450 2A3 may contribute to the metabolism of NNN in the rat nasal mucosa, it appears that another P450 also plays a role.
In conclusion, the extrahepatic rat P450, 2A3, preferentially catalyzes the metabolism of racemic NNN by 5′-hydroxylation and exclusively catalyzes the 5′-hydroxylation of (S)-NNN. However, the hydroxylation of (R)-NNN by this P450 occurs preferentially at the 2′-carbon. This is in contrast to what was previously reported for the metabolism of (R)- and (S)-NNN by cultured rat esophagus (McIntee and Hecht, 2000), and thus is inconsistent with a significant role for P450 2A3 in the catalysis of NNN α-hydroxylation in the rat esophagus.
Footnotes
-
Send reprint requests to: Sharon E. Murphy, University of Minnesota Cancer Center, 420 Delaware St. SE, Minneapolis, MN. E-mail: murph062{at}tc.umn.edu
-
This study was supported by Grant CA-74913 (to S.E.M.) from the National Cancer Institute and Grant ES-07462 (to X.D.) from the National Institute of Environmental Health Sciences.
- Abbreviations used are::
- NNN
- N′-nitrosonornicotine
- HPB
- 4-hydroxy-4-(3-pyridyl)-1-butanone
- lactol
- 5-(3-pyridyl)-2-hydroxytetrahydrofuran
- P450
- cytochrome P450
- RT-PCR
- reverse transcription-polymerase chain reaction
- Received June 16, 2000.
- Accepted July 25, 2000.
- The American Society for Pharmacology and Experimental Therapeutics
