Abstract
The objectives of this study were to determine the contributions of CYP2A13 and CYP2A6 to 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) metabolism in human peripheral lung microsomes and to determine the influence of the genetic polymorphism, CYP2A13 Arg257Cys, on NNK metabolism. 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), the keto-reduced metabolite of NNK, was the major metabolite produced, ranging from 0.28 to 0.9%/mg protein/min. Based on total bioactivation of NNK and NNAL by α-carbon hydroxylation, subjects could be classified as either high (17 subjects) or low (12 subjects) bioactivators [(5.26 ± 1.23) × 10-2 and (6.49 ± 5.90) × 10-3% total α-hydroxylation/mg protein/min, P < 0.05]. Similarly, for detoxification, subjects could be grouped into high (9 subjects) and low (20 subjects) categories [(2.03 ± 1.65) × 10-3 and (2.50 ± 3.04) × 10-4% total N-oxidation/mg protein/min, P < 0.05]. When examining data from all individuals, no significant correlations were found between levels of CYP2A mRNA, CYP2A enzyme activity, or CYP2A immunoinhibition and the degree of total NNK bioactivation or detoxification (P > 0.05). However, subgroups of individuals were identified for whom CYP2A13 mRNA correlated with total NNK and NNAL α-hydroxylation and NNAL-N-oxide formation (P < 0.05). The degree of NNAL formation and CYP2A13 mRNA was also correlated (P < 0.05). Subjects (n = 84) were genotyped for the CYP2A13 Arg257Cys polymorphism, and NNK metabolism for the one variant (Arg/Cys) was similar to that for other subjects. Although results do not support CYP2A13 or CYP2A6 as predominant contributors to NNK bioactivation and detoxification in peripheral lung of all individuals, CYP2A13 may be important in some.
Lung cancer is the leading cause of cancer-related death in the world, and it is estimated that cigarette smoking accounts for approximately 90% of lung cancer cases (Hecht, 2003). The tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) is thought to play a major role in human tobacco-related cancers (Hecht, 2003). NNK is the most prevalent pulmonary carcinogen in tobacco smoke and the most potent cancer-causing tobacco-specific nitrosamine in all animal species tested (Hecht, 1998). NNK selectively induces lung adenocarcinoma in animals (Hecht, 1998) and is believed to be a causal agent in the induction of human lung adenocarcinoma, which is now the leading form of lung cancer (Hoffmann et al., 1996; Thun et al., 1997).
To induce carcinogenesis, NNK and its keto-reduced metabolite, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL), require metabolic activation via α-carbon hydroxylation (Fig. 1). Hydroxylation of the α-methylene carbons of NNK and NNAL leads to the formation of DNA-methylating species, whereas hydroxylation of the α-methyl carbons of NNK and NNAL results in the formation of DNA-pyridyloxobutylating and -pyridylhydroxybutylating species, respectively. The formation of both types of adducts is believed to be important in the induction of carcinogenicity by NNK because less potent nitrosamines will only either methylate or alkylate DNA (Hecht, 1999). α-Carbon hydroxylation of NNK and NNAL also results in the formation of four endpoint metabolites, keto acid, keto alcohol, hydroxy acid, and diol. These endpoint metabolites are used in assessing the degree of NNK bioactivation, since their formation is indicative of the formation of the DNA-reactive metabolites. The detoxification of NNK and NNAL occurs mainly through pyridine N-oxidation, which results in the formation of excretable N-oxides (Hecht, 1998).
Cytochromes P450 (P450s), specifically CYP2A6/2A13, CYP2B6, CYP3A4/3A5, and CYP2E1 (Smith et al., 1992, 1995, 1999, 2003; Hecht, 1998), and prostaglandin H synthase (Smith et al., 1995, 1999) but not lipoxygenases (Bedard et al., 2002) have been implicated in human pulmonary NNK metabolism. Of particular interest are the CYP2A isozymes CYP2A13 and CYP2A6. CYP2A13 is expressed predominantly in the human respiratory tract, including the nasal mucosa, trachea, and peripheral lung (Su et al., 2000), and heterologously expressed CYP2A13 exhibits higher catalytic activity for NNK activation than do other P450 isoforms examined (Smith et al., 1992; Patten et al., 1996; Su et al., 2000). A functional polymorphism, resulting from a C/T transition in exon 5 of the CYP2A13 gene that leads to an Arg/Cys amino acid substitution at residue 257, significantly reduces the enzyme's activity toward several different substrates, including NNK (Zhang et al., 2002). A potential protective effect against NNK-induced carcinogenesis for individuals possessing this variant allele is supported by an epidemiological study which found that individuals with variant CYP2A13 genotype (CT and TT) had a reduced risk of lung adenocarcinoma compared with individuals with wild-type (CC) genotype (Wang et al., 2003). The other CYP2A isozyme, CYP2A6, is the main CYP2A isoform in liver (Su et al., 2000) but is also present in human nasal mucosa, trachea, and lung (Su et al., 2000). CYP2A6 is a major catalyst of nicotine and coumarin metabolism (Fernandez-Salguero and Gonzalez, 1995; Messina et al., 1997) and is thought to be one of the major P450 isoforms responsible for NNK activation (Yamazaki et al., 1992; Smith et al., 2003). The role of CYP2A6 in lung cancer is not clear, as studies that have examined the relationship between CYP2A6 polymorphisms and lung cancer risk have been conflicting. In one study, CYP2A6 deficiency was correlated with reduced lung cancer risk (Miyamoto et al., 1999), whereas another study showed no relationship (Wang et al., 2003). Because of their established involvement in NNK bioactivation and their expression in human lung, it can be suggested that CYP2A13 and CYP2A6 may contribute largely to human pulmonary NNK metabolism.
In the present study, lung tissue from a relatively large sample size of individuals (n = 29) was used to assess NNK biotransformation among individuals and, specifically, the importance of both CYP2A13 and CYP2A6 in these pathways. In addition, the influence of the CYP2A13 Arg257Cys genetic polymorphism on NNK metabolism was assessed.
Materials and Methods
Chemicals. Chemicals were obtained as follows: [5-3H]NNK (2.4-11.0 Ci/mmol; >98% pure) from Chemsyn Science Laboratories (Lenexa, KS) and Moravek Biochemicals (Brea, CA); NNK, NNAL, 4-(methylnitrosamino)-1-(3-pyridyl-N-oxide)-1-butanone (NNK-N-oxide), 4-(methylnitrosamino)-1-(3-pyridyl-N-oxide)-1-butanol (NNAL-N-oxide), 4-hydroxy-1-(3-pyridyl)-1-butanone (keto alcohol), 1-(3-pyridyl)-1,4-butane diol (diol), 1-(3-pyridyl)-1-butanone-4-carboxylic acid (keto acid), 1-(3-pyridyl)-1-butanol-4-carboxylic acid (hydroxy acid), from Toronto Research Chemicals (North York, ON, Canada); Uniscint BD radioflow scintillation cocktail from National Diagnostics (Atlanta, GA); glucose-6-phosphate dehydrogenase, coumarin, 7-hydroxycoumarin, and 7-hydroxy-4-methylcoumarin from Sigma-Aldrich (St. Louis, MO). All other chemicals were reagent grade and were obtained from common commercial suppliers.
Tissue Procurement. Sections of peripheral human lung tissue devoid of visible tumors were obtained from patients undergoing clinically indicated lobectomy at Kingston General Hospital. Tissue specimens were cut into 1.5-cm3 pieces, wrapped in aluminum foil, frozen in liquid N2, and stored at -80°C until microsome preparation (Smith et al., 2003). Histological analysis was performed on tissues to confirm the absence of microscopic tumors. Data regarding surgical diagnosis, gender, smoking history, potential occupational carcinogen exposure, and drug treatments for the month before surgery were collected to identify possible confounders including the possible inductive/inhibitory effects of certain drug treatments on biotransformation enzymes. Patients were classified as former smokers if smoking termination was reported to be >2 months before surgery (McLemore et al., 1990).
Preparation of Human Lung Microsomes. Human whole peripheral lung microsomes were prepared as described previously (Smith et al., 2003). Briefly, tissue specimens were thawed on ice for 15 min. They were then rinsed, chopped, and homogenized in 0.1 M potassium phosphate buffer containing 1.15% KCl (pH 7.4) using a Polytron homogenizer, and microsomes were prepared by differential centrifugation (Donnelly et al., 1996). Protein concentration was determined by the method of Lowry et al. (1951) using bovine serum albumin as a standard.
Incubations with NNK. Before use, the purity of [5-3H]NNK was assessed, since high purity (>98%) was needed to ensure that impurities would not interfere with quantification of metabolites. If purity was <98%, [5-3H]NNK was purified by high-performance liquid chromatography as described previously (Smith et al., 1999). Incubation mixtures were prepared as described previously (Smith et al., 2003), using 4.2 μM [5-3H]NNK and 1.0 mg of microsomal protein in a total volume of 1.0 ml. In addition to samples that contained the complete incubation mixture, samples without the NADPH-generating system and samples bubbled with carbon monoxide (CO) were also prepared to assess the overall contribution of P450s. All metabolite analyses were performed in duplicate. NNK metabolite formation by human lung microsomes has been demonstrated to be linear for at least 30 min (Smith et al., 2003).
Assessment of NNK Biotransformation. NNK metabolites were quantified by reverse-phase gradient high-performance liquid chromatography with radiometric detection (Smith et al., 2003). For immunoinhibition, 1.0 mg of microsomal protein and 200 μg of anti-CYP2A6/13 (BD Gentest, Woburn, MA) were incubated on ice for 20 min before use in incubations, as recommended by the supplier. For each metabolite, the amount produced was expressed as a percentage of the total radioactivity recovered from [5-3H]NNK plus metabolites per milligram of protein per minute, to account for differences in recovery of [5-3H]NNK between incubates. Metabolite peaks were quantified only if they were at least twice background radioactivity levels.
CYP2A mRNA Expression. Total RNA was isolated from human lung tissue (n = 28) using the Qiagen RNeasy Mini Kit (Qiagen, Valencia, CA), with an additional on-column DNase treatment step in accordance with the manufacturer's instructions. The quality of the RNA samples was determined by electrophoretic analysis of 3 μg of RNA on a denaturing gel. Ethidium bromide staining of the gel detected distinct 28S and 18S rRNA bands with an intensity ratio of 28S:18S of at least 1.5. The UV absorbance ratio (260 nm/280 nm) ranged from 1.8 to 2.1 for all RNA samples. cDNA was synthesized from 5 μg of total RNA in a reaction volume of 50 μl using the High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA). For quantitative real-time PCR, 2-μl aliquots of cDNA were amplified using CYP2A13 and CYP2A6 TaqMan primer and probes sets (Applied Biosystems; Assay ID Hs00426372_m1 and Hs00868409_s1, respectively) according to the manufacturer's recommendations. Amplification, detection, and analysis were performed using Smart Cycler II instrumentation and software (Cepheid, Sunnyvale, CA). CYP2A13 and CYP2A6 gene expression was normalized using glyceraldehyde-3-phosphate dehydrogenase as an endogenous reference. For mRNA quantitiation, the comparative CT method was used since the relative PCR efficiencies for target and reference amplification were approximately equal. PCR product specificity was verified by the presence of a single PCR product after performing agarose gel electrophoresis with ethidium bromide staining. Due to limitations in tissue availability, CYP2A13 and CYP2A6 gene expression could not be assessed for all individuals.
Coumarin 7-Hydroxylase Activity. Currently, there are no CYP2A13 or CYP2A6 isozyme-specific substrates available. Because CYP2A13 and CYP2A6 both metabolize coumarin at the 7-position, coumarin 7-hydroxylation is often used as a probe activity for the CYP2A enzymes (von Weymarn et al., 2005). 7-Hydroxycoumarin formation was assessed in isolated human lung microsomes (n = 28) and in human CYP2A6 microsomes (BD Biosciences, Woburn, MA) as a positive control, according to the manufacturer's protocol, but with some modifications. Reaction mixtures containing 0.2 mM coumarin in a 250-μl total reaction volume were preincubated at 37°C for 5 min and then incubated for an additional 25 min after addition of 0.3 mg of microsomal protein. Levels of 7-hydroxycoumarin and the internal standard, 7-hydroxy-4-methylcoumarin were assessed using a Shimadzu chromatographic system (Mandel Scientific, Guelph, ON, Canada) with RF-10Axl fluorescence detector and CLASS-VP (v. 7.2.1 SP1) software as described (http://www.cypex.co.uk/) with some modifications. The mobile phase consisted of 75:25 (v/v) 0.05% orthophosphoric acid/acetonitrile, delivered isocratically at 1.0 ml/min through a Supelcosil 5-micron LC-18 (15 cm × 4.6 mm) column (Sigma-Aldrich) at room temperature. Fluorescence was determined with excitation at 324 nm and emission at 458 nm. Levels of 7-hydroxycoumarin formation were quantified by subtracting the fluorescence of boiled microsome samples and comparing it to a standard curve of 7-hydroxycoumarin to 7-hydroxy-4-methylcoumarin peak area ratios as a function of 7-hydroxycoumarin concentration. 7-Hydroxycoumarin formation was linear up to 1 mg of microsomal protein and for 30 min.
Polymorphism Analysis. Genomic DNA was isolated from peripheral lung tissue by protease digestion followed by standard phenol/chloroform extraction and ethanol precipitation (Devereux et al., 1993). Genotypes (n = 84) for CYP2A13 at the C3375T (Arg257Cys) site were assessed by PCR-restriction fragment length polymorphism analysis (Wang et al., 2003) with some modifications. Reaction conditions were as follows: 15 min at 95°C, 13 cycles of 30 s at 94°C, 30 s at 63°C (step-down 0.5°C/cycle), and 50 s at 72°C, followed by 20 cycles of 30 s at 94°C, 30 s at 57.5°C, and 50 s at 72°C, and a final elongation step of 5 min at 72°C. The PCR resulted in a 375-bp product. Ten microliters of PCR product were digested with 2.9 units of HhaI (New England Biolabs, Inc., Beverly, MA) and restriction products were separated on a 3% agarose gel. The wild-type C allele had a HhaI restriction site that resulted in two bands (217 and 158 bp), and the variant T allele resulted in the elimination of the restriction site, producing a single 375-bp band.
Statistical Analysis. For microsomal NNK biotransformation, individual metabolite values represent the mean of duplicate incubations. When grouped, microsomal data are presented as means ± S.D. Statistically significant differences in grouped microsomal data were determined by Student's t test and Student's t test with Welch's correction if heterogeneity of variance was present. All correlation analyses were performed using Pearson product moment correlation analysis (GraphPad Prism 4 software; GraphPad Software Inc., San Diego, CA). P < 0.05 was considered statistically significant in all cases. To determine whether NNK bioactivation and detoxification group data were normally distributed, an Anderson-Darling Normality test was used. Group data with P < 0.05 were considered to have a non-normal distribution.
Results
Patient Demographics. Microsomes were prepared from sections of peripheral lung obtained from 30 human subjects (17 males and 13 females) from eastern Ontario, aged 62.6 ± 10.5 years (Table 1). Due to limitations in tissue availability, microsomes from all 30 subjects could not be used in all analyses. Based on reported smoking histories, 16 individuals were current smokers, 12 were former smokers, and information for 2 individuals was not available.
NNK Biotransformation in Lung Microsomes. NNAL was the major metabolite produced from NNK, with formation ranging from 0.280 to 0.900%/mg protein/min (representing ∼8-27 pmol NNAL/mg protein/min). Total bioactivation, represented by the sum of the four α-carbon hydroxylation endpoint metabolites, ranged from <3.33 × 10-4 to 7.50 × 10-2% total α-hydroxylation/mg protein/min (representing ∼0.01-2.3 pmol α-hydroxylation/mg protein/min). Total N-oxidation, represented by the sum of the two N-oxides, ranged from <3.33 × 10-4 to 1.33 × 10-3% total N-oxidation/mg protein/min (representing ∼0.01-0.04 pmol N-oxidation/mg protein/min). In all metabolism analyses, individuals with metabolism below the lower limit of detection were assigned a value of 0% total metabolism/mg protein/min. Normality tests of bioactivation and detoxification group data revealed that each group was not normally distributed (P < 0.05). Subsequently, based on total bioactivation, subjects could be classified as either high (17 subjects) or low (12 subjects) bioactivators, with significantly different mean total α-hydroxylation activities [(5.26 ± 1.23) × 10-2 and (6.49 ± 5.90) × 10-3% total α-hydroxylation/mg protein/min, respectively; n = 29, P < 0.05) (Fig. 2a). Similarly, based on total detoxification, subjects could be grouped into high (9 subjects) and low (20 subjects) categories, with significantly different mean total N-oxidation activities [(2.03 ± 1.65) × 10-3 and (2.50 ± 3.04) × 10-4% total N-oxidation/mg protein/min, respectively; n = 29, P < 0.05) (Fig. 2b). No correlation between NNK bioactivation and detoxification activities for individuals was found (P > 0.05). Also, no correlation was found between the degree of NNK bioactivation or detoxification and patients' age, gender, or smoking status (P > 0.05). Smoking status was compared among high and low bioactivation groups, and it was found that 71% of high bioactivators and 36% of low bioactivators were current smokers. Potential effects of drug treatments on NNK metabolism were assessed and no observable differences were found in the NNK metabolite profiles of individuals taking drugs that may alter the activities of relevant P450s, compared with those of other subjects. NNAL formation was not significantly different between the high and the low bioactivation or detoxification groups (P > 0.05).
Formation of the NNK-derived α-hydroxylation metabolites keto acid and keto alcohol was significantly higher than formation of the NNAL-derived hydroxy acid and diol, in both the high and low bioactivation groups (P < 0.05) (Table 2). When comparing metabolite formation between groups, production of hydroxy acid, keto acid, and keto alcohol was significantly higher (∼4 times, ∼6 times, and ∼26 times, respectively) in the high compared with the low bioactivation group, whereas formation of diol, which was highly variable between individuals, was not significantly different between groups (P > 0.05). Formation of total NNK-derived metabolites (i.e., keto acid plus keto alcohol) but not total NNAL-derived metabolites (i.e., hydroxy acid plus diol) was significantly different between the high and low bioactivaton groups (P < 0.05).
Removal of the NADPH-generating system decreased total α-hydroxylation in tissues from the majority of subjects, by 3.6 to 100% (median decrease = 79.6%, n = 21), but had increased or no apparent effect in microsomes from three subjects. Treatment with CO decreased total α-hydroxylation by 54.4 to 100% (median decrease = 66.7%) in microsomes from 9 subjects and increased bioactivation by 2.12 to 140% (median increase = 34%) in microsomes from 13. Similarly, removal of the NADPH-generating system decreased total N-oxidation by 25.0 to 100% (median decrease = 100%) in microsomes of 13 aubjects, increased N-oxidation by 13.3 to 66.7% (median increase = 59.9%) in 3 subjects, and had no effect in 8 subjects. CO decreased total N-oxidation by 25.1 to 100% (median decrease = 100%) in 14 subjects' microsomes, increased N-oxidation by 1.01 to 300% (median increase = 41.7%) in 6 subjects, and had no apparent effect on NNK and NNAL N-oxidation in microsomes from 4 subjects. CO treatment eliminated NNAL formation in microsomes from 2 subjects and increased it by 3.77 to 329% (median increase = 165%) in 22 subjects. In contrast, removal of the NADPH-generating system completely eliminated NNAL formation in microsomes from all subjects (n = 24).
CYP2A13 mRNA Expression and NNK Biotransformation. Regardless of whether high and low bioactivators and detoxifiers were considered separately or pooled, there was no significant correlation between CYP2A13 mRNA levels and the degree of total NNK bioactivation (Fig. 2a) or detoxification (Fig. 2b). Also, no significant associations were found between CYP2A13 mRNA levels and formation of individual α-hydroxylation metabolites. However, examination of the CYP2A13 mRNA and total α-hydroxylation scatter plot revealed the existence of a subgroup (n = 4) with both high levels of CYP2A13 mRNA and a high degree of total α-hydroxylation, for whom CYP2A13 levels positively correlated with the degree of NNK bioactivation (r = 0.967, P < 0.05). Analysis of the two N-oxides (detoxification products) independently revealed no significant correlations between CYP2A13 mRNA expression and levels of NNAL-N-oxide (r = 0.294, P > 0.05) or NNK-N-oxide (r =-0.022; P > 0.05). However, when subjects who had no detectable NNAL-N-oxide formation were excluded, a significant association was found between CYP2A13 mRNA expression and levels of NNAL-N-oxide (n = 5, r = 0.925, P < 0.05). Of these five individuals, three were high bioactivators. A statistically significant correlation occurred between the extent of NNAL formation and CYP2A13 mRNA expression (P < 0.05) (Fig. 2c). However, when individuals (n = 3) with relatively high levels of CYP2A13 mRNA and high levels of NNAL formation were excluded from the analysis, this correlation was no longer significant (r =-0.0658, P > 0.05).
CYP2A6 mRNA Expression and NNK Biotransformation. When high and low bioactivators and detoxifiers were considered separately or pooled, there was no significant correlation between CYP2A6 mRNA levels and the degree of total NNK bioactivation (Fig. 3a), detoxification (Fig. 3b), or NNAL formation (Fig. 3c). Similarly, no correlation was found between levels of CYP2A13 and CYP2A6 mRNA among individuals (n = 24, r = 0.114, P > 0.05).
Coumarin 7-Hydroxylase Activity and NNK Biotransformation. Regardless of whether high and low bioactivators and detoxifiers were considered separately or pooled, there was no significant correlation between coumarin 7-hydroxylation and the degree of total NNK bioactivation (Fig. 4a) or detoxification (Fig. 4b). 7-Hydroxycoumarin formation correlated with formation of hydroxy acid (r = 0.512, P < 0.05) but not with other individual α-hydroxylation metabolites. Correlation analysis between the degree of NNK bioactivation and coumarin 7-hydroxylation for the high bioactivation subgroup could not be carried out because of limited tissue availability for two subjects. There was no association between NNAL formation and 7-hydroxycoumarin formation (P > 0.05). However, when individuals with relatively high levels of 7-hydroxycoumarin formation (>20 fmol/mg protein/min, n = 2) were excluded from analyses, a significant correlation was found (r = 0.520, P < 0.05) (Fig. 4c). There was no difference in mean coumarin 7-hydroxylation activity between current and former smokers [7.25 ± 14.5 fmol/mg protein/min (95% CI, -0.488 to 15.00 fmol/mg protein/min) and 8.55 ± 7.02 fmol/mg protein/min (95% CI, 3.84-13.3 fmol/mg protein/min), respectively]. However, there was a positive relationship between age and coumarin 7-hydroxylation activity among individuals (n = 27, r = 0.387, P < 0.05), and males had significantly higher activity than did females [11.6 ± 15.0 fmol/mg protein/min, n = 15 (95% CI, 3.25-19.9 fmol/mg protein/min) and 3.07 ± 2.00 fmol/mg protein/min, n = 12 (95% CI, 1.80-4.33 fmol/mg protein/min), respectively, P < 0.05].
To determine the consistency between enzymatic activity as reflected by coumarin hydroxylation and CYP2A expression as assessed by mRNA levels, correlation analysis was performed. When results from all individuals were pooled, no significant correlations were found between 7-hydroxycoumarin formation and either CYP2A13 mRNA levels (n = 26, r =-0.0685, P > 0.05) or CYP2A6 mRNA levels (n = 22, r =-0.354, P > 0.05) (Fig. 5).
Immunoinhibition of NNK Biotransformation. The presence of the CYP2A6/13 inhibitory antibody decreased total α-hydroxylation by 0.35 to 98.3% (median decrease = 10.1%) in microsomes from 16 subjects, increased it by 1.52 to 57.5% (median increase = 5.42%) in 8 subjects, and had no apparent effect in microsomes from 6 subjects. It decreased total N-oxidation by 22.2 to 100% (median decrease = 85.0%) in microsomes from 14 subjects, increased N-oxidation by 25.3 and 200% in 2 subjects, and had no effect in 14 subjects. NNAL formation was decreased by 1.00 to 38.8% (median decrease = 8.10%) in microsomes from 19 subjects and increased by 4.33 to 50.5% (median increase = 14.6%) in 10 subjects. Regardless of whether high and low bioactivators and detoxifiers were considered separately or together, there was no significant correlation between the percentage change in NNK bioactivation or detoxification in the presence of a CYP2A6/13 inhibitory antibody and either the degree of NNK metabolism or formation of individual α-hydroxylation or N-oxidation metabolites (Table 3). There also was no significant correlation between NNAL formation and the percentage change in NNAL formation in the presence of a CYP2A6/13 inhibitory antibody (Table 3).
When assessing the consistency between CYP2A immunoinhibition-mediated changes in NNK metabolism and the measures of assessing CYP2A expression or activity, no significant correlations were found between levels of CYP2A13 or CYP2A6 mRNA and CYP2A6/13 immunoinhibition-mediated changes in total NNK bioactivation, total NNK detoxification, or NNAL formation. Similarly, no significant correlations were found between 7-hydroxycoumarin formation and CYP2A6/13 immunoinhibition-mediated changes in total NNK bioactivation, total NNK detoxification, or NNAL formation (Table 3).
Effects of CYP2A13 Arg257Cys Polymorphism on NNK Biotransformation. After PCR-restriction fragment length polymorphism analysis, homozygous wild-type (C/C) samples produced the anticipated products of 217 bp and 158 bp (Wang et al., 2003), whereas heterozygous variant (C/T) samples produced products of 375 bp, 217 bp, and 158 bp. The resultant genotype frequencies were 83 of 84 (98.8%) C/C, 1 of 84 (1.2%) C/T, and 0 of 84 (0%) T/T. NNK metabolite profile for the one heterozygous variant (0.047% total α-hydroxylation/mg protein/min; 0.00133% total N-oxidation/mg protein/min; 0.316% NNAL formation/mg protein/min) did not differ from those of the C/C subjects.
Discussion
Although we have previously used inhibitors to assess CYP2A6 contributions to NNK metabolism in adult human lung (Smith et al., 2003), this is, to our knowledge, the first study to assess CYP2A13 contributions. Examination of CYP2A13-mediated NNK metabolism was of interest because this human P450 is expressed predominantly in the lung (Su et al., 2000) and is the most catalytically active P450 isoform in the metabolic activation of NNK (Su et al., 2000). Our grouped results suggest no correlation between the degree of NNK bioactivation or detoxification and CYP2A gene expression and activity. However, a subgroup of individuals was identified for whom high levels of NNK bioactivation correlated with high CYP2A13 mRNA levels (Fig. 2a), suggesting that CYP2A13 may contribute substantially to NNK bioactivation in some, but not all, individuals. Similarly, for NNK detoxification, a subgroup of individuals was identified for whom CYP2A13 mRNA expression correlated with detectable levels of NNAL-N-oxide. However, the pooled results suggest that CYP2A13 is not the sole enzyme contributing to NNK metabolism in peripheral human lung microsomes. The apparent discrepancy between our results and those reported for fetal nasal microsomes (Wong et al., 2005) may reflect the fact that levels of CYP2A13 mRNA (Su et al., 2000) and CYP2A13 immunoreactive protein (Zhu et al., 2006) are considerably higher in the nasal mucosa than in the lung. Hence, in areas of the respiratory system where CYP2A13 levels are lower, other enzymes may contribute substantially to NNK metabolism.
Based on the interindividual variability in NNK metabolism, individuals were classified into high or low bioactivation and detoxification groups. The observation of distinct groups is consistent with our previous study of seven subjects (Smith et al., 2003). The distribution of subjects between bioactivation categories is not consistent with reported frequencies of established genetic polymorphisms of enzymes that have been characterized with respect to NNK metabolism, suggesting the possibility that environmental factors, rather than genetics, make the largest contribution to interindividual variability in NNK metabolism.
Although the levels of total NNK-derived metabolites were markedly different between the high and low bioactivation groups, the levels of total NNAL-derived metabolites were not significantly different (Table 2). This was unexpected because it was anticipated that the NNAL-derived metabolites would vary between the groups to the same extent as the NNK-derived metabolites, since both the high and low bioactivation groups formed equal amounts of NNAL. The differences in NNK and NNAL metabolism may be due to differences in the affinity of different P450 isoforms for NNK versus NNAL as a result of differences in polarity between the keto group of NNK and the hydroxyl group of NNAL. In fact, Jalas et al. (2003) found that CYP2A enzymes are more efficient catalysts of NNK metabolism than of NNAL metabolism.
P450 involvement in human pulmonary microsomal NNK metabolism was supported by NADPH dependence, but, consistent with previous studies (Smith et al., 1992, 1995), sensitivity to CO inhibition was variable. The fact that removal of the NAPDH-generating system and CO treatment did not completely inhibit NNK metabolism in some individuals suggests that non-P450 enzymes may also be involved in the metabolism.
Consistent with previous results (Maser et al., 2000; Smith et al., 2003), NNAL formation from NNK was NADPH-dependent for all subjects, but our results are the first to suggest a role for CYP2A13 in catalyzing the carbonyl reduction of NNK to form NNAL. This role is supported by our finding that NNAL formation positively correlated with CYP2A13 mRNA levels, although this correlation was dependent upon inclusion of individuals with high levels of CYP2A13 mRNA. It is also supported by the observation that NNAL formation positively correlated with the degree of 7-hydroxycoumarin formation in all individuals with the exception of those with relatively high coumarin 7-hydroxylation activity. It is possible that coumarin 7-hydroxylation is mainly CYP2A6-mediated in individuals with relatively high levels of 7-hydroxycoumarin formation. It has been suggested that the microsomal enzyme, 11β-hydroxysteroid dehydrogenase, rather than P450s, is the major catalyst of this reaction (Hecht, 1998). However, in microsomes from human lung, NNAL formation was insensitive to an 11β-hydroxysteroid dehydrogenase inhibitor (Breyer-Pfaff et al., 2004) and sensitive to P450 inhibitors. In contrast to results from the present study, an anti-CYP2A antibody did not inhibit NNAL formation in human fetal nasal microsomes (Wong et al., 2005). However, since fetal tissues had been pooled, interindividual variability in CYP2A13 contributions to NNAL formation could not be assessed. As well, age is known to affect P450 activity (Day et al., 2006).
Results presented here are also the first to support involvement of CYP2A13 in NNAL detoxification in human lung tissue for some individuals. Although no significant correlations were found between CYP2A13 mRNA and either total NNK detoxification or NNK-N-oxide levels, an association was found between levels of NNAL-N-oxide and CYP2A13 mRNA. This result is consistent with the finding that heterologously expressed CYP2A13 is capable of catalyzing NNAL N-oxidation but not NNK N-oxidation (Jalas et al., 2003).
No correlations were found between the degree of total NNK bioactivation or detoxification and 7-hydroxycoumarin formation, the latter of which is expected to reflect overall CYP2A activity. However, assessing associations between CYP2A13 or CYP2A6 activity and NNK metabolism is difficult since available substrates, with the exception of the drug phenacetin, do not discriminate between CYP2A13 and CYP2A6, and because CYP2A6 is 10 times more active than CYP2A13 at catalyzing coumarin 7-hydroxylation (Su et al., 2000). Although recently it was shown that CYP2A13 can efficiently metabolize phenacetin and CYP2A6 has virtually no catalytic activity toward this drug (Fukami et al., 2007), phenacetin is also a CYP1A2 substrate. Because CYP1A2 can also potentially bioactivate NNK (Smith et al., 1996) and is expressed in human lung (Wei et al., 2001), using phenacetin to assess CYP2A13 activity in human lung microsomes would also be problematic. When examining associations between individual α-hydroxylation metabolites and 7-hydroxycoumarin formation, a significant correlation was found between the degree of hydroxy acid formation and 7-hydroxycoumarin. This finding is consistent with both CYP2A13 and CYP2A6 being more efficient catalysts of α-methylene hydroxylation compared with α-methyl hydroxylation (Su et al., 2000; Jalas et al., 2003). Because neither CYP2A13 nor CYP2A6 mRNA correlated with hydroxy acid formation, the relative contribution of these isoforms is not known.
The absence of significant correlations between the degree of NNK bioactivation, detoxification, or keto reduction and CYP2A immunoinhibition may also reflect the fact that specific contributions of either CYP2A13 or CYP2A6 to NNK metabolism are difficult to assess since commercially available inhibitory antibodies do not discriminate between CYP2A13 and CYP2A6 (Su et al., 2000). Moreover, an attempt was made to determine the relative CYP2A protein levels among individuals by immunoblotting; however protein levels were too low for detection. Although our results do not suggest uniform involvement of CYP2A13 in NNK metabolism across all individuals, the possibility exists that more specific measures of enzyme contributions would reveal that a correlation exists for a greater proportion of individuals.
In some microsomal samples, the presence of the CYP2A6/13 inhibitory antibody increased levels of NNK metabolism. It has been suggested that nonuniform constitution of human microsomes between incubates and cigarette smoke particulate matter might affect how NNK biotransformation activities are expressed in different incubates (Smith et al., 2003).
No statistically significant correlations were found between CYP2A13 or CYP2A6 mRNA levels and 7-hydroxycoumarin formation when tissues from all patients were considered (Figs. 4 and 5). The lack of associations may be due to the fact that 7-hydroxycoumarin formation is a measure of overall CYP2A enzyme activity, not specifically CYP2A13 or CYP2A6 (Su et al., 2000). In addition, nicotine is a mechanism-based inactivator of both CYP2A13 and CYP2A6 (von Weymarn et al., 2006), so that current smokers could potentially have had diminished CYP2A enzyme activity. However, a higher proportion of current smokers were actually high bioactivators compared with low bioactivators, and levels of 7-hydroxycoumarin formation did not differ between current and former smokers. There also were no significant correlations between the CYP2A immunoinhibition of NNK metabolism and CYP2A13 mRNA levels or CYP2A6 mRNA levels or 7-hydroxycoumarin formation (Table 3). Inconsistency between these parameters could be attributed to a lack of specificity of the antibody for either CYP2A isoform (e.g., according to information from the supplier, the antibody reacts with CYP2A6, CYP2A13, and mildly with CYP2E1).
Genotype frequencies for the CYP2A13 Arg257Cys polymorphism were consistent with those previously published for a Caucasian population [96.2% (C/C), 3.8% (C/T), and 0% (T/T)] (Zhang et al., 2002). Although NNK metabolism for the one heterozygous variant did not differ from that of other subjects, it is not possible to accurately determine the actual contribution of this polymorphism to NNK metabolism, since only one variant was found. The current study did not assess the influence of CYP2A6 polymorphisms on NNK metabolism since no clear associations have been found between CYP2A6 polymorphisms and lung cancer risk (Raunio et al., 2001; Wang et al., 2003).
In summary, examination of the interindividual differences in the pulmonary microsomal metabolism of NNK revealed the existence of high and low bioactivation and detoxification groups. Differences in bioactivation and detoxification between groups correlated with differences in the oxidation of NNK but not of NNAL. If the basis for these differences could be determined, identification of high and low categories for bioactivation and detoxification might be useful for predicting susceptibility of individuals to NNK-induced carcinogenesis. Results from this study are the first, to our knowledge, to suggest a role for CYP2A13 in NNAL formation. Although results do not support a role for either CYP2A13 or CYP2A6 as a major contributor to NNK bioactivation or detoxification in all individuals, results from subgroups of individuals suggest that CYP2A13 contributes to NNK metabolism, but this contribution varies between individuals.
Acknowledgments
We thank Carole Fargo for administrative assistance.
Footnotes
-
This work was supported by Canadian Institutes of Health Research Grant MT10382.
-
doi:10.1124/dmd.107.017343.
-
ABBREVIATIONS: NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNAL, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol; NNK-N-oxide, 4-(methylnitrosamino)-1-(3-pyridyl N-oxide)-1-butanone; NNAL-N-oxide, 4-(methylnitrosamino)-1-(3-pyridyl N-oxide)-1-butanol; keto alcohol, 4-hydroxy-1-(3-pyridyl)-1-butanone; diol, 1-(3-pyridyl)-1,4-butane diol; keto acid, 1-(3-pyridyl)-1-butanone-4-carboxylic acid; hydroxy acid, 1-(3-pyridyl)-1-butanol-4-carboxylic acid; P450, cytochrome P450; CO, carbon monoxide; RT-PCR, reverse transcription-polymerase chain reaction; CI, confidence interval.
- Received June 20, 2007.
- Accepted August 20, 2007.
- The American Society for Pharmacology and Experimental Therapeutics