Potent mutagenicity of 3-methylindole requires pulmonary cytochrome P450-mediated bioactivation: a comparison to the prototype cigarette smoke mutagens B(a)P and NNK

Abstract

3-Methylindole (3MI) is a preferential pneumotoxicant found in cigarette smoke. A number of lung-expressed human cytochrome P450 enzymes, including 1A1, 2F1, and 2A13, catalyze the metabolism of 3MI to reactive intermediates that fragment DNA, measured with the Comet assay to assess DNA damage, in a cytochrome P450-dependent manner in primary normal human lung cells in culture, but the mutagenesis of 3MI has been controversial. In the present study, the mutagenic potential of 3MI was compared to the prototypical cigarette smoke carcinogens benzo(a)pyrene (B(a)P) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). 3MI, B(a)P, and NNK were incubated with the Salmonella typhimurium strain TA98, which is known to detect the most common subtype of cigarette smoke-induced mutagenicity, frameshift mutations in DNA, and with Salmonella typhimurium strain TA100, which detects base pair substitution mutants, with five sources of P450-mediated bioactivation: rat liver S9, human lung microsomes, recombinant CYP2A13, purified CYP2F3, and recombinant CYP1A1. Only B(a)P was mutagenic in TA100, and it was bioactivated by human lung microsomes and rat liver S9 sources of P450s. However, with the TA98 strain, CYP1A1, CYP2A13, CYP2F3, and human lung microsomes bioactivated 3MI to highly mutagenic intermediates, whereas neither human nor rat liver S9 subcellular fractions formed mutagenic intermediates from 3MI. Quantitative Western blot analysis verified that all three respiratory enzymes were present in human lung microsomes in widely varying amounts. These results indicate that metabolism of 3MI by human lung-expressed cytochrome P450 enzymes but not hepatic P450s elicits equivalent or higher mutagenicity than the prototype cigarette smoke mutagens B(a)P and NNK and indicates that 3MI is a likely human pulmonary carcinogen.

Scheme 1. Illustration of 3MI Metabolism, Putative Bioactivation, and Nucleoside Adducts

The dehydrogenated intermediate, 3-methyleneindolenine, is efficiently and selectively produced by the lung P450 enzymes CYP2F1, CYP2A13, and CYP1A1. Three putative electrophilic intermediates of 3MI bioactivation are shown in blue. Three nucleoside adducts (shown in green) of 3-methyleneindolenine with DNA have been previously identified (20), are produced from hydrolysis of 3MI-alkylated DNA, and are putative markers of the mutagenic activity of 3MI. Nucleoside adducts of the 2,3-epoxide or 3-hydroxy-3-methylindolenine have not been identified. R = deoxyribose

Figure 1. B(a)P Causes Base Pair Substitution Mutations (Salmonella typhimurium strain TA100), but NNK and 3MI do not

Incubation of 10 and 100 μM B(a)P in the presence of NADPH (green bars) with either 0.5 mg human lung microsomes (pooled sample from 10 donors with an EROD activity = 0.9 pmol/mg/min) or 2 mg rat liver S9 caused significant increases in the number of revertants per plate, whereas NNK (blue bars) and 3MI (red bars) failed to cause mutagenesis at any concentration. In the absence of cytochrome P450 activation (no NADPH - black bars), the number of revertants per plate was not significantly different from the vehicle-treated control (DMSO - gray bar), indicating that B(a)P-mediated mutagenicity was cytochrome P450-dependent. * indicates p < 0.05 compared to DMSO-treated control. Revertants/plate represents the average number of colonies counted from 3 plates per treatment group from duplicate experiments.

Figure 2. 3MI is the Most Mutagenic Chemical with Human Lung Microsomes

Incubation of μM concentrations of benzo(a)pyrene (B(a)P - green bars), 4-(methylnitrosamino)- 1-(3-pyridyl)-1-butanone (NNK - blue bars), and 3-methylindole (3MI - red bars) with Salmonella typhimurium strain TA98 in the presence of 0.5 mg of human lung microsomes (pooled sample from 10 donors with an EROD activity = 0.9 pmol/mg/min) with NADPH indicated that 3MI was the most potent cigarette smoke mutagen. The analysis was repeated with a second batch of human lung microsomes from 10 different donors, and the results were essentially identical. In the absence of cytochrome P450 activation (no NADPH - black bars), the number of revertants per plate was not significantly different from the vehicle-treated control (DMSO - gray bar), indicating that mutagenicity was cytochrome P450-dependent. Neither incubations with 2 mg human liver S9 fractions, nor incubations with 2 mg rat liver S9 fractions, mutated the bacteria. * indicates p < 0.05 compared to DMSO-treated control. Revertants/plate represents the average number of colonies counted from 3 plates per treatment group from duplicate experiments.

Figure 3. Lung P450s Selectively Bioactivate B(a)P, NNK, and 3MI to Mutate Bacteria

Incubation of 10 μM B(a)P (green bars), NNK (blue bars), or 3MI (red bars) with Salmonella typhimurium strain TA98 in the presence of NADPH and 50 pmol of CYP1A1 supersomes, recombinant CYP2A13, and recombinant CYP2F3 indicated that all three lung cytochrome P450 enzymes bioactivate 3MI, with CYP2A13 causing the highest levels of mutagenicity. In the absence of NADPH (black bars), mutagenesis did not occur. Neither incubations with 2 mg human liver S9 fractions, nor incubations with 2 mg rat liver S9 fractions, mutated the bacteria, indicating specific metabolism by lung cytochrome P450 enzymes. * indicates p < 0.05 compared to DMSO-treated control. Revertants/plate represents the average number of colonies counted from 3 plates per treatment group from duplicate experiments.

Figure 4. Inhibition of CYP2A13 Activity Inhibits 3MI- and NNK-Mediated Mutagenicity with Human Lung Microsomes, but Does Not Affect B(a)P-Mediated Mutagenicity

Incubation of the inhibitor of CYP2A13 activity, 1 μM 8-methoxypsoralen (8-MOP), in combination with 100 μM 3MI (red bar with black stripes) and 100 μM NNK (blue bar with black stripes) with Salmonella typhimurium strain TA98 in the presence of 0.5 mg of human lung microsomes (pooled sample from 10 donors with an EROD activity = 0.9 pmol/mg/min) with NADPH significantly decreased mutagenicity as compared to either 3MI (red bar) or NNK (blue bar) incubation alone. However, mutagenicity remained significantly increased above DMSO treated control (gray bar), indicating that CYP2A13 was responsible for some, but not all, of the mutagenicity induced by these compounds. Inhibition of CYP2A13 had no effect on B(a)P-mediated mutagenicity. In the absence of cytochrome P450 activity (no NADPH - black bars) no mutagenesis occurred. 8-MOP incubation alone (white bar with black stripes) had no effect on mutagenesis. * indicates p < 0.05 compared to DMSO-treated control, ‡ indicates p < 0.05 as compared to 3MI treatment, # indicates p < 0.05 compared to NNK treatment. Revertants/plate represents the average number of colonies counted from 3 plates per treatment group from duplicate experiments.

Figure 5. CYP450 Activity is Required for 3MI, NNK, and B(a)P-Mediated Mutagenicity with Human Lung Microsomes

Incubation with the suicide substrate inhibitor of virtually all cytochrome P450 activity, 0.5 mM 1-aminobenzotriazole (ABT), in combination with 3MI (red bar with black stripes), NNK (blue bar with black stripes), and B(a)P (green bar with black stripes) with Salmonella typhimurium strain TA98 in the presence of 0.5 mg of human lung microsomes (pooled sample from 10 donors with an EROD activity = 0.9 pmol/mg/min) in the presence of NADPH decreased mutagenesis levels until they were no longer significantly elevated above either DMSO control (gray bars) or treatments in the absence of NADPH (black bars), indicating that mutagenicity of all three compounds was dependent upon cytochrome P450 activity. Treatment with ABT alone (white bar with black stripes) had no effect on mutagenesis. * indicates p < 0.05 compared to DMSO-treated control. Revertants/plate represents the average number of colonies counted from 3 plates per treatment group from duplicate experiments.

Figure 6. Quantitative Analysis of Cytochrome P450 Expression in Human Lung Microsomes

Densitometric analysis of quantitative western blots indicated that CYP1A1, CYP2F1, and CYP2A13 were all expressed in the pooled human lung microsomal samples. The results shown for CYP2A13 are most likely for the total expression of CYP2A6 and CYP2A13, because the antibody used does not differentiate between these two highly homologous enzymes, and they are both expressed in human lung microsomal samples (28). CYP1A1 showed the highest levels of expression in the range of 0.5 - 1 pmol CYP1A1/30 μg microsomes, whereas CYP2A13 expression was approximately one order of magnitude lower at 0.05 - 0.1 pmol CYP2A13/30 μg microsomes.

Figure 6. Quantitative Analysis of Cytochrome P450 Expression in Human Lung Microsomes

Densitometric analysis of quantitative western blots indicated that CYP1A1, CYP2F1, and CYP2A13 were all expressed in the pooled human lung microsomal samples. The results shown for CYP2A13 are most likely for the total expression of CYP2A6 and CYP2A13, because the antibody used does not differentiate between these two highly homologous enzymes, and they are both expressed in human lung microsomal samples (28). CYP1A1 showed the highest levels of expression in the range of 0.5 - 1 pmol CYP1A1/30 μg microsomes, whereas CYP2A13 expression was approximately one order of magnitude lower at 0.05 - 0.1 pmol CYP2A13/30 μg microsomes.

Figure 6. Quantitative Analysis of Cytochrome P450 Expression in Human Lung Microsomes

Densitometric analysis of quantitative western blots indicated that CYP1A1, CYP2F1, and CYP2A13 were all expressed in the pooled human lung microsomal samples. The results shown for CYP2A13 are most likely for the total expression of CYP2A6 and CYP2A13, because the antibody used does not differentiate between these two highly homologous enzymes, and they are both expressed in human lung microsomal samples (28). CYP1A1 showed the highest levels of expression in the range of 0.5 - 1 pmol CYP1A1/30 μg microsomes, whereas CYP2A13 expression was approximately one order of magnitude lower at 0.05 - 0.1 pmol CYP2A13/30 μg microsomes.