Role of Human Aldo-Keto Reductases and Nuclear Factor Erythroid 2-Related Factor 2 in the Metabolic Activation of 1-Nitropyrene via Nitroreduction in Human Lung Cells

Abstract

1-Nitropyrene (1-NP) is a constituent of diesel exhaust and classified as a group 2A probable human carcinogen. The metabolic activation of 1-NP by nitroreduction generates electrophiles that can covalently bind DNA to form mutations to contribute to cancer causation. NADPH-dependent P450 oxidoreductase (POR), xanthine oxidase (XO), aldehyde oxidase (AOX), and NAD(P)H/quinone oxidoreductase 1 (NQO1) may catalyze 1-NP nitroreduction. We recently found that human recombinant aldo-keto reductases (AKRs) 1C1-1C3 catalyze 1-NP nitroreduction. NQO1 and AKR1C1-1C3 are genes induced by nuclear factor erythroid 2-related factor 2 (NRF2). Despite this knowledge, the relative importance of these enzymes and NRF2 to 1-NP nitroreduction is unknown. We used a combination of pharmacological and genetic approaches to assess the relative importance of these enzymes and NRF2 in the aerobic nitroreduction of 1-NP in human bronchial epithelial cells, A549 and HBEC3-KT. 1-NP nitroreduction was assessed by the measurement of 1-aminopyrene (1-AP), the six-electron reduced metabolite of 1-NP, based on its intrinsic fluorescence properties (λ_ex and λ_em). We found that co-treatment of 1-NP with salicylic acid, an AKR1C1 inhibitor, or ursodeoxycholate, an AKR1C2 inhibitor, for 48 h decreased 1-AP production relative to 1-NP treatment alone (control) in both cell lines. R-Sulforaphane or 1-(2-cyano-3,12,28-trioxooleana-1,9(11)-dien-28-yl)-1H-imidazole (CDDO-Im), two NRF2 activators, each increased 1-AP production relative to control only in HBEC3-KT cells, which have inducible NRF2. Inhibitors of POR, NQO1, and XO failed to modify 1-AP production relative to control in both cell lines. Importantly, A549 wild-type cells with constitutively active NRF2 produced more 1-AP than A549 cells with heterozygous expression of NFE2L2/NRF2, which were able to produce more 1-AP than A549 cells with homozygous knockout of NFE2L2/NRF2. Together, these data show dependence of 1-NP metabolic activation on AKR1Cs and NRF2 in human lung cells. This is the second example whereby NFE2L2/NRF2 is implicated in the carcinogenicity of diesel exhaust constituents.

Figure 1:

Time and concentration-dependence of 1-AP generation in (A) A549 cells and (B) HBEC3-KT cells. N = 6 independent experiments performed in triplicate for each for (A) and n = 3 independent experiments performed in triplicate for each for (B). All error bars represent mean ± SEM.

Figure 2:

Effect of AKR1C inhibitors on the ability of (A) A549 cells and (B) HBEC3-KT cells to produce 1-AP following 1-NP exposure. N = 5 independent experiments performed in triplicate for each for (A) and n = 4 independent experiments performed in triplicate for each for (B). Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with one-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h for (A-B). Abbreviations: 1-NP, 1-nitropyrene; FA, flufenamic acid; SA, salicylic acid; UD, ursodeoxycholate.

Figure 3:

Effect of AKR1C3-specific inhibitors on the ability of (A) A549 cells and (B) HBEC3-KT cells to produce 1-AP following 1-NP exposure. N = 4 independent experiments performed in triplicate for each for (A) and (B). Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with one-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h for (A-B). Abbreviations: 1-NP, 1-nitropyrene; ASP9521, (4-(2-hydroxy-2-methylpropyl)piperidin-1-yl)(5-methoxy-1H-indol-2-yl)methanone; BMT4-159, 3-4-trifluoromethyl-phenylamino-benzoic acid; INDO, indomethacin.

Figure 4:

Effect of NRF2 modulators on the ability of (A) A549 cells and (B-C) HBEC3-KT cells to produce 1-AP following 1-NP exposure. N = 6 independent experiments performed in triplicate for each for (A) and n = 4 independent experiments performed in triplicate for each for (B) and (C). Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with one-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h for (A-C). Abbreviations: 1-NP, 1-nitropyrene; SFN, R-sulforaphane; CDDO-Im, 1-(2-cyano-3,12,28-trioxooleana-1,9(11)-dien-28-yl)-1H-imidazole; ML385, N-[4-[2,3-dihydro-1-(2-methylbenzoyl)-1H-indol-5-yl]-5-methyl-2-thiazolyl]-1,3-benzodioxole-5-acetamide; ATRA, all-trans retinoic acid.

Figure 5:

Effect of inhibitors of additional nitroreductases on the ability of (A) A549 cells and (B-C) HBEC3-KT cells to produce 1-AP following 1-NP exposure. N = 6 independent experiments performed in triplicate for each for (A), n = 4 independent experiments performed in triplicate for each for (B), and n = 5 independent experiments performed in triplicate for each for (C). Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with one-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h for (A-C). MD was not included for HBEC3-KT cells because a false positive signal developed with MD in the HBEC3-KT medium. Abbreviations: 1-NP, 1-nitropyrene; CEES, 2-chloroethyl ethyl sulfide; MD, menadione; DIC, dicoumarol; ALP, allopurinol; ROT, rotenone; CDDO-Im, 1-(2-cyano-3,12,28-trioxooleana-1,9(11)-dien-28-yl)-1H-imidazole.

Figure 6:

Effect of POR inhibitors on the ability of (A) A549 cells and (B) HBEC3-KT cells to produce 1-AP following 1-NP exposure. N = 4 independent experiments performed in triplicate for each for (A-B). Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with one-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h for (A-B). Abbreviations: 1-NP, 1-nitropyrene; DPI, diphenyleneiodonium sulfate; OCA, n-octylamine.

Figure 7:

Effect of NFE2L2/NRF2 gene editing on the ability of A549 cells to produce 1-AP following 1-NP exposure. N = 6 independent experiments performed in triplicate for each. Each individual data point represents the average of technical replicates within one experiment. Statistical analysis was performed with two-way ANOVA followed by Tukey’s post-hoc multiple comparison of means. Groups with non-shared superscript letters are significantly different (p < 0.05); shared superscript letters show no statistical significance between compared groups. Error bars represent mean ± SEM. Treatment durations were 48 h. Abbreviations: 1-NP, 1-nitropyrene; MD, menadione; ROT, rotenone.

Figure 8:

Scheme of the dependence of nitroreduction of 1-NP on NRF2 and AKR1C1-1C3 in a system of constitutive NRF2 versus inducible NRF2. We found that AKR1C1-1C3 inhibitors reduce 1-AP production in each cell type, and effects of flufenamic acid, salicylic acid, and ursodeoxycholate were particularly potent in HBEC3-KT cells, each causing over 70% inhibition to support that AKR1Cs are the major 1-NP nitroreductases in HBEC3-KT cells. Whereas NRF2 activators were effective at increasing 1-AP production in HBEC3-KT cells that have inducible NRF2, NRF2 inhibitors were effective at decreasing 1-AP production in A549 cells that have constitutively active NRF2. By contrast, we did not find POR inhibitors, a NQO1 inhibitor, or a XO inhibitor to reduce 1-AP production in either cell type. Our results with rotenone imply a role of ROS or mitochondrial nitroreductases in the nitroreduction of 1-NP, particularly for A549 cells where AKR1C inhibition could not explain all contribution to 1-NP nitroreduction.