Pathways for the mutagenesis of 1-nitropyrene and dinitropyrenes in the human hepatoma cell line HepG2.

The mutagenicity, metabolism, DNA adduction and induction of unscheduled DNA synthesis (UDS) of 1-nitropyrene and 1,8-dinitropyrene were investigated in the human hepatoma cell line HepG2. Previous results had demonstrated that 1-nitropyrene was both mutagenic at the hgprt locus and induced UDS in these cells. In the present study, we find that the dinitropyrenes, although highly mutagenic in Salmonella typhimurium, are not mutagenic and do not induce UDS in the HepG2. Although the rate of 1,8-dinitropyrene nitroreduction was less than that of 1-nitropyrene nitroreduction, this did not explain the lack of mutagenicity and UDS induction by the dinitropyrenes. Therefore, it is proposed that the arylhydroxylamine O-esterificase is not expressed in these cells. Since cytochrome P450-mediated C-oxidation is the predominant metabolic pathway in vivo, we sought to determine if an increase in the ratio of cytochrome P450-mediated C-oxidation over nitroreduction would result in increased or decreased DNA adducts in the HepG2. The administration of 2.5 microM 3-methylcholanthrene to the HepG2 increased the ratio of C-oxidation/nitroreduction from 2.8 +/- 1.9 to 50.4 +/- 46.1. This was accompanied by a decrease in the C8-guanyl adduct of 1-nitropyrene (via nitroreduction) from 18.7 +/- 7.0 to 4.8 +/- 1.7 fmoles/micrograms DNA, without any further increase in other 1-nitropyrene DNA adducts. These results suggest that the cytochrome P450-mediated metabolism of 1-nitropyrene to epoxides, phenols, and dihydrodiols is not an activation pathway in the HepG2 cells, and may explain the weak carcinogenicity of 1-nitropyrene in vivo, where cytochrome P450-mediated C-oxidation predominates. ImagesFigure 4.


Introduction
The method for assessing the human risk of a chemical is to use the in vivo and in vitro data on the mutagenicity and tumorigenicity of the chemical, and to determine the likelihood for human genotoxicity or adverse health effects. Risk assessment evaluations take into consideration in vitro data from both prokaryotic and eukaryotic cell studies, as well as in vivo toxicity studies from several species. It would be most advantageous to study the metabolism and genotoxicity of chemicals in vitro in human cells that metabolically resemble the cell of interest in humans. While studying the metabolism and genotoxicity of chemicals in vitro ignores the interaction of chemicals This paper was presented at the Fifth International Conference on Carcinogenic and Mutagenic N-Substituted Aryl Compounds held 18-21 October 1992 in Wurzburg, Germany.
This work was supported in part by grant ES03648 from the National Institute of Health. The authors thank L. King and J. Lewtas for the gift of the 1-nitropyrene-4,5-epoxide modified DNA. Address correspondence to P.C. Howard, Division of Biochemical Toxicology, HFT-1 10, National Center for Toxilogical Research, Jefferson, AR 72079. Telephone (501) 543-7000. and metabolites between cells or organs (e.g. enterohepatic circulation), it may present a simulated environment in which to study the generation of reactive metabolites within a human cell.
The human hepatoma cell holds good promise as an in vitro candidate for studies on xenobiotic metabolism. The isolation of primary human hepatocytes is restricted by a lack of availability of tissue, and poses the additional problem of interindividual variation. Several hepatoma cell lines have been isolated, with the most promising being the HepG2 cell line. This cell line was derived from a primary hepatoblastoma isolated from an 11-year-old Argentinean male (1). These cells retain many characteristic enzyme pathways of hepatocytes (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15), and have been cultured successfully for more than 100 passages.
The HepG2 has been shown by several groups to possess the enzymes necessary for the activation of many chemicals. For instance, Diamond and coworkers (16) demonstrated that when X-ray-irradiated HepG2 was cocultured for 48 (17) demonstrated that the HepG2 was capable of activating cyclophosphamide to induce sister chromatid exchanges (SCEs), and demonstrated that the content of cytochrome P450 is very low in the HepG2 cells (a phenomenon that has been reported by several laboratories). Other compounds that have been reported to be activated by the HepG2 include benzo-[a] pyrene (18,19), 7,1 2-dimethylbenz- [a]anthracene (20,21), aflatoxin B1, (22,23), several N-nitroso compounds (24), benzidine (23,25) (21).
In S. typhimurium, 1-nitropyrene is mutagenic through nitroreduction to the corresponding nitroso (structure 12) then the hydroxylamino (structure 13) derivative, which has been shown to form a C8guanyl adduct (32). This adduct also is responsible for the mutagenicity of 1nitropyrene in CHO cells (34), and cultured human diploid fibroblasts (39)(40).
However, the pathways responsible for this activation were not known. Moreover, both cytochrome P450-mediated C-oxidized metabolites and nitroreduced metabolites (1-aminopyrene) were detected (41). Therefore, we were unable to deduce the pathway or the DNA adduct responsible for the mutagenesis of 1-nitropyrene in HepG2.
Therefore we sought to determine whether 1,8-dinitropyrene, like 1-nitropyrene, was mutagenic in HepG2 cells, and to describe the pathway involved in the metabolic activation of 1-nitropyrene and 1,8-dinitropyrene in HepG2.

Synthesis in HepG2 Cells
The mutagenesis of the chemicals by selection of mutations at the hgprt locus and the quantitation of induction of unscheduled DNA synthesis (UDS) were essentially as described in Eddy et al. (41).

Mletabolism of [3H] 1-nitropyrene and [3H] 1,8-dinitropyrene
For the metabolism studies, HepG2 cells were plated at 1.5 x 10 cells/100 mm culture plates and incubated overnight with MEM + 10% hi-FBS. Immediately prior to adding the radiolabeled compounds, the medium was changed to MEM + 2% hi-FBS. Either 4 pM [4,5,9, Chemsyn, Inc.) were added to the media and incubated for up to 24 hr. The metabolism of the compounds was terminated by decanting and cooling the media on ice, removal of the cells with trypsin, and extraction of the parent compounds and metabolites with chloroform:methanol (2:1) followed by chloroform. Additionally, cells and media were incubated at 370C overnight in the presence of 0-glucuronidase (Sigma) or arylsulfatase (Sigma), followed by extraction with chloroform.
The analysis of the metabolites of [3H] 1-nitropyrene and [3H] 1,8-dinitropyrene was essentially as described for 1nitropyrene (30) using HPLC (Varian Instr., Walnut Creek, CA) and reversephase columns (Waters Assoc., Milford, MA). Nonradiolabeled standards were routinely coinjected to verify retention times of the metabolites. The conversion of the parent compound was quantified using a flow-through scintillation counter (Flo-One, Radiomatic Instruments). Represen  aN-methyl-N'-nitro-N-nitrosoguanidine. The com-0 20 40 pounds were added to HepG2 and mutagenicity at the minutes hgprt gene was determined as described in the text using 6-thioguanine. 32P-Posdabeling ofDNA from HepG2 DNA adducts were assayed by 32P-postlabeling on DNA by the n-butanol enrichment and contact transfer procedures as indicated in Smith et al. (48). The adducts were quantified by comparison to DNA standards that were modified to a known extent with 1 -nitropyrene.
In Table I  Another method for detecting genotoxic damage to the HepG2 cells is by monitoring the repair of DNA adducts (unscheduled DNA synthesis). The results of incubation of 1-nitropyrene and the dinitropyrenes and determination of the UDS is shown in Table 2. The presence of 4 pM 1-nitropyrene resulted in a 52% increase in UDS. As with the mutation results, there was not a dose-dependent increase in UDS with 1,3-dinitropyrene, although UDS was increased approximately 16% above the background values. Neither 1,6or 1,8-dinitropyrene induced UDS above the background values. These results suggest that the highly mutagenic and carcinogenic 1,6-and 1,8-dinitropy- In order to determine if the lack of genotoxicity of 1,8-dinitropyrene resulted from a lack of metabolic activation, the nitroreduction of 1-nitropyrene and 1,8dinitropyrene were contrasted. Both of these compounds have been shown to be metabolically activated by nitroreduction to arylhydroxylamines that bind to DNA. In prokaryotic cells, the metabolic activation of 1,6-and 1,8-dinitropyrene additionally requires the esterification of the arylhydroxylamine to an acyloxy ester via acyltransferase enzymes (52,53), while this step is not required for 1-nitropyrene or 1,3-dinitropyrene. Figure 3  in HepG2 cannot only be attributed to the loss of nitroreduction, but additionally results from a lack of arylhydroxylamine 0esterificase activity.
One possible explanation for the weak tumorigenicity of 1 -nitropyrene and high tumorigenicity of 1,6-and 1 ,8-dinitropyrene could be a lack of activation of 1nitropyrene in vivo and a predominance of activation for 1,6-and 1 ,8-dinitropyrene in vivo. The metabolism of 1 -nitropyrene in vitro with microsomes and in vivo is dominated by cytochrome P450-mediated Coxidation to phenols or dihydrodiols. However, no cytochrome P450-mediated C-oxidation has been reported for 1,6or 1 ,8-dinitropyrene, leaving nitroreduction as the sole pathway for metabolism and removal of the compound. This would obligate cells to metabolize the dinitropyrenes through a pathway that results in mutagenicity in cells in vitro.
A method for testing the hypothesis that nitroreduction of I1-nitropyrene is responsible for the mutagenesis of 1nitropyrene in the HepG2, and that cytochrome P450-mediated C-oxidation is not an activation pathway, would be to vary the ratio of cytochrome P450-mediated C-oxidation to nitroreduction in HepG2, and then determine the effect on DNA adduction. In Table 3  To determine the effect of the altered ratio of C-oxidation/nitroreduction on the DNA adduct formation, control and 3methylcholanthrene-treated HepG2 cells were exposed to 10 pM 1 -nitropyrene. The DNA was isolated, hydrolyzed, and analyzed for the presence of DNA adducts by the 32 P-postlabeling method (Figure 4). When S. typhimurium TA98 is exposed in suspension to 1-nitropyrene, the C8guanyl adduct of 1-nitropyrene (dG-C8-AP), C8-guanyl adduct of either 1,6or 1,8-dinitropyrene (dG-C8-ANP), and a polar adduct presumed to be the ringopened product of dG-C8-AP ("a") are detected ( Figure 4A). While no adducts appear in HepG2 in the absence of added compounds ( Figure 4B), the inclusion of 10 pM 1 -nitropyrene induced two adducts (Figure 4 C). One of the adducts comigrated with dG-C8-AP and the other with "ca" of panel A. Since we used 1-nitropyrenes of identical purity (-99.5%), the lack of DNA adducts in HepG2 from the dinitropyrenes that contaminate 1 -nitropyrene, and the presence of these adducts in the S. typhimurium TA98, argue in favor of our conclusion that HepG2 lack the arylhydroxylamine O-esterificase necessary for the activation of the dinitropyrenes. Pretreatment of HepG2 cells with 3methylcholanthrene and exposure to 1nitropyrene resulted in the formation of less dG-C8-AP, and the formation of 3- aThe total C-oxidation was calculated as the sum of dihydrodiol and phenol metabolites produced. bThe ratio of the formation of 1-nitropyren-6-ol plus 1-nitropyren-8-ol divided by 1-nitropyren-3-ol formation. The results are the sum of five experiments and are presented as the mean ± SD. methylcholanthrene DNA adducts ( Figure  4D). There was a lack of formation of other DNA adducts that could be attributable to 1-nitropyrene K-region epoxides. The DNA from the HepG2 treated with 1nitropyrene contained 18.7 ± 7.0 fmoles/pg DNA of dG-C8-AP, while the cells pretreated with 3-methylcholanthrene then treated with 1-nitropyrene had 4.8 ± 1.7 fmoles dG-C8-AP per pg DNA.
These results demonstrate that when the ratio of cytochrome P450-mediated Coxidation of 1-nitropyrene is increased over the nitroreduction pathway, there is a decrease in the formation of dG-C8-AP DNA adducts, which arise from the nitroreduction pathway, and that this decrease in adducts is not associated with the formation of DNA adducts arising from cytochrome P450-mediated C-oxidation, e.g., the 4,5-epoxide DNA adduct.
Nitroreduction is not a favored pathway in vivo in tissues where cytochrome P450 concentrations are significantly higher than in HepG2 cells. Additionally, the epoxides of 1-nitropyrene are quickly hydrolyzed by epoxide hydrolase in human liver tissue (54). These results, along with the weak tumorigenic response of rodents to 1nitropyrene, indicate that 1-nitropyrene may not pose a significant tumorigenic risk to the human population.