Human hepatocyte-mediated mutagenesis and DNA repair activity

Methods of in vitro toxicology

In vitro methods are common and widely used for screening and ranking chemicals, and have also been taken into account sporadically for risk assessment purposes in the case of food additives. However, the range of food-associated compounds amenable to in vitro toxicology is considered much broader, comprising not only natural ingredients, including those from food preparation, but also compounds formed endogenously after exposure, permissible/authorised chemicals including additives, residues, supplements, chemicals from processing and packaging and contaminants. A major promise of in vitro systems is to obtain mechanism-derived information that is considered pivotal for adequate risk assessment. This paper critically reviews the entire process of risk assessment by in vitro toxicology, encompassing ongoing and future developments, with major emphasis on cytotoxicity, cellular responses, toxicokinetics, modelling, metabolism, cancer-related endpoints, developmental toxicity, prediction of allergenicity, and finally, development and application of biomarkers. It describes in depth the use of in vitro methods in strategies for characterising and predicting hazards to the human. Major weaknesses and strengths of these assay systems are addressed, together with some key issues concerning major research priorities to improve hazard identification and characterisation of food-associated chemicals.

A simple method for the isolation and culture of epithelial and stromal cells from benign and neoplastic prostates

OBJECTIVES

Current primary prostate cell culture techniques use an overnight digestion or extensive media preparation. In this report, we describe a method for the culture of benign and neoplastic cells from human prostatectomy specimens that is rapid and contains no undefined factors in the medium.

METHODS

Characterization of the human cultured prostate cells was performed using immunohistochemical methods and monoclonal antibodies AE1/AE3 and cytokeratin 8, as well as monoclonal antibodies against prostate-specific antigen (PSA). Polymerase chain reaction was used to measure the exclusive epithelial and stromal cell products, c-met and hepatocyte growth factor (HGF), respectively. Electron microscopy was performed to assess the cell junctions and morphologic features of epithelial cells. Optimum cell growth in different media was tested using a cell replication assay.

RESULTS

Microscopic evidence revealed that the cells demonstrate typical epithelial morphology, with polyhedral cells forming tight junctions in a continuous monolayer. Desmosomes were present in electron micrographs of epithelial cells. The cultured epithelial cells described in this report also demonstrate positive cytokeratin staining. The epithelial cells reacted positively with PSA antibody, indicating that the cells retain their secretory role in cell culture for a limited period. Epithelial cells expressed the HGF receptor, c-met; stromal cells secreted HGF. Insulin, transferrin, and selenium increased the growth of cells in the chemically defined media, compared with minimum essential media (MEM) and Ham's F12.

CONCLUSIONS

In summary, essentially pure cultures of prostate stromal or epithelial cells have been established using simple isolation and culture methods. These cells will be useful for the investigation of related growth factors, such as insulin-like growth factor I and insulin-like growth factor II, and in understanding the basis for stromal-epithelial cell interactions.

Detection of genotoxic effects in human gastric and nasal mucosa cells isolated from biopsy samples

To assess genotoxic burdens from chemicals, it is necessary to relate observations in experimental animals to humans. The success of this extrapolation would be increased by including data on chemical activities in human tissues. Therefore, we have developed techniques to assess DNA damage in human gastric and nasal mucosa (GM, NM) cells. Biopsy samples were obtained during gastroscopy from macroscopically healthy tissue of the stomach or from healthy nasal epithelia during surgery. The specimens were incubated for 30-45 min at 37 degrees C with a digestive solution. We obtained 1.5-8 x 10(6) GM cells and 5-10 x 10(5) NM cells per donor, both with viabilities of 80-95%. The cells were incubated in vitro for 1 hr at 37 degrees C with the test compounds added in their appropriate solvents. In GM cells, we studied N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), sodium dichromate (Na2Cr2O7), nickel sulphate (NiSO4), cadmium sulphate (CdSO4), and lindane. In NM cells, lindane was investigated. Each compound was assessed for DNA damaging activity in cells of at least three different human donor samples using the microgel single cell assay. Similar studies were performed with GM and NM cells obtained from Sprague-Dawley rats. We have found human GM cells to be more sensitive to the genotoxic activity of MNNG than rat GM cells (low effective concentration [LEC] = 0.16 and 0.625 micrograms/ml for human and rat, respectively). Human cells were also more sensitive to the cytotoxic/genotoxic activity of NiSO4 (LEC = 5 and 19 mumoles/ml for human and rat, respectively). CdSO4 was genotoxic in human GM cells (LEC = 0.03-0.125 mumoles/ml), whereas no dose-related genotoxicity was observed in rat GM at concentrations up to 0.5 mumoles/ml. In contrast, approximately equal responses regarding genotoxicity and cytotoxicity were observed in rat and human GM for Na2Cr2O7 (0.25-1 mumoles/ml). Lindane, however, was genotoxic in three out of four rat GM but not in human GM cells (0.5-1 mumoles/ml), whereas it was active in both rat and human NM cells. Together with other recently published in vivo findings, our results with lindane can be interpreted according to a parallelogram approach. In view of possible human exposure situations and the sensitivities of the two target tissues from both species, the data imply that lindane will pose a health risk to humans by inhalation but not by ingestion.

Carcinogen-DNA adducts in cultures of rat and human hepatocytes

Exposure to chemical carcinogens can often be identified by detection of DNA adduct lesions. Primary cultures of isolated rat and human hepatocytes were exposed to 2-acetyl-aminofluorene (AAF), 4-aminobiphenyl (ABP), or benzo[a]pyrene (BP). The isolated DNA from the exposed cells was analyzed using the 32P-post-labeling assay. A greater total of carcinogen-DNA adducts, 2-12-fold, were observed in human hepatocytes than rat hepatocytes at the same concentrations. The predominant DNA adducts for each carcinogen were the same between rat and human cells. The N-(deoxyguanosin-8-yl)-2-aminofluorene (dG-C8-AF) was the major AAF-DNA adduct. The N-(deoxyguanosin-8-yl)-4-aminobiphenyl (dG-C8-ABP) was the major ABP-DNA adduct. In the rat N2-[10 beta-(7 beta, 8 alpha, 9 alpha-trihydroxy-7,8,9,10-tetrahydrobenzo[a]pyrene)yl] deoxyguanosine (dG-N2-BP) and two unidentified adducts were nearly equivalent in amount, while the major BP-DNA adducts in the humans was the dG-N2-BP. The rat hepatocyte in vitro results are comparable to the predominant adducts found with rats exposed in vivo. The two different cultures of human hepatocytes demonstrated qualitative and quantitative differences in specific DNA adducts from rat hepatocytes. This study and others using human hepatocyte cultures demonstrate that this in vitro system can provide useful information for assessing human carcinogenic hazards.

Liver slices in dynamic organ culture. II. An in vitro cellular technique for the study of integrated drug metabolism using human tissue

1. Precision cut human liver slices in dynamic organ culture have been used to study the integrated metabolism of 7-ethoxycoumarin and the conjugation of 7-hydroxycoumarin. 2. The metabolism of 7-ethoxycoumarin and 7-hydroxycoumarin was monitored for 6 h. For both substrates there was a time-dependent increase in metabolites present in the incubation medium. The low levels of free 7-hydroxycoumarin found in the medium when 7-ethoxycoumarin was the substrate suggests good coupling of phase I and phase II metabolism. 3. With suitable incubation conditions, i.e. change of medium containing new substrate every 2 h, the metabolism of both 7-ethoxycoumarin and 7-hydroxycoumarin by human liver slices was found to proceed at similar rates for up to 24 h. This was demonstrated using five separate human liver preparations. 4. Human liver slices also metabolized mono-chlorobenzene and o-, m- and p-dichlorobenzene to aqueous soluble metabolites. There was a time-dependent increase in the appearance of aqueous soluble metabolites present in the incubation medium. Metabolites were not retained by the liver slices. 5. A cold-storage transit buffer has been described and used to maintain the levels of drug metabolism in both rat and human tissue for periods of up to 6 h. 6. The use of human liver slices in dynamic organ culture as a suitable method for the direct assessment of integrated hepatic drug metabolism is proposed.

Induction of cytochrome P(1)450 RNA and benzo[a]pyrene metabolism in primary human hepatocyte cultures with benzanthracene

Exposure of cells to microsomal enzyme inducers can modify the potency of many carcinogens. We have examined the steady-state level of RNA from the P(1)450 gene and the metabolism of benzo[a]pyrene (BP) in primary cultures of human hepatocytes exposed for up to 4 days to 12.5 microM benzanthracene (BA), and in uninduced control cultures. While the steady-state levels of RNA from the P(1)450 gene were nondetectable in uninduced (DMSO only) human hepatocytes, 12.5 microM BA-induced AHH activity, BP metabolism, and/or P(1)450-specific RNA in hepatocytes from seven human cases were investigated. RNA levels specific for the P(1)450 gene appeared maximal at 24 hr following exposure to BA, whereas, the protein, as determined by AHH enzyme activity from BA-induced hepatocytes, continued to increase up to the last time point examined, 72 hr. BA induction for 96 hr increased metabolism of BP (initial concentration of BP, 10 microM) over a time course of 3, 6, 12, and 24 hr of incubation with BP compared with that of controls. The major metabolites of BP produced by human hepatocytes in culture were the unidentified polar BP metabolite(s), possibly polyhydroxylated. BA induction caused approximately a twofold increase in these metabolites. BA-induced cultures showed an increase in glutathione conjugation compared to that in controls. The percentage of glucuronide and sulfate conjugates remains similar in all cultures. Total binding of tritium label BP to DNA was 1.3-fold to fivefold greater in induced cultures, and related more to total metabolism than to production of a specific metabolite. Exposure of human hepatocytes in vitro to BA leads to a large increase in the steady-state level of the RNA specific for the P(1)450 gene and an increase metabolism of BP.

A comparison of the inhibition of deacetylase in primary cultures of rat and human hepatocytes effecting metabolism and DNA-binding of 2-acetylaminofluorene

The metabolism of 2-acetylaminofluorene (AAF) in primary cultures of rat and human hepatocytes was investigated to determine if the activation of this well-studied chemical carcinogen proceeds via similar routes of metabolism between species. The total level of AAF metabolite(s) bound to hepatocellular DNA was determined in the presence of deacetylase inhibitors, diethyl(p-nitrophenyl) phosphate (paraoxon) or bis(p-nitrophenyl) phosphate (BPNPP). These compounds are known to inhibit deacetylase and to decrease the mutagenicity of AAF. Experiments with rat and human hepatocytes demonstrated inhibition in the deacetylation of AAF (5 x 10(-4) M) with paraoxon or BPNPP. The BPNPP (5 x 10(-4) M inhibited 99% of the AF formation in the human hepatocytes and 88% inhibition in the rat hepatocytes. Paraoxon at 10(-4) M demonstrated a 98% inhibition of deacetylation with humans and a 92% inhibition with rats. The rat hepatocytes also showed a 53% decrease in DNA binding in the presence of paraoxon. In contrast with human hepatocytes, while paraoxon decreased the AF metabolite by greater than 97%, there was no change in total DNA binding.

Genotoxicity of N-nitrosochlordiazepoxide in cultured mammalian cells

Chlordiazepoxide, a benzodiazepine derivative commonly used for the treatment of anxiety, was found to react with sodium nitrite in HCl aqueous solution yielding, at pH ranging from 0.5 to 5,N-nitrosochlordiazepoxide (NO-CDE). In the absence of a metabolic activation system, a dose-dependent frequency of DNA single-strand breaks was revealed by the alkaline elution technique in V79 cells exposed to subtoxic NO-CDE concentrations ranging from 33 to 330 microM. DNA lesions were only partially repaired within 48 hr, and their promutagenic character was demonstrated by the induction of 6-thioguanine resistance in the same cells. The genotoxicity of NO-CDE was confirmed by results obtained in metabolically competent primary cultures of both rat and human hepatocytes, which displayed similar dose-related amounts of DNA fragmentation and of DNA repair synthesis after treatment with concentrations ranging from 33 to 1000 microM. In conclusions similar to those which might occur in the stomach of a patient taking chlordiazepoxide the concentration of NO-CDE in the reaction mixture (50 microM) was of the same order as the concentrations found to induce a genotoxic effect in cultured mammalian cells.

Genotoxicity of carcinogens in human hepatocytes: application in hazard assessment

Evaluation of chemical genotoxicity has been used in assessing human cancer hazard, based on the observation that most human carcinogens are known to be DNA-reactive. The availability of data on the DNA-reactivity of compounds in metabolically competent human cells would assist hazard assessment by providing direct information of human genotoxicity. To evaluate the reliability of human hepatocytes for this purpose, the induction of DNA repair by DNA-reactive carcinogens of several structural classes and related noncarcinogens was studied. All the carcinogens elicited DNA repair synthesis, whereas the noncarcinogens did not. These studies provide additional support for the use of human hepatocytes in a DNA repair test in the investigation of genotoxicity. The demonstration of genotoxicity in human cells is suggested to provide important information for hazard assessment.

Induction of unscheduled DNA synthesis in primary cultures of rat, mouse, hamster, monkey, and human hepatocytes

Variation in hepatic metabolism between species may be an important factor in the differences observed in chemical carcinogenesis. We examined 6 chemicals representative of 4 chemical classes in the in vitro hepatocyte DNA repair assay using cells isolated from the Fischer-344 rat, B6C3F1 mouse, Syrian golden hamster, cynomolgus monkey and from human liver. Hepatocytes were isolated by in situ or biopsy liver perfusion and incubated with [3H]-thymidine and the test chemical. Unscheduled DNA synthesis (UDS) was measured as net grains/nucleus (NG) by quantitative autoradiography. Qualitative and quantitative differences in UDS responses were observed for every chemical. Liver cultures isolated from the rat, mouse, hamster, human, and monkey and treated with aflatoxin B1 or dimethylnitrosamine all yielded dose-related increases in NG. Human, rat, and hamster hepatocyte cultures yielded positive responses following exposure to the aromatic amines 2-acetylaminofluorene, 4-aminobiphenyl, and benzidine, whereas cultures isolated from the monkey and mouse yielded less than 0 NG. Treatment with benzo[a]pyrene (BAP) produced strong positive responses in monkey and human hepatocyte cultures, weak positive responses in hamster cultures, and equivocal or negative responses in rat and mouse hepatocyte cultures. Hepatocyte function was assessed by measurement of DNA content, glutathione content, BAP hydroxylase activity, p-nitroanisole-O-demethylase activity, p-nitrophenol conjugation, and urea synthesis rates. The functional capabilities of isolated hamster, monkey, and human hepatocyte cultures do not appear to correlate with UDS responses observed for any compound; however, they indicate that the cultures were metabolically competent at the time of chemical exposure. These studies suggest that rat hepatocytes are a suitable model for human hepatocytes, whereas mouse and male monkey hepatocytes may be insensitive to aromatic amines.

Induction of DNA fragmentation and DNA repair synthesis in human and rat hepatocytes by diethylstilbestrol

The synthetic estrogen diethylstilbestrol (DES), a known human carcinogen, was examined for cytotoxicity, and the induction of DNA damage and repair in primary cultures of human and rat hepatocytes. In both species concentrations of DES ranging from 5.6 to 18 micrograms/ml constantly produced reduction of cell viability and DNA fragmentation in dose-related amounts. However, large individual quantitative differences in the sensitivity to the cytotoxic and DNA-damaging activities of DES were observed among cultures derived from the 5 human donors. DES capability of eliciting DNA-excision repair was weak but statistically significant in both human and rat hepatocytes. Taken as a whole these results contribute to support the hypothesis of a genotoxic mechanism in DES-induced carcinogenesis.

The hepatocyte primary culture/DNA repair test using hepatocytes from several species

The hepatocyte primary culture/DNA repair test, originally validated with rat hepatocytes, has been extended to use hepatocytes from other species including mouse, hamster, guinea pig, rabbit, monkey and human. Both qualitative and quantitative differences have been observed when chemicals are examined in the hepatocyte primary culture/DNA repair test using hepatocytes from more than one species. Examples are discussed that illustrate that the genotoxicity of a chemical can be a species-specific response and that multi-species testing permits a more complete assessment of genotoxicity.

Cell and species differences in metabolic activation of chemical carcinogens

Metabolic activation and DNA binding of aflatoxin B1 (AFB1), N-nitrosodimethylamine (DMN) and benzo[a]pyrene (B[a]P) were compared in human, rat and mouse hepatocytes and human pulmonary alveolar macrophages (PAM). The degree of carcinogen activation by hepatocytes and PAM was measured by cell-mediated mutagenesis assays in which co-cultivated Chinese hamster V79 cells were used to monitor mutagenic metabolites. Hepatocytes from human, mouse and rat metabolized DMN and released the active metabolites to induce either ouabain- or 6-thioguanine-resistant mutation. The mutation frequencies mediated by hepatocytes of the 3 animal species were approximately 3-9 mutants/10(5) survivors at a concentration of 0.2 mM DMN. The variations of radioactivity bound to liver cell DNA were relatively small in cultured mouse, rat, and human hepatocytes exposed to 14C label DMN (0.5 mM) and the binding values were in a range of 6-12 X 10(3) pmoles/mg DNA. However, rat hepatocytes were at least 10-fold more effective than either human or mouse hepatocytes in generating mutagenic metabolites of AFB1 and also had a much higher AFB1 metabolite DNA-binding value. The AFB1 DNA-binding levels were 4.1, 12-27 (range), 120 pmoles/mg DNA respectively in mouse, human, and rat liver cells following AFB1 (3.3 microM) exposure for 20 h. Hepatocytes from the 3 animal species were unable to mediate mutation in the presence of 4 microM B[a]P; PAM activated B[a]P and effectively mediated mutation in the co-cultivated V79 cells. In contrast to results with hepatocytes, PAM failed to generate enough mutagenic metabolites of AFB1 (3.3 microM) and the mediation of mutations was seen only at very high concentration of DMN (80 mM). The genotoxic effects of the 3 carcinogens on hepatocytes from different species in vitro were in agreement with the in vivo animal experiments in that mice are relatively resistant to AFB1 carcinogenesis whereas rats are sensitive; B[a]P is not effective as a complete liver carcinogen in adult rat and mouse whereas DMN induces liver cancer.

Use of the human liver cell line Hep G2 in a modified Salmonella reversion assay

The human hepatoma cell line Hep G2 was used to activate promutagenic chemicals to mutagens in a modified Salmonella typhimurium reversion assay. Hep G2 cells mediated positive mutagenic responses in tester strain TA98 with 5 and 25 micrograms/plate of 2-aminofluorene, but these responses were consistently lower than those seen using primary rat hepatocytes. In addition, 3 and 6 X 10(6) Hep G2 cells per assay produced positive mutagenic responses with 2-aminoanthracene, benzidine, acetylbenzidine and aflatoxin B1, while benzo[a]pyrene, 7,12-dimethylbenz[a]anthracene, 3-methylcholanthrene, 4-aminobiphenyl and 4- and 11-aminobenzo[a]pyrene were nonmutagenic with Hep G2-cell activation. These results indicate that Hep G2 cells may be a useful intact cellular metabolizing system of human origin for predicting the genotoxicity of promutagenic agents, but that the use of Salmonella as a target cell may limit the classes of mutagens detected.

A study of the potential genotoxicity of cimetidine using human hepatocyte primary cultures: discrepancy from results obtained in rat hepatocytes

The genotoxicity of cimetidine, a drug widely used in the treatment of peptic ulcer, was examined in human hepatocyte primary cultures. No induction of unscheduled DNA synthesis, as detected by autoradiography, or of DNA fragmentation, as measured by alkaline elution, was seen in metabolically competent human hepatocytes exposed for 20 h to cimetidine concentrations ranging from 0.33 to 9 mM. These findings, which are in contrast with the previously observed capability of cimetidine to induce DNA damage and repair in rat hepatocyte primary cultures, suggest that for some chemicals the rat hepatocyte model might be an inappropriate predictor of potential genotoxic effects in the analogous human cells.

DNA fragmentation by N-nitrosodimethylamine and methyl methanesulfonate in human hepatocyte primary cultures

The ability of N-nitrosodimethylamine (DMN) and methyl methanesulfonate (MMS) to induce DNA damage in primary cultures of human hepatocytes was examined by the alkaline elution technique. Both the agents induced a dose-dependent increase in DNA elution rate, but appreciable differences in the degree of response to the procarcinogen DMN were observed among cultures obtained from the livers of four patients. A comparative analysis of DNA fragmentation indicated a substantial similarity between human and concurrently studied rat hepatocytes in their response to both DMN and MMS.

Drug metabolising activity of freshly isolated human hepatocytes

Hepatocytes have been isolated from samples of adult human liver by removal of extracellular calcium followed by perfusion with collagenase. The hepatocytes were isolated with a yield of up to 39 X 10(6) cells/g and with a viability of up to 74%. The cells were active in the oxidation of aldrin and 7-ethoxycoumarin. They also catalysed the conjugation of 7-hydroxycoumarin. Monooxygenase activity of the hepatocytes was linear for at least 60 min. Maintenance of the hepatocytes in suspension at 4 degrees C for 19 h resulted in a 15% loss in viability. This was accompanied by a 50% decrease in monooxygenase activity expressed per viable cell. It is concluded that human hepatocytes can be isolated in sufficient yield and with satisfactory viability for use in a range of studies on drug metabolism and toxicity.