Unscheduled DNA synthesis caused by norethindrone and related contraceptive steroids in short-term male rat hepatocyte cultures

Novel genotoxicity assays identify norethindrone to activate p53 and phosphorylate H2AX

Norethindrone is a commonly used drug for contraception and hormone replacement therapy, whose carcinogenic potential is still controversial. We applied a novel and particularly sensitive method to screen for DNA damage with special attention to double-strand breaks (DSBs) and identified norethindrone to be likely genotoxic and therefore potentially mutagenic: a p53-reporter assay served as a first, high-throughput screening method and was followed by the immunofluorescent detection of phosphorylated H2AX as a sensitive assay for the presence of DSBs. Norethindrone at concentrations of 2-100 microg/ml activated p53 and phosphorylated H2AX specifically and in a dose-dependent manner. No p53 activation or H2AX phosphorylation was detected using a panel of structurally/functionally related drugs. The overall amount of DNA damage induced by norethindrone was low as compared with etoposide and ionizing radiation. Consistently, norethindrone treatment did not cause a cell cycle arrest. DSBs were not detected with the neutral comet assay, a less sensitive method for DSB assessment than H2AX phosphorylation. Our findings in the p53-reporter and gamma-H2AX assays could not be ascribed to common DSB-causing artifacts in standard genotoxicity screening, including drug precipitation, high cytotoxicity levels and increased apoptosis. Therefore, our study suggests that norethindrone induces DSBs in our experimental setting, both complementing and adding a new aspect to the existing literature on the genotoxic potential of norethindrone. As the effective concentrations of norethindrone used in our assays were approximately 100- to 1000-fold higher than therapeutical doses, the significance of these findings with regard to human exposure still remains to be determined.

Genotoxic potential of ethinylestradiol in cultured mammalian cells

Ethinylestradiol, a steroidal estrogen, is widely used with various progestogens in oral contraceptives formulations. There are sufficient evidences for the carcinogenicity of ethinylestradiol in experimental animals. The reports on the genotoxic potential of ethinylestradiol are contradictory. Here in the present study we have tested the genotoxicity of ethinylestradiol in human lymphocytes using chromosomal aberrations (CAs), mitotic index (MI) and sister chromatid exchanges (SCEs) as a parameter. The study was carried out in the absence, as well as in the presence, of rat liver microsomal fraction, with and without NADP. Ethinylestradiol was studied at three different concentrations (1, 5 and 10 microM) and was found non-genotoxic in the absence of metabolic activation (S9 mix) and in S9 mix without NADP. Ethinylestradiol was found to be genotoxic at 5 and 10 microM in the presence of S9 mix with NADP. To study the possible mechanism of the genotoxicity of ethinylestradiol, superoxide dismutase (SOD) and catalase (CAT) were used separately and in combination along with 10 microM of ethinylestradiol at different doses. SOD treatment increased CAs and SCEs and decreases MI as compared with treatment with 10 microM of ethinylestradiol alone in the presence of S9 mix with NADP at both of the tested doses. CAT treatment decreased the frequencies of CAs and SCEs and increased MI, as compared with treatment with 10 microM of ethinylestradiol alone in the presence of S9 mix with NADP. CAT treatment in combination with SOD also decreased the frequencies of CAs and SCEs and increased MI suggesting a possible role of reactive oxygen species for the genotoxic damage.

Genotoxicity of hormonal steroids

Hormonal steroids have a widespread use in medicine and their side effects are continuously debated. The possible genotoxic activity of steroids has been the subject of many investigations. The natural estrogens estradiol, estrone and estriol are generally negative in the ICH core battery of tests, but several positive results have been obtained when using additional endpoints of genotoxicity. The genotoxic activity of the 4-hydroxy metabolites of estradiol and estrone is well established. The synthetic steroidal estrogens have a comparable profile of negative and positive test results. Cyproterone acetate and some of its analogues have a special position within the group of progestins. Their genotoxic potential has been established. Other progestins are generally negative in the routine tests. Anti-glucocorticoids, anti-progestins, corticosteroids, androgens, anabolics and anti-androgens appear to be devoid of genotoxic activities. The genotoxic potential of estradiol, estrone and cyproterone acetate with its analogues may play no role under normal physiological and therapeutic conditions. The metabolic conditions that are needed for the formation of DNA-reactive metabolites and oxygen radicals may not be present in humans. Epidemiological cancer data seem to support this view. The importance of thresholds in the dose-effect-relationship of genotoxicity data and their use in risk assessment is discussed.

Species, sex and inter-individual differences in DNA repair induced by nine sex steroids in primary cultures of rat and human hepatocytes

Sex steroids, due to the generally negative responses observed in routinely employed standard genotoxicity assays, are considered epigenetic carcinogens. Some doubts on this conviction are raised by the results of recent studies providing evidence that cyproterone acetate and two structural analogues, chlormadinone acetate and megestrol acetate, are genotoxic in female rats but only for the liver, and in primary human hepatocytes from donors of both genders. The experimental evidence suggests that the metabolic activation of these molecules to reactive species and the consequent formation of DNA adducts occur only in the intact hepatocyte. Since the possibility that other sex steroids cause a liver-specific genotoxic effect cannot be ruled out a priori, we investigated nine drugs of this family for their ability to induce DNA repair synthesis in primary cultures of rat and human hepatocytes. Each steroid was tested in cultures from at least two male and two female donors of each species. Hepatocytes were exposed for 20h to sub-toxic concentrations ranging from 1 to 50 micro M, and DNA repair induction was measured by quantitative autoradiography. In primary rat hepatocytes, induction of DNA repair indicative of a frankly positive response was detected in cultures from: 2/2 males and 3/3 females with drospirenone, 2/2 males and 1/2 females with ethinylestradiol, 1/2 males and 1/2 females with oxymetholone, 1/2 males with norethisterone, 1/4 females with progesterone, and 1/4 males with methyltestosterone. Consistent negative responses were obtained with testosterone and stanozolol. A few inconclusive responses were observed in rat hepatocytes exposed to progesterone, medroxyprogesterone, norethisterone, methyltestosterone and oxymetholone. In contrast, under the same experimental conditions the nine sex steroids provided frankly negative responses in the large majority of cultures of primary hepatocytes from both male and female human donors; the only exceptions being the inconclusive responses obtained in cultures from two of the donors exposed to norethisterone and to ethinylestradiol, and from one of the donors exposed to testosterone, methyltestosterone, and stanozolol. These results and previous findings concerning cyproterone and its structural analogues suggest that sex steroids differ for their ability to induce DNA repair, and that their genotoxicity may be: (i) different in rat and human hepatocytes, (ii) dependent on the sex of the donor, and (iii) affected by inter-individual variability.

Are some progestins genotoxic liver carcinogens?

Progestins (progestogens) are classified by the International Agency for Research on Cancer (IARC) as possibly carcinogenic to humans. In the last decade evidence has shown that a synthetic drug of this family, cyproterone acetate, is activated to a reactive species by the liver, and forms DNA adducts and elicits DNA repair in hepatocytes from both rats and humans. The response is similar in humans of both genders but markedly higher in female than in male rats. The promutagenic character of DNA lesions is indicated by the increase in liver of female rats of the frequency of micronucleated cells, of mutations, and of enzyme-altered preneoplastic foci. Two other synthetic progestins, chlormadinone acetate and megestrol acetate, and an aldosterone antagonist, potassium canrenoate, share with cyproterone acetate the 17-hydroxy-3-oxopregna-4,6-diene structure. While less extensively studied, results so far obtained indicate that they are capable of inducing genotoxic effects qualitatively similar to those of cyproterone acetate. The majority of progestins have not been systematically tested for genotoxicity and the generally negative responses obtained with the standard battery of genotoxicity tests might be the consequence of the use of inappropriate target cells and/or metabolic activation systems. Cyproterone acetate, is activated by the hepatocytes to reactive species of such short half-life that they react only with the DNA of the cell in which are formed. Therefore, it cannot be excluded a priori that other progestins will not display genotoxic effects when tested adequately. This hypothesis is supported by the knowledge that estrogen-progestin combinations used as oral contraceptives are classified by the IARC as carcinogenic to humans due to the increased risk of hepatocellular carcinoma. This risk should probably be ascribed to the progestin component, since estrogens are carcinogenic to humans due to the increased risk of endometrial and possibly of breast cancer but not liver cancer.

Induction of micronuclei and of enzyme-altered foci in the liver of female rats exposed to progesterone and three synthetic progestins

Progesterone (PG) and three structurally similar synthetic progestins-norethisterone (NE), allylestrenol (AE), and dydrogesterone (DG)-have been compared for their ability to induce the formation of micronuclei and of enzyme-altered foci in the liver of female rats. In the micronucleus assay, carried out in rats given a single p.o. dose of 100 mg kg-1 3 days before partial hepatectomy and sacrificed for cell sampling 2 days later, the frequency of micronucleated hepatocytes was 3.5-fold higher than in controls with PG, 2.8-fold with DG, 2.2-fold with NE and 2.1-fold with AE, but the increase was statistically significant only for PG. In the liver foci assay, performed to evaluate the tumor initiating activity of p. o. dosing with 100 mg kg-1 once a week for 6 successive weeks, the values of the number and area of gamma-glutamyltranspeptidase-positive foci were, as compared to controls, 15.9- and 100-fold higher with NE, and 13.9- and 52-fold higher with AE, but only the increase of area produced by NE was statistically significant; PG and DG did not display in this test any activities. Considered together with previous findings, these results suggest that NE might be biotransformed in the liver into reactive species and thus behave as a weak genotoxic agent.

Recommended protocols based on a survey of current practice in genotoxicity testing laboratories: I. Unscheduled DNA synthesis assay in rat hepatocyte cultures

A protocol based primarily on current laboratory practices in the performance of the unscheduled DNA synthesis (UDS) assay with primary rat hepatocyte cultures has been developed. These guidelines were developed using tabulated responses to a detailed questionnaire completed by North American and European governmental, university and contract laboratories involved with the UDS test. This report identifies those modifications to previously described methodologies which are used on a regular basis and also serves to clarify confusing or inconsistent practices. Although this protocol pertains specifically to the use of primary rat hepatocyte cultures, it can be modified to incorporate other types of cells in which certain aspects remain the same.

Fluorescence-activated sorting of rat hepatocytes based on their mixed function oxidase activities towards diethoxyfluorescein

The formation of ethoxyfluorescein and fluorescein from diethoxyfluorescein by isolated rat hepatocytes has been used as a basis for separating such cells dependent on their mixed function oxidase activities by fluorescence-activated flow cytometry. Five equal fractions defined by computer-generated regions were isolated. Non-viable cells with low fluorescence (region 1) represented 10-15% of the population, while the remainder with higher mixed function oxidase activities (regions 2-5), were greater than 95% viable by Trypan Blue exclusion. In region 1, 30% of the viable cells were binucleate, 67% diploid while in region 5, 13% were binucleate and 69% tetraploid. At 3 h after sorting, following attachment to glass coverslips, exposure of cells to methyl methanesulphonate, retrorsine or norethindrone resulted in unscheduled DNA synthesis which was 2-fold higher in the tetraploid-rich region 5, while aflatoxin B1, benzo[a]pyrene or 2-acetylaminofluorene caused a 5-fold increase in unscheduled DNA synthesis in these cells, relative to the diploid-rich hepatocytes in region 2.

Destruction of cytochrome P-450 and formation of green pigments by contraceptive steroids in rat hepatocyte suspensions

The contraceptive steroid norethindrone caused a rapid time and dose-dependent loss of cytochrome P-450 from rat hepatocytes in suspension cultures. Up to 30% of this cytochrome was lost in the first 5 min of incubation; longer incubations resulted in little further loss even though not all the steroid was metabolised and the cells remained viable. Such cultures were used to investigate the formation of N-alkylated porphyrins (green pigments) which could be extracted from cell incubation mixtures following exposure to norethindrone and separation by HPLC or TLC. The number of N-alkylated porphyrins formed was dependent both on the time of incubation and the concentration of steroid. After 1 min, 1 major green pigment (GP1) was resolved using either high (0.3 mM) or low (0.03 mM) norethindrone concentrations. With longer incubation times (60 min), at high steroid concentrations, only one additional polar adduct (GP2) was formed. At lower steroid levels, 3 more polar components (GP2, 3 and 4) were seen. As judged by HPLC or TLC, GP1 corresponds to the pigment formed in microsomal preparations incubated with norethindrone in vitro, while GP2, 3 and 4 correspond to the pigments extracted from the livers of rats administered this steroid in vivo. Pretreatment of rats with either phenobarbitone or 3-methylcholanthrene induced cytochrome P-450s. Relative to controls, phenobarbitone pretreatment also resulted in a greater accumulation of green pigments in hepatocytes incubated with norethindrone, the more polar forms of green pigments (GP3 and 4), showing a disproportionate increase in concentration. The mixed function oxidase inhibitor SKF 525-A or high concentrations of steroid not containing an ethynyl function, e.g. norethandrolone, when added to cell cultures containing norethindrone, preferentially inhibited the formation of GP3 and 4. When purified green pigments were added to cell incubation mixtures in the absence of norethindrone, preferentially inhibited the formation of GP3 and 4. When purified green pigments were added to cell incubation mixtures in the absence of norethindrone, no interconversion of one form to another could be demonstrated. The results suggest that the more polar norethindrone-protoporphyrin IX adducts (GP2, 3 and 4) arise as a result of metabolic modification of norethindrone rather than the protoporphyrin IX moiety, either prior to or after activation of the ethynyl function. The formation of several green pigment components in hepatocyte suspensions was not unique to norethindrone but occurred with a number of other 17-ethynyl-substituted steroids.