Toshio Kasamatsu

IWGT report on quantitative approaches to genotoxicity risk assessment II. Use of point-of-departure (PoD) metrics in defining acceptable exposure limits and assessing human risk ☆

This is the second of two reports from the International Workshops on Genotoxicity Testing (IWGT) Working Group on Quantitative Approaches to Genetic Toxicology Risk Assessment (the QWG). The first report summarized the discussions and recommendations of the QWG related to the need for quantitative dose-response analysis of genetic toxicology data, the existence and appropriate evaluation of threshold responses, and methods to analyze exposure-response relationships and derive points of departure (PoDs) from which acceptable exposure levels could be determined. This report summarizes the QWG discussions and recommendations regarding appropriate approaches to evaluate exposure-related risks of genotoxic damage, including extrapolation below identified PoDs and across test systems and species. Recommendations include the selection of appropriate genetic endpoints and target tissues, uncertainty factors and extrapolation methods to be considered, the importance and use of information on mode of action, toxicokinetics, metabolism, and exposure biomarkers when using quantitative exposure-response data to determine acceptable exposure levels in human populations or to assess the risk associated with known or anticipated exposures. The empirical relationship between genetic damage (mutation and chromosomal aberration) and cancer in animal models was also examined. It was concluded that there is a general correlation between cancer induction and mutagenic and/or clastogenic damage for agents thought to act via a genotoxic mechanism, but that the correlation is limited due to an inadequate number of cases in which mutation and cancer can be compared at a sufficient number of doses in the same target tissues of the same species and strain exposed under directly comparable routes and experimental protocols.

Genotoxicity studies on green tea catechin.

The beneficial effects of tea catechins are well documented. We evaluated the genotoxic potential of a green tea catechin preparation using established genotoxicity assays, including a bacterial reverse mutation assay (Ames test), a chromosomal aberration assay in cultured Chinese hamster lung cells (CHL/IU), a mouse lymphoma L5178Y/tk assay, and a bone marrow micronucleus (MN) assay in ICR CD mice and SD rats. No significant increases in the number of revertant colonies were observed in the Ames test, but positive responses were observed in two in vitro assays: the chromosomal aberration assay and mouse lymphoma L5178/tk assay. However, the in vivo study demonstrated no significant increase in micronucleated polychromatic erythrocytes (MNPCE) in the bone marrow of both ICR CD mice and SD rats administered a high dose of the green tea catechin preparation up to 2000mg/kg. Combined with favorable epidemiological information suggesting a chemopreventive effect of tea catechins on carcinogenesis, we conclude that green tea catechin presents no significant genotoxic concern under the anticipated conditions of use. These results are consistent with other genotoxicity studies of tea catechins, which show minimal, if any, genotoxic potential.

Involvement of hydrogen peroxide in chromosomal aberrations induced by green tea catechins in vitro and implications for risk assessment.

Catechins, which are polyphenol compounds found in abundance in green tea, have elicited high interest due to their beneficial effects on health. Catechins have also been demonstrated to induce chromosomal aberrations in vitro, although no clastogenicity was confirmed in studies in vivo. We investigated the mechanism of catechin-induced chromosomal aberrations in CHL/IU cells. Addition of catalase suppressed chromosomal aberrations, indicating involvement of hydrogen peroxide (H2O2). We confirmed that substantial amounts of H2O2 are generated when catechins are incubated under in vitro culture conditions, whereas, interestingly, extremely low amounts of H2O2 were detected when catechins were incubated at the same concentration in water. Generation of H2O2 increased steeply above pH 6, indicating that pH is a key factor in determining how much H2O2 is generated via catechins in vitro. Our assessment indicates that humans have practically non-existent exposure to H2O2 when catechins are ingested in a beverage. Polyphenols, including catechins, are known to act as antioxidants due to their reducing potential. However, under in vitro culture conditions, catechins are thought to act primarily as pro-oxidants by reducing ambient or dissolved oxygen to form H2O2. Based on the above observations, we conclude that in vitro culture conditions as currently employed are inappropriate to address genotoxicity concerns regarding polyphenols, including catechins.

Measurement of glycidol hemoglobin adducts in humans who ingest edible oil containing small amounts of glycidol fatty acid esters.

Hemoglobin (Hb) adducts are frequently used to address and/or monitor exposure to reactive chemicals. Glycidol (G), a known animal carcinogen, has been reported to form Hb adducts. Here, we measure G adduct levels in humans who daily ingest DAG oil, an edible oil consisting mainly of diacylglycerol. Since DAG oil contains a small amount of glycidol fatty acid esters (GEs), possible exposure to G released from GEs has been raised as a possible concern. For measurement of Hb adducts, we employed the N-alkyl Edman method reported by Landin et al. (1996) using gas chromatography-tandem mass spectrometry with minor modifications to detect G-Hb adducts as N-(2,3-dihydroxy-propyl)valine (diHOPrVal). Blood samples were collected from 7 DAG oil users and 6 non-users, and then G-Hb adduct levels were measured. G-Hb adducts were detected in all samples. The average level of diHOPrVal was 3.5±1.9pmol/g globin in the DAG oil users and 7.1±3.1pmol/g globin in the non-users. We conclude that there is no increased exposure to G in individuals who daily ingest DAG oil.

Glycidol exposure evaluation of humans who have ingested diacylglycerol oil containing glycidol fatty acid esters using hemoglobin adducts.

Glycidol fatty acid esters (GEs) have been found as impurities in refined edible oils including diacylglycerol (DAG) oil, and concerns of possible exposure to glycidol (G), a known animal carcinogen, during digestion have been raised. We previously measured N-(2,3-dihydroxy-propyl)valine (diHOPrVal), a G hemoglobin adduct, for DAG oil exposed and non-exposed groups and showed there was no significant difference between them. In the present study, we conducted an additional analysis to verify the outcome of the previous report. The first experiment was designed as a matched case-control study to adjust variables with an increased sample size. The average levels of diHOPrVal were 6.9 pmol/g-globin (95%CI: 4.9-9.0) for 14 DAG oil exposed subjects and 7.3 pmol/g-globin (95%CI: 6.1-8.5) for 42 non-exposed volunteers, and no significant difference in levels was found between the two groups. In a second experiment, we compared the adduct levels of 12 DAG oil exposed subjects before and after discontinuing use of DAG oil, and found there was no significant change in diHOPrVal levels (from 7.1±1.1 to 7.5±1.4 pmol/g-globin). These results suggest that there was no increased exposure to G for humans who ingested DAG oil daily, although the evaluated population was limited.

Characterization of glycidol-hemoglobin adducts as biomarkers of exposure and in vivo dose.

Hemoglobin adducts have been used as biomarkers of exposure to reactive chemicals. Glycidol, an animal carcinogen, has been reported to form N-(2,3-dihydroxy-propyl)valine adducts to hemoglobin (diHOPrVal). To support the use of these adducts as markers of glycidol exposure, we investigated the kinetics of diHOPrVal formation and its elimination in vitro and in vivo. Five groups of rats were orally administered a single dose of glycidol ranging from 0 to 75mg/kg bw, and diHOPrVal levels were measured 24h after administration. A dose-dependent increase in diHOPrVal levels was observed with high linearity (R(2)=0.943). Blood sampling at different time points (1, 10, 20, or 40days) from four groups administered glycidol at 12mg/kg bw suggested a linear decrease in diHOPrVal levels compatible with the normal turnover of rat erythrocytes (life span, 61days), with the calculated first-order elimination rate constant (kel) indicating that the diHOPrVal adduct was chemically stable. Then, we measured the second-order rate constant (kval) for the reaction of glycidol with N-terminal valine in rat and human hemoglobin in in vitro experiments with whole blood. The kval was 6.7±1.1 and 5.6±1.3 (pmol/g globin per μMh) in rat and human blood, respectively, indicating no species differences. In vivo doses estimated from kval and diHOPrVal levels were in agreement with the area under the (concentration-time) curve values determined in our earlier toxicokinetic study in rats. Our results indicate that diHOPrVal is a useful biomarker for quantification of glycidol exposure and for risk assessment.

Modified Ames test using a strain expressing human sulfotransferase 1C2 to assess the mutagenicity of methyleugenol

IntroductionSeveral alkenylbenzenes, including methyleugenol (ME), are present in a wide range of botanicals and exhibit carcinogenic and mutagenic properties. Negative results are generally obtained for alkenylbenzenes in standard in vitro genotoxicity tests, including the Ames test. A lack of mutagenicity observed in such tests is thought to result from impaired metabolic activation of alkenylbenzenes via hydroxylation, with subsequent sulfoconjugation to its ultimate mutagenic or carcinogenic form. Although recent studies have reported the mutagenicity of hydroxylated ME metabolites in the Ames test using modified TA100 strains expressing human sulfotransferases (SULTs), to our knowledge, the detection of ME mutagenicity has not yet been reported.FindingsUsing strain TA100-hSULT1C2, which expresses human SULT1C2, we optimized the protein content of S9 Mix and the pre-incubation time required to promote metabolic activation in the Ames test. This procedure enabled us to obtain a positive response with ME.ConclusionsWe established Ames-test conditions enabling the detection of ME-induced mutagenicity, using a strain expressing human SULT1C2 in the presence of induced-rat S9 Mix. This simple approach will help assess the mutagenicity of other alkenylbenzenes and related chemicals.

The reconstructed skin micronucleus assay in EpiDerm™: reduction of false-positive results - a mechanistic study with epigallocatechin gallate.

The high rate of false-positive or misleading results in in vitro mammalian genotoxicity testing is a hurdle in the development of valuable chemicals, especially those used in cosmetics, for which in vivo testing is banned in the European Union. The reconstructed skin micronucleus (RSMN) assay in EpiDerm™ (MatTek Corporation, USA) has shown promise as a follow-up for positive in vitro mammalian genotoxicity tests. However, few studies have explored its better predictive performance compared with existing in vitro assays. In the present study, we followed the protocol of the RSMN assay and used eight chemicals to compare micronucleus (MN) induction with EpiDerm™ with that in normal human epidermal keratinocytes (NHEKs), both derived from human skin. The assessments of EpiDerm™ conformed to those of in vivo MN assay, whereas those of NHEKs did not. The effect of cell differentiation status on MN induction was further addressed using a model compound, epigallocatechin gallate (EGCG), which is a major component of green tea extract that shows positive results in in vitro mammalian genotoxicity assays via oxidative stress and negative results in in vivo MN studies. RSMN assay in an underdeveloped epidermal model, EpiDerm-201™ (MatTek Corporation), showed a negative result identical to that in EpiDerm™, indicating that the barrier function of keratinocytes has limited impact. Analysis of the gene expression profile of both EpiDerm™ and NHEKs after EGCG treatment for 12h revealed that the expression of genes related to genotoxic response was significantly induced only in NHEKs. Conversely, antioxidative enzyme activities (catalase and glutathione peroxidase) in EpiDerm™ were higher than those in NHEKs. These results indicate that EpiDerm™ has antioxidant properties similar to those of a living body and is capable of eliminating oxidative stress that may be caused by EGCG under in vitro experimental conditions.

A retrospective evaluation method for in vitro mammalian genotoxicity tests using cytotoxicity index transformation formulae

Although in vitro chromosomal aberration tests and micronucleus tests have been widely used for genotoxicity evaluation, false-positive results have been reported under strong cytotoxic conditions. To reduce false-positive results, the new Organization for Economic Co-operation and Development (OECD) test guideline (TG) recommends the use of a new cytotoxicity index, relative increase in cell count or relative population doubling (RICC/RPD), instead of the traditionally used index, relative cell count (RCC). Although the use of the RICC/RPD may result in different outcomes and require re-evaluation of tested substances, it is impractical to re-evaluate all existing data. Therefore, we established a method to estimate test results from existing RCC data. First, we developed formulae to estimate RICC/RPD from RCC without cell counts by considering cell doubling time and experiment time. Next, the accuracy of the cytotoxicity index transformation formulae was verified by comparing estimated RICC/RPD and measured RICC/RPD for 3 major chemicals associated with false-positive genotoxicity test results: ethyl acrylate, eugenol and p-nitrophenol. Moreover, 25 compounds with false-positive in vitro chromosomal aberration (CA) test results were re-evaluated to establish a retrospective evaluation method based on derived estimated RICC/RPD values. The estimated RICC/RPD values were in good agreement with the measured RICC/RPD values for every concentration and chemical, and the estimated RICC suggested the possibility that 12 chemicals (48%) with previously judged false-positive results in fact had negative results. Our method enables transformation of RCC data into RICC/RPD values with a high degree of accuracy and will facilitate comprehensive retrospective evaluation of test results.

Repeated-dose liver and gastrointestinal tract micronucleus assays with CI Solvent Yellow 14 (Sudan I) using young adult rats.

The in vivo genotoxicity of CI Solvent Yellow 14 (Sudan I) was examined using repeated-dose liver and gastrointestinal tract micronucleus (MN) assays in young adult rats. Sudan I is a mono-azo dye based on aniline and 1-amino-2-hydroxynaphthalene. This dye was demonstrated as a rat liver carcinogen in a National Toxicology Program (NTP) bioassay, and genotoxicity was noted in a rat bone marrow micronucleus (BMMN) assay. In the present study, Sudan I was administered orally to rats for 14-days, and the MN frequency in the liver, stomach, colon, and bone marrow were analyzed. The frequency of micronucleated hepatocytes (MNHEPs) was not significantly increased by the administration of the Sudan I. Gastrointestinal tract MNs were also not induced. However, in the BMMN assay, a significant increase in micronucleated immature erythrocytes (MNIMEs) was observed in a dose-dependent manner. While Sudan I has been reported to lack hepatic genotoxicity, it has also exhibited tumor-promoting activities. These results are consistent with the lack of induction of MN in the hepatocytes. The lack of MN induction in cells of the gastrointestinal tract was also logical because azo-compounds are reported to be unlikely to induce DNA damage in the rat gut. The repeated-dose rat liver and gastrointestinal tract MN assays have the potential to be used in the evaluation of the genotoxicity of a chemical in each organ in accordance with its mode of action.