4,4'-Methylenedianiline (MDA) is used as a hardener in the manufacture of plastics and polyurethanes. MDA has been classified as a carcinogen in animals and is a suspected human carcinogen. Assuming that MDA would yield similar DNA adducts to other arylamines, we synthesized the following C-8 guanine adducts: N'-acetyl-N-(deoxyguanosin-8-yl)-MDA, N-(deoxyguanosin-8-yl)-MDA, N-(deoxyguanosin-8-yl)-4MA, and their corresponding 3'-monophosphate derivatives. We developed methods to identify these adducts of MDA in liver DNA using 32P-postlabeling, HPLC, and GC-MS techniques. Liver DNA was obtained from rats treated with radiolabeled MDA (1.11 and 116.5 mumol/kg body weight). The total radioactivity bound to the DNA corresponded to 0.06 and 2.7 adducts per 10(7) nucleotides [covalent binding index (CBI = (mumol of adduct per mol of nucleotide)/(mmol of compound per kg body weight)) of 1.05 and 2.3]. This DNA-binding potency is in the range of weakly genotoxic compounds. The liver DNA was analyzed for the presence of the synthesized adducts by the following methods: (I) HPLC analysis of nucleotides and purines after enzymatic and acid hydrolysis, and (II) 32P-postlabeling after enzymatic hydrolysis. The major adducts found in vivo did not correspond to the synthesized standards. Further work was carried out to determine the structure of the unidentified adducts. It was possible to release MDA and MDA-d4 from DNA of rats dosed with MDA and/or MDA-d4 and from the synthesized adducts using strong base hydrolysis. Liver of two female Wistar rats given 500 mumol/kg MDA-2HCl was hydrolyzed in 0.1 M NaOH overnight at 110 degrees C. GC-MS analysis of the heptafluorobutyric anhydride derivatized dichloromethane extracts detected 428 +/- 40 fmol of MDA/mg of DNA. In the control animals no MDA was found. The experiment was repeated with livers from animals dosed 500 mumol/kg MDA-d4.2DCl. In these rats 488 +/- 19 fmol MDA-d4 was found to be bound at liver DNA. Taking into account a 68% yield of the method, the CBI found in these cases was 0.82 and 1.0, respectively.

mutational processes must be more important for the carcinogenicity of HCH. IntroductionHexachlorocyclohexane (HCH)* comprises of a group of isomers of which the gamma-isomer, later called lindane, has very useful pesticidal activity (1). HCH have become of great public concern because the lindane batches used in the late forties contained appreciable concentrations of alpha-and beta-isomer. The alpha-isomer was found to induce liver tumors in rats and mice (2), the beta-isomer was found to have very low biodegradability and to be deposited in animal fat. Although the lindane batches used since the fifties were at least 99% pure gamma-isomer, a new discussion arose from controversial findings of a liver tumor-inducing potential of lindane itself.Chemically-induced tumors are now thought to be the result of a DNA damage succeeded by appropriate promotion (3). Most chemicals exert their activity by covalent interaction of a reactive metabolite with DNA in the target organ and are therefore called genotoxic. The metabolism of HCH involves the formation of olefins (1) and a subsequent epoxidation could result in the generation of an electrophilic species.Another group of tumor-enhancing agents, viz cocarcinogens and promoters, do not themselves react with DNA but apparently modulate one or several out of a variety of biochemical and biological steps related to the process of tumor formation. Such activities are also discussed for HCH. For instance, alpha-HCH was found to enhance the proliferation of putative preneoplastic cells in rat liver (4), and all HCH isomers are known to be inducers of drug-metabolizing enzymes (1), the alpha-isomer being more potent than lindane.It was the aim of this study to provide more information about the mechanism of tumor induction by HCH. For this reason it was examined whether the isomers of HCH can be metabolized in vivo to reactive metabolites able to reach and bind to liver DNA or whether the hepatocarcinogenicity is rather due to non-genotoxic effects. It seemed especially worthwhile to investigate whether the clear difference between the alpha-and the gamma-isomer with respect to biological effects and tumor induction (2) was reflected in their ability to bind to DNA, and whether the apparent difference in susceptibility of different strains of mice to the carcinogenicity of gamma-HCH (5 -7) can be based upon different levels of DNA binding.

Synergistic interactions have been reported in the carcinogenicity of two epoxy resin components to mouse skin. A mixture of bisphenol A diglycidylether and bis epoxycyclopentylether was highly carcinogenic, despite the fact that neither compound gave positive results when applied individually. To elucidate the mechanism of this synergistic interaction we have investigated the effects of bis epoxycyclopentylether upon the hydrolysis and DNA-binding of bisphenol A diglycidylether. This glycidylether was rapidly hydrolysed by microsomal and cytosolic fractions of mouse liver and skin. In three different mouse strains the specific epoxide hydrolase activities were 28.3-48.5; 33.0-38.8; 7.9-10.2 and 0.85-0.98 nmol/mg protein/min for liver microsomal and cytosolic and skin microsomal and cytosolic fractions respectively. This is the first demonstration of an epoxide hydrolase activity in skin cytosolic fractions. Bis epoxycyclopentylether inhibited the microsomal activities. This inhibition appeared to be slightly more effective with microsomal fractions from liver. The effect of this inhibition upon the binding of bisphenol A diglycidylether to mouse skin DNA was investigated using bisphenol A diglycidylether radiolabelled at two different positions. When high doses of bisphenol A diglycidylether were applied to the mouse skin one major DNA adduct was observed which was identified as a glycidaldehyde adduct. This adduct was not detectable at the lowest bisphenol A diglycidylether dose tested, unless bis epoxycyclopentylether was applied simultaneously. These findings suggest that glycidaldehyde may be formed from bisphenol A diglycidylether. At low doses, however, the epoxide groups are hydrolysed before glycidaldehyde can be formed, unless the epoxide hydrolase is inhibited. Such inhibition and the associated increased production of glycidaldehyde may account for the potentiation of the carcinogenic response in the epoxide mixture.

No abstract

The formation of a glycidaldehyde-DNA adduct in skin of C3H mice treated with [14C]bisphenol A diglycidylether has been previously reported and it was assumed that the modification occurred on guanine residues. We were interested in elucidating the structure of this glycidaldehyde-DNA adduct by using a non-radioactive approach. Male C3H mice were treated with a single topical dose of 2 mg bisphenol A diglycidylether in acetone for 48, 96 or 192 h. An additional two mice were treated with 2 mg glycidaldehyde in acetone for 24 h. Epidermal DNA was isolated and enzymatically digested to nucleoside-3'-monophosphates. Aliquots of the DNA hydrolysates were separated on HPLC using a reverse-phase column with a potassium dihydrogen phosphate/methanol gradient. Fluorescence analysis of the eluent indicated the presence of a fluorescent DNA adduct, which was identified as hydroxymethylethenodeoxyadenosine-3'-monophosphate by comparison with a synthetic reference standard. Amounts of adducts were determined by fluorescence measurements using a calibration curve obtained with the authentic adduct standard. Irrespective of duration of exposure, all DNA hydrolysates of treated mice contained similar amounts of the deoxyadenosine adduct. The alkylation frequency was 0.1-0.8 and 166 adducts/10(6) normal nucleotides for the treatment with bisphenol A diglycidylether and glycidaldehyde respectively. The limit of detection using 500 micrograms DNA samples for analysis was approximately 0.03 adducts/10(6) normal nucleotides.