Identification of nucleic acid adducts from trans-4-acetylaminostilbene

The structure and function of the H-ras proto-oncogene are not altered in rat liver tumors initiated by 2-acetylaminofluorene, 2-acetylaminophenanthrene and trans-4-acetylaminostilbene

Liver tumors were generated in Wistar rats in an initiation-promotion experiment. 2-Acetylaminofluorene (AAF), 2-acetylaminophenanthrene (AAP), and trans-4-acetylaminostilbene (AAS) were administered to newborn animals as initiators, and phenobarbital as a promoter was added to the drinking water after weaning. Livers were examined after 26, 52, 78, and 104 weeks. Tumors were present in all groups except for at the first time point. The potency of the initiators decreased in the order AAS > AAP > AAF. DNA from tumors of all groups and of control livers was analyzed for mutations in the H-ras gene, but no mutations could be found. The sequence of almost the entire H-ras gene was determined and was compared to other H-ras genes. There are some differences with the sequence in other rat strains, particularly in intron D containing the alternative splicing site. The expression of the H-ras gene has also been studied by various methods in enzyme altered foci and tumors, but no alterations could be found. It is, therefore, concluded that structural of functional alterations of this proto-oncogene are not involved in the generation of liver tumors in Wistar rats by the three genotoxic arylamines.

The tentative identification of DNA-adducts generated by trans-4-dimethylaminostilbene and trans-4-acetylaminostilbene in rats

Tritium-labeled trans-4-dimethylaminostilbene (DAS) and trans-4-acetylaminostilbene (AAS) were administered orally to female Wistar rats. RNA and DNA were isolated from livers after 24 h and 28 days. Hydrolysates were analyzed by gel filtration and HPLC. Total binding to nucleic acids and elution profiles of hydrolysates were very similar for both aminostilbene derivatives. The large polar fraction eluting early in Sephadex LH20 chromatograms accounting for 60-80% of total DNA-bound radioactivity could not be assigned to individual base adducts and most likely is not due to incomplete hydrolysis but rather to cross-links between bases or proteins and bases. Of the total radioactivity bound to nucleic acids 6-7% in RNA and 3-5% in DNA could be tentatively identified as four isomeric, cyclic guanine adducts (predominantly (S,S)- and (R,R)-guanine-N2,beta-N3,alpha-N-acetylaminobibenzyl) by cochromatography with synthetic reference compounds. The most abundant single adduct accounting for 20-30% of RNA-bound radioactivity (fraction G in Sephadex LH20 chromatography) could not be identified. The long-term experiment revealed different persistence of DNA adducts: the polar material decreased to about 2/3, the cyclic guanine adducts (fraction d-B) to about 1/3 to 1/2 within 4 weeks, whereas one of the unidentified DNA-adducts (fraction d-E) persisted completely. AAS labeled in the acetyl group was administered in an additional experiment. The presence of the acetyl group could be demonstrated in most of the adducts, but non-acetylated adducts were found also. The ratio of non-acetylated:acetylated cyclic B-adducts in RNA was 1:2 from DAS and 1:13 from AAS, in DNA 1:3 and 1:10, respectively.

Roles of etheno-DNA adducts in tumorigenicity of olefins

Cyclic DNA adducts bearing an "etheno" structure have been described to occur after interaction with metabolites of halogenated olefins. Extensive work has been published on adducts of vinyl chloride, both in vitro and in vivo. The major DNA adduct of vinyl chloride is 7-(2-oxoethyl)guanine, but an important minor adduct appears to be N2,3-ethenoguanine. Other etheno adducts, i.e., 1, N6-ethenoadenine and 3, N4-ethenocytosine, are readily formed with DNA, vinyl chloride, and a metabolizing system in vitro and with RNA in vivo, but usually are not detected as DNA adducts in vivo. Other compounds that have been studied with respect to possible formation of etheno DNA adducts are vinyl bromide (which is more or less completely analogous to vinyl chloride), acrylonitrile, vinyl acetate and vinyl carbamate. Proposals of possible structures of DNA adducts with an etheno structure have been promutagenic potential of these lesions which may lead to misincorporation of wrong DNA bases in newly synthesized DNA.

Reaction of trans-4-N-acetoxy-N-acetylaminostilbene with guanosine and deoxyguanosine in vitro: the primary reaction product at N2 of guanine yields different final adducts

The model ultimate carcinogen, trans-4-N-acetoxy-N-acetylaminostilbene (N-acetoxy-AAS), was reacted with guanosine (Guo) and deoxyguanosine (d-Guo) and the resulting adducts were purified by Sephadex LH-20 chromatography and HPLC for structure identification. A number of new adducts was identified by mass and 1H-NMR spectroscopy. The generation of all known adducts can now be explained by a common mechanism. The electrophile formed from the hydroxamic acid ester at C-beta reacts in a first step predominantly with N2 of guanine (Gua). The resulting quinone-imide intermediate reacts in a second step with either one of three nucleophiles: (1) predominantly with N3 of Gua to yield the previously described angular cyclic adducts ((5R,6R)/(5S,6S)-9-oxo-5,6,7,9-tetrahydro-imidazo(2,1-b)purines); (2) with N1 of Gua to yield linear cyclic adducts ((6R,7R)/(6S,7S)-9-oxo-5,6,7,9-tetrahydro-imidazo(1,2-a)purines); (3) with water to yield the open ring (1R,2R)/(1S,2S)-2-(N2'-guanyl)-1-hydroxyethanes. To some minor extent (1:8-1:9) the electrophile reacts first with N1 or N3 of guanine which leads to the formation of two pairs of the corresponding regioisomeric cyclic adducts. This reaction mechanism may also explain the formation of cross-links between different bases.

Reaction of trans-4-N-acetoxy-N-acetylaminostilbene with guanosine, deoxyguanosine, RNA and DNA in vitro: predominant product is a cyclic N2,N3-guanine adduct

The model ultimate carcinogen trans-4-N-acetoxy-N-acetylaminostilbene was reacted with guanosine, deoxyguanosine, RNA and DNA using differently labeled reactants. The nucleoside as well as the deoxynucleoside yielded predominantly four cyclic guanine adducts: (S,S)- and (R,R)-guanine-N2,beta-N3,alpha-N-acetyl-aminobibenzyl and the regioisomers with the N2,alpha-N3,beta-attachment in a ratio of 9:9:1:1. The same adducts predominate in RNA and DNA which demonstrates that guanine reacts most avidly among the bases. The stability of the N-glycosidic bond is quite different between ribosides and deoxyribosides. Under neutral conditions, the riboside derivatives are stable, whereas deoxyribose is cleaved off rather readily. As a consequence DNA depurinizes to some extent during the in vitro reaction and during enzymatic digestion. On the other hand, N2,N3-attachment of the acetylaminostilbene moiety to guanine appears to impair the activity of nucleases for steric reasons. This could explain the incomplete enzymatic hydrolysis of modified nucleic acids. The results provide an important basis for further investigations to identify the nucleic acid adducts generated in vivo.

Induction of single-strand breaks and interstrand cross-links in liver DNA after the administration of 2-acetylaminofluorene and trans-4-acetylaminostilbene to rats

2-Acetylaminofluorene (AAF) or trans-4-acetylaminostilbene (AAS) was orally or intraperitoneally administered to female Wistar rats. DNA from liver cells was analyzed for single-strand breaks by the alkaline elution assay. Only borderline effects were observed with doses (100 mumol/kg) used in animal carcinogenesis experiments. Even high doses of AAF (1,000 mumol/kg) were not effective. Methyl methanesulfonate (MMS) in vivo and gamma irradiation in vitro were shown to produce dose-dependent DNA single-strand breaks (positive control). Only a marginal effect was obtained with 100 mumol/kg MMS. The elution rate of DNA was increased by a factor of 34 in liver cells in vitro with 400 rad of gamma irradiation. Only a fraction of this rate could be demonstrated immediately after irradiation in vivo, and no lesions were found two hours later. This strongly indicates the rapid repair of single-strand breaks. Additional experiments showed that AAS, a nonhepatocarcinogen, produced more interstrand cross-links in the rat liver DNA than did AAF.

The role of DNA damage in chemical carcinogenesis of aromatic amines

Many findings support the notion that the generation of DNA adducts by aromatic amines is causally related to carcinogenesis. Adducts have been identified in most cases and representative examples are reviewed. However, extent and persistence of DNA adducts (DNA dose) does not correlate satisfactorily with the tumor response of different tissues. Distribution of DNA damage, repair, indirect and secondary DNA damage are discussed as possible explanations for the observed noncorrelations. In addition, however, it is proposed to pay attention to specific mechanisms such as receptor mediated cellular effects which are not related to the generation of electrophiles. The effects of trans-4-aminostilbene and 2-amino-fluorene derivatives on rat liver are compared. It is concluded that trans-4-acetylamino-stilbene is a strong liver tumor initiator but an incomplete liver carcinogen lacking tumor promoting properties, and that 2-acetylaminofluorene is a complete liver carcinogen with initiating and promoting properties.