ENZYMATIC OXIDATION OF DRUGS TO TOXIC AND CARCINOGENIC METABOLITES

海湾地区环氧二醇代谢产物的致癌性多环芳烃

Bay-region diol epoxide metabolites of carcinogenic polycyclic aromatic hydrocarbons are believed to initiate cell transformation by covalent modification of DNA. We are interested in determining the relationships between the DNA adducts formed from specific diol epoxides and their biological effects, namely, mutagenesis and tumorigenesis. The most tumorigenic of the four bay-region diol epoxides formed metabolically from benzo[a]pyrene (BaP) is the 7R,8S-diol 9S,10R-epoxide-2 (benzylic hydroxyl group and epoxide oxygen trans). This diol epoxide forms DNA adducts almost exclusively at the exocyclic N-2 amino groups of deoxyguanosine (dG) residues. In contrast, two of the bay-region diol epoxides of enzo[c]phenanthrene (BcPh), the 4R,3S-diol 2S,1R-epoxide-2 and the 4S,3R-diol 2S,1R-epoxide-1 (benzylic hydroxyl group and epoxide oxygen cis) are highly tumorigenic on mouse skin. Furthermore, the BcPh diol epoxides react extensively with the exocyclic N-6 amino groups of deoxyadenosine (dA) residues. Work during the past year has focused on comparison of the carcinogenic diol epoxides from these two hydrocarbons with respect to both their chemistry and biology. (1) The effect of DNA structure on the partitioning of the BaP and BcPh diol epoxides between dG and dA adduct formation was investigated by comparing the products of their reactions with DNA and with an equimolar mixture of the nucleotides dAMP and dGMP. For three of the four BcPh diol epoxide isomers, differences in the dA/dG adduct distributions on reaction with native DNA in vitro and with the nucleotide mixture were small: 40-60% of the total exocyclic N-adducts were derived from dA upon reaction with either native DNA or the nucleotides. An exception was the 4S,3R-diol 2S,1R-epoxide-1 isomer, whose dA selectivity was much higher with native DNA (85%) than with the free nucleotides (50%). In the presence of the nucleotide mixture, the BaP diol epoxides exhibited a slight to moderate preference for dG adduct formation (60-77% of total exocyclic adducts formed). For three of the isomers, and most markedly for the carcinogenic BaP 7R,8S-diol 9S,10R-epoxide-2, the preference for dG was greatly enhanced by DNA (97% dG adducts as contrasted with 66% in the presence of the nucleotide mixture). We conclude that the dG selectivity of the BaP diol epoxides upon reaction with DNA results from a strong effect of DNA structure that contrasts with a relatively small effect on the dA selectivity of the BcPh diol epoxides. Reaction of denatured relative to native DNA generally resulted in decreased dA selectivity of the BcPh diol epoxides, and had little effect on selectivity for exocyclic N-adduct formation from BaP diol epoxides. (2) Modification of mouse skin DNA in vivo by the four BcPh diol epoxide isomers was assayed by P-32 postlabeling. For three of the four isomers, the total extent of binding to dA and dG was parallel to their tumorigenic response. Strikingly, however, the highly tumorigenic 4S,3R-diol 2S,1R-epoxide-1 exhibited 3- to 4-fold lower overall binding to DNA than the 4R,3S-diol 2S,1R-epoxide-2 isomer, which has comparable tumorigenic activity. Thus, specific lesions formed by 4S,3R-diol 2S,1R-epoxide-1 may have a greater likelihood of producing a tumorigenic response. The relationship of this finding to the very high dA selectivity (see above) of this isomer is intriguing but remains to be explained. (3) Previous studies of mutational profiles in Chinese hamster V-79 cells had shown that the distribution of mutations produced by the most carcinogenic BaP diol epoxide isomer is strongly dose-dependent. We have now extended these studies to the highly tumorigenic BcPh 4R,3S-diol 2S,1R-epoxide-2 and its relatively inactive 4S,3R-diol 2R,1S-epoxide-2 enantiomer. With a high, cytotoxic dose of the more tumorigenic diol epoxide isomer, 69% of the observed mutations were at A-T base pairs, whereas with a low, noncytotoxic dose the relative proportion of mutations at dA was increased to 89%. In contrast, both high and low doses of the less active isomer gave a 50:50 distribution of mutations at A-T and G-C base pairs. These results parallel those previously reported for the corresponding stereoisomers of BaP diol epoxide-2, which had also demonstrated a marked increase in mutations at dA relative to dG at low doses (such as those expected to result from environmental exposure to the hydrocarbon) for the tumorigenic but not for the nontumorigenic enantiomer. Taken together, all of these results suggest that lesions at dA residues of DNA may prove to be highly significant to the eventual understanding of carcinogenesis induced by polycyclic aromatic hydrocarbons.