Mutagenicity and tumorigenicity of dihydrodiols, diol epoxides, and other derivatives of benzo(f)quinoline and benzo(h)quinoline.

The mutagenic activities of benzo[f]quinoline, benzo[h]quinoline, and a number of their derivatives, including dihydrodiols, K-region oxides, diol epoxides, and tetrahydroepoxides, were assessed in strain TA 100 of Salmonella typhimurium. The dihydrodiol derivatives of benzo[f]quinoline and benzo[h]quinoline were also tested for tumorigenic activity in newborn mice. Benzo[f]quinoline was metabolically activated in the presence of rat liver S-9 preparation to products mutagenic to the bacterial system to a greater extent than was benzo[h]quinoline. However, trans-7,8-dihydro-7,8-dihydroxybenzo[f]quinoline was less mutagenic compared to trans-7,8-dihydroxy-7,8-dihydrobenzo[h]quinoline in the presence of rat liver homogenate. The data on the mutagenic activity of the dihydrodiol derivatives of benzoquinolines were consistent with the intrinsic mutagenicity of the corresponding epoxide derivatives, in that the bay-region diol epoxides and tetrahydroepoxide of benzo[h]quinoline exhibited considerably higher mutagenic activities compared to those of the corresponding derivatives of benzo[f]quinoline at equivalent doses. The K-region oxides of benzo[f]quinoline and benzo[h]quinoline were significantly less mutagenic than their corresponding bay-region diol epoxide and tetrahydroepoxide derivatives. The demonstration that benzo[f]quinoline is significantly more mutagenic than trans-7,8-dihydro-7,8-dihydroxybenzo[f]quinoline, a precursor to the weakly mutagenic bay-region diol epoxide, suggests that the bay-region diol epoxide formation is not the principal pathway for the metabolic activation of benzo[f]quinoline to a mutagen. On the other hand, the isomeric benzo[h]quinoline appears to exert its mutagenic effect via the formation of its bay-region diol epoxide. These results indicate that the position of a nitrogen heteroatom in phenanthrene (the analogous carbocyclic aromatic hydrocarbon) not only has a marked effect on the mutagenic activities of the diol epoxide derivatives, but also can alter the metabolic activation pathways of the parent hydrocarbon. Benzo[f]quinoline, benzo[h]quinoline, and their dihydrodiol derivatives were not tumorigenic in newborn mice.

Mutagenicity of dihydrodiols and diol epoxides of dibenz[a, h]acridine in bacterial and mammalian cells.

Bay-region diol epoxides are ultimate carcinogenic metabolites of a number of polycyclic aromatic compounds. Dibenz[a, h]acridine can form two diastereomeric pairs of these diol epoxides which are not positionally equivalent as a result of the nitrogen atom at position 7. We have assessed the structure-activity relationships resulting from heterocyclic nitrogen substitution by examining the mutagenic activity of these four bay-region diol epoxides of dibenz[a,h]acridine in both bacterial and mammalian cells. In strains TA98 and TA100 of Salmonella typhimurium, the diastereomeric 10,11-diol-8,9-epoxides were 20 to 40 times more mutagenic than the corresponding 3,4-diol-1,2-epoxides. Furthermore, in strain TA100, dibenz[a,h]acridine 10,11-dihydrodiol, the expected metabolic precursor of the 10,11-diol-8,9-epoxide, was metabolically activated by rat hepatic microsomes up to a 12-fold greater extent than the 3,4-dihydrodiol. In Chinese hamster V79 cells, the 10,11-diol-8,9-epoxide diastereomers were 20 to 80 times more mutagenic than their 3,4-diol-1,2-epoxide counterparts. Quantum mechanical calculations of the predicted ease of benzylic carbocation formation at C-1 and C-8 from the diol epoxides indicate that the 3,4-diol-1,2-epoxides should be less reactive due to resonance destabilization of the C-1 carbocation as a result of the electronegative nitrogen atom. Decreased chemical reactivity of 3,4-diol-1,2-epoxides may explain their decreased mutagenic activity.

Comparative metabolism of phenanthrene and benzo[f]quinoline by rat liver microsomes.

The metabolism of benzo[f]quinoline (BfQ) and its carbon analog phenanthrene has been compared in incubations with liver microsomes from control, 3-methylcholanthrene (3-MC)- and phenobarbital (PB)-pretreated rats. The rates of phenanthrene metabolism by the three types of microsomes were 0.7, 4.1 and 1.5 nmol/mg protein per min, respectively; the values for BfQ were 0.5, 3.7 and 2.5, respectively. Besides N-oxidation, the metabolism of BfQ by all the above microsomes was almost exclusively at the benzo-ring (49-69%) while that of phenanthrene was predominantly at the K-region (50-71%). Phenanthrene-1,2-dihydrodiol, a precursor of the bay-region diol epoxide of phenanthrene, was produced many times more than phenanthrene-3,4-dihydrodiol by both 3-MC- and PB-induced microsomes. While BfQ-7,8-dihydrodiol, the precursor of the bay-region diol epoxide of BfQ, was the predominant metabolite with 3-MC-induced microsomes, it was a minor metabolite with PB-induced microsomes. The benzo-ring oxidation of BfQ, but not of phenanthrene, was position-specific, i.e. predominantly 7,8-oxidation by 3-MC-induced microsomes and 9,10-oxidation by PB-induced microsomes, and implies that aza-substitution results in a site-specific attack by different cytochromes P-450.

Formation and persistence of DNA adducts in the liver of brown bullheads exposed to benzo[a]pyrene.

The formation and persistence of benzo[a]pyrene (BP)-DNA adducts in the liver of brown bullheads (Ictalurus nebulosus) treated with the hydrocarbon (20 mg/kg body wt, i.p.) was investigated using the 32P-postlabeling assay. The highest level of covalent binding of BP to liver DNA (188 fmol BP adducts/mg DNA) was observed 25-30 days following treatment. After 70 days, the adduct level in liver DNA had declined to approximately 26% of the maximum adduct level. One major BP-DNA adduct and several minor ones were detected in the liver. The major adduct co-chromatographed with anti-BP-7,8-diol-9,10-epoxide-deoxyguanosine (anti-BPDE-dGuo) adduct. The data suggest that brown bullheads metabolically activate BP by the same mechanism as the mammalian systems susceptible to carcinogenic effects of the hydrocarbon.

Hepatic DNA adduct formation in rats treated with benzo[f]quinoline.

The formation of hepatic DNA adducts in male Sprague-Dawley rats following i.p. administration of benzo[f]quinoline (BfQ) was examined using a 32P-post-labeling assay. BfQ exhibited a low binding (11-27 amol adducts/microgram DNA) to liver DNA. Two BfQ-nucleoside adducts (one major and one minor) were detected. The BfQ-DNA adducts formed in vivo were chromatographically distinct from the adducts formed by the reaction of calf thymus DNA in vitro with BfQ-5,6-oxide, syn-7 beta,8 alpha-dihydroxy-9 beta,10 beta-epoxy-7,8,9,10-tetrahydroBfQ, anti-9 alpha,10 beta-dihydroxy-7 alpha,8 alpha-epoxy-7,8,9,10-tetrahydroBfQ, or anti-7 beta,8 alpha-dihydroxy-9 alpha,10 alpha-epoxy-7,8,9,10-tetrahydroBfQ-N- oxide. These results suggest that the bay-region diol epoxide of BfQ, unlike the bay-region diol epoxide derivatives of polynuclear aromatic hydrocarbons, is not involved in the covalent binding of BfQ to DNA.

Induction of hepatic microsomal cytochrome P-448-mediated oxidases by 3,3'-dichlorobenzidine in the rat.

Intraperitoneal administration of the hepatocarcinogen 3,3'-dichlorobenzidine (4,4'-diamino, 3,3'-dichlorobiphenyl) to adult male rats caused the induction of hepatic microsomal ethoxycoumarin O-deethylase and p-nitrophenetole O-deethylase activities comparable in magnitudes to those induced by 3-methylcholanthrene; neither aniline hydroxylase nor aminopyrine N-demethylase activity was affected by the pretreatment. The induction was not accompanied by a significant increase in content of hepatic microsomal cytochrome P-450; however, a shift in the absorption maximum of the reduced + CO spectrum of the cytochrome to 448 nm and an increase in the ratio of the 455 nm:430 nm peaks of the reduced + ethylisocyanide spectrum of the hemoprotein was effected. Arylhydrocarbon hydroxylase activity was stimulated 5-fold by dichlorobenzidine pretreatment in comparison with a 12-fold stimulation following 3-methylcholanthrene pretreatment. However, enzymically mediated covalent binding of benzo[a]pyrene to microsomal protein was greater in microsomes from dichlorobenzidine-pretreated rats than in those from methylcholanthrene-pretreated rats. All of the dichlorobenzidine-induced enzymic activities were inhibited by alpha-naphthoflavone but not by SKF-525A. Hepatic microsomes from dichlorobenzidine-pretreated rats appeared to have a higher capacity for metabolizing dichlorobenzidine than those from untreated animals; both sets of microsomes elicited the Type II spectral change on combination with the compound, albeit with different binding affinities and capacities. The results show that dichlorobenzidine, although only a dihalogenated biphenyl derivative, is a potent inducer of cytochrome P-448.

Metabolism of benzo[a]pyrene and (-)-trans-benzo[a]pyrene-7,8-dihydrodiol by freshly isolated hepatocytes of brown bullheads.

The metabolism of [3H]benzo[a]pyrene (BP) and (-)-trans-[14C]7,8-dihydroxy-7,8-dihydrobenzo[a]pyrene (BP-7,8-diol) was studied in freshly isolated hepatocytes of the wild benthic fish, brown bullhead (Ictalurus nebulosus). Bullhead hepatocytes incubated with 40 microM [3H]BP for 1 h metabolized BP to water soluble metabolites which were separated on silica gel t.l.c. plates to reveal conjugates with glucuronic acid, glutathione, and sulfate (51%, 14% and 4% of total metabolites, respectively). Additional metabolites that were extractable with ethyl acetate were separated by reversed phase HPLC to reveal only two major metabolites: BP-9,10-dihydrodiol and BP-7,8-diol (13% and 2.6% of total metabolites, respectively). Hepatocytes isolated from individual fish displayed an 11-fold variability in the rates at which they metabolized BP (756 +/- 167 pmol x mg dry wt-1 x h-1), which correlated negatively (r = -0.7, P less than 0.01) with an 18-fold variability in the glycogen content of the cells. Hepatocytes isolated from the same fish, in parallel incubations under the same optimum conditions, metabolized BP-7,8-diol 4.5-fold faster than they metabolized BP. The variability in the rate of BP-7,8-diol metabolism was about 7-fold. Major metabolites included glutathione conjugates, glucuronides and sulfates (35%, 25% and 30% of total metabolites, respectively). These conjugates, like those formed from BP, were degradable with gamma-glutamyltransferase, beta-glucuronidase and arylsulfatase, respectively. Ethyl acetate extractable metabolites were predominantly isomeric benzo-ring tetrahydrotetrols (9% of total metabolites). In summary, this study indicates that during short-term incubations bull-head hepatocytes metabolize BP and BP-7,8-diol primarily to conjugated derivatives. The usefulness of thin-layer chromatography for the convenient determination of the rate of BP-7,8-diol metabolism is demonstrated.

32P-Postlabeling analysis of lipophilic DNA adducts resulting from interaction with (+/-)-3-hydroxy-trans-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo [a]pyrene.

Bay-region diol epoxides are considered the putative ultimate carcinogens of polynuclear aromatic hydrocarbons. However, the results of studies on tumorigenesis and DNA binding of benzo[a]pyrene (BP) and its bay-region diol epoxide, (+)-trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyren e [(+)-anti-BPDE] suggest that, in addition to anti-BPDE, other reactive metabolite(s) of BP may also be involved in BP-induced carcinogenesis. Recent studies have demonstrated that 3-hydroxy-trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a ]pyrene (anti-BPTE) is another highly reactive metabolite of BP. In order to identify syn- and anti-BPTE-derived DNA adducts and their base selectivity, we synthesized both compounds by two different methods and reacted in vitro with calf thymus DNA and individual nucleotides. The resultant adducts were analyzed by nuclease P1-enhanced 32P-postlabeling. Anti-BPTE produced three major and several minor adducts with DNA; dAp and dGp were the preferred substrates, while dCp and dTp were the least reactive. In contrast, syn-BPTE produced two major adducts each with DNA and dGp; dAp generated only one adduct. Co-chromatography of anti-BPTE-derived DNA adducts with those of mononucleotide adducts revealed that the major adducts in DNA were guanine derived. Further, co-chromatographic results revealed that the anti-BPTE-DNA adducts were distinctly different from that of anti-BPDE-DNA adducts. These observations indicate that both syn- and anti-BPTE can react with DNA bases and these DNA adducts may also contribute to BP-induced carcinogenesis.

Interaction of PAH-related compounds with the alpha and beta isoforms of the estrogen receptor.

The ability of several 4- and 5-ring polycyclic aromatic hydrocarbons (PAHs), heterocyclic PAHs, and their monohydroxy derivatives to interact with the estrogen receptor (ER) alpha and beta isoforms was examined. Only compounds possessing a hydroxyl group were able to compete with 3H-labeled 17beta-estradiol (E2) for binding to either a glutathione-S-transferase and human ERalpha D, E, and F domain fusion protein (GST-hERalphadef) or to the full-length human ERbeta. Competitive binding was comparable for both isoforms, with IC(50) values ranging from 20 to 300 nM (E2 IC(50) approximately 3 nM). However, several compounds were able to induce reporter gene expression preferentially through mERbeta, using MCF-7 cells transiently transfected with either a Gal4-human ERalphadef or Gal4-mouse ERbetadef construct, as well as a Gal4-regulated reporter. These data extend the number and type of PAH-related compounds capable of interacting with ERalpha and ERbeta, and provides additional evidence that even though some compounds may possess a similar affinity for both ER isoforms, the capacity for transcriptional activation can still be isoform-specific.

Mutagenicity of dibenz[a,c]anthracene and its derivatives in Salmonella typhimurium TA100.

The mutagenic activities of dibenz[a,c]anthracene (DB[a,c]A), and its 11 derivatives, including 3 diols, 6 phenols and 2 oxepines, were studied in the TA100 strain of Salmonella typhimurium at doses varying from 0 to 20 micrograms/plate in the presence of a rat-liver S9 (9000 x g) preparation. Among the diols of DB[a,c]A tested DB[a,c]A-10,11-diol was the most mutagenic compound. However, it was consistently less mutagenic than the parent hydrocarbon. Oxepine-1 and oxepine-2 which are believed to be the photoisomerized products of DB[a,c]A-1,2 oxide and DB[a,c]A-3,4-oxide, respectively, were also less mutagenic than DB[a,c]A. In contrast to these results, 4-hydroxyDB[a,c]A was almost twice as active as DB[a,c]A, and 2-hydroxy- and 3-hydroxyDB[a,c]A were even more (4-6-fold) mutagenic than DB[a,c]A. The remaining phenols were relatively inactive or weakly active in this mutagenicity assay. These results provide initial evidence that the bay-region theory may not be applicable to the mutagenesis of DB[a,c]A, and that the angular ring substituted phenols of DB[a,c]A may be involved in the metabolic activation of this highly mutagenic hydrocarbon.