Mutagenicity and tumorigenicity of the four enantiopure bay-region 3,4-diol-1,2-epoxide isomers of dibenz[a,h]anthracene

Abstract

Each enantiomer of the diastereomeric pair of bay-region dibenz[a,h]anthracene 3,4-diol-1,2-epoxides in which the benzylic 4-hydroxyl group and epoxide oxygen are either cis (isomer 1) or trans (isomer 2) were evaluated for mutagenic activity. In strains TA 98 and TA 100 of Salmonella typhimurium, the diol epoxide with (1S,2R,3S,4R) absolute configuration [(–)-diol epoxide-1] had the highest mutagenic activity. In Chinese hamster V-79 cells, the diol epoxide with (1R,2S,3S,4R) absolute configuration [(+)-diol epoxide-2] had the highest mutagenic activity. The (1R,2S,3R,4S) diol epoxide [(+)-diol epoxide-1] also had appreciable activity, whereas the other two bay-region diol epoxide enantiomers had very low activity. In tumor studies, the (1R,2S,3S,4R) enantiomer was the only diol epoxide isomer tested that had strong activity as a tumor initiator on mouse skin and in causing lung and liver tumors when injected into newborn mice. This stereoisomer was about one-third as active as the parent hydrocarbon, dibenz[a,h]anthracene as a tumor initiator on mouse skin; it was several-fold more active than dibenz[a,h]anthracene as a lung and liver carcinogen when injected into newborn mice. (–)-(3R,4R)-3β,4α-dihydroxy-3,4-dihydro-dibenz[a,h]anthracene [(–)-3,4-dihydrodiol] was slightly more active than dibenz[a,h]anthracene as a tumor initiator on mouse skin, whereas (+)-(3S,4S)-3α,4β-dihydroxy-3,4-dihydro-dibenz[a,h]anthracene [(+)-3,4-dihydrodiol] had only very weak activity. The present investigation and previous studies with the corresponding four possible enantiopure bay-region diol epoxide enantiomers/diastereomers of benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[c]phenanthrene, dibenz[c,h]acridine, dibenz[a,h]acridine and dibenz[a,h]anthracene indicate that the bay-region diol epoxide enantiomer with [R,S,S,R] absolute stereochemistry has high tumorigenic activity on mouse skin and in newborn mice.

Introduction

The first demonstration of a pure chemical that induces neoplasia was reported in 1930 when Kennaway and Hieger demonstrated that the polycyclic aromatic hydrocarbon dibenz[a,h]anthracene (DBA) was carcinogenic on mouse skin (1). Numerous studies since the early work by Kennaway and Hieger have indicated that polycyclic aromatic hydrocarbons and other cancer-causing chemicals exert their carcinogenic activity only after metabolism to highly reactive products (ultimate carcinogens) capable of covalent binding to macromolecules (2–5).

Evidence for the metabolic activation of a polycyclic aromatic hydrocarbon to reactive intermediates was first reported in 1951 by E.C.Miller. She applied benzo[a]pyrene to mouse skin and found covalent binding of this hydrocarbon to skin protein (6). The first demonstration of proximate and ultimate carcinogenic metabolites of a polycyclic aromatic hydrocarbon came from finding that racemic benzo[a]pyrene 7,8-dihydrodiol and benzo[a]pyrene 7,8-diol-9,10-epoxide-2 were, respectively, about 12- and 40-fold more tumorigenic than benzo[a]pyrene in newborn mice (7,8).

Bay-region diol epoxides are the major known ultimate carcinogenic metabolites of polycyclic aromatic hydrocarbons (9–14). The pathway for metabolic formation of bay-region diol epoxides (Scheme 1) involves oxidation of a terminal angular benzo ring of the polycyclic aromatic hydrocarbon (A) by a cytochrome P450 monooxygenase to form an arene oxide (B), hydration of the arene oxide by epoxide hydrolase to yield a trans-dihydrodiol (C) and subsequent oxidation of the bay-region alkene bond of the dihydrodiol by cytochrome P450 to form the diol epoxides (D). The metabolic sequence resulting in the formation of the diol epoxides 1 and 2 from DBA are shown by the partial structures in Scheme 1. The chemically synthesized 3,4-arene oxide enantiomers (B) of DBA (A) were found to spontaneously racemize at ambient temperature in vitro. It is probable that the initially formed arene oxide metabolite of DBA in vivo will also undergo a very rapid racemization process via the corresponding oxepine valence tautomer (15). This racemization process could in turn have an important influence on the formation of both enantiomers of the trans-3,4-dihydrodiol (60%ee), which was found to be present as the principal metabolite (16), and formation of the corresponding ultimate diol epoxide carcinogens.

Scheme 1.

Metabolism of DBA (A) via arene oxide (B) and trans-dihydrodiol (C) intermediates to yield diol epoxides (D).

The diol epoxides exist as a pair of diastereomers in which the benzylic hydroxyl group is either cis (diol epoxide-1) or trans (diol epoxide-2) to the epoxide oxygen. Because each diastereomer can be resolved into a pair of enantiomers, a total of four bay-region diol epoxide stereoisomers (enantiomers/diastereomers) with different absolute configurations are possible following regioselective oxidation at one bond within each angular polycyclic aromatic hydrocarbon discussed herein. Mutagenicity and tumorigenicity studies have revealed markedly different biological activities among the enantiopure bay-region diol epoxide isomers of polycyclic aromatic hydrocarbons for benzo[a]pyrene, chrysene, benz[a]anthracene and benzo[c]phenanthrene (17–25), as well as for the nitrogen heterocycles dibenz[c,h]acridine and dibenz[a,h]acridine (26–28). For each of these hydrocarbons, the bay-region diol epoxide isomer with [R,S,S,R]-absolute configuration had high tumorigenic activity, whereas the other three bay-region stereoisomers from each hydrocarbon were essentially non-tumorigenic at the doses tested except for (+)-(IR, 2S, 3R, 4S)-3,4-dihydroxy-1,2-epoxy-1,2,3,4-tetrahydrobenzo(c)phenanthrene that had substantial tumorigenic activity on mouse skin (25). The [R,S,S,R]-diol epoxide is also the metabolically predominant isomer formed from benzo[a]pyrene and chrysene (29–31).

In earlier studies, we found that racemic DBA trans -3,4-dihydrodiol was metabolically activated to bacterial mutagens by rat liver microsomes to a much greater extent than the racemic trans-1,2- and 5,6-dihydrodiols of DBA (32). In additional studies, we found that racemic DBA trans-3,4-dihydrodiol was strongly active as a tumor initiator on mouse skin and in causing lung tumors when injected into newborn mice (33). The trans-1,2- and 5,6-dihydrodiols had little or no tumorigenic activity (33). Hydrogenation of the double bond at the 1,2-position of DBA trans-3,4-dihydrodiol resulted in the formation of a trans-tetrahydrodiol with markedly less tumorigenic activity than DBA trans-3,4-dihydrodiol on mouse skin (33). These results suggest that the DBA trans-3,4-dihydrodiol is a proximate carcinogenic metabolite and that a DBA 3,4-diol-1,2-epoxide is an ultimate carcinogenic metabolite of DBA. In this study, we have evaluated the mutagenic activities of the four 3,4-diol-1,2-epoxide enantiomers/diastereomers of DBA in Salmonella typhimurium and in Chinese hamster V-79 cells. We have also evaluated the tumorigenic activity of the enantiopure diol epoxides on mouse skin and in newborn mice as well as the tumorigenic activities of the two DBA trans-3,4-dihydrodiol enantiomers on mouse skin. The results indicate strong mutagenic activity in Chinese hamster V-79 cells and strong tumorigenic activity on mouse skin and in newborn mice for the R,S,S,R isomer of the bay-region DBA 3,4-diol-1,2-epoxide.

Materials and methods

Chemicals

DBA was obtained from Eastman Organic Chemicals, Rochester, NY. The pure (+)- and (–)-enantiomers of DBA trans-3,4-dihydrodiol and both enantiomers from the diastereomeric pair of DBA 3,4-diol-1,2-epoxides of each trans dihydrodiol were synthesized by a similar procedure to that used previously for benz[a]anthracene and chrysene (34,35). All stereoisomers were >99% chemically and enantiomerically pure. Complete structures and absolute configurations of the compounds studied are illustrated in Figure 1. Compounds were dissolved in dimethyl sulfoxide (DMSO) or acetone and stored in amber vials at −80°C. All manipulations of the compounds were under subdued light. Spectral grade acetone was obtained from Burdick and Jackson, Muskegon, MI, and DMSO was distilled from calcium hydride under reduced pressure and stored under an atmosphere of argon in amber bottles.

Fig. 1.

Absolute configuration of the four enantiopure isomers of the diastereomeric bay-region diol epoxides and two enantiopure isomers of the 3,4-dihydrodiols of DBA.

Mutagenesis assay with bacteria

Strains TA98 and TA100 of S. typhimurium (36) were obtained from Dr B. Ames (University of California, Berkeley) and cultured as described (37). The diol epoxides were added in 15 μl of anhydrous DMSO to 2 × 10⁸ bacteria suspended in 0.5ml of phosphate-buffered saline (5mM potassium phosphate−150mM sodium chloride, pH 7.0). Assay mixtures were incubated for 5min at 37°C before plating on minimal agar medium. Mutations from histidine-dependence to histidine-independence were assessed 48h after plating by counting the macroscopic colonies of bacteria on the Petri dishes. Mutation frequencies, obtained from three replicates at four different dose levels of epoxide in each experiment, were calculated as described previously (38).

Mutagenesis assays with mammalian cells

The Chinese hamster cell line V-79-4, which lacks the capacity to oxidatively metabolize polycyclic aromatic hydrocarbons to mutagenic products, was obtained from the American Type Culture Collection (Rockville, MD). Resistance to the lethal effects of the purine analog, 8-azaguanine, was used as the mutagenic marker. Procedures for inducing 8-azaguanine-resistant variants were adapted from Chu (39), and the conditions used were as described previously (37). Diol epoxides were added in 20 μl of anhydrous DMSO to cell monolayers that were growing in 5ml of culture medium. In each of three separate experiments for each compound, 4 and 16 replicate dishes were used to assess cell survival and 8-azaguanine resistance, respectively.

Tumorigenicity studies on mouse skin

Female CD-1 mice at 6 weeks of age (Charles River Breeding Laboratories, Kingston, NY) were housed in polycarbonate boxes with corn cob bedding and were fed a commercial diet (Purina Laboratory Chow 5001; Ralston Purina Co., St Louis, MO) and water ad libitum. At 7 weeks of age, the mice were shaved on the dorsal surface with electric clippers. Two days later, 30 mice in each treatment group were given a single topical application of compound (50 or 250 nmol) in 200 μl of acetone. Control mice received solvent. The tumor promoter 12-O-tetradecanoylphorbol-13-acetate (16 nmol/200 μl of acetone) was applied topically twice weekly to each mouse, beginning 9 days after application of the initiator or solvent. Development of skin tumors was recorded every 2 weeks, and papillomas >2mm in diameter were included in the cumulative total when they persisted for 2 weeks or longer.

Tumorigenicity studies in newborn mice

CD-1 pregnant mice (Charles River Breeding Laboratories, Kingston, NY) were housed in plastic cages on corn cob bedding. They delivered their litters from 5 to 8 days after arrival. Within 24h of birth, 10 pups in each litter were given an intraperitoneal injection of compound (one-seventh of the total dose). Subsequent injections were given on the 8th and 15th days of life (two-seventh and four-seventh of the total dose, respectively). The mice were administered a total dose of 50 or 150 nmol of compound. Control mice were given three injections of DMSO (5,10 and 20 μl). The mice were weaned at 25 days of age and killed at 32–36 weeks of age. At necropsy, the major organs of each animal were examined grossly, tumors were counted and tissues were fixed in 10% phosphate-buffered formalin. Visible nodules on the lung and liver were counted as tumors as described previously (21,23–25). A representative number of pulmonary tumors and all hepatic tumors were examined histologically. Pathology of the lung tumors (adenomas) and liver tumors (hepatomas) was the same as described previously (40–42).

Results

Mutagenic activity of enantiopure DBA diol epoxides in bacterial and mammalian cells

The mutagenicity of the four enantiopure diol epoxides of DBA in strains TA98 and TA100 of S. typhimurium and in Chinese hamster V-79 cells is illustrated in Table I. In strain TA98, the (–)-diol epoxide-1 induced an average of 780 revertants/nmol, whereas the (+)-diol epoxide-1, (–)-diol epoxide-2 and (+)-diol epoxide-2 isomers had 51, 44 and 20% of this activity, respectively. In strain TA100, the (–)-diol epoxide-1 was again the most active compound and induced 2400 revertants/nmol. The (+)-diol epoxide-1, (–)-diol epoxide-2 and (+)-diol epoxide-2 isomers were 57, 24 and 41% as mutagenic, respectively. In Chinese hamster V-79 cells, (+)-diol epoxide-2 (the R,S,S,R isomer) was the most mutagenic compound as it induced the largest number (28.8) of 8-azaguanine resistant variants/10⁵ survivors/nmol. The activity of the (+)-diol epoxide-2 isomer was ~40% higher than the activity observed with (+)-diol epoxide-1 and ~8-fold higher than the activity observed with the other two diol epoxide isomers.

Table I.

Mutagenicity of enantiopure stereoisomers of bay-region diol epoxides

Compound	Mutagenic activity
	TA98a	TA100a	V-79b
(–)-[1S,2R,3S,4R]- diol epoxide-1	780 (100%)c	2400 (100%)c	3.8 (13%)c
(+)-[1R,2S,3R,4S]-diol epoxide-1	397 (51%)	1368 (57%)	20.5 (71%)
(–)-[1S,2R,3R,4S]-diol epoxide-2	343 (44%)	567 (24%)	3.3 (11%)
(+)-[1R,2S,3S,4R]-diol epoxide-2	156 (20%)	984 (41%)	28.8 (100%)

^aIn dose–response studies with two strains of histidine-dependent S. typhimurium, the number of His⁺ revertants per nanomole of diol epoxide was determined. Each value represents the mean from two experiments. Similar results (relative activities of the compounds) were obtained in each experiment.

^bIn dose–response studies with Chinese hamster V-79 cells, the number of 8-azaguanine resistant colonies per 10⁵ surviving cells per nanomole of diol epoxide was determined. Each value represents the mean from three experiments. Similar results (relative activities of the compounds) were obtained in each experiment.

^cRelative mutagenic activity.

Tumor-initiating activity of DBA trans-dihydrodiol enantiomers and diol epoxide enantiomers/diastereomers on mouse skin

The tumor-initiating activity of DBA, the (+)- and (–)-enantiomers of DBA trans-3,4 dihydrodiol and the (+)- and (–)-enantiomers of the two diastereomeric DBA 3,4-diol-1,2-epoxides 1 and 2 on mouse skin are shown in Table II. A single initiating dose of 50 or 250 nmol of DBA produced an average of 0.90 and 2.83 tumors per mouse, respectively, and a 41 and 69% tumor incidence, respectively. At the same initiating doses, (–)-DBA trans-3,4-dihydrodiol induced skin tumors in 50 and 63% of the mice, respectively, and it produced 1.40 and 3.53 tumors per mouse, respectively. (–)-DBA trans-3,4-diol was at least 10-fold more active as a tumor initiator than its (+)-enantiomer, and it had about the same activity as its parent hydrocarbon, DBA.

Table II.

Tumor-initiating activity of DB[a,h]A and its derivatives on mouse skin

Initiator	Dose (nmol)	Percent of mice with tumors	Tumors per mouse
Solvent		3	0.03±0.03a
DB[a,h]A	50	41*	0.90±0.26**
	250	69*	2.83±0.67**
(–)-DB[a,h]A (3R,4R)-dihydrodiol	50	50*	1.40±0.30**
	250	63*	3.53±0.92**
(+)-DB[a,h]A (3S,4S)-dihydrodiol	50	10	0.20±0.12**
	250	17*	0.17±0.07
(±)-DB[a,h]A 3,4-diol-1,2-epoxide-1	50	13	0.13±0.06
	250	10	0.14±0.08
(±)-DB[a,h]A 3,4-diol-1,2-epoxide-2	50	11	0.21±0.13
	250	24*	0.31±0.11**
(–)-(1S,2R,3S,4R)-diol epoxide-1	50	10	0.10±0.06
	250	10	0.10±0.06
(+)-(1R,2S,3R,4S)-diol epoxide-1	50	3	0.03±0.03
	250	3	0.07±0.07
(–)-(1S,2R,3R,4S)-diol epoxide-2	50	3	0.03±0.03
	250	3	0.03±0.03
(+)-(1R,2S,3S,4R)-diol epoxide-2	50	21*	0.28±0.11**
	250	53*	1.13±0.27**

The indicated dose of compound was applied once to the dorsal skin of female CD-1 mice and 16 nmol of 12-O-tetradecanoyl phobol-13-acetate were administered topically twice weekly for 20 weeks beginning 10 days after initiation treatment group consisted of 30 mice and at least 28 mice in each group survived to termination of the study.

^aMean ± SE.

^* P < 0.05 compared with solvent-treatment controls as determined by Fisher’s Exact test.

^** P < 0.05 compared with solvent-treatment controls as determined by Student’s t-test.

A comparison of the tumor-initiating activity of the four enantiopure bay-region 3,4-diol-1,2 epoxide isomers of DBA revealed that only (+)-DBA 3,4-diol-1,2-epoxide-2 had significant tumorigenic activity (Table II). A single dose of 50 or 250 nmol of (+)-DBA 3,4-diol-1,2-epoxide-2 produced skin tumors in 21 and 53% of the mice, with an average of 0.28 and 1.13 tumors per mouse, respectively (Table II). The tumor-initiating activity of this diol epoxide was about one-third of that of the parent hydrocarbon. The remaining three enantiopure diol epoxide isomers had no substantial activity at these doses. The racemic bay-region diol epoxides were also evaluated for tumor-initiating activity on mouse skin. Racemic DBA diol epoxide-1 had little or no tumorigenic activity, and the racemic diol epoxide-2 had activity, which was lower than the parent hydrocarbon.

Tumorigenic activity of DBA diol epoxide enantiomers in newborn mice

The tumorigenic activities of DBA and the four enantiomers/diastereomers of the bay-region diol epoxides of DBA in newborn mice are shown in Table III. At the termination of the study, 7% of control mice had developed pulmonary tumors, with an average of 0.07 tumors per mouse. When tested at a total dose of 50 or 150 nmol, DBA produced a significant incidence of lung tumors (65 and 78%, respectively) with an average of 1.41 and 4.19 tumors per mouse, respectively. Of the four stereoisomers of DBA 3,4-diol-1,2 epoxide, only the (+)-diol epoxide-2 with [R,S,S,R] absolute stereochemistry had high pulmonary tumorigenic activity. Treatment of the mice with 50 or 150 nmol of this bay-region [R,S,S,R]-diol epoxide isomer produced an 89–100% incidence of pulmonary tumors and an average of 3.8 tumors per mouse for the low-dose group and 20.9 tumors per mouse for the high-dose group. (–)-Diol epoxide-2 and (+)-diol epoxide-1 were weakly active for causing pulmonary tumors. In newborn mice, the (+)-diol epoxide-2 was ~3- to 5-fold more active than DBA and >10-fold more tumorigenic than the other 3-diol epoxide isomers when lung tumor multiplicities were compared (Table III). DBA and (+)-diol epoxide-2 were the only compounds tested that also produced a significant incidence of hepatic tumors in male mice, and (+)-diol epoxide-2 was many-fold more active than DBA in causing liver tumors (Table III).

Table III.

Tumorigenicity of DB[a,h]A and its derivatives in newborn mice

Compound	Total dose (nmol)	Number of mice injected (day 1)	Number of mice at weaning (day 25)	Sex and number of mice alive at termination		Pulmonary tumors		Hepatic tumors
						Percent of mice with tumors	Tumors/ mouse	Percent of mice with tumors	Tumors/ mousea
Solvent		60	59	Female	29	10	0.10±0.31	0	0
				Male	26	4	0.04±0.19	0	0
				Total	55	7	0.07±0.26
DB[a,h]A	50	60	58	Female	25	72	1.44±1.68	0	0
				Male	29	59	1.38±1.85	21*	0.38
				Total	54	65**	1.41±1.77
	150	60	57	Female	22	73	4.00±5.24	0	0
				Male	29	86	4.34±4.41	72**	5.55
				Total	51	78**	4.19±4.79
(–)-DB[a,h]A (1S,2R,3S,4R)-diol epoxide-1	50	60	53	Female	25	15	0.28±0.85	0	0
				Male	22	4	0.05±0.21	0***	0
				Total	47	10†	0.17±0.63
	150	60	43	Female	19	5	0.05±0.22	0	0
				Male	22	5	0.05±0.21	0†	0
				Total	41	5†	0.05±0.21
(+)-DB[a,h]A (1R,2S,3R,4S)-diol epoxide-1	50	60	56	Female	27	22	0.33±0.72	0	0
				Male	27	22	0.22±0.42	4††	0.04
				Total	54	22†,†††	0.28±0.59
	150	60	52	Female	25	32	0.44±0.60b	0	0
				Male	22	14	0.18±0.40b	9†	0.09
				Total	47	23†,†††	0.32±0.60b
(–)-DB[a,h]A (1S,2R,3R,4S)-diol epoxide-2	50	60	57	Female	27	11	0.15±0.45	0	0
				Male	24	13	0.51±2.00	0‡	0
				Total	51	12†	0.31±1.42
	150	60	55	Female	31	26	0.38±0.74	0	0
				Male	24	17	0.83±3.02	8†	0.08
				Total	55	22†,†††	0.57±2.07
(+)-DB[a,h]A (1R,2S,3S,4R)-diol epoxide-2	50	60	53	Female	25	92	4.64±3.51	0	0
				Male	23	86	2.93±2.27	74**,†	16.22
				Total	48	89**,‡‡	3.78±3.09
	150	70	55	Female	20	100	22.5±13.6	5	0.15
				Male	24	100	19.6±13.9	96**,‡‡‡	14.63
				Total	44	100**,§	20.9±13.8

Groups of 60 or 70 CD-1 mice received intraperitoneal injections of compounds on the 1st, 8th and 15th day of life.

Control animals received injections of DMSO (solvent). The animals were killed at 32–36 weeks of age.

Data are expressed as the mean ± SE. Statistical analysis of pulmonary tumor data was done for combined males and females since there was no sex difference. For hepatic tumors, only males responded. The two-sample unequal varience t-test was used to test the mean difference in tumor numbers between two treatments. The Pearson Chi-square test was used to compare the difference of percentages of mice with tumors.

^aStandard error values were not recorded.

^bStandard error values for these animals were lost and are an estimate.

^* P = 0.014 compared with vehicle control.

^** P < 0.001 compared with vehicle control.

^*** P = 0.0231 compared with the DBA group.

^†P < 0.001 compared with DBA group.

^†† P = 0.0548 compared with DBA group.

^††† P < 0.05 compared with vehicle control.

^‡ P = 0.0180 compared with the DBA group.

^‡‡ P = 0.0032 compared with DBA group.

^‡‡‡ P = 0.0238 compared with the DBA group.

^§ P = 0.0011 compared with DBA group.

Discussion

The results obtained with the (+)- and (–)-enantiomers of trans-DBA 3,4-dihydrodiol and the enantiomers of the diastereomeric DBA 3,4 diol-1,2-epoxides on mouse skin and in newborn mice indicated a marked difference in their tumorigenic activities. Of these benzo-ring derivatives of DBA, only (–)-DBA trans-3,4-dihydrodiol and its major metabolite [(+)-DBA 3,4-diol-1,2-epoxide-2] had high tumor-initiating activity on mouse skin. (–)-DBA trans-3,4 dihydrodiol was 3- to 5-fold more active than (+)-DBA 3,4-diol-1,2-epoxide-2 as a tumor initiator on mouse skin. Of the four bay-region diol epoxide enantiomers/diastereomers, only (+)-DBA 3,4-diol-1,2-epoxide-2 with R,S,S,R absolute stereochemistry had high tumorigenic activity on mouse skin and in newborn mice at the doses tested. This diol epoxide had many-fold greater tumor-initiating activity on mouse skin than did the other three stereoisomers and was >10 times more potent in inducing lung and liver tumors in newborn mice than the other three isomers. The (+)-DBA trans-3,4-diol-1,2-epoxide-2 isomer was only one-third as tumorigenic as its parent hydrocarbon on mouse skin, but this isomer was at least 3-fold more potent than DBA for the induction of lung and liver tumors in newborn mice.

The racemic bay-region 3,4-diol-1,2-epoxide-2 had tumor-initiating activity that was considerably less than its active enantiomer, (+)-3,4-diol-1,2-epoxide-2 at the doses tested. As expected, the related racemic diol epoxide-1 isomer, in which the benzylic hydroxyl group and the oxirane ring were cis was inactive. The very low tumor-initiating activity of racemic DBA 3,4-diol-1,2-epoxide-2 in comparison with the (+)-3,4-diol-1,2-epoxide-2 suggests a possible inhibitory interaction between the (+)- and (–)-enantiomers.

Among the four bay-region diol epoxide enantiomers/diastereomers of DBA, the isomer with [1R,2S,3S,4R] absolute configuration had strong mutagenic activity in Chinese hamster V-79 cells, whereas the other three isomers had little or no activity except for the (+)-[1R,2S,3R,4S]-diol-epoxide-1, which had ~70% of the mutagenic activity of the (+)-[1R,2S,3S,4R]-3,4-diol-1,2-epoxide of DBA toward Chinese hamster V-79 cells (Table I). This article and previous studies on the biological activities of the four enantiopure bay-region diol epoxides of benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[c]phenanthrene, dibenz[a,h]acridine and dibenz[c,h]acridine indicate strong mutagenic activity toward Chinese hamster V-79 cells, strong tumorigenic activity on mouse skin and/or strong tumorigenic activity in newborn mice for the R,S-diol-S,R-epoxide (17–28). In almost all cases, the other three enantiopure diol epoxides had little or no activity. The relationship between the structures of the highly active [R,S,S,R] bay-region diol epoxides of benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[c]phenanthrene, dibenz[a,h]anthracene, dibenz[a,h]acridine and dibenz[c,h]acridine are shown in Figure 2.

Fig. 2.

Structures of the highly tumorigenic diol epoxide-2 isomers of benzo[a]pyrene (B[a]P), chrysene, benz[a]anthracene (BA), benzo[c]phenanthrene (B[c]PH), dibenz[a,h]acridine (DB[a,h]ACR), dibenz[c,h]acridine (DB[c,h]ACR) and DBA with [R,S,S,R]-absolute configuration.

Why the RSSR diol epoxide enantiomer of bay-region diol epoxides are more carcinogenic in mice than the other three possible enantiomers with only minor exceptions has not been adequately explained. Possible reasons include differences in the stability and metabolic inactivation of the four enantiomers as they move from the site of administration to their target as well as differences in the covalent binding of the enantiomers to DNA and differences in the repair of the DNA adducts. Whalen and his colleagues observed acid-catalyzed and spontaneous solvolysis of bay-region diol epoxides, and different compounds had different rates of solvolysis (43). The carcinogenic (±)-7β,8α-dihydroxy-9α,10β-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (diol epoxide 2) was ~10-fold more stable than the non-carcinogenic (±)-7β,8α-dihydroxy-9β,10β-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (diol epoxide-1) in tissue culture medium (44). Increased stability for the diol epoxide-2 isomer relative to that for the diol epoxide-1 isomer was also observed for diol epoxides of benz[a]anthracene and dibenz[a,j]anthracene in dioxane/water (1:10) (45). In addition, bay-region diol epoxides 1 and 2 from benzo[a]pyrene and other polycyclic hydrocarbons are refractory to metabolism by epoxide hydrolase (44,46). The comparative stability and metabolic inactivation of the four bay-region diol epoxide enantiomers of polycyclic hydrocarbons by epoxide hydrolase, glutathione transferase or other detoxifying enzymes has not yet been evaluated. Jerina and his colleagues discussed some of the complexity associated with nucleophilic ring opening of the epoxide ring of diol epoxides to give both cis and trans adducts with the exocyclic amino groups of adenine and/or guanine in DNA which would occur with any of the four diol epoxide stereoisomers from a polycyclic hydrocarbon (16 possible adducts) in competition with the major ring opening pathway associated with hydrolysis of the epoxides to give tetraols (47,48). There could certainly be a difference in reactivity between the chiral DNA reactant and each of the chiral diol epoxides reflected by the relative proportion of adducts.

Jerina and his colleagues determined the ability of the four enantiopure bay-region diol epoxides from benzo[a]pyrene, benz[a]anthracene, benzo[g]chrysene, dibenz[a,j]anthracene and benzo[c]phenanthrene to bond covalently to calf thymus DNA (49). The RSSR isomer showed the highest percentage of binding relative to solvolysis for each of the five hydrocarbons. However, the differences in extent of bonding to DNA between isomers were not large enough to account for much larger differences in tumorigenic response. Additional studies also indicated no obvious correlation between DNA reactivity of individual enantiopure bay-region diol epoxides and their known tumorigenic effects (50,51). Buterin showed that high tumorigenicity may partly be explained by slow nucleotide excision repair of stable base adducts derived from the reaction of the individual diol epoxides with DNA (52). As indicated from the above discussion, it is likely that multiple factors contribute to the high carcinogenicity of the RSSR bay-region diol epoxides.

During the course of our studies with bay-region diol epoxide enantiomers/diastereomers of DBA (Tables I–III) and the six other polycyclic aromatic hydrocarbons described above (17–20,28), we observed that the mutagenic activity of the diol epoxide enantiomers in Chinese hamster V-79 cells but not in S. typhimurium strains TA98 or TA100 was a good (but not perfect) predictor of tumorigenic activity on mouse skin and in the newborn mouse.

The stereochemical factors involved in the metabolic transformation of benzo[a]pyrene, chrysene, benz[a]anthracene, benzo[c]phenanthrene and DBA to trans-dihydrodiol precursors of bay-region diol epoxides is predominately to the (–)-enantiomer with [R,R]-absolute configuration (16,30,31,53–56). For all five hydrocarbons, the [R,R]-dihydrodiol is more tumorigenic than is the [S,S]-dihydrodiol. The [7R,8R]-, [1R,2R]-, [3R,4R]- and [3R,4R]-dihydrodiol of benzo[a]pyrene, chrysene, benz[a]anthracene and DBA, respectively, are each stereoselectively converted to the bay-region diol epoxide with [R,S,S,R]-absolute configuration [(+)-diol epoxide-2] which is also the most tumorigenic enantiomer of the bay-region diol epoxides (21–27). In all four cases, the highly tumorigenic (+)-diol epoxide-2 isomer has the same absolute [R,S,S,R] configuration (Figure 2). Our results indicate that a high degree of stereoselectivity is required for the initiation of carcinogenesis by polycyclic aromatic hydrocarbon metabolites. Similar studies on the stereochemical metabolism of benzo[a]pyrene were also done by Gelboin et al. (57–59). Based on the knowledge that the highly tumorigenic bay-region diol epoxide isomers of benzo[a]pyrene, chrysene, benz[a]anthracene, benzo[c]phenanthrene, DBA, dibenz[a,h]acridine and dibenz[c,h]acridine] all have the same [R,S,S,R] absolute configuration and the same conformation (pseudodiequatorial hydroxyl groups), it seems likely that the critical cellular targets required for the initiation of carcinogenesis are dependent on the same stereochemical relationship for all seven polycyclic aromatic hydrocarbons.

Funding

Hoffmann-La Roche Inc.; Department of Chemical Biology at Rutgers; National Institutes of Health (CA49765, ES005022).

Glossary

Abbreviations:

DBA

dibenz[a,h]anthracene

DMSO

dimethyl sulfoxide.