Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

4-Vinylcyclohexene (VCH¹) is formed by the spontaneous dimerization of two molecules of 1,3-butadiene during the rubber-curing process (Rappaport and Fraser, 1976; International Agency for Research on Cancer, 1994). VCH is also an intermediate in the synthesis of styrene and vinylcyclohexene diepoxide (VCD) for epoxy resin formation. Repeated exposure of mice to VCH causes premature ovarian failure by depletion of ovarian primordial and primary follicles (Collins and Manus, 1987; Smith et al., 1990a). This premature ovarian failure may be associated with the ovarian neoplasms that develop in mice chronically exposed to VCH (National Toxicology Program, 1986; Collins et al., 1987). Cytochrome P450 (CYP)-catalyzed bioactivation of VCH to metabolites VCH-1,2-epoxide, VCH-7,8-epoxide, and ultimately VCD (Fig. 1) is necessary for VCH-induced ovotoxicity to occur (Smith et al., 1990b, Doerr and Sipes, 1996). Interestingly, female Fischer 344 (F-344) rats are resistant to the ovarian toxicity caused by treatment with VCH. This resistance in female rats is at least partially related to a low capacity to bioactivate VCH to epoxide metabolites (Smith et al., 1990c).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Schematic of all possible stereoisomers of VCH and its toxic metabolites. Bioactivation of VCH to epoxides by CYP. Fundamentals of stereochemistry prove that it is impossible for (R)- and (S)-VCH to form the same epoxide metabolite stereoisomers.

VCH is most commonly formed as a racemic mixture of (R)-VCH and (S)-VCH. Numerous examples exist in which one enantiomer of a compound is more toxic than the racemic mixture or the other enantiomer (Braun et al., 1986; D'Arcy and Griffin, 1994). Furthermore, CYP enzymes have been shown to possess varying degrees of stereoselectivity in catalyzing epoxidation reactions in compounds such as polycyclic aromatic hydrocarbons (Yang, 1988). Therefore, it is important to determine whether the enantiomers of VCH also demonstrate selective bioactivation and, ultimately, differential toxicity. The effects of stereochemistry on the bioactivation of the related compound 1,3-butadiene have recently been investigated. This prochiral compound can form multiple mono- and diepoxide stereoisomers or diastereoisomers through CYP-catalyzed epoxidation, as can VCH. Nieusma et al. (1997, 1998) demonstrated stereoselectivity of 1,3-butadiene bioactivation in both mouse and rat hepatic microsomes, as well as with purified human CYP2E1 protein. VCH may also undergo stereoselective bioactivation in cytochrome P450 systems. Sixteen possible monoepoxide and diepoxide diastereomers can be formed from a racemic mixture of VCH (Fig. 1). Perhaps only one or two of these possible metabolites are formed in vivo and are responsible for VCH-induced toxicity.

The studies reported here compared the epoxidation of (R)-VCH and (S)-VCH in microsomes obtained from female B6C3F₁ mice and F-344 rats. The effects of pretreatment with various CYP-inducing agents on the epoxidation of the enantiomers of VCH were also examined in both species. Furthermore, since initiating a toxic response in mice with VCH requires several days of repeated exposure (Doerr-Stevens et al., 1999), microsomes prepared from mice or rats pretreated with racemic VCH for 10 days were also incubated with (R)-VCH or (S)-VCH to determine how this would affect stereoselective bioactivation. Lastly, several human CYP isoforms were incubated with (R)-VCH and (S)-VCH to gain insight into how humans metabolize VCH.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Treatments. Female B6C3F₁ mice and F-344 rats (approximately 28-38 days old) were obtained from Harlan Sprague-Dawley, Inc. (Indianapolis, IN). The animals were housed in cages with sawdust bedding and had free access to food (Teklad, Harlan Sprague-Dawley, Inc., Madison, WI) and water. Animals were maintained on a 12-h light/dark cycle and acclimated to this environment for at least 7 days before dosing and/or preparation of hepatic microsomes. Treated animals were dosed with either racemic VCH (7.5 mmol/kg i.p. per day for 10 days, as previously described by Doerr-Stevens et al., 1999), phenobarbital (PB; 80 mg/kg i.p. per day for 5 days), or acetone (1% in the drinking water for 5 days).

Chemicals. (R,S)-4-Vinylcyclohexene, vinylcyclohexene 1,2-monoepoxide, vinylcyclohexene diepoxide, methylcyclohexene, and acetone were purchased from Aldrich Chemical Co. (Milwaukee, WI). NADP⁺, G6PDH, and G6P were purchased from Sigma Chemical Co. (St. Louis, MO). Cyclohexene oxide was produced by National Toxicology Program/RTP (Research Triangle Park, NC). Vinylcyclohexene 7,8-monoepoxide was synthesized by the method of Watabe et al. (1981). (R)-VCH and (S)-VCH were synthesized as previously reported (Mash and Gregg, 1995). (Caution: VCH and its epoxides are either potential carcinogens or are carcinogens in animals and should be handled with appropriate precautions).

Subcellular Preparation and Characterization. Animals were euthanized by inhalation of carbon dioxide 24 h after final dosing. Livers were excised and homogenized in a 50 mM Tris-HCl buffer (pH 7.6) using a drill motor and Teflon-glass homogenizer. Microsomes were prepared by differential centrifugation of the homogenate, as described by Guengerich (1989). Mouse microsomes were prepared and pooled as four mice/group. Protein concentrations were determined by using a Bicinchoninic Acid kit (Pierce, Rockford, IL). Total P450 concentrations (nanomoles per milligram of microsomal protein) were determined by the carbon monoxide-binding spectra as described by Omura and Sato (1964).

Capillary Gas-Liquid Chromatographic Conditions for Epoxide Analysis. Analyses were performed on a Hewlett-Packard HP 5890A gas chromatograph equipped with a DB-624 capillary column (J & W Scientific, Folsom, CA) and a flame ionization detector. The nitrogen carrier gas flow rate was 1 ml/min. The flame ionization detector gas flow rates for H₂, N₂, and air were 42, 35, and 400 ml/min, respectively. Splitless injection was used with the purge off from time 0 to 1.0 min, with a 2-µl injection volume. The injection and detector temperatures were 200 and 250°C, respectively. The oven temperature was held at 60°C for 10 min and then increased to 230°C at a rate of 12°C/min. The final temperature was held for 3 min to ensure elution of the diepoxide. Retention times were 11.9 min for methylcyclohexene, 15.2 min for cyclohexene oxide, 16.6 min for VCH, 20.4 min for VCH-1,2-epoxide, 21.4 min for VCH-7,8-epoxide, and 25.0 min for VCD. Formation of VCH-1,2-epoxide, VCH-7,8-epoxide, and VCD (nanomoles per milligram of microsomal protein) were quantified by comparing the peak areas to those in standard curves prepared with known amounts of the epoxides. Detection limits for the epoxides were each approximately 1 nmol.

Microsomal Incubations. Microsomal protein (final concentration of 1 mg/ml in 1-ml total volume) was added to a 50 mM HEPES/EDTA buffer solution containing a recycling NADP⁺ system (0.5 mM NADP⁺, 1 unit of G6PDH/ml, and 10 mM G6P), 2 mM cyclohexene oxide (an epoxide hydrolase inhibitor; Guest and Dent, 1980), and 1 mM (R,S)-, (R)-, or (S)-VCH. Previous studies were conducted that demonstrated that reactions were linear up to 1 mg/ml protein using the individual enantiomers of VCH as well as racemic VCH. Samples were incubated in a 37°C water bath. After the appropriate amount of time (ranging from 0-60 min), the CYP reactions were terminated by submersion in liquid nitrogen. VCH and its epoxide metabolites (VCH-1,2-epoxide, VCH-7,8-epoxide, and VCD) were extracted with ethyl acetate containing 1 µg/ml methylcyclohexene as an internal standard. The epoxide metabolites were identified and quantified using gas chromatography (DB-624 capillary column). Blanks included incubations containing denatured microsomes (preheated at 60°C for 30 min before incubating) and incubations lacking glucose 6-phosphate. Data are presented as nanomoles of epoxide formed per milligram of microsomal protein. The extraction efficiencies of VCH-1,2-epoxide, VCH-7,8-epoxide, and VCD were 96, 92, and 76%, respectively. Reported values were corrected for recovery.

Incubations of "Supersomes" Containing Human CYP2A6, CYP2B6, CYP2E1, or CYP3A4. To investigate whether individually expressed human CYP protein can catalyze the epoxidation of (R)- or (S)-VCH to the epoxide metabolites, the Supersome system was used (human CYP + P450 reductase + cytochrome b₅; GENTEST Corporation, Woburn, MA). Supersomes containing either human CYP2A6, CYP2B6, CYP2E1, or CYP3A4 (100 pmol of total P450/ml) were combined with 1 mM (R)-VCH or (S)-VCH and a recycling system for NADPH (0.5 mM NADP⁺, 1 unit G6PDH/ml, and 10 mM G6P) for a total volume of 500 µl. Samples were incubated at 37°C for up to 30 min. Epoxide metabolites were identified and quantified using gas chromatography.

Statistical Analysis. A two-sample inference was made using a Student's t test. Means were considered significantly different at p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Biotransformation of (R)- and (S)-VCH in Mouse Hepatic Microsomes. There were clear differences in epoxidation of (R)-VCH and (S)-VCH in mouse hepatic microsomes. After 60 min incubations with control mouse microsomes, (R)-VCH formed more than twice the amount of VCH-1,2-epoxide as did (S)-VCH (72.5 ± 9.0 and 29.8 ± 3.7 nmol, respectively), but significantly less VCH-7,8-epoxide (30.4 ± 4.0 and 56.3 ± 7.3 nmol, respectively) (Fig. 2, A and C). Neither enantiomer formed the diepoxide at detectable levels. Comparative incubations with (R,S)-VCH demonstrated epoxidation to the monoepoxides at rates approximately average of that seen with (R)-VCH and (S)-VCH (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   (R)-VCH and (S)-VCH bioactivation in mouse and rat microsomes. Comparison of VCH-1,2-epoxide formation (A and B) and VCH-7,8-epoxide formation (C and D) from (R)-VCH () and (S)-VCH () in hepatic microsomes obtained from female B6C3F1 mice (A and C) and Fischer 344 rats (B and D). Data are expressed as mean ± S.D. (n = 3-4 in duplicate; mouse microsomes were prepared from four mice per group). a, significantly different amount of epoxide formed than from (S)-VCH in the same species, at p < 0.05 (Student's t test). b, significantly different from epoxide formed from same enantiomer in the rat, at p < 0.05 (Student's t test).

Biotransformation of (R)- and (S)-VCH in Rat Hepatic Microsomes. Parallel studies performed in hepatic microsomes from adult female F-344 rats demonstrated lower overall bioactivation of (R)- and (S)-VCH compared with the mouse. The metabolic profiles in the control rat microsomes were stereochemically similar to those in the mouse, as (R)-VCH was converted to approximately twice as much VCH-1,2-epoxide as was (S)-VCH after 60 min but to equal amounts of VCH-7,8-epoxide (Fig. 2, B and D). Neither enantiomer was converted to the diepoxide at detectable levels in control microsomes. All pretreatments caused relatively nonstereoselective induction of VCH bioactivation in vitro, but induction overall was still much less in the rat compared with the mouse (Table 1). Furthermore, none of the pretreatments in the rat resulted in VCD formation from either enantiomer in vitro.

Biotransformation of (R)- and (S)-VCH in Human CYP2E1 Supersomes. Human CYP2E1 formed both epoxides when incubated with (R)- or (S)-VCH (Fig. 3, A and B). By 30 min, (R)- and (S)-VCH were converted to 0.7 ± 0.0 and 0.4 ± 0.0 nmol of VCH-1,2-epoxide/pmol of CYP2E1, respectively, and 0.3 ± 0.0 and 0.5 ± 0.0 nmol of VCH-7,8-epoxide/pmol of CYP2E1, respectively. Neither enantiomer was converted to VCD at detectable levels when incubated with human CYP2E1. There was no epoxide formation in parallel incubations with control insect microsomes lacking CYP2E1.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   (R)-VCH and (S)-VCH bioactivation in Supersomes containing human CYP2E1. Comparison of VCH-1,2-epoxide (A) and VCH-7,8-epoxide (B) formation from 1 mM (R)-VCH () or (S)-VCH () by human CYP2E1 Supersomes (100 pmol/ml). a, significantly different amount of epoxide formed than from (S)-VCH, at p < 0.05.

Biotransformation of (R)- and (S)-VCH in Human CYP2A6 Supersomes. Human CYP2A6 Supersomes were limited in their ability to form epoxides from either enantiomer of VCH (Fig. 4). Furthermore, VCH-7,8-epoxide was not formed in detectable levels from (R)- or (S)-VCH by human CYP2A6. (R)-VCH was converted to more VCH-1,2-epoxide than (S)-VCH at later time points, and neither enantiomer formed VCD at detectable levels. Activity of the human CYP2A6 Supersomes was confirmed by coumarin hydroxylation.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   (R)-VCH and (S)-VCH bioactivation in Supersomes containing human CYP2A6. Comparison of VCH-1,2-epoxide formation from 1 mM (R)-VCH () or (S)-VCH () by human CYP2A6 Supersomes (100 pmol/ml). a, significantly different amount of epoxide formed than from (S)-VCH, at p < 0.05.

Biotransformation of (R)- and (S)-VCH in Human CYP2B6 Supersomes. Human CYP2B6 Supersomes demonstrated a greater ability than human CYP2A6 to form epoxides from (R)- and (S)-VCH. After 30 min, 0.4 ± 0.0 and 0.2 ± 0.0 nmol of VCH-1,2-epoxide/pmol of CYP2B6 and 0.9 ± 0.1 and 0.7 ± 0.0 nmol of VCH-7,8-epoxide/pmol of CYP2B6 were formed from (R)- and (S)-VCH, respectively (Fig. 5, A and B). Thus, human CYP2B6 is a more selective catalyst for epoxidation of (R)- and (S)-VCH to the epoxide at the 7,8 position over the 1,2 position. Neither enantiomer formed VCD at detectable levels when incubated with human CYP2B6. There was no epoxide formation in parallel incubations with control insect microsomes lacking CYP2B6.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   (R)-VCH and (S)-VCH bioactivation in Supersomes containing human CYP2B6. Comparison of VCH-1,2-epoxide (A) and VCH-7,8-epoxide (B) formation from 1 mM (R)-VCH () or (S)-VCH () by human CYP2B6 Supersomes (100 pmol/ml). a, significantly different amount of epoxide formed than from (S)-VCH, at p < 0.05.

Biotransformation of (R)- and (S)-VCH in Human CYP3A4 Supersomes. Neither (R)- or (S)-VCH were converted to epoxide metabolites at detectable levels by human CYP3A4.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

These studies clearly demonstrated enantioselectivity and regioselectivity in the bioactivation of VCH in rodent hepatic microsomes as well as in expressed human CYP enzymes. In all microsomal incubations (from both treated and nontreated rodents), (R)-VCH formed significantly more of the 1,2 monoepoxide, while (S)-VCH generally formed more of the 7,8 monoepoxide. In general, since the 1,2 monoepoxide was the major metabolite, more of the (R)-enantiomer was converted to monoepoxide compared with the (S)-enantiomer.

In both rodent species, the various CYP-inducing agents caused greater induction of VCH-1,2-epoxide formation compared with VCH-7,8-epoxide formation from both enantiomers. In the mouse, phenobarbital pretreatment increased VCH-1,2-epoxide formation 6-fold and VCH-7,8-epoxide only 2-fold; acetone pretreatment increased VCH-1,2-epoxide formation 2-fold and did not affect VCH-7,8-epoxide formation; VCH pretreatment increased VCH-1,2-epoxide formation approximately 3-fold and VCH-7,8-epoxide 2-fold. Since all pretreatments resulted in induction of bioactivation that was similar with both (R)- and (S)-VCH, it appears that these inductions did not result in increased enantioselective bioactivation. Although there was no formation of VCD, the ovotoxic metabolite of VCH, in any microsomal incubation from nontreated mice and rats, it was detected in incubations with microsomes obtained from mice that had been treated with the inducing agents. In both phenobarbital and acetone pretreatment groups, (R)-VCH formed more VCD than did (S)-VCH. This indicates that when there are elevated levels of certain CYP isoforms such as CYP2E1, CYP2A, or CYP2B in the liver, (R)-VCH may be more responsible than (S)-VCH for bioactivation to VCD. It is interesting, however, that VCH pretreatment in the mouse resulted in approximately the same microsomal capacity to form VCD from each VCH enantiomer. Therefore, VCH may induce isoforms other than those induced by acetone or PB, which catalyze the reactions forming VCD from (S)-VCH.

Induction of CYPs typically increases the concentration of CYP protein. Compounds such as phenobarbital achieve this by increasing transcription of mRNA encoding certain CYP genes (Waxman and Azaroff, 1992), whereas others (e.g., acetone, ethanol) stabilize the CYP protein (Forkert et al., 1991) or increase the translational efficiency of mRNA into protein (Tsutsumi et al., 1993). Research in our laboratory has shown that repeated dosing with racemic VCH increases CYP2A and CYP2B protein levels in the mouse, but not in the rat (Doerr-Stevens et al., 1999). Since VCH pretreatment results in relatively the same amounts of VCD formation from each VCH enantiomer, it may be inducing both (R)- and (S)-VCH-metabolizing CYPs equally through one or more than one of the CYP-inducing pathways mentioned above. This is important stereochemically, since repeated exposure to VCH is required to cause ovotoxicity (Doerr-Stevens et al., 1999).

Human microsomes have been shown to metabolize VCH, but at rates lower than those of mouse or rat hepatic microsomes (Smith and Sipes, 1991). To assess which CYP isoforms epoxidize VCH, the two separate VCH enantiomers were incubated with individual expressed human CYP proteins. Two other papers currently in preparation show that of eight different human CYP isoform Supersomes tested (CYP1A1, -2A6, -2B6, -2C9, -2E1, -3A4, -4A11, and Aromatase), only CYP2E1, -2A6, and -2B6 are capable of bioactivating racemic VCH (unpublished observations). In these studies, CYP2A6, CYP2B6, and CYP2E1 metabolized the enantiomers of VCH to monoepoxides. Although human and rodent CYP isoforms cannot be directly compared, the human CYP2A and -2E1 protein systems used in these studies stereochemically mimicked metabolic profiles as seen with the mouse and rat microsomal systems. For example, in the CYP2E1 Supersome system, (R)-VCH preferentially formed more VCH-1,2-epoxide but less VCH-7,8-epoxide as compared with (S)-VCH. It is interesting that CYP2E1-inducing acetone pretreatment in both rodent species resulted in 2-fold increases in VCH-1,2-epoxide formation from both enantiomers but no increases in VCH-7,8-epoxide formation from either enantiomer, and (R)-VCH still formed more VCH-1,2-epoxide but less VCH-7,8-epoxide as compared with (S)-VCH.

Since (R)-VCH preferentially forms VCH-1,2-epoxide and (S)-VCH forms more VCH-7,8-epoxide in microsomal samples, it is important to know whether one of these monoepoxides forms the toxic diepoxide at a more rapid rate. Mouse hepatic microsomal incubations with either diastereometric VCH-1,2-epoxide or diastereometric VCH-7,8-epoxide demonstrated that the monoepoxides are bioactivated at comparable rates to form VCD (unpublished observations). However, it is important to note that those studies measured metabolism of each monoepoxide from diastereometric mixtures [2(R)- and 2(S)-monoepoxides in total] rather than one diastereomer of the monoepoxide, so VCD formation represents only an average from both (R)- and (S)-VCH monoepoxides.

VCH causes ovotoxicity in mice by depleting primordial and primary follicles through an apoptotic-like mechanism (Kao et al., 1999). Structure-activity studies have shown that the formation of the diepoxide is essential to cause ovotoxicity (Doerr and Sipes, 1996). Diepoxides such as butadiene diepoxide are highly electrophilic and may cause toxicity by cross-linking biomolecules (Rydberg et al., 1996; Hartley et al., 1999). Such cross-linking may require certain configurations of the diepoxides, as biomolecules themselves have specific topographies. Therefore, it is possible that a limited number of the possible eight VCD diastereoisomers formed are responsible for the ovarian toxicity.

These studies represent the first important step in examining the stereochemistry of VCH bioactivation and toxicity. There were differences between the bioactivation of (R)- and (S)-VCH in rodent hepatic microsomes as well as in individual expressed human CYP isoforms. These differences in metabolic profiles of (R)- and (S)-VCH may be due to, but are not limited to, different CYP isoforms involved in (R)- and (S)-VCH bioactivation for which there may be differences in binding affinities, different modes of binding based on how the molecule enters the active site, and different kinetic rates for each binding mode for each isoform. It is important to recognize, however, that in vitro bioactivation studies may not be entirely indicative of the physiological induction of ovarian toxicity by the VCH enantiomers. For example, one of the two VCH enantiomers may have a different distribution or partitioning in vivo than the other. Additionally, the epoxide metabolites of one VCH enantiomer may be more readily detoxified in the liver or ovary compared with the epoxide metabolites of the other enantiomer. Future studies that compare the ovarian toxicity of (R,S)-VCH with that of (R)- or (S)-VCH will further address whether stereochemistry is critical in VCH-induced ovarian toxicity.