Materials and Methods.

Liver microsomes and cytosol from male Sprague-Dawley rats and male B6C3F₁ mice were prepared as described previously (6-8). Human liver microsomes and cytosol were prepared from frozen liver samples of two donors obtained from SRI International (Menlo Park, CA); H10 was from a 15-year-old Caucasian female, and H11 was from a 36-year-old Caucasian male. Microsomes from a human lymphoblastoid cell line containing cDNA-expressed human microsomal epoxide hydrolase were obtained from Gentest (Woburn, MA). All microsomal and cytosol samples were stored at −80°C until use. For in vivo studies, male Sprague-Dawley rats (230–290 g) and male B6C3F₁ mice (6–8 weeks) were used. Ethyl acetate (carbonyl-free) was purchased from American Burdick & Jackson (Muskegon, MI). Racemic BMO, styrene oxide, and 1-phenyl-1,2-ethanediol were obtained from Aldrich Chemical Co. (Milwaukee, WI). R,S-BDD was purchased from Eastman Kodak (Kingsport, TN).

Enzymatic Hydration of BMO.

In a final reaction volume of 0.5 ml, assays were conducted by incubating 2.5 μl BMO made in ethanol (0.25–10 mM = final concentration for rat and mouse liver; 0.1–10 mM = final concentration for human liver) with 0.25 ml microsomal (0.4–1.5 mg) or cytosolic (2.2–11.0 mg) protein and 0.25 ml buffer (0.1 M KH₂PO₄:0.15 M KCl:1.5 mM EDTA, pH 7.4) in 8-ml vials capped with a Teflon-lined septum. Kinetic reactions were conducted for 15 min. All reactions were performed in a 37°C water bath and terminated by cooling slightly in a dry ice/acetone bath. Each sample was vortexed after the addition of NaCl (0.5 g) and ethyl acetate (1.5 ml), and the samples were then centrifuged for 5 min in a Beckman TJ-R tabletop centrifuge at 3000 rpm. The samples were then placed on dry ice until the aqueous layer was frozen. The ethyl acetate layer was transferred to a microcentrifuge tube and concentrated to 100 μl under vacuum in a Savant SC110 Speed Vac (Holbrook, NY). The resulting samples were stored on dry ice until GC analysis was conducted using flame ionization detection. Controls not containing protein were also performed as described previously to determine the extent of nonenzymatic hydrolysis of BMO under the assay conditions. Experiments using cDNA-expressed epoxide hydrolase were conducted in a similar manner as described previously, except that the incubation time was 60 min and the BMO concentration used was 5 mM. Similar experiments were also performed with styrene oxide (5 mM), a known substrate, to ensure the enzyme was active and to compare the catalytic activities of the two substrates. Protein concentrations of the mouse, rat, and human microsomal and cytosolic samples were determined by the method of Lowryet al. (9). Protein concentrations of the cDNA-expressed human microsomal epoxide hydrolase samples were determined by the bicinchoninic acid method (Pierce, Rockford, IL). Bovine serum albumin was used as standard for both assays.

In Vivo Hydration of BMO.

Rats were housed individually in Nalgene metabolic cages (Rochester, NY), whereas mice (4/cage) were housed in cages adapted for use with mice. Animals were kept on a 12-hr light/dark cycle, and allowed food and water ad libitum. Doses of 14.3, 71.5, 143, and 285 μmol/kg BMO in 3 ml/kg corn oil or corn oil alone were given intraperitoneally to the rats or mice. Rat urine was collected at 8 and 24 hr. The 8-hr urine samples were diluted to 5 ml with deionized water and stored at 0°C until analysis, whereas 24-hr urine was stored at 0°C without dilution. Mouse urine was collected at 24 hr only, after which it was diluted to 4 ml with deionized water and stored as described previously. Extraction of the urine (0.5 ml) was conducted by vortexing the sample with 1.5 ml ethyl acetate. The organic layer was then separated from the aqueous phase, dried with a Whatman Na₂SO₄ sample drying device (Clifton, NJ), and concentrated 6- or 15-fold with speed vacuum lyophilization. GC analyses were performed as described herein. The BDD peak was quantitated by comparing peak height with the standard curve of reference BDD, subjected to a similar extraction procedure from urine.

GC Analysis.

For BDD analyses, an initial oven temperature of 95°C was held for 3 min, then increased at a rate of 70°/min to 175°C where it was held for an additional 1.5 min. The injector temperature was 140°C and the injection volume was 3 μl. The instrument, capillary column, and flows were as described previously (6-8). The retention time of eitherR- or S-BDD was 2.2 min under these conditions. For 1-phenyl-1,2-ethanediol analyses, an initial oven temperature of 180°C was held for 3 min and then increased at a rate of 10°C/min to 200°C. A final oven temperature of 230°C was reached at a rate of 30°C/min, where it was held for 1 min. The injector temperature was 200°C. The other instrument parameters and injection volume were the same as for BDD analysis. The retention time for 1-phenyl-1,2-ethanediol was 3.1 min under these conditions. Quantitations of BDD and 1-phenyl-1,2-ethanediol were conducted by comparison of peak heights with standard curves prepared from reference materials. Limits of detection for BDD and 1-phenyl-1,2-ethanediol were 24 μM and 15 μM, respectively. The extraction efficiency of BDD (100–600 μM) from mouse, rat, and human liver microsomal incubation mixtures was nearly 20–25%, and extraction efficiency from mouse and human liver cytosolic incubations was nearly 30%; results were corrected for the extraction efficiency.

GC-MS.

Incubations were performed as described previously using rat liver microsomes and 2.5 mM BMO. The ethyl acetate extracts from 20 vials were combined and concentrated under N₂ to nearly 1 ml. GC/MS was obtained on a Hewlett-Packard Series 6890 gas chromatograph and mass spectrometer. The column used was a HP-5 (30 m × 0.25 mm, 0.25 μm film thickness; Hewlett-Packard, Palo Alto, CA). The temperature gradient used was 80°C for 3 min that was then increased at a rate of 70°/min to 225°C for 10 min. Injector temperature was 225°C, and ion source temperature was 280°C. Ionization energy used was 70 eV. The retention times of both the authentic standard and the peak generated from the microsomal incubation were 3.0 min.

Statistics.

Statistical significance of the data was determined viaStudent’s t test, using p ≤0.05 as the criterion for significance.

Results and Discussion.

When BMO (0.25 mM) was incubated in the presence of mouse, rat, or human liver microsomes, a new peak was detected at 2.2 min that coeluted with authentic BDD; formation of this peak was linear to at least 20 min (data not shown). Identification of this peak as BDD from rat liver microsomes was confirmed by comparing its mass spectrum to that of reference BDD; the spectrum revealed a base peak atm/z 57 that corresponds to [M-CH₂OH]⁺ and a significant peak atm/z 70, indicating loss of H₂O. These fragmentation patterns matched well with patterns previously reported for BDD (10).

When BMO (0.25–5.0 mM) was incubated for 15 min at 37°C in the absence of microsomal protein, no BDD was detected; at 10 mM BMO concentrations, only a trace amount of BDD was detected, thus indicating insignificant nonenzymatic hydration rates under assay conditions. Incubations using rat liver cytosol were also ineffective in producing BDD, even when the protein concentration was increased nearly 5-fold over the microsomal protein concentrations used to produce BDD. However, when mouse or human liver cytosolic fractions were used as the enzyme source, BDD was formed, although the levels detected were lower than those detected with microsomes (table1). When cDNA-expressed human microsomal epoxide hydrolase was incubated with BMO or styrene oxide (5 mM), similar amounts of BDD (198 nmol/mg protein/hr) and 1-phenyl-1,2-ethanediol (143 nmol/mg protein/hr) were detected, whereas no metabolites were detected when human lymphoblast control microsomes were used. The previous results provide strong evidence for microsomal epoxide hydrolase being the major catalyst of BMO hydration in the mouse, rat, and human liver microsomes. In addition, soluble epoxide hydrolases present in mouse and human liver cytosol can also metabolize BMO to yield BDD, although the amounts of BDD detected in the microsomes were higher than those detected in the cytosolic fractions. Similar to the results presented in table 1, human liver microsomes were previously shown to hydrolyze styrene oxide and benzo(a)pyrene-4,5-oxide at rates that were much higher than those of cytosol (11, 12).

BDD formation in mouse, rat, and human liver microsomes was dependent on the BMO concentration (table 2); kinetic constants were determined using a curve-fitting method as described previously (13). TheV_max/K_m ratio (catalytic efficiency) of H11 was ∼4-fold, 8-fold, and 3-fold higher the ratios obtained with H10, and mouse and rat liver microsomes, respectively. Furthermore, human liver microsomes are at least as efficient as the rat liver microsomes in converting BMO to BDD, whereas mouse liver microsomes exhibited lower rates. Thus, mouse and rat differences in BMO hydration to yield BDD may contribute to the biochemical basis for the higher sensitivity of the mouse to BD-induced carcinogenicity, and BMO hydration may represent a significant detoxification reaction for humans.

Rats given BMO (71.3–285.0 μmol/kg) excreted <1% of the BMO dose as BDD into their urine within 24 hr after BMO administration, whereas no BDD was detected at the low BMO dose (14.3 μmol/kg). Trace amounts of BDD were detected in the urine of some of the mice given the 285.0 μmol/kg dose; however, no BDD was detected at the lower doses of BMO given to mice. Thus, for the mouse, <0.2% of the high dose and <3% of the low dose, if any, were excreted as BDD based on the limit of detection of the assay. Thus, the in vivo data show that <1% of BMO was excreted in urine as BDD after either rats or mice were given BMO. At similar BMO dosages, rats and mice excreted nearly 7–26% of the BMO dose as the corresponding mercapturic acids (14). Thus, the in vivo results suggest a limited role for epoxide hydrolases in BMO metabolism in vivo. Alternatively, BDD may be subject to further metabolism, which could explain the low levels of BDD detected in the mouse and rat urine. Recent work from our laboratory has shown that BDD can function as a substrate for alcohol dehydrogenases and cytochrome P450 enzymes (13, 15), providing evidence for the latter hypothesis.

In summary, the results presented herein provide direct evidence for BMO hydration by mouse, rat, and human liver microsomes and mouse and human liver cytosol. Results suggest that epoxide hydrolases play an important role in species differences in BD metabolism and that it is a significant pathway for BMO metabolism in humans in vitro. The extent to which BMO hydration occurs in vivo requires further investigation, because only small amounts of BDD were excreted in the urine of rats or mice given BMO.