Genetic polymorphism of microsomal epoxide hydrolase activity in the mouse.

Hepatic microsomal epoxide hydrolase activity (EC 3.3.2.3), assayed using styrene oxide as the substrate, has a pH optimum of 9.5 from C57BL/6J mice and a pH optimum of 8.7 from DBA/2J mice. In cross and back-cross matings between C57BL/6J and DBA/2J mice, this phenotypic difference in epoxide hydrolase activity is inherited as a single autosomal trait, with co-dominant expression in heterozygotes. Heating liver microsomes from C57Bl/6J mice at 62 degrees C for 30 min produced a slight decrease in enzyme activity, whereas the same treatment of DBA/2J microsomes reduced enzyme activity to less than 3% of its initial value. Twenty-six inbred strains of mice were examined and separated into two phenotypic classes on the basis of differences in pH optima and heat sensitivity of microsomal epoxide hyrolase activity. Eph-1 is proposed as the locus symbol of the structural gene for microsomal epoxide hydrolase, with superscripts b and d designating the alleles carried by C57BL/6J and DBA/2J mice, respectively. Using 24 recombinant inbred strains derived from C57BL/6J and DBA/2J mice, Eph-1 was found to be linked to two loci on Chromosome 1.

Hepatic microsomal epoxide hydrolase activity (EC 3.3.2.3), assayed using styrene oxide as the substrate, has a pH optimum of 9.5 from C57BL/6J mice and a pH optimum of 8.7 from DBA/SJ mice. In cross and backcross matings between C57BL/6J and DBA/2J mice, this phenotypic difference in epoxide hydrolase activity is inherited as a single autosomal trait, with co-dominant expression in heterozygotes. Heating liver microsomes from C57BL/6J mice at 62°C for 30 min produced a slight decrease in enzyme activity, whereas the same treatment of DBA/23 microsomes reduced enzyme activity to less than 3% of its initial value. Twenty-six inbred strains of mice were examined and separated into two phenotypic classes on the basis of differences in pH optima and heat sensitivity of microsomal epoxide hydrolase activity. Eph-1 is proposed as the locus symbol of the structural gene for microsomal epoxide hydrolase, with superscripts b and d designating the alleles carried by C57BL/6J and DBA/2J mice, respectively. Using 24 recombinant inbred strains derived from C57BL/6J and DBA/2J mice, Eph-1 was found to be linked to two loci on Chromosome 1.
Many synthetic and some naturally occurring alkene and arene compounds are oxidized by the microsomal monooxygenase enzymes to epoxide metabolites (1). Some of these epoxides are reactive electrophiles which can bind covalently to cellular macromolecules and produce toxic, mutagenic, and/or carcinogenic effects (2). Epoxides can be converted enzymatically to dihydrodiols by epoxide hydrolase or conjugated with glutathione either nonenzymatically or catalytically by the glutathione-S-transferases (3). Thus, these enzymes play an important role in controlling the level of potentially hazardous epoxide compounds.
Microsomal epoxide hydrolase activity (EC 3.3.2.3), although present at highest levels in the liver, is found in almost all tissues that have been examined (4). The hepatic enzyme activity is increased by the administration of a variety of * This research was supported by National Institute of Environmental Health Grant R01-ES-01884, National Cancer Institute Program Project Grant Pol-CA-22484, and National Institute of General Medical Sciences Grant GM18684. The Jackson Laboratory is fully accredited by the American Association for Accreditation of Laboratory Animal Care. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$Supported by United States Public Health Service National Research Service Award T22 CA-09135.  (8). In mice, BHA' is particularly effective, producing up to an 11-fold increase in hepatic microsomal epoxide hydrolase activity.
Genetic differences in the ability to metabolize drugs have been noted in humans and rodents (9), and several single locus polymorphisms are known in mice, e.g. the Ah locus (lo), and the Coh locus (11). Differences in hepatic epoxide hydrolase activity between inbred strains of mice have been noted, but not further characterized (12,13). In light of the important role hepatic microsomal epoxide hydrolase plays in drug metabolism, we undertook the present study to look for strain differences in this activity. We now report a genetic polymorphism in this enzyme activity.
BHA, Tween 80, and bovine serum albumin were purchased from Sigma Chemical Co., St. Louis, MO. MOPS buffer was purchased from Calbiochem, La Jolla, CA. Thin layer chromatography plates (Whatman LK5DF) were purchased from Kontes Glass Co., Vineland, NJ. A 1% BHA diet was prepared by Teklad Test Diets, Madison, WI, by adding BHA to powdered Wayne Mouse Breeder Blox (Allied Mills, Chicago, IL) and then repelleting the diet. A control diet was prepared by the same method but without BHA.
Animals-Mice were obtained from the production and research stocks of the Jackson Laboratory, Bar Harbor, ME. The recombinant inbred (RI) strains of mice (designated BXD) were derived by inbreeding the F2 generation of C57BL/6J and DBA/2J mice (14). The mice used in these studies were 9 to 12 weeks old female mice unless otherwise noted. In experiments involving BHA, the mice were placed on the 1% BHA diet for an 8-day period prior to killing. AU mice were given food and water ad libitum.
Preparation of Microsomes-Mice were killed by cervical dislocation. Livers were removed, weighed, and homogenized in 3 volumes of 0.15 M KC1 with a glass homogenizer fitted with a Teflon pestle. The homogenate was then centrifuged a t 10, OOO X g for 20 min a t 4°C. The resulting postmitochondrial supernatant was centrifuged for 1 h a t 100,000 X gat 4°C. The microsomal pellet was rinsed twice and then resuspended in 0.25 M sucrose to a concentration equivalent to 1 g wet weight liver/ml of sucrose. Microsomes were either assayed immediately for epoxide hydrolase activity or frozen at -70°C for glycine) with 0.1% Tween 80 added. These buffers were titrated with 1 N sodium hydroxide to the required pH measured at 37°C.
Assays of Epoxide Hydrolase Actiuity-Microsomal epoxide hydrolase activity was assayed routinely using [7-3H]styrene oxide as a substrate by the method of Oesch et al. (5) with modifications of Seidegard et al. (15) and Jerina et al. (16). The assay was performed in 10-ml round bottom, stoppered glass test tubes to prevent evaporation of the substrate. The reaction mixture consisted of 25 1-11 of buffer, 50 p1 of a suspension of microsomes (containing 20 to 150 pg of protein), and 5 p1 of tetrahydrofuran containing 200 nmol of [7-3H]styrene oxide (100,OOO dpm/assay) (final concentration 2.5 m). The reaction was started by the addition of the substrate and the tubes were then incubated in a shaking water bath a t 37°C for 15 min. Sample blanks received an additional 25 pl of tetrahydrofuran prior to addition of substrate. The reaction was stopped by the addition of 5 ml of ice-cold petroleum ether (b.p. 40-60') followed by rapid blending on a Vortex mixer for 15 s. The product, [7-"H]styrene glycol, is then extracted by the method of Oesch et al. (5) and counted by liquid scintillation spectrometry. Product formation is linear with respect to both time and protein concentration. Values obtained are not corrected for recovery of styrene glycol in the extraction procedures. Enzyme activity is expressed as nanomoles of styrene glycol per mg of protein per min. To assure a low zero time blank, the [7-3H]styrene oxide (stored in petroleum ether) is extracted with water periodically to remove water-soluble degradation products which accumulate during storage (15).
Epoxide hydrolase activity was measured using benzo(a)pyrene-4,5-oxide as the substrate by the thin layer chromatography method of Jerina et al. (16) with the following modifications. The buffers used in this assay were the same ones used in the styrene oxide assay (see "Buffers"). Benzo(a)pyrene-4,5-oxide was dissolved in tetrahydrofuran instead of acetonitrile. The reaction was run for 5 min during which product formation was linear with both time and protein concentration.
Protein Measurements-Proteins were determined by the method of Lowry et al. (17) using bovine serum albumin as a standard.
Heat Stability Experiments-Liver microsomes from BHA-treated mice were diluted with 225 mM sucrose, 10 mM potassium phosphate buffer (pH 7.1) and then put into test tubes. The microsomes were heated in a water bath at 62°C for varying time periods, then placed immediately in an ice water bath. After cooling, the microsomes were added to the glass-stoppered test tubes, the pH was adjusted to pH 8.7 with Buffer A and epoxide hydrolase activity was determined. Control microsomes were treated in the same manner except they were not heated.

RESULTS
The hydration of styrene oxide to styrene glycol, catalyzed by hepatic microsomal epoxide hydrolase, has a pH optimum of 9.5 in C57BL/6J mice and of 8.7 in DBA/2J mice (Fig. LA). Feeding a 1% BHA diet to these mice increases the level of enzyme activity in both strains 6 to 9-fold, but does not alter the pH optima. B6D2F1/J mice have a pH curve that is a composite of the parent strains with an intermediate pH optimum of 9.1 (data not shown).
The pH curves of hepatic microsomal epoxide hydrolase activity in six other inbred mouse strains are seen in Fig. 1B.
To determine whether the low enzyme activity in DBA/ZJ microsomes at pH 9.5 resulted from denaturation of the enzyme, we incubated the microsomes at pH 9.5 (Buffer C) for 30 min at 37"C, and then adjusted the pH to 8.7 (Buffer A). Epoxide hydrolase activity was the same as that in microsomes maintained at pH 8.7, indicating that the decline in enzyme activity seen when shifting from pH 8.7 to 9.5 is reversible. With benzo[a]pyrene-4,5-oxide as the substrate, the hepatic epoxide hydrolase activity from both C57BL/6J and DBA/2J mice had broad pH optima from pH 7.1 to 8.3, but the activity in DBA/ZJ microsomes fell off more rapidly at a more alkaline pH range, 9.1 to 9.5, than did the activity from C57BL/6J microsomes (data not shown).
Genetic Segregation of Microsomal Epoxide Hydrolase Actiuity-The inheritance pattern of hepatic microsomal epoxide hydrolase activity in the BHA-treated offspring of genetic crosses of C57BL/6J, DBA/2J, and B6D2F1/J mice is shown in Fig. 2. Hepatic microsomal epoxide hydrolase activity was measured in each animal at both pH 9.5 and at pH 8.7 and is expressed as the ratio of enzyme activity at pH 9.5/pH 8.7. T h i s ratio value is typically 1.2 to 1.7 for C57BL/6J mice, w -10-  Heat Stability-Differences in the primary structure of a protein are most likely to be detected as changes in thermo-TABLE I Strain distribution of microsomal epoxide hydrolase alleles Hepatic microsomes from individual mice were phenotyped for pH optimum by either doing a complete pH curve (as in Fig. 1) or by determining the ratio of enzyme activity at pH 9.5/pH 8.7. Heat sensitivity of microsomal epoxide hydrolase activity was measured using liver microsomes from individual mice as described under "ExDerimentd Procedures."   Mice were phenotyped by determining the ratio of enzyme activity at pH 9.5/pH 8.7 as described in the text. "B" and "D" are used as generic symbols for alleles inherited from C57BL/6J and DBA/ZJ mice, respectively. The C57BL/6J genotype is Sas-1 Ltw-4' Eph-1 Mls'/Sas-l" Ltw-4' Eph-1' Mls' and the DBA/2J genotype is Sasl o Ltw-4" Eph-ld Mls"/Sas-I" Ltw-4" Eph-ld Mls". Regions where cross-overs have resulted in recombination of the parental alleles in the BXD strains are denoted by an X. Animals of both sexes were used in this experiment.

Sas-I Ltw-4 Eph-1 Mls
Mouse strain lability (18). The relative heat stability of epoxide hydrolase activity (measured with styrene oxide as substrate) in hepatic microsomes from C57BL/6J, DBA/2J, and B6D2Fl/J mice is seen in Fig. 3A. Heating for 30 min at 62°C produced only a slight decrease in epoxide hydrolase activity from C57BL/6J microsomes, but the enzyme activity from DBA/2J microsomes fell to less than 3% of its initial value. In B6D2F1/J microsomes, the enzyme activity showed a heat sensitivity pattern that was approximately intermediate to that of the parent strains, and similar to the inactivation pattern seen with a 1:l mixture (v/v) of C57BL/6J and DBA/2J microsomes (Fig. 3A). The relative heat stability of epoxide hydrolase activity in hepatic microsomes from C57BL/6J, DBA/2J, and B6D2FI/J mice was similar when either styrene oxide or benzo[a]pyrene-4,5-oxide was used as the substrate (Fig. 3B).
Strain Distribution-The distribution of these two phenotypes (pH optimum and thermolability) in 26 inbred strains is summarized in Table I. We propose Eph-1 as the locus symbol for microsomal epoxide hydrolase, with superscripts b and d designating the alleles inherited from C57BL/6J and DBA/2J mice, respectively.
Linkage of Microsomal Epoxide Hydrolase Activity to Mouse Chromosome 1-The data on microsomal epoxide hydrolase activity in liver microsomes from C57BL/6J, DBA/ 25, and 24 recombinant inbred (RI) mouse strains derived from inbreeding the F2 generation of the C57BL/6J X DBA/ 25 cross (14,19) is summarized in Table 11. The mice were phenotyped for epoxide hydrolase alleles by determining the ratio of enzyme activity at pH 9.5/pH 8.7. Thirteen of the twenty-four RI strains have inherited the "B" allele of epoxide hydrolase from C57BL/6J while 11 strains have inherited the "D" allele from DBA/2J. This approximately 1:l distribution of alleles among the 24 RI strains is consistent with the inheritance of a randomly assorting single gene. Table I1 shows the co-segregation of alleles of Eph-I with three markers on the distal portion of Chromosome 1. They are: Sas-1, serum antigenic substance (20); Ltw-4, a soluble liver protein variant identified by two-dimensional gel electrophoresis (21); and Mls, the mouse mixed lymphocyte stimulatory locus (22). The distribution of alleles a t these three loci among the BXD RI strains had previously been determined (23,24).' A highly significant ( This gene order does not require the postulation of any double cross-overs in the Sas-1-Ltw-4-Eph-1 region and further supports this gene arrangement. The data would not preclude the placement of Eph-1 distal to Mls, but such an arrangement would represent a less likely interpretation of the data. A partial linkage map of the distal end of mouse Chromosome 1 is shown in Fig. 4. The recombination frequency, estimated by the method of Taylor et al. (25), between Eph-I and Ltw-4 is 3.9 +. 2.6%; between Eph-1 and MLs this value is also 3.9 k 2.6%; and between Sas-1 and Ltw-4 this value is 7.6 f 4.4%. These values are in agreement with previously estimated map distances for this portion of the chromosome.

DISCUSSION
We have presented evidence that there is a genetic polymorphism of hepatic microsomal epoxide hydrolase activity *R. W. Elliott, C. Romejko, and C. Hohman (1980) manuscript submitted for publication.

Murine Polymorphism of Microsomal Epoxide Hydrolase
among inbred mouse strains. Two phenotypes have been distinguished by the criteria of pH optima and thermolability. Our results are consistent with the hypothesis that these phenotypes are a result of allelic variants of a single structural gene for microsomal epoxide hydrolase which is located near the end of Chromosome 1. Other investigators (12, 13) have noted genetic variation in hepatic epoxide hydrolase activity between inbred strains of mice, but suboptimal assay conditions prevented their identification of this polymorphism.
Hepatic microsomal epoxide hydrolase has been purified from a number of species including mouse (26), but has been studied most extensively in the rat. Rabbit antibody to the purified rat protein gave a single immunoprecipitin band to both solubilized microsomal and nuclear membrane fractions of rat liver (27) and to solubilized hepatic microsomes from C57BL/6J and DBA/BJ mice (28). This suggests that there is a single species of this enzyme in rat microsomes, and this species cross-reacts with the mouse enzyme. However, recent evidence (29,30) suggest that there may be several very closely related species of microsomal epoxide hydrolase in rat liver. At present the problem of whether or not multiple species of epoxide hydrolase exist remains unresolved. Our data indicate that the Eph-I locus codes for two allelic forms of microsomal epoxide hydrolase in mice, but we cannot state that this is the only structural gene for this enzyme.
A cytosolic epoxide hydrolase enzyme with catalytic properties that are markedly different from the microsomal enzyme has been described r e~e n t l y .~ Styrene oxide, a good substrate for the microsomal enzyme, is not metabolized at all by the cytosolic enzyme. Differences in substrate specificity, pH optimum, and molecular weight (31) indicate that the cytosolic enzyme is not a solubilized form of the microsomal enzyme and therefore is most likely coded for by a separate structural gene.
The only other microsome drug-metabolizing enzyme polymorphism to be mapped in the mouse is coumarin hydroxylase, whose controlling locus (Coh) is located near the centromere of Chromosome 7 (32). Both enzyme activities are induced by phenobarbital, but not by 3-methylcholanthrene. It is of interest to note that Chromosomes 1 and 7 have been hypothesized to have been derived from a common ancestral chromosome by the process of tetraploidization (33). Although the evidence presented in support of this hypothesis is weak, the present assignment of Eph-I to Chromosome 1 would be consistent with the hypothesis if coumarin hydroxylase and epoxide hydrolase were proven to be paralogous. In any case, it appears that microsomal drug-metabolizing enzymes are controlled by genes located at multiple sites in the mouse genome.
In this report we have not examined whether this polymorphism has any consequences in viuo; i.e. if a mouse with one allelic form of epoxide hydrolase is more susceptible to the toxic effects of reactive epoxides. Indirect evidence suggests that it is unlikely that differences in this enzyme will play a critical role in the metabolism of epoxides. Epoxides a.re inactivated by several other routes, including enzymatic and nonenzymatic conjugation with glutathione (34,35). We examined the metabolism of styrene oxide in vitro, at pH 6.9 (intracellular pH), and found hepatic microsomal epoxide hydrolase activity from C57BL/6J mice had a VmaX of 4.9 nmol/mg/min and K , of 0.13 m~, and from DBA/W mice a V,,, of 8.6 nmol/mg/min, and K,,, of 0.32 m~. We think that these small differences in K,,, and Vmax are unlikely to be significant in viuo.