Epoxides as Obligatory Intermediates in the Metabolism of Olefins to Glycols*

SUMMARY In the presence of rat liver microsomes and NADPH n-l-octene, n-4-octene and 3-ethyl-2-pentene were converted to the glycols with no trace of epoxides. Increased substitution of ethylenic hydxogen atoms by alkyl groups was found to retard. the rate of biological oxid.ation but to enhance that of epoxidation by perbenzoic acid. in chloroform. Microsomes without cofactors hydrolyzed the monosubstituted ethylene oxide more rapidly than the di- or trisubstituted derivatives. The relative rates were in the opposite order of those pre-dicted for acid-catalyzed hydrolysis. The epoxides were found capable of inhibiting epoxide hydrolase. Incubation of microsomes and NADPH with 1 mM n-1-octene in the presence of 20 m&r 4,5-epoxy-n-octane yielded both 1,2-epoxy-n-octane and n-octane-l ,2-diol. However, in the presence of 20 mM 1,2-epoxy-n-octane, 1 mM n-4-octene yielded 4,5-epoxy-n-octane but no n-octane-4,5-diol.


Epoxides as Obligatory
Intermediates in the Metabolism of Olefins to Glycols* (Received for publication, October 20, 1969) E. W. n/IAYNERT, R. L. FOREMAN, AND T. WATABE$ From the Department of Pharmacology, College of Medicine, University of Illinois, Chicago, Illinois 60680 SUMMARY In the presence of rat liver microsomes and NADPH n-loctene, n-4-octene and 3-ethyl-2-pentene were converted to the glycols with no trace of epoxides.
Increased substitution of ethylenic hydxogen atoms by alkyl groups was found to retard. the rate of biological oxid.ation but to enhance that of epoxidation by perbenzoic acid. in chloroform. Microsomes without cofactors hydrolyzed the monosubstituted ethylene oxide more rapidly than the di-or trisubstituted derivatives. The relative rates were in the opposite order of those predicted for acid-catalyzed hydrolysis. The epoxides were found capable of inhibiting epoxide hydrolase. Incubation of microsomes and NADPH with 1 mM n-1-octene in the presence of 20 m&r 4,5-epoxy-n-octane yielded both 1,2epoxy-n-octane and n-octane-l ,2-diol. However, in the presence of 20 mM 1,2-epoxy-n-octane, 1 mM n-4-octene yielded 4,5-epoxy-n-octane but no n-octane-4,5-diol. The complete replacement of n-octane-4,5-diol by 4,5-epoxy-noctane in the presence of the inhibitor indicates that the epoxide is an obligatory intermediate in the conversion of n-4-octene to the glycol.
A few compounds containing a carbon-carbon double bond are metabolized to epoxides (l-3), but the usual products are glycols (4-6).
At the beginning of the present investigation, the role of epoxides in the biological formation of glycols was uncertain.
It seemed possible that the glycols might arise from either enzymatic or spontaneous hydrolysis of epoxides.
However, direct dihydroxylation of the ethylenic moiety could not be ruled out. An attractive approach to this problem was to focus attention on a few series of olefins, epoxides, and glycols in which the epoxides are quite stable in water.
The conditions for sedimentation of the various subcellular fractions are specified elsewhere in the text. In all experiments the aliquots were equivalent to 2 g of liver. The composition of the incubation medium and other experimental details are presented in the legends of appropriate figures and tables.
Analytical Methods-The same analytical procedure was used for the quantification of the metabolites of the three olefins. The chilled reaction mixture plus 5 g of NaCl was extracted with 25 ml of ether. A 20-ml portion of the ether layer was removed after centrifugation and cautiously evaporated to minimize the loss of volatile metabolites.
The chilled residue was dissolved in 1 ml of acetone, and aliquots of this solution were analyzed in a model 5000 Barber-Colman gas chromatograph equipped with an ionization detector. The composition of the column and other experimental conditions are specified in the legend of Fig.  1. When added to boiled subcellular fractions, the glycols (0.1 to 1 pmole) were recovered almost quantitatively (90%). The recovery of 1,2-epoxy-n-octane or 2,3-epoxy%ethylpentane averaged 90%, but that of 4,5-epoxy-n-octane was only 70%. All data were corrected for recovery losses.
With each olefin the identity of the epoxide and glycol metabolites was checked by both gas-liquid and thin layer cochromatography.

Nonenzymatic Synthesis
and Nydrolysis of Epoxides-Prior to the initiation of the biochemical work it seemed worthwhile to obtain some data on the influence of branching on the nonenzymatic conversion of olefins to epoxides. Table I shows that the rate constants for perbenzoic acid oxidation increased strikingly with increased substitution of the ethylenic hydrogen atoms. Thus, in this reaction, the inductive effects of the alkyl groups outweigh steric hindrance. This finding is in agreement with the observations reviewed by Swern (9). Nonenzymatic hydrolysis of epoxides was not investigated beyond demonstrating that at pH 3, 7, or 10 saturated aqueous solutions of 1,2-epoxy-n-octane, 4,5-epoxy-n-octane, or 2,3epoxy-3-ethylpentane were stable at 37". Incubation of 0.5 M solutions of these epoxides in 40% acetone for 6 hours resulted in less than 0.01% conversion to the glycols. Information on the behavior of epoxides in neutral solutions has been scarce, but both acids and bases have been shown to catalyze cleavage of the ring. The study of Prichard and Long (10) indicates that electron-releasing substituents, including alkyl groups, facilitate acid-catalyzed hydrolysis of ethylene oxides. On the other hand, alkaline hydrolysis appears to be influenced by steric as well as inductive effects, and generalizations are not yet possible.
Emymatic Oxidation of Olejins-Incubation of n-1-octene, trans-n-4-octene, and 3-ethyl-2-pentene with the NADPHenriched 9000 x g supernatant of a rat liver homogenate produced the corresponding glycols, but the epoxides could not, be detected (Fig. 1). The limit for the gas chromatographic detection of the epoxides was 0.3 nmoles or a 0.06% yield based on the initial amount of olefin.
The identity of the glycols was confirmed by thin layer and gas cochromatography. The relative yields of the glycols (11.3%, 4.0'%, 0.12%) indicate that increasing substitution of the ethylenic moiety by alkyl groups decreases the rate of the reaction. Thus, here, steric hindrance from the alkyl groups appears to outweigh their inductive effects.
The product from n-1-octene contained a trace of an unknown metabolite with a retention time (ET) lower than that of the epoxide. Both of the other olefins yielded at least three or four 1. Gas chromatograms of t'he products from n-l-octene (A), trans-n-4-octene (B), and 3-ethyl-2-pentene (C) after incubation with the 9000 X g supernatant fluid of a rat liver homogenate.
In each panel the left arrow indicates the expected position of the epoxide, and the right arrow, the glycol; the large peaks ore the left represent acetone and unchanged olefin. Mixtures of 10 Gmoles of an olefin dissolved in ethanol (0.2 ml), 100 pmoles of MgC&, 50 moles of nicotinamide, 3 pmoles of NADP, 50 kmoles of glucose g-phosphate, supernatant from a homogenate of 2 g of liver centrifuged at 9000 X g for 20 min, and sufficient 70 mM phosphate (pEI 7.4) to bring the volume to 9.5 ml were incubated for 60 min at 37' in closed vessels containing an oxygen atmosphere. Extracts were prepared as described under "Methods." The chromatographic column (6 feet X 3 mm, inner diameter) packed with 20yo Apiezon 1, on 100 to 120 mesh Gas Chrom Q was eluted with argon (60 ml per min). The column temperature was held at 80" (60' for the products from 3-ethyl-2-pentene) for 5 min and then increased to 180" at a rate of 10" per min. The yields of the glycols from 10 @molesof the olefins were 11.3% in A, 4.0% in B, and 0.12% in C. compounds in addition to the glycols. The most abundant metabolite from n-4-octene (RT = 10.7 min), like the glycol (RT = 14.9 min), increased linearly with time of incubation.
The following evidence suggests that this compound is an octenol. First, its RF is similar to that of 2-octanol (10.9 min). Second, treatment of the extract with an alcohol reagent (acetyl chloride) caused this substance to disappear and a new compound with a shorter RT to appear. Third, the addition of bromine to the extract converted the metabolite to a derivative with lower volatility. The formation of the other two unknown products of n-4-octene was not clearly related to incubation time. No attempt was made to identify these substances or the unidentified metabolites of 3-ethyl-2-pentene.
The enzymatic conversion of the olefins to glycols was found to require microsomes and an NADPH-generating system (Table II). Twice-washed microsomes plus NADPH produced The rate of hydrolysis of 4,5-epoxy-n-octane by microsomes phate dehydrogenase was added to the medium. was slower than that of 1,2-epoxy-n-octane but faster than that of 2,3-epoxy-3-ethylpentane. ' Thus, this reaction diiers from L of the new product from 4,5-epoxy-n-octane was 13.1 min, -and from 2,3-epoxy-3-ethylpentane, 14.0 min (cj. Fig. 1). In the former case, the peak height was only 30% as large as that because they could not be detected in crude synthetic products The substrate was 10 pmoles of 1,2-epoxy-n-octane, 4,5-epoxy-expected to contain all possible isomers. They may represent n-octane, or 2,3-epoxy-3-ethylpentane added to the incubation alcoholic derivatives of the epoxides. In respect to its relative mixture dissolved in 0.2 ml of ethanol.
Methods and conditions activity in hydrolyzing 1,2-and 4,5-epoxy-n-octane, the enzyme were t,he same as those described in Table II and Fig. 1 except in the 165,000 X g supernatant behaved much the same as that that. the incubation time was 15 min. The experiment without in the microsomes. This observation indicates that the two cofactors involved the omission of nicotinamide and MgClz as well as the NADPH-generating system. enzymes may be the same.

Inhibition of Epoxide Hydrolase-Recognition that microsomes
Fraction n-Oc&e-1,2-n-octfi-4,s 3-Ethylpen-0 tane-Z+Iiol contain an epoxide hydrolase suggested that inhibition of this enzyme would provide useful information on the role of epoxides p?wles umoles J&moles in the conversion of olefins to glycols. Among a number of 9,000 X g supernatant + epoxides and glycols examined for such inhibitory action, only NADPH . . 9.95 4, B-epoxy-n-octane was found capable of revealing 1,2-epoxy-n-Boiled 9,000 X g supernatant + octane as a metabolite of n-1-octene in a system containing NADPH 0.00 0.00 0.00 microsomes and NADPH ( pounds was practically the same as the amount of glycol formed in the absence of the inhibitor (0.63 versus 0.64 pmole).

Control experiments established the requirement of NADPH and
the glycols in about the same amounts as did the 9000 x g super-gaseous oxygen for the appearance of the epoxide. Although natant. Moreover, the products from the individual olefins apparently ineffective as inhibitors of epoxide hydrolase, heptadid not differ appreciably from those shown in Fig. 1. Thus, chlor epoxide and dieldrin reduced the formation of n-octaneit would appear that soluble enzymes are not responsible for the 1 , 2-diol. It seems possible that these epoxides may have an formation of any of the unidentified metabolites. inhibitory effect on the epoxidase involved in the conversion of Enzymatic Hydrolysis of Epoxides-The absence of epoxides n-1-octene to 1,2-epoxy-n-octane. However, this question was and the presence of glycols in the products from the biological not investigated.
The data in Table  microsomes of 1,2-epoxy-n-octane and 4,5-epoxy-n-octane were III show that incubation of 10 pmoles of 1,2-epoxy-n-octane for found to be 695 and 78 nmoles per min per g of liver, respectively.
15 min with a 9,000 x g supernatant equivalent to 2 g of rat The substrate concentrations were 1 mM, and the enzyme concentration, 100 mg of liver per ml. Both rate curves were linear liver resulted in a quantitative conversion to n-octane-l ,2-diol. for at least 10 min.  OH OH E$P( 1 of potential inhibztors oj' epoxitle /yd~olase on metabolism of n-l -0ctene by rat liver mcro.somes The sltbstrate was 10 pmoles of the olefin added to the incubation mixtllre in 0.2 ml of ethanol in Tvhich the inhibitor u'as also dissolved or, in the case of heptachlor eposide and dieldrin, SKpended.
The medium colitnined the XADPH-generating system.
The incubation time was 30 min. Other methods and conditions mere the same as those described in Fig. 1 and Table  II. In several other experiments employing the 4,5-epoxide, the yields of the 1,2-epoxide and the 1,2-din1 equalled that of the glycol in the absence of the inhibitor. showing the inhibitory effect of 20 mnr 1,2-epoxy-n-octane on the metabolism of n-4-octene to n-octane-4,5-diol by rat liver microsomes. The left arrow indicates the peak from 4,5-epoxy-n-octane, and t.he right arrow, the expect'ed position of n-oct.ane-4,5-diol.
In separate experiments all the peaks specified were identified by cochromatography of authentic compounds and the extract. Methods and conditions were the same as those in Table  IV. The slight differences in retention times from those in Fig. 1 can be att,ributed to aging of t'he chromatographic column.
suggested that a monosubstituted ethylene oxide might provide maximal blocking activity.
The quantity of 4,5-epoxy-n-octane in the product was approximately equivalent to the amount of glycol formed in the absence of the inhibitor.
An exact numerical comparison was precluded by the magnification of a trace (0.2yC) of the 4,5-epoxide in the 1,2-epoxide to substantial proportions as a result of the high concentraOion of inhibitor used in this experiment.
Direct evidence that these results should be attributed to inhibition of epoxide hydrolase was obtained by employing an epoxide as the substrate.
,4 low concentration (0.3 m$ of 1,2-epoxy-n-octane inhibited completely for 10 min the hydrolysis of 5 rn>I 4,5-epoxy-n-octane by washed microsomes at a concentration of 25 mg of liver per ml.
Recently, Leibman and Ortiz (11) reported the dihydroxylation of indene and styrene by a similar microsomal preparation.
Thus, it would appear that hepatic microsomes contain the necessary enzymes for the conversion of both aliphatic and aromatic double bonds to diols. The capacity of microsomes and NADPH to form stable epoxides from cyclodiene insecticides such as aldrin and heptachlor has been recognized for several years (13, 14).
Inasmuch as the conversion of an epoxide to a glycol ha:: classically been formulated as a hydrolytic reaction, the appropriate name for the enzyme would appear to be epoxide hydrolase.
Like other hydrolytic enzymes such as esterases and amidases, it does not require NADPH or magnesium ion as a cofactor.
Presumably, the bulk of the alkyl groups hinders the approach of the epoxide to the surface of the enzyme. The facile cleavage of the oxirane ring by the hydrolase could account for the fact that epoxides have only rarely been detected as end products of metabolism.
Heptachlor epoxide and dieldrin, which are stable enough to be excreted in substantial amounts, Epoxide Intermediates in Olefin Metabolism Vol. 245,No. 20 may be presumed to be poor substrates for the enzyme, although rats can convert dieldrin to the trans form of the corresponding glycol (19). In the present investigation, the metabolism of heptachlor epoxide and dieldrin was not examined beyond demonstrating that they did not inhibit the action of microsomal epoxide hydrolase on 1, Z-epoxy-n-octane.
The discovery that high concentrations of some epoxides could inhibit the microsomal hydrolase provided the means for examining the role of epoxides in the conversion of olefins to glycols. In the presence of 20 mM 4,5-epoxy-n-octane, the product from n-1-octene contained both 1,2-epoxy-n-octane and n-octane-1 ,2-diol, whereas in the absence of the inhibitor only the glycol could be detected.
This result could be explained in two ways. (a) The only route from the olefin to the glycol involves the epoxide, but inhibition of the hydrolase was incomplete.
(b) Glycol may be formed by a mechanism other than epoxidation, for example, by direct dihydroxylation (Fig. 3). A similar experiment in which the substrate was n-4-octene and the inhibitor was 20 mM 1 ,2-epoxy-n-octane gave unequivocal results. In the presence of the inhibitor, the product contained the epoxide but not the glycol, whereas, in the absence of the inhibitor, the glycol was present, but the epoxide was not. This observation appears to prove that the epoxide is an obligatory intermediate in the conversion of n&octene to the corresponding glycol.
It now seems likely that the biological conversion of both aliphatic and aromatic carbon-carbon double bonds proceeds through epoxides.
In connection with aromatic compounds, the evidence of Holtzman, Gillette, and Milne (20) is strong, although indirect.
They found that trans-1 ,2-dihydro-1 , Z-dihydroxynaphthalene formed from naphthalene in the presence of microsomes, NADPH, and gaseous l8O2 contained only 1 atom of heavy oxygen.
Recently, Jerina et al. (17) reported the detection of 1 ,2-naphthalene oxide as a metabolite of naphthalene and suggested its role as an obligatory intermediate in the formation of the dihydrodiol as well as 1-naphthol and a premercapturic acid. Whether a single microsomal hydrolase acts on both aliphatic and aromatic epoxides is not yet clear. Studies involving inhibitors such as 1, Z-epoxy-n-octane should answer this question.