Stereoselectivity of the epoxidation of 7,8-dihydrobenzo[a]pyrene by prostaglandin H synthase and cytochrome P-450 determined by the identification of polyguanylic acid adducts.

The stereoselectivity of the oxidation of 7,8-dihydrobenzo[a]pyrene (H2BP) to 9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (H4BP-epoxide) by prostaglandin H (PGH) synthase and cytochrome P-450 has been studied using microsomal preparations from ram seminal vesicles and rat liver. Incubations were performed in the presence of polyguanylic acid and the adducts formed between H4BP-epoxide and guanosine were isolated following the recovery and hydrolysis of the poly(G). When (+/-)-H4BP-epoxide was reacted with poly(G), four diastereomeric adducts were formed by the cis and trans addition of the exocyclic amino group of guanine to the benzylic carbon of the epoxide enantiomers. Each diastereomer was identified by a combination of ultraviolet, nuclear magnetic resonance, circular dichroism, and mass spectroscopy. Under comparable conditions, ram seminal vesicle microsomes in the presence of arachidonic acid triggered the binding of H2BP to poly(G) to a greater extent than rat liver microsomes from untreated and phenobarbital- and methylcholanthrene pretreated animals in the presence of NADPH. Quantitation of the (-)-cis- and (+)-cis-guanosine adducts revealed the degree of stereoselectivity of epoxidation. The ratio of (-)/(+) adducts was 54:46 for PGH synthase and 89:11 (control), 62:38 (phenobarbital), and 69:31 (methylcholanthrene) for cytochrome P-450-catalyzed reactions. PGH synthase catalyzed the epoxidation of H2BP with little or no stereoselectivity in contrast to cytochrome P-450. The utility of the poly(G) binding technique for the elucidation of the stereoselective generation of chiral electrophiles is discussed along with the mechanistic implications of the results.

The stereoselectivity of the oxidation of 7,8-dihydrobenzo[u]pyrene (H2BP) to 9,10-epoxy-7,8,9,10-tetrahydrobenzo[u]pyrene mBP-epoxide) by prostaglandin H (PGH) synthase and cytochrome P-450 has been studied using microsomal preparations from ram seminal vesicles and rat liver. Incubations were performed in the presence of polyguanylic acid and the adducts formed between &BP-epoxide and guanosine were isolated following the recovery and hydrolysis of the poly(G). When (-+)-&BP-epoxide was reacted with poly(G), four diastereomeric adducts were formed by the cis and tmns addition of the exocyclic amino group of guanine to the benzylic carbon of the epoxide enantiomers. Each diastereomer was identified by a combination of ultraviolet, nuclear magnetic resonance, circular dichroism, and mass spectroscopy. Under comparable conditions, ram seminal vesicle microsomes in the presence of arachidonic acid triggered the binding of HzBP to poly(G) to a greater extent than rat liver microsomes from untreated and phenobarbital-and methylcholanthrene pretreated animals in the presence of NADPH. Quantitation of the (-)-cis-and (+)-cis-guanosine adducts revealed the degree of stereoselectivity of epoxidation. The ratio of (-)/(+) adducts was 54:46 for PGH synthase and 89:ll (control), 62:38 (phenobarbital), and 69:31 (methylcholanthrene) for cytochrome P-450-catalyzed reactions. PGH synthase catalyzed the epoxidation of HzBP with little or no stereoselectivity in contrast to cytochrome P-450. The utility of the poly(G) binding technique for the elucidation of the stereoselective generation of chiral electrophiles is discussed along with the mechanistic implications of the results.
The stereochemistry of epoxidation of bay region double bonds of polycyclic hydrocarbons is of particular importance since different enantiomers of diolepoxides differ in their reactivity toward nucleic acids and in their mutagenic and carcinogenic potency (29)(30)(31)(32)(33)(34)(35)(36). In the case of BP-7,8-diol, the stereochemistry of epoxidation has been elucidated by metabolizing optically pure enantiomers resolved from (+)-BP-7,8diol and analyzing the enantiomeric epoxide hydrolysis products formed (15-17, 37). We have developed an alternative method for determining the stereochemistry of epoxidation which utilizes a chiral nucleophile, poly(G), to trap enantiomeric epoxides (38). After hydrolysis to nucleosides, the resulting diastereomeric adducts are separated and quantitated by HPLC (39). Using this technique, we have shown that PGH synthase oxidizes both enantiomers of (+)-BP-7,8-diol to an equal mixture of enantiomers of diolepoxide 2 (38).
HzBP is an achiral molecule, but can be converted to enantiomeric KBP-epoxides and, therefore, is enantiotopic (Equation 2). Since the starting material cannot be resolved,

HJ3P Epoxidation by PGH Synthase and Cytochrome P-450
HZBP I 9 S , 10R ) ( 9 R , 1 0 S ) our chiral nucleophile approach is uniquely suited to define the stereochemistry of the intermediate epoxides.
The present study was undertaken in order to compare the stereochemistry of epoxidation of H2BP by PGH synthase and by the mixed function oxidase RLM. As a result of this study, we are able to assign not only the relative stereochemistry of oxidation by both enzymes but also the absolute stereochemistry of the nucleoside adducts.

EXPERIMENTAL PROCEDURES
Materials-HzBP and HdBP-epoxide were synthesized by a modification of published procedures (26,40 Microsomes from induced animals were prepared from animals injected intraperitoneally with MC (25 mg/kg) in corn oil or PB (80 mg/kg) in saline for 3 consecutive days prior to death.
Preparation of 9-Acetoxy-lO-anilino-7,8,9,10-tetrahydrobenzo[a] pyrene-9-Acetoxy-10-anilino-7,8,9,I0-tetrahydrobenzo[a]pyrene was prepared using the method of Yagi et al. (43). Fifty mg of H,BPepoxide was dissolved in 50 ml of dry t-butyl alcohol. Twenty-five mg of aniline in 2 ml of tetrahydrofuran was added dropwise and the mixture was stirred at room temperature under argon for 24 h. After diluting with 200 ml of water, the reaction mixture was extracted with chloroform. The chloroform extract was washed with ice-cold 0.05 M HCl and then with water, dried over anhydrous magnesium sulfate, and the solvent evaporated. The residue was purified by chromatography on a Varian 5000 HPLC equipped with a Valco injector and Varichrome detector. The samples were injected onto an Ultrasphere ODS column (10 X 250 m m ) and eluted with a linear gradient of 70-95% aqueous methanol in 30 min at a flow rate of 3 ml/min. The two major peaks (retention times 19 min and 25 mi n, respectively) were collected, dried, and acetylated with pyridine/acetic anhydride at room temperature for 18 hr. The resulting monoacetates were purified by HPLC under conditions described above for 9-hydroxy-10-anilino- Large Scale Preparation of HSP-epoxide-Guanosine Adducts-Twenty-five mg of poly(G) was dissolved in 10 ml of water and the pH of the solution was adjusted to 7.0. Twenty five mg of H,BPepoxide was added in 10 ml of acetone and the mixture was incubated at 37 "C for 1 h. The resulting solution was extracted with ethyl acetate (5 X 10 ml) and the poly(G) was reisolated as described before (38). The poly(G) was hydrolyzed with 1.0 M KOH at 37 "C for 18 h and then digested to guanosine adducts with alkaline phosphatase by the method of Moore et al. (44). After hydrolysis, the solution was lyophilized and the residue was introduced onto a Sephadex LH-20 column (1 X 20 cm) as a suspension in water. The column was washed with water to remove protein, salts, and unreacted nucleosides. The nucleoside adducts were eluted with methanol. The methanol eluent was reduced to a small volume and analyzed by HPLC.
HPLC Separation of Hap-epoxide-Guanosine Adducts-The samples were injected onto an Ultrasphere ODS column (4.6 x 250 m m ) and eluted with a linear gradient of 45-70% aqueous methanol over a time span of 50 min at a flow rate of 1 ml/min. The eluent was monitored at 344 nm. For large scale separation of adducts, samples were injected onto an Ultrasphere ODs, preparative column (10 x 250 m m ) and eluted with the same gradient used in analytical separation except that the flow rate was changed to 3 ml/min. The adducts were separated into four peaks with retention times of 31,36,38, and 44 mi n. The retention times were approximately the same on the analytical or preparative column.
Acid Hydrolysis of HSP-epoxide-Guanosine Adducts-The diastereomeric adducts (-100 nmol) were individually hydrolyzed in 0.05 M aqueous HCl by heating at 90 "C for 30 min. The solution was neutralized with aqueous sodium bicarbonate solution and extracted with ethyl acetate. The ethyl acetate extract was dried over anhydrous magnesium sulfate and the solvent was evaporated. Analysis of the residue by HPLC on an Ultrasphere ODS column under conditions described for KBP-epoxide-guanosine adduct gave two peaks with retention times of 37 and 51 min.
Attempted Base Hydrolysis of Hap-Guanosine Adducts-A mixture of purified H4BP-epoxide-guanosine adducts (containing -5 nmol of each adduct) was heated in 1 ml of 1 M KOH at 100 "C for 1 h. After cooling to room temperature, the solution was neutralized with 1.0 M HCl to pH 7.0 and extracted with ethyl acetate. The extract was dried and evaporated. The residue was analyzed by HPLC under identical conditions as those used for H4BP-epoxide-guanosine adducts. The four peaks eluted had the same retention time as those for H4BP-epoxide-guanosine adducts. Traces of minor degradation products were eluted at longer retention times.
Determination of pK, Values of HSP-epoxide-Guanosine Adducts-The pK, values of the KBP-epoxide adducts were determined by the method of Moore and Koreeda (45). The adducts (-0.3 A) were partitioned between 1-ml buffer solutions of pH 1.0-12.0 and 1 ml of 25% 1-butanol in ethyl acetate. The amounts of adduct in the aqueous phase were measured spectrophotometrically (45).
CD Spectra-CD spectra of KBP-epoxide-guanosine adducts were taken in methanol on a Jasco-40 spectropolarimeter equipped with a Nova-3 data processor and a Tracor Northern digital signal processor.
Mass Spectra-The field desorption mass spectrum of an underivatized H4BP-epoxide-guanosine adduct (peak B) was recorded on a Varian Mat CH-5 double-focussing mass spectrometer. Metabolism of HzBP and Poly(G) Binding in the Presence of RSVM-Incubation mixtures contained 500 nmol of HzBP, 5 mg of poly(G), -5 mg of RSVM protein, and 20 1. 11 of acetone in a total volume of 5 ml of 0.01 M Na2HPO4 (pH 7.5). The reaction was initiated by the addition of 204 (50 PM) and the incubation was carried out for 30 min at 37 "C. The reaction was stopped by the addition of 5 ml of phenol reagent (phenol (500 g), m-cresol (70 ml), water (50 ml), and 8-hydroxyquinoline (0.5 g)). Two ml of 0.10 M NaCl was added and the samples were extracted as before (38). The aqueous layers were combined and extracted with ethyl acetate (3 X 5 ml). The poly(G) was reisolated and digested to guanosine adducts according to the procedure of Moore et al. (44). The adducts were purified by Sephadex LH-20 chromatography and analyzed by HPLC under conditions described for the separation of H4BP-epoxide-guanosine adducts. Duplicate incubations were combined for LH-20 separation and HPLC analysis. The individual peaks were collected and UV-Vis spectra were recorded. The adducts were quantitated from their absorbance at 344 nm using an extinction coefficient of 37,000 for the tetrahydrobenzo[a]pyrene moiety (44). The peaks were separately co-injected with standard adducts to establish their identity.
Metabolism of H2BP and Poly(G) Binding in the Presence of Rat Liver Microsomes-Incubation mixtures contained 500 nmol of H2BP, 5 mg of poly(G), 5-6 mg of microsomal protein (MC-induced, PB-induced, or control), and 20 pl of acetone in a total volume of 5 ml of 50 mM Tris-HC1, pH 7.5, containing 10 m~ MgCh and 1 O p~ MnC12. The reaction was initiated by the addition of NADPH (0.5 mM). The incubations were carried out for 30 min at 37 "C. The reactions were stopped by the addition of phenol reagent and worked up as indicated. The reisolated poly(G) was hydrolyzed and adducts were quantitated as stated above.
Metabolism reisolated from the reaction mixture as before. It was hydrolyzed to guanosine adducts and purified by Sephadex LH-20 column chromatography. Standard guanosine adducts derived from poly(G) and diolepoxides were added prior to LH-20 purification (38). The guanosine adducts were analyzed according to a previously published procedure (39).

RESULTS
H4BP-epoxide was found to bind to poly(G) at neutral pH. Binding was characterized by the UV absorption pattern of the 7,8,9,10-tetrahydrobenzo[a]pyrene moiety in the 320-355 nm region and was similar to that observed for the binding of poly(G) to diolepoxides (44). Digestion of the hydrocarbonbound poly(G) to nucleosides followed by HPLC separation on an Ultrasphere ODS column gave four peaks which are shown in Fig. 1. The peaks do not co-chromatograph with the diol hydrolysis products of H4BP-epoxide. The HPLC profile in Fig. 1 shows that the products formed consist of two groups based on their peak area and suggests that the peaks of equal intensity are formed by trans or cis opening of the enantiomers of H4BP-epoxide. The structures of the four peaks in Fig.  1 were deduced by spectral and chemical methods.
Samples of the four peaks were collected and their UV-Vis spectra were recorded in methanol. The peaks gave identical spectra which were similar to those of &BP-diols and are characteristic of the presence of a tetrahydrobenzo[a]pyne chromophore.
Analysis of the peaks by low resolution field desorption mass spectrometry provided information on their molecular weight. Fig. 2 is the field desorption mass spectrum of peak B which shows prominent ions at m/e 576 (84%), 558 (14%), and 444 (100%). The ion at 576 represents a KBP-epoxide-guanosine adduct plus sodium. The fragment ions at 558 and 444 correspond to the loss of water and the loss of ribose with the transfer of a proton, respectively. The latter provides additional evidence for a guanosine derivative. Another ion seen at m/e 426 (16%) is consistent with the loss of a guanine moiety from the molecule. Other minor ions at m/e 306 (2%) l'BOl  and m/3a70 (2%) correspond to guanosine + Na and loss of guanosine and a proton from the molecule, respectively.
The CD spectra of the four peaks are shown in Fig. 3. The spectra of pairs of peaks A and B and C and D are essentially mirror images. On this basis, it can be reasonably concluded that the four peaks A to D are diastereomers. The spectra are very similar to those of diolepoxide-guanosine adducts (44). This gives further evidence that peaks A to D are formed by the reaction of guanosine with bBP-epoxide. The pairs of diastereomers with mirror image CD spectra would result from the cis or trans opening of the oxirane ring at C-10 of H,BP-epoxide. Thus, the pairs AB and CD should represent either the cis or trans pair of possible guanosine adducts formed from KBP-epoxide.
The cis/trans stereochemistry of the four adducts was assigned by NMR spectroscopy. Fig. 4 displays the NMR spectra of adducts A and D in the region 65.5-9 ppm. The pairs of diastereomers gave virtually identical NMR spectra in this region. The major differences in the NMR spectra of A and D are due to differences in the chemical shift value for the (3-10 proton. The C-10 proton signal in the NMR spectrum of A (65.95 ppm) is shifted downfield in the spectrum of D (66.41 ppm). Table I gives a comparison of chemical shift values for the C-10 proton of guanosine adducts, aniline adducts, and cis-and trans-KBP-diols. In the case of H,BPdiols, the C-10 proton of the cis-isomer is shifted downfield compared to the trans and this has been confirmed by authentic synthesis of cis-and trans-&BP diols (26). The downfield shift for the (2-10 proton in the cis isomer is in agreement with previous studies on NMR spectra of guanosine, adenosine, methanol, phenol, and aniline adducts derived from diolepoxides (43,46,47).
The position of substitution of the guanosine moiety in compounds A to D is deduced from the NMR spectra and by chemical methods. The NMR spectra of A and D in acetone (Fig. 4) show a doublet for the N2 proton at 6.48 and 6.54 ppm, respectively, due to coupling to the C-10 proton. The N2 proton signal and its coupling to the C-10 proton, is eliminated by deuterium exchange in CD30D. The possibility of substitution at C-8 of guanine was excluded by the presence of the C-8 proton signal at 7.9 ppm.
In order to provide further evidence for the N2 substitution, compounds A to D were treated with 0.05 M HC1. They were completely hydrolyzed in 30 min. This is in agreement with the data of Moore et al. (44) on the acid hydrolysis of diolepoxide-guanosine adducts. The products of hydrolysis were identified as cis-and trans-H4BP-diols and a small amount of KBP-9-one. This suggests that the site of attachment of guanosine to the H4BP moiety is through an oxygen or nitro-  cis-9,lO-Diol" 5.57 trans-9,10-Diolb
Levin et al. (16) have shown that in the presence of the purified MC-inducible P-450 monoxygenase system (+)-BP-7,8-diol is oxidized to give 42% diolepoxide 1-derived products and 58% diolepoxide 2-derived products. They have also shown that (+)-BP-diol gives predominantly diolepoxide 1 (97.5%) and the (-)-BP-7,8-diol gives diolepoxide 2 (82%). Deusch et al. (15, 37) and Belvedere et al. (50) have shown that (-)-BP-7,8-diol is oxidized to diolepoxide 2 by certain forms of PB-or P-naphthoflavone-inducible purified rabbit liver microsomes. These results are in good agreement with the data obtained from our poly(G) binding studies in the presence of MC-RLM and NADPH. The peaks in Fig. 5B with retention times at 30 min and 45 min are (-)-cis and (+)-trans adducts, respectively, from (+)-diolepoxide 2 and peaks with retention times 43 min and 36 min are (+)-cis and (-)-trans adducts, respectively, from (+)-diolepoxide 1. This demonstrates that our poly(G) trapping method gives results which are comparable to those obtained using conventional methods. Fig. 6a is the HPLC profile of the guanosine adducts formed when poly(G) was added to incubation mixtures containing H,BP, RSVM, and 204. The profile shows that the pairs of diastereomeric adducts are formed in almost equal amounts. In other words, both enantiomers of the epoxide are formed from the hydrocarbon. This is in agreement with our previous frndings on the poly(G) binding to BP-7,8-diol in the presence of PGH synthase (38). Fig. 6b represents the HPLC profile of the guanosine adducts resulting from incubation of poly(G) with HzBP in the presence of microsomal preparations from control rat liver and NADPH. It shows that the predominant guanosine adduct formed is the (-)-cis isomer (co-chromatographic with peak C in Fig. 1). The adduct profiles obtained from the reaction of HzBP with poly(G) in the presence of MC-induced and PB-induced RLM are qualitatively similar to the profile in Fig. 6b. This clearly demonstrates that NADPH-dependent mixed function oxidases oxidize HzBP to a specific enantiomer of HsBP-epoxide. A comparison of the HPLC profiles in Fig.  6 graphically illustrates the differences in the stereochemistry of epoxidation by PGH synthase and cytochromes P-450.
The amounts of guanosine adducts obtained from HzBP and poly(G) were quantitated from their absorption spectra. In all cases, the cis-and trans-diastereomers are formed in a ratio of 3:l. In some cases the trans adducts are formed in very small amounts and in such cases a combination of peak height and absorbance at 344 nm were used for quantitation. Since (-)-cis-and (+)-cis-diastereomers are produced from the two enantiomers of the epoxide, a comparison of the percentage of the cis-isomer is sufficient to demonstrate the stereochemistry of epoxidation. Table I11 is a comparison of the percentages of cis-isomers formed in the presence of PGH synthase and cytochromes P-450. The results indicate that the mixed function oxidases epoxidize HzBP stereoselectively while PGH synthase-dependent epoxidation is nonstereoselective.
A comparison of the relative amount of adducts formed in the presence of PGH synthase and cytochromes P-450 under identical incubation conditions, H2BP concentration, and microsomal protein concentration is also presented in Table 111. The highest amounts of adducts were isolated when HzBP was metabolized in the presence of RSVM and 20:4. The data in Table I11 show that in the presence of NADPH, MCinduced RLM oxidizes HzBP to a greater extent than PBinduced RLM and noninduced RLM. The difference in the amounts of adducts formed in the presence of RSVM and RLM may be due to differences in the levels of epoxide hydrolase activity. H4BP-epoxide has been shown to be a substrate for epoxide hydrolase and the specific and total activities of epoxide hydrolase are higher in RLM than in RSVM (26,27). These data demonstrate that the extent of epoxidation of dihydroaromatic hydrocarbons by PGH synthase can be comparable to that catalyzed by classical drug metabolizing enzymes. Although identical protein and substrate concentrations were employed in these in vitro experiments, it is not possible to extrapolate the results to the in vivo situation. Comparative studies intended to quantitate the relative contributions of PGH synthase and mixed function oxidases to xenobiotic metabolism must be performed in cellular preparations derived from target tissues.

I1
Comparison to literature precedents of the stereochemistry of (+)-BP-7,8-diol epoxidation by MC-induced RLM determined by the poly(G) method Poly(G) adducts isolated from incubation of (+)-BP-7,8-diol in the presence of MC-induced rat liver microsomes and NADPH.

BP-7,8-diol
Diastereomeric gua-Relative '% of the Relation between guanosine Relative % of Diolepoxides formed nosine adducts total adducts from adducts and optically active tetraols isolatedd formed" the diolepoxideb diolepoxide' The percentage is calculated from radioactivity co-eluting under the diastereomeric adduct peaks. Radioactivity was determined by collecting fractions every 30 s and counting in dioxane mixture. Quenching was corrected using a quench curve. "Relative yield of adducts are obtained by setting the lowest amount of adduct formed as 1 and represents the amount of adducts formed relative to that formed in the presence of PB-induced RLM and NADPH.

DISCUSSION
We have utilized poly(G) as a chiral nucleophile to trap enantiomeric epoxides generated during the oxidative metabolism of HZBP. When racemic KBP-epoxide is reacted with poly(G) and the reisolated nucleic acid hydrolyzed, four nucleoside adducts are isolated. Chemical and spectral evidence indicates that they are diastereomers formed by the cis and trans addition of the exocyclic amino group of guanine to the benzylic carbon of the enantiomeric epoxides. NMR and CD spectroscopy suggests that the fiist pair of compounds eluting from the reversed phase column (A and B) are trans and the second pair (C and D) cis adducts. When diolepoxides are reacted with poly(G), analogous adducts are formed (38, 39,44). The principal difference between the adducts formed from H4BP-epoxide and those formed from diolepoxides is the ratio of cis to trans adducts. The major adducts from H4BPepoxide are the cis adducts, whereas the major adducts from the diolepoxides are the trans adducts (38, 44). This may be due to differences in the conformation of the tetrahydrobenzo ring of H,BP-epoxide relative to the diolepoxides (51).
Kinoshita et al. (52) have recently characterized the deoxynucleoside adducts formed by the reaction of (+)-H,BPepoxide with calf thymus DNA. The major adducts arise by the addition of the exocyclic amino group of guanine to the benzylic epoxide carbon. All four cis and trans adducts are formed and the adducts from opposite enantiomers are formed in roughly equivalent amounts (52). The spectral properties of the individual deoxyguanosine adducts are very similar to those of the guanosine adducts and the relative order of elution from the reversed phase HPLC column is identical. In addition, the cis-deoxyguanosine adducts predominate (52). Thus, the reaction of H4BP-epoxide with the guanine residues of DNA appears to be very similar to its reaction with poly(G) even though DNA has more extensive secondary structure than poly(G).
The chiral nucleophile trapping approach has been used by others to elucidate the stereochemistry of the formation of polycyclic hydrocarbon epoxides (53,54). Nonenzymatic reaction of exogenous glutathione has been utilized to determine the stereoselectivity of the epoxidation, by a reconstituted mixed function oxidase, of BP and benzo[a]anthracene (53,54). Diastereomeric glutathione conjugates of BP-4,5-oxide and benzo[a]anthracene-5,6-, and 8,9-oxides were separated on HPLC and quantitated in order to estimate the enantiomeric composition of the epoxides generated. In addition, the diastereomeric composition of the glutathione adducts has revealed the regiochemistry and stereochemistry of the conjugation of BP-4,5-oxide by purified glutathione transferases (55).
An advantage of using poly(G) as the nucleophilic trapping agent is that it can be used with impure enzyme preparations. The polymeric adducts can be isolated and purified prior to hydrolysis to the nucleoside adducts, thereby removing contaminants present in the enzyme preparation. Furthermore, poly(G) is the most reactive of the synthetic polynucleotides toward polycyclic hydrocarbon epoxides (56). A major potential drawback of its use is that its secondary structure may cause preferential reaction with one of the enantiomers of a racemic mixture. This would compromise its utility for quantitative studies of stereoselective epoxide generation. We have not observed such preferential reactivity in our previous studies (38,39) but as a control we have quantitated the enantiomeric composition of the epoxides generated from (&)-BP-7,8diol by MC-induced RLM. Previous studies have shown that this system generates enantiomers of diolepoxides 1 and 2 with 9S,lOR absolute configuration in high optical purity (16,17). The data in Table I1 indicate that these two enantiomers are practically the only products detected in the incubations. This demonstrates that the poly(G) trapping method may be a useful general technique with which to defme the stereochemistry of the enzymatic generation of chiral electrophiles capable of diffusing into solution.
The stereoselectivity of HzBP epoxidation by PGH synthase and cytochrome P-450 is very similar to the stereoselectivity of BP-7,8-diol epoxidation by the same enzymes (16,17,37,38,50). Both enantiomers of XBP-epoxide are formed in equal amounts from HzBP by PGH synthase implying that there is no stereoselectivity in the introduction of the epoxide oxygen. In contrast, a high degree of stereoselectivity is observed in the NADPH-dependent epoxidation of H2BP by microsomal preparations from control, PB-, and MC-treated rats. From 6249% of the cis adducts are the (-)-diastereomer (adduct C). The degree of stereoselectivity of epoxidation by cytochrome P-450 appears to be slightly lower for H2BP than for BP-7,8-diol.
Jerina et al. (57) have recently proposed a model for the orientation of polycyclic hydrocarbons at the active site of cytochrome P-45Oc with respect to the locus of oxygen introduction. When this model is applied to H2BP, it predicts the preferential generation of KBP-epoxide with 9S,10R absolute configuration. Since MC-induced RLM contain >70% cytochrome P-45Oc (58), we can tentatively assign the absolute stereochemistry of the major trans and cis adducts and hence The dramatic differences in the stereoselectivity of epoxidation of HzBP may be indicative of differences in the mechanisms of epoxidation by cytochrome P-450 and PGH synthase. Oxygen insertion by cytochrome P-450 is believed to occur from a heme iron-oxo complex (59). Although epoxidation may be stepwise, the intermediates do not appear to diffuse away from the active site of the protein (60). The detailed mechanism of epoxidation by PGH synthase is not known. However, the source of the oxygen, stereoselectivity, hydroperoxide specificity, and sensitivity to antioxidants are analogous to the same features of epoxidations catalyzed by hematin in the presence of detergent (61). Hematin-catalyzed epoxidation appears to involve peroxy radicals derived from the unsaturated moiety of the hydroperoxide as oxidizing agent (61). If the mechanism of the PGH synthase-catalyzed epoxidation is similar, the polycyclic hydrocarbon substrate may be in a region removed from the heme prosthetic group or even in the bulk phase. As a result, the approach of the epoxidizing agent (the peroxy radical) may not be restricted to either of the enantiotopic faces, leading to nonstereoselective oxidation.
The stereoselectivity of enzymatic reactions is a reflection of the orientation of the substrate at the active site and provides information about the chirality of the active site and/or the mobility of the substrate. This is graphically illustrated by our comparison of the stereoselectivity of epoxidation of H2BP by PGH synthase and cytochrome P-450. Since these enzymes can be differentiated on the basis of their protein composition (62), substrate specificity (19), inhibitor sensitivity (21,63), and mechanism (23, 64), our findings provide an additional line of evidence which indicates that these two xenobiotic-metabolizing enzymes are complementary.
The present report demonstrates the utility of poly(G) trapping as a method for determination of the enantiomeric composition of enzymatically generated chiral electrophiles. Since enantiomeric electrophiles can exhibit significant differences in chemical reactivity and biological activity, information about the major enantiomeric metabolites of chiral or prochiral carcinogens can provide important clues to the identity of ultimate carcinogens. Conversely, it may also help explain why certain analogs of known carcinogens are not carcinogenic.