ARCHIVES

OF BIOCHEMISTRY

Vol. 298, No. 2, November

AND

BIOPHYSICS

1, pp. 697-702, 1992

Cytochrome P450,,,-Catalyzed Oxidation of a Hypersensitive Radical Probe Vaughn

P. Miller,

Department

Julia

A. Fruetel,

of Pharmaceutical

Chemistry,

and Paul R. Ortiz School

of Pharmacy,

de Montellano’

University

of California,

San Francisco,

California

94143-0446

Received June 2, 1992, and in revised form July 16, 1992

trans-1-Phenyl-2-vinylcyclopropane, a hypersensitive radical probe, is oxidized by cytochrome P450,,, (CYPlOl) to a diastereomeric mixture of the corresponding epoxide (Sl%), (trans-2-phenylcyclopropyl)acetaldehyde (6%), and trans-5-phenyl-2-penten1,5-diol(l3%). trans-5-Phenyl-2-penten-1-oland (trans2-phenylcyclopropyl)ethane1,2-diol are not detectably formed. Authentic standards of all the products have been synthesized and used to establish the identities (or the absence) of the metabolites. Studies with [“C]H,O demonstrate that the oxygens at positions 1 and 5 in the rearranged diol derive from molecular oxygen and water, respectively. Catalytic turnover of the enzyme is required for product formation from the olefin, but incubation of the epoxide metabolite with the enzyme, or with buffer alone, yields both the aldehyde and the rearranged diol products. The absence of trans-5-phenyl-2-pentenl-01 implies that the lifetime of the putative radical intermediate is so short that its existence as a discrete entity is questionable. A cationic intermediate is unlikely but cannot be excluded because the same metabolites are formed in a secondary reaction, even at pH 8.0, from the epoxide. The results provide no evidence for the involvement of radicals or cations in the epoxidation reaction, in agreement with results on the oxidation of olefins in organic solvents by metalloporphyrin catalysts. o 1992 Academic

Press,

Inc.

Olefin epoxidation by cytochrome P450 is thought to be mediated by a ferry1 (FeIV = O)/porphyrin (or protein) radical species analogous to that involved in the epoxidations catalyzed by iron porphyrins (l-lo). Cytochrome P450-catalyzed olefin oxidations usually give no products other than epoxides, although oxidation of the double bond to a - CH,CO - function is observed as a minor process with styrene (11) and 1-phenylbutene (12) and 1 Author to whom correspondence 000%9861/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

should be addressed.

as a major one with halogenated olefins (13). The oxidation of terminal olefins also results in N-alkylation of the prosthetic heme group (14). A number of epoxidation mechanisms have been proposed, including: (a) direct, concerted oxygen transfer to give the epoxide, (b) formation of a metallaoxetane that rearranges by internal ligand transfer to the epoxide, (c) electron transfer to give a radical cation that is trapped by the ferry1 oxygen to give a cation intermediate, and (d) addition of the ferry1 oxygen to the a-bond to give a radical intermediate (l10). Bruice and his co-workers have proposed from model studies that the key step is formation of a charge-transfer complex that subsequently decays to one or more of the above intermediates (8, 9). The metalloporphyrin-catalyzed oxidation of (Z)-1,2-bis(trans-2,trans-3-diphenylcyclopropyl)ethene was found to give products that could be rationalized by oxidation of the olefin to a radical cation but not to the terminally oxygenated neutral radical shown below (P = porphyrin) (15, 16): P(+.)Fe’“=

0 + CH2=CH-R

+

PFe’V-O-CH2-CH-R This led Castellino and Bruice to propose that a neutral radical is not a discrete intermediate in the epoxidation of olefins by metalloporphyrins (15, 16). In agreement with this view, the metalloporphyrin-catalyzed oxidation in organic solvents of trans-1-phenyl-2-vinylcyclopropane (I), a probe based on the (trans-2-phenylcyclopropyl)carbinyl radical rearrangement timed by Newcomb at k = 1.8 X 1On s-l(25’C) (17), was recently reported to yield the epoxide without the detectable formation of rearrangement products (18). We report here that 1 is oxidized by purified, reconstituted cytochrome P450,,,’ to2 the epoxide without the detectable formation of cata’ Abbreviations used: cytochrome P450,.,, CYPlOl according to the nomenclature of Nebert et al. (19); GLC/MS, gas-liquid chromatography/mass spectrometry; BSTFA, his-(trimethylsilyl)trifluoroacetamide. 697

Inc. reserved.

698

MILLER,

FRUETEL,

AND

lytically generated ring-opened products, although rearranged products are formed from the epoxide by secondary reactions with the aqueous media. EXPERIMENTAL

PROCEDURES

Materials. Cytochrome P450,,,, putidaredoxin, and putidaredoxin reductase were cloned and purified as described previously (20). Camphor-free cytochrome P450,,, was prepared immediately before use by passing it over a Sephadex G-15 column equilibrated with 50 mM potassium phosphate buffer (pH 7.0). No shoulder was present at 391 nm in the spectrum of the protein after passage through the column, confirming the absence of residual camphor-bound protein. Catalase, NADH, and bi.-(trimethylsilyl)trifluoroacetamide (BSTFA) were purchased from Sigma (St. Louis, MO). Hs’*O was purchased from Icon (Summit, NJ). Analytical methods. Gas-liquid chromatography was performed on a Hewlett-Packard 5890A gas chromatograph equipped with a flame ionization detector and interfaced to a Hewlett-Packard 3365 Chemstation (DOS series). Mass spectra were obtained on a VG-70 mass spectrometer equipped with a Hewlett-Packard 5890A gas chromatograph. NMR spectra (300 MHz) were recorded in C’HCl, on a General Electric QE-300 instrument. Chemical shift values are reported in parts per million. trans-I-Phenyl-2-uinylcyclopropane (I). The title compound was obtained by a three-step sequence involving Simmons-Smith cyclopropanation of 3-phenyl-2-propen-l-01, pyridinium chlorochromate oxidation of the cyclopropyl alcohol, and Wittig conversion of the aldehyde to the vinyl group. To a vigorously stirred solution of cinnamyl alcohol (1.9 ml, 14.9 mmol) in 100 ml of hexanes at -20°C was added diethylzinc (30 ml of 1.0 M solution, 29.8 mmol). Vigorous mechanical stirring and strict adherence to the order of addition is required to avoid a potentially explosiue mixture. Methylene iodide (13.2 g, 49.2 mmol) was added dropwise to the stirred solution and the mixture was stirred at -20°C for 6 h and at 25°C overnight before it was poured into cold aqueous ammonium chloride. Repeated extraction with diethyl ether, washing of the combined ether layers with sodium thiosulfate and water, and concentration in uacuo gave an oil that was purified by flash chromatography (20% CH2ClZ in hexanes). (truns-2-Phenylcyclopropyl)methanol (1.9 g, 86%) was thus obtained (21): ‘H NMR 7.23-7.00 (5H, m, aryl H), 3.50 (2H, d,J = 6.7 Hz, CH,OH), 3.07 (lH, bs, OH), 1.77-1.71 (lH, m, CH), 1.37-1.33 (lH, m. CH), and 0.91-0.83 ppm (2H, m, CH,). To a stirred suspension of pyridinium chlorochromate (3.33 g, 15.4 mmol) in 100 ml CHPClz was added the (trans-2-phenylcyclopropyl)methanol (1.9 g, 12.8 mmol). After 2 h at 25’C, the mixture was filtered through a pad of silica gel and the filtrate was concentrated in uucuo. The (tram-2-phenylcyclopropyl)formaldehyde (1.9 g, 95% yield) was purified by flash chromatography (10% CH,Cls in hexanes) (20): ‘H NMR 9.30 (lH, d, J = 4.6 Hz, CHO), 7.31-7.09 (5H, m, aryl H), 2.65-2.58 (lH, m, CH), 2.18-2.14 (lH, m, CH), 1.75-1.69 (lH, m, CH,), and 1.55-1.48 ppm (lH, m, CH,); 13C NMR 199.6 (CHO), 138.7, 128.3, 128.2, 126.5, 125.9 (aryl carbons), 33.6,26.3, 16.2 (cyclopropyl carbons). To 3.84 g (10.7 mmol) of methyltriphenylphosphonium bromide in 50 ml of tetrahydrofuran at 0°C was added dropwise 4.3 ml of n-butyllithium (2.5 M in hexane, 10.7 mmol). After the mixture was stirred at 0°C for 1 h, the (trans-2-phenylcyclopropyl)formaldehyde (1.06 g, 7.2 mmol) was added dropwise and the mixture was heated at 50°C overnight. Water was then added slowly until the precipitate dissolved and the mixture was partitioned between CH,CIP and water. The organic extracts were combined, dried over anhydrous sodium sulfate, and concentrated by distillation. Flash chromatography with hexane yielded 300 mg (28%) of the desired trams-l-phenyl-2-vinylcyclopropane (18): ‘H NMR 7.26-7.03 (5H, m, aryl H), 5.57-5.49 (lH, m, pCH=CHP), 5.09 (lH, d, J = 16.9 Hz, CH=CH,); 4.92 (lH, d, J = 10.4 Hz, CH=CH,), 1.9331.87 (lH, m, CH), 1.73-1.64 (lH, m, CH), 1.21-1.16 (lH, m, CH,), and 1.14-1.05 ppm (lH, m, CH,); 13C NMR 142.3 (aryl), 140.6 (C=C), 128.3, 125.6 (aryl), 112.5 (C=C), 27.4, 25.2, and 16.7 ppm (cyclopropyl carbons).

ORTIZ

DE MONTELLANO

(bans-2-Phenylcyclopropyl)ethylene oxide (2). m-Chloroperbenzoic acid (105 mg, 0.61 mmol) was slowly added to a magnetically stirred biphasic mixture of trans-l-vinyl-2-phenylcyclopropane (1) (73 mg, 0.51 mmol) in 6 ml of CH2C12 and 2 ml of aqueous 0.5 M sodium bicarbonate. The mixture was stirred overnight. The layers were then separated and the organic layer was extracted with water, dried over anhydrous sodium sulfate, and concentrated in vacua. Flash chromatography (2:l hexanes: CH&l,) yielded 20 mg (25%) of an equal mixture of the two epoxide diastereomers (18, 22): ‘H NMR 7.28-7.05 (5H, m, aryl H), 2.97-2.94 (lH, m, -CH-0), 2.80-2.77 (lH, m, -CH-O), 2.60-2.58 (lH, m, -CH-0), 1.97-1.89 (lH, m, CH), 1.28-1.21 (lH, m, CH), 1.06-0.88 ppm (2H, m, CH,); 13C NMR 142.0, 141.9, 128.3, 125.9, 125.7 (aryl carbons), 53.1, 52.8 (epoxide carbon), 46.9, 46.6 (epoxide carbon), 23.4, 23.1, 20.2, 19.8, 12.4, 11.9 ppm (cyclopropane carbons). (tran.s-2-Phenylcyclopropyl)-l,2-ethanediol(3). Osmium tetroxide (40 mg, 0.16 mmol) was added to a solution of tram+l-phenyl-2-vinylcyclopropane (10 mg, 0.07 mmol) and the mixture was stirred overnight at 25°C. Sodium bisulfite (100 mg) and 2 ml of water were then added and the mixture was stirred for 30 min before it was extracted with CH,Cl,. The combined organic layers were washed with saturated aqueous copper sulfate, dried over sodium sulfate, and evaporated to dryness in vacua. Flash chromatography (2% methanol in CH&&) yielded 3.5 mg (29%) of (tram-2-phenylcyclopropyl)ethane-1,2-diol as an equal mixture of two diastereomers: ‘H NMR 7.28-7.04 (5H, m, aryl H), 3.84-3.78 (lH, m, CH,OH), 3.67-3.61 (lH, m, CH,OH), 3.34-3.29 (lH, m, CHOH), 2.01-1.94 (0.5H, m, PhCH), 1.90-1.84 (0.5H, m, PhCH), 1.30-1.22 (lH, m, cyclopropyl CH), and 1.09-0.94 ppm (2H, m, cyclopropyl CH,); *Y! NMR 128.4,125.8 (aryl carbons), 75.8,75.6 (CH,OH), 66.5,66.3 (CHOH), 25.4, 25.2, 20.7, 20.4, 13.1, and 12.9 ppm (cyclopropyl carbons). Exact mass calculated for CuHr402, 178.0994; observed, 178.0987. (trans.2-Phenylcyclopropyl)acetaldehyde (4). This compound was obtained from trans-4-phenyl-3-buten-l-01 by Simmons-Smith cyclopropane formation and oxidation of the alcohol to the aldehyde. Diethylzinc (4 ml of a 1.0 M solution, 4.0 mmol) was added at 0°C to a stirred solution of trans-4-phenyl&buten-l-01 (300 mg, 2.0 mmol) in 10 ml of hexanes. Methylene iodide (0.54 ml, 6.7 mmol) was then added dropwise to the stirred solution and the final mixture was stirred at 25°C overnight. The reaction was poured into cold aqueous ammonium chloride and the mixture was extracted repeatedly with diethyl ether. The ether layers were washed with sodium thiosulfate and water before they were concentrated in uucuo. Flash chromatography (CH,Cl,) of the residual oil gave 150 mg (50%) of 2-(trans.2-phenylcyclopropyl)ethanol (23): ‘H NMR 7.36-7.02 (5H, m, aryl H), 3.73 (2H, t, J = 6.4 Hz, CH,OH), 1.81 (lH, bs, OH), 1.75-1.57 (3H, m, CH? and PhCH), 1.12-1.05 (lH, m, CH), 0.95-0.89 (lH, m, cyclopropyl CH2), and 0.84-0.78 ppm (lH, m, cyclopropyl CH,); r3C NMR 143.5, 128.4, 125.4 (aryl carbons), 62.8 (CH,OH), 37.4 (CH,), 22.9, 20.4, 15.8 (cyclopropyl carbons). 2-(trans-2-Phenylcyclopropyl)ethanol (50 mg, 0.31 mmol) in 2 ml of CH&12 was added to a suspension of pyridinium chlorochromate (80 mg, 0.37 mmol) in 5 ml of CH&l;! and the mixture was stirred for 3 h at 25°C before it was diluted with 3 ml of diethyl ether. After stirring for 10 min, the mixture was filtered through a pad of silica gel and the filtrate was concentrated on a rotary evaporator. (trans.2-Phenylcyclopropyl)acetaldehyde (28 mg, 56%) was obtained by flash chromatography (50% CH#& in hexanes): ‘H NMR 9.82 (lH, t, J = 1.8 Hz, CHO), 7.317.07 (5H, m, aryl protons), 2.45-2.41 (2H, m, CH,), 1.77-1.72 (lH, m, PhCH), 1.34-1.32 (lH, m, CH), 1.10-1.03 (lH, m, cyclopropyl CH,), and 0.90-0.84 ppm (lH, m, cyclopropyl CH,); r3C NMR 201.6 (CHO), 142.2, 128.4, 125.8, 125.7 (aryl carbons), 48.1 (CH,), 22.6, 16.1, and 15.2 ppm (cyclopropyl carbons). Exact mass calculated for C,,H,,O, 160.0888; observed, 160.0877. trans-5Phenylpent-2-en-I-ol(5). To a suspension of NaH (0.06 g, 2.43 mmol) in tetrahydrofuran at 0°C was added 1.04 g (2.43 mmol) of (carbethoxymethyl)triphenylphosphonium bromide. After stirring the mixture for 1 h, hydrocinnamaldehyde (0.25 ml, 1.88 mmol) was added. After refluxing the mixture overnight, water was added at O”C, the mixture was extracted with CH2C12, and the combined extracts were dried

OXIDATION

OF A RADICAL

PROBE

over anhydrous sodium sulfate and concentrated in uocuo. Flash chromatography with 21 hexanes:CH,Cl, yielded 194 mg (51%) of ethyl trans-5-phenyl-2-pentenoate (24): ‘H NMR 7.28-7.15 (5H, m, aryl H), 7.13-6.96 (lH, m, =CHCO-) 5.82 (lH, d, J = 15.0 Hz, CH=C), 4.14 (2H, dd, J = 7.0, 13.87 Hz, CO&H,), 2.75-2.70 (2H, m, CH*), 2.502.45 (2H, m, CH2), and 1.24 ppm (3H, t, J = 7.3 Hz, Me); 13C NMR 166.2 (CO,R), 147.7 (=C-CO), 140.5, 128.2, 128.0, and 125.9 (aryl carbons), 121.6 (=CCH.J, 59.9 (OCH2), 34.1 (CH.J, 33.6 (CH,), and 14.0 ppm (CHJ. To a solution of ethyl trans-5-phenyl-2-pentenoate (70 mg, 0.34 mmol) in 9 ml of tetrahydrofuran at 0°C was added diisopropylaluminum hydride (0.9 ml as a 1.4 M solution in hexane, 1.23 mmol). After stirring at 25°C for 5 h, the mixture was poured into an aqueous saturated ammonium chloride solution and the organic layer was separated, dried over sodium sulfate, and concentrated in uocuo. The desired trons-5phenylpent-2-en-l-01 was isolated by flash chromatography with CH,Cl, as the solvent (18 mg, 33%) (25): ‘H NMR 7.30-7.16 (5H, m, aryl H), 5.76-5.63 (2H, m, C=), 4.06 (2H, d, J = 5.1 Hz, CH,OH), 2.73-2.65 (2H, m, CH&, and 2.41-2.31 ppm (2H, m, CH.J; 13C NMR 141.6, 129.5, 128.6, 128.4 (aryl carbons), 125.8, 132.2 (C=C), 63.6 (CH,OH), 35.5, and 33.9 (CHJ. trans-5-Phenylpent-2-en-l,5-diol (6). This diol was prepared by a three-step sequence involving addition of lithium acetylide to styrene oxide, base-catalyzed condensation with paraformaldehyde, and reduction of the triple bond with lithium aluminum hydride. Thus, lithium acetylide ethylenediamine complex (0.84 g, 9.2 mmol) was added to a solution of styrene oxide (1.0 g, 8.3 mmol) in 10 ml of dimethyl sulfoxide and the mixture was stirred overnight at 25°C. The mixture was then partitioned between brine and diethyl ether and the combined organic layers were washed with water, dried over anhydrous sodium sulfate, and concentrated in uacuo. The desired 1-phenylbut-3-yn-l-01 (0.97 g, 80%) was purified by flash chromatography (CH,Cl,) (26): ‘H NMR 7.40-7.28 (5H, m, aryl H), 4.85 (lH, t, J = 6.4 Hz, CHO), 2.85 lH, bs, OH), 2.63 (2H, dd, J = 6.2, 2.4 Hz, CH2), and 2.07 ppm (lH, s, C=CH); 13C NMR 142.4, 128.3, 127.8, 125.7 (aryl carbons), 80.7 (C3), 72.1 (CHOH), 70.8 (C4), and 29.1 ppm (CH,). To a solution of l-phenylbut-3-yn-l-01 (200 mg, 1.37 mmol) in 10 ml of tetrahydrofuran at 0°C was slowly added 1.15 ml of n-butyllithium (2.5 M in hexane, 2.88 mmol). After stirring for 5 min, paraformaldehyde (123 mg) was added and the mixture was stirred overnight at 25°C before it was partitioned between saturated aqueous ammonium chloride and CH2C12. The combined extracts were dried over sodium sulfate and concentrated in uocuo. Flash chromatography (2% methanol in CH,Cl,) yielded94 mg (39%) of 5-phenylpent-2-yn-1,5-diol. If allowance is made for the recovery of 100 mg of starting material, the yield is 95%: ‘H NMR 7.35-7.26 (5H, m, aryl H), 4.83 (lH, t, J = 6.2 Hz, CHOH), 4.20 (2H, s, C&OH), 3.47 (lH, bs, OH), 3.20 (lH, bs, OH), and 2.64 ppm (2H, m, CH,); 13C NMR 142.6, 128.4, 127.9, 125.7 (aryl carbons), 82.6, 80.8 (C=C), 72.4 (CHOH), 50.9 (CH,OH), and 29.6 ppm (CH,). Exact mass calculated for CllHlzOz, 176.0837; observed, 176.0867. A solution of 5-phenylpent-2-yn-1,5-diol (30 mg, 0.17 mmol) in 3 ml of tetrahydrofuran was added to a suspension of lithium aluminum hydride (0.26 ml of a 1 M solution in ether, 0.26 mmol) and sodium methoxide (28 mg, 0.52 mmol) in 10 ml of tetrahydrofuran and the mixture was refluxed for 8 h. The excess lithium aluminum hydride was carefully quenched with water and the resulting white precipitate was removed by filtration. The filtrate was extracted with diethyl ether and the extract was dried over sodium sulfate and concentrated in vacua. Flash chromatography (2% methanol in CHZC12 yielded trans-5-phenylpent-2-en1,5-diol (11.7 mg, 40%) (27): ‘H NMR 7.35-7.26 (5H, m, aryl H), 5.765.69 (2H, m, HC=CH), 4.73 (lH, t, J = 6.2 Hz, PhCH), 4.09 (2H, d, J = 4.9 Hz, CH,OH), 2.52-2.47 (2H, m, CH,), and 2.07 ppm (2H, bs, OH); 13C NMR 143.8, 128.4, 127.6, 125.7 (aryl carbons), 132.7, 128.3 (C=C), 73.5 (PhCHOH), 63.3 (CH,OH), and 42.1 ppm (CH,). Enzyme incubations. Solutions containing 2 nmol cytochrome P450-, 16 nmol putidaredoxin, 4 nmol putidaredoxin reductase, 0.5 ~1 substrate, and 0.8 pmol EDTA in 1 ml total volume of 50 mM potassium phosphate

BY CYTOCHROME

P450

699

buffer (pH 7.0) were preincubated for 2 min at 25°C. In some experiments, the pH of the buffer was 8.0. NADH (5 mM) was then added and the tubes were incubated in a shaking water bath for 2 h at 25°C. Additional NADH was added periodically during the incubation. The metabolites were extracted with 0.5 ml of CHzClz and were analyzed by gas chromatography on a DB-1 fused silica column (0.25 mm i.d. X 30 m) programmed to remain at 60°C for 2 min, then to rise at 4”C/min to a temperature of 185’C, and finally to hold at the latter temperature for 20 min. The retention times of the products were as follows: truns2-phenyl-l-vinylcyclopropane, 18.3 min; (truns-2-phenylcyclopropyl)acetaldehyde, 24.4 min; (trons-2-phenylcyclopropyl)ethylene oxide, 25.0 and 25.1 min (two diastereomers); trans-5-phenylpent-2-en-l-01, 26.1 min; trans-5-phenylpent-2-en-1,5-diol, 32.1 min; and (trans-2phenylcyclopropyl)-1,2-ethanediol, -32.1 min. The identity of each of the metabolites was established by coelution with an authentic standard and by electron impact GLC/MS. Control incubations were modified by addition of catalase, either HzOz (0.5 mM) or buffer in place of NADH, or the (truns-2-phenylcyclopropyl)ethylene oxide metabolite (50 nmol) in place of the olefin substrate. Incubations of (truns-2-phenylcyclopropyl)ethylene oxide (50 nmol) in 1 ml of buffer, buffer pretreated with Chelex resin (Bio-Rad, Richmond, CA) to remove transition metals, or deionized, double-distilled water were also run and analyzed under the same conditions as the enzyme incubations. For the studies with H2180, incubation mixtures were prepared as described above except for the addition of substrate and NADH. The mixtures were frozen and lyophilized overnight, then reconstituted in “O-labeled water (96% H2’*0) to a volume of 0.8 ml. The substrate was added neat, and the NADH was added as a 50 mM solution in HslsO (two 0.1.ml aliquots were added over the 2-h incubation period). The metabolites were extracted as already described and were analyzed by GLC/MS. To obtain a molecular ion for the truns-5-phenylpent-2-en1,5-diol metabolite it was necessary to derivatize it with BSTFA and to carry out the GLC/MS analysis under chemical ionization conditions. BSTFA derivatization was accomplished by taking 100 ~1 of the CHzClz extract, placing it in a l-ml Reactivial, adding l-2 ~1 of BSTFA, and heating the solution at 55’C for 30 min. Under these conditions, the (M + NH,)’ ion is the most prominent species in the mass spectrum of trons-5-phenylpent-2-en-1,5-diol. RESULTS

Synthesis of the substrate and putative metabolites. trams-1-Phenyl-2-vinyl-cyclopropane (l), the probe for these studies, was obtained from 3-phenyl-2-propenl-01 by Simmons-Smith cyclopropanation (86% yield), pyridinium chlorochromate oxidation of the cyclopropyl alcohol to the aldehyde (95% yield), and Wittig conversion of the aldehyde to the vinyl group (28% yield). In order to facilitate the search for, and identification of, metabolites, authentic samples were prepared of the most likely metabolites. Epoxide 2, the normally expected metabolite, was synthesized from 1 by reaction with meta-chloroperbenzoic acid. (tram-2-PhenylcyclopropyU-ethane-1,2-diol (3), which could arise by hydrolysis of epoxide 2 (Fig. l), was made by osmium tetroxide oxidation of 1. Precedent exists for direct, cytochrome P450-catalyzed oxidation of olefins to the acetaldehyde derivatives, so aldehyde 4 was prepared by Simmons-Smith cyclopropanation of trans4-phenyl-3-buten-l-01 and oxidation of the alcohol to the aldehyde with pyridinium chlorochromate. Ring opening of the cyclopropane ring could yield a radical that is quenched by hydrogen atom abstraction, yielding trans-

MILLER,

700

Ph

FRUETEL,

AND

ORTIZ

DE MONTELLANO

-OH

10

5 FIG.

1.

Hypothetical

8

mechanisms for the oxidation

5-phenylpent-2-en-l-01(5), or that is oxidized to a cation and is then trapped by water, yielding 6 (Fig. 1). Alcohol 5 was made by reduction of trans-5-phenyl-2-pentenoate and dio16 by addition of lithium acetylide to styrene oxide, condensation with paraformaldehyde, and reduction of the triple bond with lithium aluminum hydride. The spectroscopic and analytical properties of all the synthetic products were entirely consistent with the assigned structures. Cytochrome P450,,, -catalyzed oxidation of trans-lphenyl-2-vinylcyclopropane. Gas-liquid chromatographic analysis of the products extracted from incubation of trans-1-phenyl-2-vinylcyclopropane with a reconstituted cytochrome P450,,, system shows that the olefin is oxidized to (% of total products) epoxide 2 (81%), (trans-2-phenylcyclopropyl)acetaldehyde 4 (6%), trans5-phenylpent-2-en-1,5-dio16 (13%), and a minor unidentified polar product (Fig. 2). Two peaks are observed for the epoxide because it is composed of two diastereomers that are resolved in the chromatographic system employed. No trace is detected of (trans-2-phenylcyclopropyl)ethane-1,2-diol (3) or, more importantly, trans-5phenyl-2-penten-l-01 (5). The structures of the metabolites were confirmed by coelution and mass spectrometric comparisons with authentic standards. Control experiments establish that product formation requires both NADH and a functional cytochrome P450 system. The reaction is not inhibited by catalase, nor can the NADH be replaced by HzOz. These latter control experiments are essential because earlier work has shown that cytochrome P450,, turnover can be highly uncoupled with abnormal substrates toward the production of HzOz (28). The internal oxygen in 6 was shown to derive quantitatively (within +5%) from water by mass spectrometric

of 1 by nonconcerted

mechanisms.

analysis of the diol isolated from incubations of 1 with cytochrome P450,,, in Hz’s0 (not shown). These experiments also confirmed, as expected, that the terminal oxygen in the diol derives from a source other than water (i.e., molecular oxygen). Chemical properties of (trans-2-phenylcyclopropyl)ethylene oxide. Studies of the stabilities of the metabolites show that epoxide 2 is unstable in aqueous media. Indeed, incubation of epoxide 2 in the incubation buffer (pH 7.0 or 8.0), in the presence or absence of the enzymatic components of the reaction, results in the formation of aldehyde 4, ring-opened diol 6, and the unidentified product in approximately the same ratios and

3000 3 i

2000

f d

1000

Time (minutes) FIG. 2. Gas-liquid chromatogram of the products obtained in the oxidation of 1 by cytochrome P450,,. The identities of the peaks are indicated in the figure. The incubation and chromatographic conditions are given in the text. All the peaks in the chromatogram except those indicated by arrows are present in chromatograms of control incubations carried out in the absence of NADH.

OXIDATION

OF A RADICAL

PROBE

BY CYTOCHROME

P450

comparable rates as are obtained by oxidation of the parent olefin in the fully functional system. The unidentified product is observed when the control incubations contain cytochrome P450,,, but not when they are carried out in buffer alone, suggesting that it is due to an impurity in the cytochrome P450,,, preparation. Mass spectrometric analysis confirmed that dio16 rather than dio13 is formed in the incubations, a necessary experiment because the two diols are not well resolved by GLC. Ph-OH

- Hz0

PhbOH

DISCUSSION

The formation of epoxide 2, aldehyde 4, and diol 6 in the cytochrome P450,,, oxidation of trans-2-phenyl-lvinylcyclopropane appears, at first sight, to provide support for a nonconcerted epoxidation mechanism. Although epoxide 2, the expected product, provides no mechanistic information, aldehyde 4 undoubtedly derives from carbocation 7 by a reaction process for which there is precedent in the cytochrome P450 literature (11-13). Rearranged dio16 also clearly derives from 7 or its free radical analogue 8 (Fig. 1). The absence of alcohol 5, however, which would also be expected if the free radical 9 were involved, suggests that dio16 derives from cation 7 rather than radical 8. This inference is strengthened by the finding that the internal oxygen in 6 derives from water rather than, like the terminal oxygen, from molecular oxygen. This rules out addition of molecular oxygen to radical 9 to give a hydroperoxy radical that is subsequently reduced to the alcohol. The oxygen labeling results do not absolutely exclude the participation of 8, however, because the rearranged cation could arise by radical ringopening to 9 followed by oxidation of the rearranged radical to cation 10 (Fig. 1). This explanation requires that oxidation of the radical to the cation occur more rapidly than addition of oxygen or abstraction of a hydrogen atom. Formation of 7 rather than 8 as the primary reaction pathway is confirmed, at the same time as direct formation of either of these intermediates from 1 is obscured, by the finding that epoxide 2 rearranges at neutral or slightly basic pH to the incubation products. The instability of the epoxide at these high pH values is surprising, but proton-dependent ring opening of the epoxide readily rationalizes the formation of both 4 and 6 (Fig. 3). Opening of the epoxide to cation 7 rather than to the alternative primary cation is expected if the ring-opening reaction is controlled (as it is in SN, reactions) by the stability of the carbocation. Formation of all the products in comparable ratios and at roughly comparable rates from both epoxide 2 and the catalytic oxidation of 1 suggests that 2 is the predominant or only enzymatic product, although it would be difficult to detect a trace of the same products formed by direct oxidation over the background of the epoxide-derived products. The complete absence of ring-opened alcohol 5, however, argues against the involvement of a free rad-

OH

6

FIG. 3. Mechanism for the formation epoxide 2 in aqueous medium.

of metabolites

4 and 6 from

ical rearrangement process. Rearrangement of 8 to the ring-opened form, from the work of Castellino and Bruice (15, 16) and Newcomb and Manek (17), should occur at a rate between lOlo and 10” s-l. Given this rate and our ability to detect 5 if formed at one-hundredth the amount of the epoxide, closure of radical 8 to epoxide 2 would have to occur at a rate on the order of 1012s1 to make 5 undetectable. This rate, as noted by Castellino and Bruice (15, 16), approaches the theoretical limit between an intermediate and a transition state. This analysis assumes, of course, that constraint of the probe in the active site does not decrease the rate of the rearrangement reaction. The results suggest that the enzymatic oxidation, like the reaction catalyzed by metalloporphyrin models (15, 16, IS), probably proceeds via a mechanism that does not involve radical 5 as a discrete intermediate. The involvement of cation 7, however, or of radical 8 if the rearranged radical 9 is exclusively oxidized to cation 10, cannot be completely ruled out because a low yield of metabolite 6 would be masked by the background provided by the epoxide-derived products. truns-2-Phenyl-1-vinylcyclopropane has been proposed as an attractive radical clock probe of the mechanistic timing of biological epoxidation reactions (18). Its relatively small size makes it more attractive for use in biological systems, where binding to an enzyme is important, than the probe of Castellino and Bruice (15, 16). Both probes have been successfully used in organic solvents to study the oxidation of olefins by metalloporphyrins. The present results show, however, that the epoxide formed from olefin 1, and presumably that produced from related probes, undergoes proton-mediated ring opening to cation-derived rearranged products at a significant rate in aqueous media. The resulting product background makes the detection of low yields of direct oxidation rearrangement products difficult and diminishes the utility of 1 as a probe of enzymatic oxidation mechanisms. ACKNOWLEDGMENTS This work was supported by National Institutes of Health Grant GM 25515. Mass spectra were obtained in the Bio-organic, Biomedical Mass

MILLER,

702 Spectrometry Facility (A. Burlingame, RR 01614 and P-50 DK 26743.

Director)

FRUETEL, supported

AND

by Grants

REFERENCES P. R. (1986) in Cytochrome P450: Structure, 1. Ortiz de Montellano, Mechanism, and Biochemistry (Ortiz de Montellano, P. R., Ed.), pp. 217-271, Plenum, New York. 2. Guengerich, 977153.

F. P. (1990) CRC Crit. Reu. Biochem. Mol. Biol. 25,

3. Groves, J. T., and Watanabe, 844338452. 4. Groves, J. T.. and Gilbert,

Y. (1988) J. Am. Chem. Sot. 110,

J. A. (1986) Inorg. Chem. 25, 123-125.

5. Groves, J. T., and Watanabe, 507-508.

Y. (1986) J. Am. Chem. Sot. 108,

6. Collman, J. P., Hampton, P. D., and Brauman, Chem.Soc. 112,2986-2998. 7. Balch, A. L., Latos-Grazynski, Chem. Sot. 107,2983-2985.

J. I. (1990) J. Am.

L., and Renner, M. W. (1985) J. Am.

8. He, G.-X., Mei, H.-Y., and Bruice, T. C. (1991) J. Am. Chem. Sot.

113,5644-5650. 9. Garrison,

J. M., Ostovic, D., and Bruice, T. C. (1989) J. Am. Chem.

Sot. 111,4960-4966. 10. Traylor, 1958.

T. G., and Xu, F. (1988) J. Am. Chem. Sot. 110, 1953-

11. Mansuy, D., Leclaire, J., Fontecave, M., and Momenteau, Biochem. Biophys. Res. Commun. 119,319-325.

M. (1984)

12. Liebler, D. C., and Guengerich, F. P. (1983) Biochemistry 5489.

22,5482-

13 Miller, 1097.

21, 1090-

R. E.. and Guengerich,

F. P. (1982) Biochemistry

ORTIZ

DE MONTELLANO

14. Ortiz de Montellano, P. R. (1985) in Bioactiuation of Foreign pounds (Anders, M. W., Ed.), pp. 121-155, Academic Press, York. 15. Castellino, A. J., and Bruice, T. C. (1988) J. Am. Chem. Sot. 7512-7519. 16. Castellino, A. J., and Bruice, T. C. (1988) J. Am. Chem. Sot. 1313-1315. 17. Newcomb, M., and Manek, M. B. (1990) J. Am. Chem. Sot.

ComNew

110, 110, 112,

9662-9663. 18. Fu, H., Look, G. C., Zhang, W., Jacobsen, E. N., and Wong, C.-H. (1991) J. Org. Chem. 56,6497-6500. 19. Nebert, D. W., Nelson, D. R., Coon, M. J., Estabrook, R. W., Feyereisen, R., Fujii-Kuriyama, Y., Gonzalez, F. J., Guengerich, F. P., Gunsalus, I. C., Johnson, E. F., Loper, J. C., Sato, R., Waterman, M. R., and Waxman, D. J. (1991) DNA Cell Biol. 10, 1-14. 20. Peterson, J. A., Lorence, M. C., and Amarneh, B. (1990) J. Biol. Chem. 265,6066-6073. 21. Baldwin, J. E., Patapoff, T. W., and Barden, T. C. (1984) J. Am. Chem. Sot. 106, 1421-1426. 22. Prestwich, G. D., Lucarelli, I., Park, S. K., Loury, D. N., Moody, D. E., and Hammock, B. D. (1985) Arch. Biochem. Biophys. 237, 361-372. 23. Yamaguchi, M., and Hirao, I. (1984) Tetrahedron L.&t. 25, 45494552. 24. Vedejs, E., and Fleck, T. J. (1989) J. Am. Chem. Sot. 111, 58615871. 25. Nunez, M. T., and Martin, V. S. (1990) J. Org. Chem. 55, 19281932. 26. Burgess, K., and Jennings, L. D. (1991) J. Am. Chem. Sot. 113, 6129-6139. 27. Meyer, F. K., Drewett, J. G., and Carlson, R. M. (1986) Synth. Commun. 16,261-265. 28. Fruetel, J. A., Collins, J. R., Camper, D. L., Loew, G. H., and Ortiz de Montellano, P. R. (1992) J. Am. Chem. Sot., in press.