6442
`
`Biochemistry 1994, 33, 6442-6449
`
`On the Mechanism of Action of Cytochrome P450: Evaluation of Hydrogen
`Abstraction in Oxygen-Dependent Alcohol Oxidation* *
`Alfin D. N. Vaz* and Minor J. Coon
`Department of Biological Chemistry, Medical School, The University of Michigan, Ann Arbor, Michigan 48109-0606
`Received February 9, 1994; Revised Manuscript Received March 28, 1994®
`
`The mechanism of oxidation of primary and secondary benzylic alcohols to the corresponding
`abstract:
`carbonyl compounds by purified rabbit liver cytochrome P450 forms 2B4 and 2E1 in a reconstituted enzyme
`system has been examined by linear free energy relationships, intramolecular and steady-state deuterium
`isotope effects, and the incorporation of an 02-derived oxygen atom or solvent-derived deuterium. The k^t
`and Km values were found to be relatively insensitive to the presence of electronic perturbations at the para
`position. The Hammett reaction constants for the oxidation of benzyl alcohols by P450s 2B4 and 2E1 are
`-0.46 and -0.37, respectively, and with 1-phenylethyl alcohols the corresponding reaction constants are
`-1.41 and -1.19, respectively. With [ l-2Ht] benzyl alcohol, P450s 2B4 and 2E1 show similar intramolecular
`deuterium isotope effects of 2.6 and 2.8, respectively, whereas with [ 1 -2H2]benzyl alcohol under steady-state
`conditions, the deuterium isotope effects on the catalytic constants are 2.8 and 1.3, respectively. No significant
`isotope effect on the catalytic constant was noted for either form of P450 with 1-phenylethyl alcohol.
`In
`D20, acetophenone formed by either form of P450 from 1-phenylethyl alcohol does not contain a deuterium
`atom at the methyl group, whereas under an atmosphere of 1802 approximately 30% of the labeled oxygen
`is incorporated into the carbonyl group with either form of the cytochrome. The results are consistent with
`a mechanism that involves stepwise oxidation of the alcohol to a carbon radical a to the alcohol function,
`followed by oxygen rebound to yield the gem-diol, dehydration of which gives the carbonyl product. However,
`the rate-determining step is dependent on the alcohol substrate and the form of cytochrome P450 that is
`examined. Carbon-hydrogen bond cleavage in benzyl alcohol is clearly rate-limiting with P450 2B4 and
`partially rate-limiting with P450 2E1, whereas in 1-phenylethyl alcohol this step is not rate-limiting with
`either cytochrome.
`
`P4501 heme proteins constitute a class of highly versatile
`biological catalysts that utilize molecular oxygen and NADPH
`to oxidize diverse organic compounds of endobiotic and
`xenobiotic origin (Coon et al., 1992). The various reactions
`result in the insertion of an atom of molecular oxygen into a
`types of oxidation at
`in other
`hydroxylated product or
`functional groups such as amines, ethers, esters, aldehydes,
`and alcohols (Guengerich, 1987). A general mechanistic
`scheme developed for such P450-catalyzed reactions accounts
`for various aspects of the catalytic cycle and for the insertion
`of a molecular oxygen-derived oxygen atom into the oxidized
`In the oxidation of alcohols
`product (White & Coon, 1980).
`to carbonyl products, however, some exceptions to the predicted
`incorporation of an atom of molecular oxygen into the carbonyl
`product have been observed by various laboratories. Partial
`or complete lack of incorporation of 02-derived oxygen into
`the carbonyl product has been observed that is apparently not
`explainable by exchange with water (Akhtar et al., 1982;
`Cheng & Schenkman, 1983; Suhara et al., 1984; Wood et al.,
`1988). This has resulted in various mechanistic hypotheses
`for the oxidation of alcohols by P450 such as oxidative
`dehydrogenation (Cheng & Schenkman, 1983; Wood et al.,
`
`t This investigation was supported by Grant AA-06221 from the
`National Institute on Alcohol Abuse and Alcoholism (to M.J.C.) and
`Grant GM-46807 from the National Institutes of Health (to A.D.N.V.).
`* Author to whom correspondence should be addressed.
`® Abstract published in Advance ACS Abstracts, May 1, 1994.
`1 Abbreviations: P450, cytochrome P450; reductase, NADPH-cyto-
`chrome P450 reductase; DLPC, dilauroylglyceryl-3-phosphorylcholine;
`HPLC, high-pressure liquid chromatography. P450 2B4 and P450 2E1
`are the currently recommended names (Nelson et al., 1993) for the rabbit
`liver microsomal isoforms originally designated LM2 and LM3a, respec-
`tively.
`
`0006-2960/94/0433-6442$04.50/0
`
`1988) or stereospecific dehydration of a transient gem-diol
`such that the inserted oxygen is specifically lost (Suhara et
`al., 1984). Ekstróm et al. (1987) have reported that cleavage
`of the Ci-H bond of ethanol appears to be a rate-determining
`step in catalysis by the ethanol-inducible form of P450.
`We have chosen to examine the oxidation of benzylic
`alcohols by P450 2E1 and 2B4 as a mechanistic model since
`reactions at benzylic positions are sensitive to electronic
`perturbations by substituents on the aromatic ring, and the
`magnitude of this effect on the rate constant is a useful indicator
`of the intermediate generated at the benzylic position (Jaffe,
`1953). Two further advantages are the large extinction
`coefficients of benzaldehydes and acetophenones that permit
`sensitive quantitative analysis by reversed-phase HPLC and
`the enolization rate, hydration equilibrium, and electron impact
`mass fragmentation pattern of acetophenone that permit a
`sensitive measurement of the incorporation of a solvent proton
`at the methyl group or of an 1802-derived oxygen atom into
`In this study, we have determined (a) the
`the carbonyl group.
`linear free energy relationship for the oxidation of a series of
`para-substituted benzyl and 1-phenylethyl alcohols to the
`aldehydes and ketones, respectively; (b) the intramolecular
`deuterium isotope effect for the oxidation of benzyl alcohol
`to benzaldehyde; (c) the steady-state deuterium isotope effect
`on the catalytic constants for the oxidation of benzyl and
`1-phenylethyl alcohols to benzaldehyde and acetophenone,
`respectively; and (d) the incorporation into acetophenone of
`a solvent-derived proton at the methyl group or an 1802-derived
`oxygen atom at the carbonyl group.
`Our results establish that the oxidation of benzyl alcohols
`by P450 2B4 proceeds by the rate-determining formation of
`© 1994 American Chemical Society
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`

`

`Alcohol Oxidation by Cytochrome P450
`a benzylic radical as an intermediate and that oxidative
`desaturation of 1-phenylethyl alcohol to an enol
`is not an
`intermediate with either P450 2B4 or 2E1. Our findings also
`indicate that both forms of P450 oxidize primary or secondary
`benzylic alcohols to the corresponding carbonyl compounds
`by the same sequence of reactions involving an intermediate
`benzyl radical and oxygen rebound to form the gem-diol,
`dehydration of which yields the carbonyl compounds. How-
`ever, the rate-limiting step in the overall reaction is dependent
`on the alcohol substrate as well as the isozyme of P450.
`MATERIALS AND METHODS
`Substrates and Reagents. NADPH and DLPC were
`obtained from Sigma and Calbiochem, respectively. Sodium
`borodeuteride, lithium aluminum deuteride, and primary and
`secondary benzyl alcohols were obtained from Aldrich. The
`commercially obtained alcohols were examined for contami-
`nation by the corresponding carbonyl compounds and were
`redistilled or recrystallized from 40% hot aqueous ethanol
`when necessary. Other alcohols were synthesized as described
`below from aldehydes or ketones obtained from Aldrich. 1802
`of 98% isotopic purity was obtained from Cambridge Isotope
`Laboratories.
`Synthesis of para-Substituted and Deuterium-Labeled
`Benzyl and 1-Phenylethyl Alcohols. To a solution of the
`in 100 mL of 20% aqueous
`aldehyde or ketone (50 mmol
`ethanol), sodium borohydride (100 mmol) was added, and
`the reaction mixture was stirred overnight at room tempera-
`ture. The solution was then concentrated at room temperature
`under reduced pressure, diluted with 50 mL of water, and
`extracted twice with 25-mL portions of CH2CI2. After the
`combined extract had been dried over anhydrous sodium
`sulfate, the solvent was removed under reduced pressure, and
`the residue was vacuum distilled (at 18 mmHg) to yield the
`desired alcohol in an overall yield that ranged from 60 to 80%.
`In all cases, the alcohol contained less than 0.01% of the starting
`aldehyde or ketone as determined by HPLC. For the synthesis
`of [l-2Hi]-l-phenylethanol and [l-2Hi]benzyl alcohol, the
`method was essentially as above, except that sodium boro-
`deuteride (98% isotopic purity, from Aldrich) was used in a
`2.5-fold equivalent excess and the acetophenone or benzal-
`dehyde was 5.0 M in 20% ethanol. For the synthesis of [ 1 -2H2] -
`benzyl alcohol, benzoic acid (24.6 mmol) was dissolved in 25
`mL of dry ether that had been freshly distilled from lithium
`aluminum hydride, and the solution was injected into 250 mL
`of dry ether containing lithium aluminum deuteride (23.8
`mmol, 98% isotopic purity, from Aldrich). The mixture was
`maintained under an atmosphere of dry nitrogen and stirred
`overnight at room temperature, after which 100 mL of ice-
`cold 2 N HC1 was added to decompose the excess reductant.
`The ether layer was separated, washed once with saturated
`aqueous sodium bicarbonate, and dried over anhydrous sodium
`sulfate. The ether was then removed under reduced pressure
`distilled
`at room temperature, and the residue was vacuum
`(at 18 mmHg) to yield the labeled alcohol in 60% yield, based
`on the starting benzoic acid.
`Enzymes. P450 forms 2B4 and 2E1 and the reductase were
`purified from rabbit liver by methods previously described by
`this laboratory (Coon et al., 1978; French & Coon, 1979;
`Koop et al., 1982). The individual preparations were
`homogeneous as judged by sodium dodecyl sulfate-polyacryl-
`amide gel electrophoresis, and the specific contents were 15.7,
`18.8, and 12.2-13.1 nmol/mg of protein, respectively. Stock
`solutions of these enzymes were 70.5, 24.0, and 53.0 µ ,
`respectively.
`Reconstitution and Enzyme Assay. For all the experiments
`reported herein, the stock solutions of reductase and P450
`
`6443
`
`Biochemistry, Vol. 33, No. 21, 1994
`were mixed in a 1:1 molar ratio, distributed into sufficient
`tubes for each assay, and maintained at -20 °C until used.
`A typical steady-state kinetic assay was as follows.
`In a final
`volume of 0.5 mL, the reaction mixture contained 25 µ     of
`potassium phosphate buffer, pH 7.4,30 µg of freshly dispersed
`DLPC, 0.1 nmol of P450 2E1 or 0.2 nmol of P450 2B4 with
`an equimolar amount of the reductase, substrate at
`the
`appropriate concentration, and 0.5 qmol of NADPH as the
`final addition. After incubation for 20 min at 30 ° C, 0.25 mL
`of 6% perchloric acid was added, and the mixture was kept
`on ice for 30 min prior to centrifugation at 5000 rpm for 10
`min. A 50-µ  aliquot of the supernatant solution was then
`analyzed by HPLC. Each substrate concentration was assayed
`in duplicate along with a blank from which the NADPH was
`omitted. Under these assay conditions, product formation
`was found to be linear for 45 min. The rate of the reaction
`was also found to be linear with respect to the level of P450
`in the range of 0.05-0.7 nmol for a period of 20 min (data not
`shown).
`HPLC Analysis. Quantitative analysis of the enzymatically
`formed aldehyde or ketone was done with a Waters µBonda-
`pack C-18 reversed-phase analytical column with use of an
`automated HPLC system consisting of a Waters WISP Model
`710 autosampler, a Model 600 solvent delivery system, and
`a Model 480 UV/visible detector set at
`the appropriate
`wavelength maximum of the carbonyl product, and a Hewlett-
`Packard Model 3600 integrator. An isocratic solvent system
`consisting of acetonitrile and water containing 0.1% trifluo-
`roacetic acid was used for all determinations, with the
`concentration of acetonitrile adjusted so that the retention
`time of the aldehyde or ketone was between 7 and 9 min.
`Solvent mixtures that led to elution of the product earlier
`than 6 min caused it to appear as a shoulder on the front of
`the NADP/NADPH peak, frequently resulting in incorrect
`recognition of the peak for automated integration. Elution
`times longer than 10 min were also undesirable, since the
`product was eluted as a broad peak, resulting in decreased
`sensitivity. Standards of the carbonyl product were run with
`each assay in the range from 20 to 600 pmol; in this range,
`the integrated area of the peak was
`found to be directly
`proportional to the amount of the standard. Typically, the
`lower limit of accurate product quantitation was 30 pmol.
`injected in duplicate, and the mean
`Each sample was
`integration value was used to quantitate the carbonyl product.
`Intramolecular Deuterium Isotope Effect with [1-2H¡]-
`Benzyl Alcohol Determined by GC/MS. A typical reaction
`mixture contained 50 µ     of potassium phosphate buffer,
`pH 7.4, 60 µg of DLPC, 0.5 nmol of P450 form 2E1 or 1.0
`nmol of form 2B4 reconstituted with reductase in a 1:1 molar
`ratio as described, 1.5 or 6.0 µ     of [l-2Hi]benzyl alcohol
`for reactions with 2E1 or 2B4, respectively, and 5 µ     of
`NADPH in a final volume of 1.0 mL. The mixture was
`incubated at 30 °C for 2 h, after which time it was extracted
`by vigorous mixing with 2 mL of CH2CI2. The CH2CI2 layer
`was dried over anhydrous sodium sulfate, reduced to a volume
`of approximately 0.5 mL under a stream of nitrogen, and
`applied to an analytical silica gel HPLC column (25 X 0.4
`cm) previously equilibrated with CH2CI2. The column was
`treated with CH2CI2 at a flow rate of 1 mL/min, and the
`fraction eluted between 4.0 and 6.0 min was collected and
`concentrated to approximately 20 µ  under a stream of dry
`(Benzaldehyde and benzyl alcohol standards were
`nitrogen.
`eluted at 4.5 and 12.2 min, respectively, under these condi-
`tions.) A 4-µ  aliquot was injected onto a 30-m DB-5 fused
`silica capillary column (l.O-µ   film thickness, 0.32 µ  
`i.d.,
`J&W Scientific) in a Finnigan gas chromatograph. The
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`

`

`6444 Biochemistry, Vol. 33, No. 21, 1994
`
`system
`
`Table 1: Components Required and Effect of Catalase and
`Superoxide Dismutase on Oxidation of Benzyl Alcohol by P450 2E1
`activity
`[nmol min-1
`(nmol of P450)'1]
`3.12 ±0.02
`complete"
`cytochrome P450 omitted
`0.0
`reductase omitted
`0.0
`NADPH omitted
`0.0
`3.20 ± 0.04
`complete + catalase (10 units)
`3.05 ± 0.04
`complete + catalase (100 units)
`3.20 ± 0.05]
`complete + superoxide dismutase (180 units)
`3.10 ± 0.04
`complete + superoxide dismutase (720 units)
`“ The complete system was as described in the text, with 1.5 µ     of
`In the other experiments, the components indicated were
`benzyl alcohol.
`selectively omitted or added to the complete system.
`
`splitless injector was maintained at 200 °C, and helium was
`used as the carrier gas at a head pressure of 10.0 psi. The
`column temperature was maintained at 50 °C for 2 min and
`then raised to 27 5 °Cat 10 °C/min. Benzaldehyde was eluted
`at 7.4 min under these conditions. The gas chromatograph
`was attached to a Finnigan 4021 mass spectrometer operating
`at 70 eV. Data aquisition and processing were done with the
`Finnigan INCOS data system.
`Solvent Deuterium Incorporation into Acetophenone. A
`reaction mixture similar to that described above for deter-
`mination of the intramolecular isotope effect with benzyl
`alcohol was used, except that the medium was 92% deuterium
`oxide and the substrate was 1 -phenylethanol at a concentration
`of 3.0 or 6.0 mM with P450 2E1 or 2B4, respectively. A
`comparable reaction mixture lacking NADPH and the
`substrate but containing 1.5 /umol of acetophenone served as
`a control for the exchange of deuterium with the methyl
`hydrogens. Acetophenone was isolated by HPLC on silica
`for
`gel and analyzed by GC/MS as described earlier
`benzaldehyde. The retention times of acetophenone and
`1-phenylethanol on the silica gel column were 6.5 and 14.0
`min, respectively; the retention time of acetophenone on the
`DB-5 fused silica capillary column was 8.5 min.
`Incorporation of Oxygen from i802 into Acetophenone. A
`typical 10-mL reaction mixture contained 0.5 mmol of
`potassium phosphate buffer, pH 7.4, 0.6 mg of DLPC, 60
`/¿mol of 1-phenylethanol, and 5.0 nmol of P450 2B4 or 2.5
`nmol of P450 2E1 reconstituted with reductase as described.
`The mixtures were made anaerobic by repeated purging with
`oxygen-free nitrogen, 18C>2 was then introduced, and NADPH
`(0.1 mmol) was injected as an aqueous anaerobic solution to
`initiate the reaction. The incubation was at 30 °C for 2.5 h,
`after which time the reaction mixtures were extracted with
`two 5-mL aliquots of CH2CI2. The extract was dried over
`anhydrous sodium sulfate and then evaporated under reduced
`pressure to 0.5 mL, and the acetophenone was purified by
`silica gel chromatography and analyzed by GC/MS as
`described above. To determine the extent of 1602 isotope
`dilution that might have occurred under the experimental
`conditions, the hydroxylation of toluene to benzyl alcohol by
`P450 2B4 was determined. GC/MS analysis of the resulting
`benzyl alcohol showed 96% 180 incorporation (data not given),
`indicating that insignificant isotopic dilution took place under
`the experimental conditions used.
`RESULTS
`Components Required. The requirements for the oxidation
`of benzyl alcohol
`to benzaldehyde by P450 2E1 in the
`reconstituted system and the effect of catalase and superoxide
`dismutase on the reaction are shown in Table 1. The formation
`
`Vaz and Coon
`
`v/[S], min-1 mM-1
`Figure 1: Typical Woolf-Augustinsson-Hofstee plot as shown for
`p-fluorophenylethyl alcohol with P450 2B4,
`from which kinetic
`constants were obtained. The plot indicates the inhibition observed
`at the highest substrate concentrations. Km and kM values were
`obtained from those concentrations at which substrate inhibition was
`not observed.
`of benzaldehyde is dependent on P450, the reductase, and
`NADPH; omission of any of these components results in no
`observable oxidation of the alcohol. Catalase and superoxide
`dismutase have no effect on product formation, indicating
`that hydrogen peroxide and superoxide, which are produced
`in the reconstituted enzyme system (White & Coon, 1980),
`are not involved in a nonenzymatic, Fenton-type reaction with
`the alcohol. Similar results were obtained with all of the
`other alcohols examined in this study, including the experi-
`ments with P450 2B4, thus showing that alcohol oxidation
`occurs within the catalytic site of the cytochrome.
`Steady-State Kinetics and Linear Free Energy Correlation
`Analysis. The activities of benzyl alcohol and seven para-
`substituted derivatives and of 1 -phenylethanol and eight para-
`substituted derivatives were examined in each case at six
`concentrations, the range of which depended on the form of
`P450 being studied. With P450 2B4, the substrate concen-
`trations varied from 0.3 to 10.0 mM, and with 2E1 they were
`from 0.1 to 3.0 mM. The kinetic constants were obtained
`from linear regression analysis of the initial rates fitted to
`Lineweaver-Burk and Woolf-Augustinson-Hofstee plots
`the latter
`for
`shown in Figure 1
`is
`(Segel, 1975);
`p-fluorophenylethyl alcohol. Some substrates showed inhibi-
`tion at high concentrations with both cytochromes. Accord-
`ingly, such results were not included in the determination of
`kinetic constants, but at least four data points were used for
`each value calculated. Both plots gave correlation coefficients
`greater than 0.99, and the kinetic constants obtained by the
`two analytical methods were in good agreement. Tables 2
`and 3 summarize the steady-state kinetic constants determined
`for the oxidation of the series of benzyl alcohols and the series
`of 1-phenylethyl alcohols, respectively, by both P450 2B4 and
`P450 2E1. The Km values vary from 0.11 to 7.3 mM for the
`benzyl alcohols and from 0.05 to 7.5 mM for the phenylethyl
`alcohols, with no obvious correlation with the partition
`coefficient, the P450 used, or the rate of oxidation. The fc<»t
`values were used for linear free energy correlation analysis as
`shown in Figure 2 A for the oxidation of para-substituted benzyl
`alcohols and in Figure 2B for the oxidation of para-substituted
`1 -phenylethyl alcohols by the two cytochromes. The reaction
`constants for the oxidation of benzyl alcohols obtained from
`these plots are -0.46 (correlation coefficient =
`-0.59) and
`-0.37 (correlation coefficient =
`-0.60) with P450 2B4 and
`2E1, respectively. With 1-phenylethyl alcohols,
`the cor-
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`

`

`Alcohol Oxidation by Cytochrome P450
`
`Biochemistry, Vol. 33, No. 21, 1994
`
`6445
`
`Table 2: Steady-State Kinetic Constants for Oxidation of
`para-Substituted Benzyl Alcohols by P450s 2E1 and 2B41
`P450 2B4
`P450 2E1
`
`Hammett
`Km
`Km
`kcal
`&cat
`constant ( ° )
`(mM)
`(mM)
`(min-1)
`(min-1)
`para-substituent
`och3
`-0.27
`4.20
`4.31
`1.20
`1.00
`ch3
`-0.17
`0.33
`0.79
`3.37
`0.31
`H
`0.00
`7.28
`0.45
`3.38
`3.59
`F
`0.32
`2.89
`0.06
`0.61
`1.28
`Br
`0.13
`0.75
`0.13
`2.08
`0.23
`Cl
`0.52
`0.13
`2.37
`0.23
`0.11
`CN
`0.49
`0.69
`0.66
`1.48
`2.31
`no2
`2.90
`0.73
`0.78
`0.43
`1.90
`“ The kc»t and Km values were obtained from Lineweaver-Burk and
`Woolf-Augustinsson-Hofstee plots and are the mean of two or
`three
`separate determinations. The Hammett substituent constants are taken
`from Hansch (1973).
`
`Table 3: Steady-State Kinetic Constants for Oxidation of
`para-Substituted 1-Phenylethyl Alcohols by P450s 2E1 and 2B4“
`P450 2B4
`P450 2E1
`
`Hammett
`Km
`Km
`(mM)
`(mM)
`constant ( ° )
`(min-1)
`(min-1)
`para-substituent
`och3
`-0.27
`0.64
`3.50
`3.12
`3.46
`ch3
`-0.17
`0.72
`0.44
`4.37
`3.38
`H
`4.04
`0.00
`0.49
`6.03
`1.88
`0.26
`0.06
`2.28
`0.65
`2.35
`F
`Br
`0.06
`0.06
`0.92
`0.23
`1.59
`Cl
`0.27
`0.05
`0.23
`1.76
`1.93
`COOH
`6.09
`7.54
`0.62
`0.45
`0.31
`CN
`0.42
`0.66
`0.97
`1.19
`1.19
`no2
`0.13
`0.78
`1.00
`1.06
`2.31
`‘ The km and Km values were obtained from Lineweaver-Burk and
`Woolf-Augustinsson-Hofstee plots and are the mean of two or
`three
`separate determinations. The Hammett substituent constants are taken
`from Hansch (1973).
`
`=
`
`=
`
`responding reaction constants are-1.41 (correlation coefficient
`-0.96) and -1.19 (correlation coefficient
`-0.85),
`respectively. The results with p-carboxy-1 -phenylethyl alcohol
`deviated significantly from linearity and were excluded from
`linear regression analysis of the data. The significance of the
`reaction constants in connection with the isotope effects to be
`presented is discussed below.
`Deuterium Isotope Effect with Benzyl and Phenylethyl
`Alcohols. Table 4 shows the isotope effect on the steady-
`state parameters for the oxidation of benzyl alcohol as
`compared to its 1-dideuterio derivative and of 1-phenylethyl
`alcohol as compared to its 1 -monodeuterio derivative by P450
`2B4 and P450 2E1. With phenylethyl alcohol and either
`cytochrome no isotope effect on the catalytic constant was
`observed, suggesting that, with these enzymes, cleavage of
`the benzylic carbon-hydrogen bond is not rate-determining
`in the overall reaction. With the dideuterio benzyl alcohol,
`however, a significant isotope effect was obtained with both
`of the cytochromes. The effect on the catalytic constant
`in
`the case of P450 2B4 is 2.8, indicating that carbon-hydrogen
`bond breakage contributes to the enzymatic rate-limiting
`In contrast, the isotope effect on this parameter with
`process.
`P450 2E1 is 1.3, suggesting that hydrogen abstraction is only
`partially rate-limiting in the overall reaction. With respect
`to the Km values, a significant isotope effect was seen only
`with P450 2B4 and benzyl alcohol; the cause of the 1.6-fold
`increase has not been studied in detail.
`Intramolecular Deuterium Isotope Effects with Benzyl
`Alcohol. The procedure used for examining the intramolecular
`isotope effect for the oxidation of monodeuterio benzyl alcohol
`is given in Materials and Methods. The mass fragmentation
`
`Figure 2: Linear free energy correlation diagrams for the P450
`2E1-catalyzed ( ) and P450 2B4-catalyzed (A) oxidation of benzyl
`alcohols (panel A) and of 1-phenylethyl alcohols (panel B). Ln k\/
`kn was calculated from the rate constants shown in Tables 2 and 3,
`and values for  °  were obtained from Hansch (1973).
`pattern of benzaldehyde at 70 eV shows a molecular ion peak
`1)+ ion peak (m/z = 105) of
`(m/z = 106) and an (M -
`approximately the same intensity (Figure 3, panel A). With
`deuterium at the aldehydic position, produced in the enzymatic
`oxidation of the dideuterio benzyl alcohol, the fragmentation
`pattern has a molecular ion peak at m/z = 107 and an (M
`- 2)+ peak at m/z = 105 of about the same intensity (panel
`B). The loss of one mass unit from benzaldehyde and two
`units from deuteriobenzaldehyde indicates that in this mass
`region the fragmentation pattern arises from the loss of the
`aldehydic hydrogen. The small m/z = 106 peak obtained
`with [2H] benzaldehyde with an intensity approximately 8%
`of that of the m/z = 107 peak is due to the inherent contribution
`by 13C (5.6%) and by the residual hydrogen present from the
`lithium aluminum deuteride (98% isotopic purity) used in the
`synthesis of [1-2H2] benzyl alcohol. Thus, in a mixture of
`the ratio of the
`benzaldehyde and deuteriobenzaldehyde,
`corrected signal intensities at m/z = 106 and 107 gives the
`relative abundance of the two species in the mixture. As shown
`by the scheme in Figure 4, benzaldehyde formed by enzymatic
`oxidation of (±) [ 1 -2Hi] benzyl alcohol by P450 would contain
`hydrogen and deuterium at the aldehydic position in amounts
`proportional to the rates of C-D and C-H bond cleavage,
`the ratio of the corrected peak
`respectively. Therefore,
`intensities at m/z =
`107 and 106 corresponds to the
`intramolecular deuterium isotope effect. Panels C and D of
`Figure 3 show the relevant ion mass fragments of benzaldehyde
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 4
`
`

`

`6446 Biochemistry, Vol. 33, No. 21, 1994
`
`Vaz and Coon
`
`Table 4: Steady-State Kinetic Deuterium Isotope Effects on Oxidation of Benzyl Alcohol and 1-Phenylethyl Alcohol by P450 Cytochromes 2E1
`and 2B4°
`
`P450 2E1
`(H)
`&cat (D)
`
`1.3
`
`Km (H)
`Km (D)
`
`1.1
`
`P450 2B4
`kca. (H)
`&cat (D)
`
`2.8
`
`Km (H)
`Km (D)
`
`1.6
`
`Km (mM)
`substrate
`Km (mM)
`kcat (min-1)
`ka,t (min-1)
`0.54
`benzyl alcohol
`4.26
`3.22
`7.82
`[1-2H2] benzyl alcohol
`0.49
`3.33
`4.93
`1.15
`1-phenylethyl alcohol
`0.50
`5.86
`3.95
`1.87
`[ 1 -2H i] -1 -phenylethyl alcohol
`4.28
`0.9
`0.53
`0.9
`5.38
`1.68
`1.1
`1.1
`0 The fcd and Km values were obtained from Lineweaver-Burk and Woolf-Augustinsson-Hofstee plots and are the mean of two or three separate
`determinations.
`
`Ph—CDO (M+1 = 107)
`
`catalyzed oxidation of [ 1 -2Hi] benzyl alcohol; and panel D, benzaldehyde obtained from P450 2B4-catalyzed oxidation of [ 1 -2Hi]benzyl alcohol.
`incorporation of an atom of deuterium into the enzymatically
`formed acetophenone would cause the m/z = 121 ion signal
`intensity to increase by approximately 92% and the m/z =
`120 intensity to decrease by an equal amount. However, as
`the intensity of m/z =
`shown in Figure 5,
`121 was not
`significantly increased relative to that of the molecular ion
`peak at m/z = 120. The results indicate that an oxidative
`desaturation mechanism does not operate with either form of
`P450 examined.
`The possible incorporation of an atom of 1S02 into
`acetophenone was determined to establish whether a gem-
`diol
`intermediate in the oxidation of
`is formed as an
`1-phenylethyl alcohol by either form of P450.
`(The rapid
`exchange of oxygen at the aldehyde function precluded the
`estimation of 180 incorporation into benzaldehyde formed
`from benzyl alcohol.) Figure 6 shows the relevant mass
`fragmentation region of acetophenone. The extent of oxygen
`incorporation was determined from the ion intensities at m/z
`122 and 120 and at m/z 107 and 105. The results indicate
`that with P450 forms 2B4 and 2E1 the incorporation of oxygen
`from 1802 into the carbonyl group was 32 ± 2 and 29 ± 2%,
`respectively. A control experiment with acetophenone in
`H2180 under similar conditions showed incorporation of 14
`± 2% lsO due to solvent exchange; thus, it is obvious that the
`observed 1802 incorporation into the product would have been
`greater without loss by such an exchange.
`
`Ph—C(H)(D)OH
`
` )
`
`Ph—CHO <M+‘ = 106)
`Figure 4: Molecular ion peak (m/z value) expected in mass
`fragmentation pattern of benzaldehyde formed by loss of hydrogen
`(kH) or of deuterium (kD) from [l-2Hi]benzyl alcohol.
`
`formed from (±) [l-2Hi]benzyl alcohol by P450 2B4 and
`P450 2E1, respectively. The similar and relatively small
`magnitude of the intramolecular isotope effect for benzyl
`alcohol oxidation by 2E1 and 2B4, 2.8 and 2.6, respectively,
`suggests a comparable geometry for bond cleavage by the two
`forms of the cytochrome, but the significance with respect to
`transition states, as discussed by More O’Ferrall (1970) for
`chemical models, is not clear.
`Isotope Incorporation into Acetophenone from Solvent or
`Molecular Oxygen in Enzymatic Oxidation of 1-Phenylethyl
`Alcohol. The ability of P450 to catalyze oxidative desaturation
`reactions (Nagata et al., 1986; Rettie et al., 1987, 1988)
`suggests the possibility of such a reaction with 1-phenylethyl
`alcohol to form the enol as a route to the carbonyl product.
`With this alcohol such a mechanism would result in the uptake
`of a solvent-derived proton into the methyl group of acet-
`ophenone. Accordingly, acetophenone formed enzymatically
`in 92% deuterium oxide from 1-phenylethyl alcohol was
`examined for deuterium uptake. The expectation was that
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 5
`
`

`

`Alcohol Oxidation by Cytochrome P450
`
`Biochemistry, Vol. 33, No. 21, 1994
`
`6447
`
`m/z
`Figure 5: Mass spectral fragmentation pattern for acetophenone in the molecular ion region. Panel A, authentic standard; panels B and C,
`acetophenone produced from 1-phenylethanol oxidation by P450 2B4 or P450 2E1, respectively, in D20; panel D, control experiment with
`P450 2B4 (or 2E1) present but with NADPH and 1-phenylethanol omitted and 1.5 µ     of acetophenone added.
`
`m/z
`Figure 6:
`ls02-derived oxygen incorporation into the carbonyl group of acetophenone, determined from the mass spectral fragmentation
`pattern in the region between m/z = 100 and 125. Acetophenone standard, panel A; acetophenone obtained from oxidation of 1-phenylethyl
`alcohol under an atmosphere of l802 by P450 2E1 or 2B4, panels B and C, respectively; acetophenone (1.5 Mmol) incubated in H2180 under
`conditions similar to those of the enzymatic reaction, but with NADPH and 1-phenylethanol omitted, panel D.
`DISCUSSION
`and 2B4 are about equally active with unsubstituted benzyl
`alcohol, and 2B4 has 50% greater activity than 2E1 with
`As with certain other types of reactions catalyzed by P450
`unsubstituted 1-phenylethanol.
`in microsomal membranes, alcohol oxidation is accomplished
`For chemical reactions where the rate-limiting step is at a
`isozymes with overlapping substrate specificity.
`by numerous
`the sign and magnitude of the reaction
`Purified ethanol-inducible P450 2E1 was originally shown to
`benzylic carbon,
`be more active than four other forms of the cytochrome in the
`constant can distinguish the anionic, free radical, or cationic
`character of the intermediate developed in the transition state
`oxidation of ethanol as well as other aliphatic alcohols (Morgan
`(Hammett, 1940). Correlation of the catalytic constants with
`et al., 1982). Additional substrates were
`subsequently
`the Hammett substituent constants for the oxidation of benzylic
`identified (Koop & Coon, 1986), including benzyl alcohols,
`which are particularly useful for mechanistic studies, as briefly
`alcohols examined in this study gave reaction constants between
`reported several years ago (Vaz & Coon, 1989). The activities
`-0.37 and -1.41 with both forms of P450. These small negative
`vary somewhat with the substrate examined, but P450s 2E1
`values indicate that the slow step in the enzymatic reaction,
`
`Auspex Exhibit 2005
`Apotex v. Auspex
`IPR2021-01507
`Page 6
`
`

`

`6448 Biochemistry, Vol. 33, No. 21, 1994
`while being relatively insensitive to electronic perturbation on
`the aromatic ring, involves an electron-deficient species. The
`lack of deviation from linearity between Hammett  °  values
`from -0.27 to +0.78 suggests a single transition state for the
`oxidation of these alcohols. The correlation diagrams in this
`study have been limited to the Hammett substituent constant
` ° . A further detailed description of the active-site mechanics
`involved in benzylic alcohol oxidations might be obtained from
`consideration of other substituent constants (Hansch, 1973)
`as reported for the hydroxylation of toluene by P450 (White
`& McCarthy, 1986), but is beyond the scope of this study.
`Deuterium isotope effects, a useful means of examining the
`cleavage of carbon-hydrogen bonds and the extent to which
`such cleavages are enzymatically rate-limiting, have been used
`previously to study various P4