Biochemistry 1998, 37, 7039-7046
`
`7039
`
`Evaluation of Cytochrome P450 Mechanism and Kinetics Using Kinetic Deuterium
`Isotope Effects†
`LeeAnn Higgins, Grace A. Bennett, Miyuki Shimoji, and Jeffrey P. Jones*
`Department of Pharmacology and Physiology, UniVersity of Rochester, Rochester, New York 14642
`ReceiVed December 4, 1997; ReVised Manuscript ReceiVed March 11, 1998
`
`ABSTRACT: In this paper two hypotheses are tested: (i) the active oxygen species is similar in energetics
`for all cytochrome P450 (CYP) enzymes and (ii) linear free-energy relationships can be used to evaluate
`the mechanism of the reaction of these enzymes. A series of intramolecular isotope effects were determined
`and compared for CYPs 1A2, 2B1, 2C9, 2E1, and P450cam. The results indicate that the isotope effects
`are very similar for each of these isoforms of P450 and that the first hypothesis is likely to be true.
`Attempts to establish a linear free-energy relationship were only moderately successful: log Vmax ) 0.11(cid:243)+
`p
`+ 1.73; r2 ) 0.588. It was determined, through the use of intermolecular isotope effects, that the rates
`of hydrogen atom abstraction are masked. Thus, the second hypothesis is found to be false. This is
`likely to be a general result for CYP reactions, and linear free-energy relationships can only be used to
`determine the mechanism under very special circumstances. In all cases, the rate-limiting step should be
`evaluated with isotope effect experiments before any mechanistic conclusions can be drawn.
`If the
`intermolecular isotope effects are found to be masked, no mechanistic conclusion can be drawn from the
`linear free-energy relationship study.
`
`The cytochrome P450 (CYP)1 enzymes comprise a su-
`perfamily of heme-containing enzymes that consists of more
`than 300 individual isoforms, and the CYP enzymes are
`found in plant, bacterial, and mammalian species (1). These
`enzymes function mainly as monooxygenases (2). In mam-
`mals, they are responsible for the metabolism of certain
`endogenous as well as exogenous compounds (3). Because
`these enzymes play an important role in xenobiotic metabo-
`lism, the abilities to predict reaction rates, identify possible
`metabolites, and understand CYP mechanisms are important
`goals. Electronic models have been developed that use
`computational (4, 5) and chemical (6, 7) approaches for the
`successful prediction of relative reaction rates and isotope
`effect profiles for some classes of small compounds.
`One of the most important considerations for a potential
`CYP model is whether the model will apply to relative rate
`predictions for several (or all) CYP isoforms or for a limited
`number of isoforms.
`In this study, we use isotope effect
`profiles (IEPs) to test the hypothesis that the active oxygen
`is conserved in multiple CYP isoforms. An IEP consists of
`
`† This research was supported by the National Institutes of Health,
`Grants ES06062 and ES09122.
`* To whom correspondence should be addressed. Department of
`Pharmacology and Physiology, Box 711, University of Rochester, 601
`Elmwood Avenue, Rochester, NY 14642. Phone:
`(716) 275-5371.
`Fax: (716) 244-9283. E-mail:
`jpj@cyp.medicine.rochester.edu.
`1 Abbreviations: CYP, cytochrome P450; LFER, linear free-energy
`relationship; IEP, isotope effect profile; KIE, kinetic isotope effect(s);
`LAD, lithium aluminum deuteride; CH2Cl2, dichloromethane; TLC,
`thin-layer chromatography; NMR, nuclear magnetic resonance; GC-
`MS, gas chromatography-mass spectroscopy; TK-, thymidine kinase
`deficient; KPi, potassium phosphate; Pd, putidaredoxin; PdR, putidare-
`doxin reductase; E. coli, Escherichia coli; NADP+, (cid:226)-nicotinamide
`adenine dinucleotide phosphate; NADH, (cid:226)-nicotinamide adenine di-
`nucleotide, reduced; MTBSTFA, N-methyl-N-(tert-butylimethylsilyl)-
`trifluoroacetamide; HP, Hewlett-Packard; D0, nondeuterated; D3, tri-
`deuterated; DV and D(V/K), kinetic deuterium isotope effects on the
`enzymatic rate constants Vmax and V/K, respectively.
`
`isotope effect measurements for a series of structurally related
`compounds that most nearly approximate intrinsic isotope
`effects. When intrinsic KIEs are available, comparable
`trends in IEPs between different systems or different isoforms
`support similar mechanisms (6-8).
`If the hypothesis that the active oxygen is very similar in
`multiple CYP isoforms is proven to be true, a single
`electronic model can be used to describe (the corresponding)
`multiple P450 enzymes. Linear free-energy relationships are
`utilized in chemistry and enzymology as a predictive tool
`and as a probe of the electronic characteristics of the
`transition state of a reaction. We test here the second
`hypothesis that LFERs can be used to predict relative rates
`of metabolism by CYP and to evaluate physical character-
`istics of CYP reactions. If relative rates of product formation
`are masked by other steps in an enzymatic cycle, LFERs
`must be interpreted with caution. Kinetic information, but
`not mechanistic information, can be deduced, if relative
`reaction rates are masked. Therefore, it is important to assess
`whether masking occurs with the substrates chosen for the
`analysis of a LFER.
`Herein, we report IEP results for a series of para-
`substituted toluenes. The data support the hypothesis of a
`conserved active oxygen species for multiple CYPs. Our
`attempts to establish LFERs for a series of para-substituted
`toluenes were unsuccessful. From isotope effect studies, we
`conclude that relative reaction rates for para-substituted
`toluenes in this study are partially masked.
`MATERIALS AND METHODS
`Chemicals and Reagents. All solvents were purchased
`from J. T. Baker, Inc. (Phillipsburg, NJ). Chemicals were
`purchased from Aldrich Chemical Co. (Milwaukee, WI) with
`the following exceptions: MTBSTFA (+1% tert-butyldi-
`methylchlorosilane) was purchased from Regis Technologies,
`
`S0006-2960(97)02986-3 CCC: $15.00 © 1998 American Chemical Society
`Published on Web 04/23/1998
`
`Downloaded via WASHINGTON STATE UNIV on April 14, 2020 at 18:45:46 (UTC).
`
`See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
`
`Apotex Ex. 1026
`
`Apotex v. Auspex
`IPR2021-01507
`
`

`

`7040 Biochemistry, Vol. 37, No. 19, 1998
`
`Inc. (Morton Grove, IL), and potassium phosphate dibasic
`and monobasic and magnesium sulfate were from EM
`Science (associate of E. Merck, Darmstadt, Germany).
`p-Chlorotoluene, p-methylanisole, and p-xylene were purified
`by distillation before use. p-Bromotoluene was crystallized
`from ethanol, and p-tolunitrile was crystallized from benzene:
`pet ether. Gibco BRL products (Gaithersburg, MD) were
`used for all cell culture methods. Biochemicals were
`purchased from Sigma (St. Louis, MO). p-Xylene-R, R, R,
`R¢, R¢, R¢-d 6 (99.8% pure by GC) was purchased from
`Aldrich Chemical Co.
`Synthesis of Compounds. Para-Substituted Benzyl-R-2H2
`Alcohols. Methyl 4-bromobenzoate, methyl 4-chloroben-
`zoate, methyl 4-cyanobenzoate, methyl 4-methylbenzoate,
`and methyl 4-methoxybenzoate were used as starting materi-
`als for the synthesis of the corresponding para-substituted
`benzyl alcohols according to the following procedure.
`Lithium aluminum deuteride (1.2 mol) was suspended in
`ether and cooled to 0 (cid:176)C in an ice bath while under N 2. The
`substituted benzoate (1 mol) was dissolved in a small amount
`of ether and added dropwise to the LAD suspension while
`keeping the temperature at 0 (cid:176) C. After the addition, the
`mixture was stirred at 0 (cid:176)C for an additional hour, or until
`complete by TLC (silica gel: 80% hexane, 20% ethyl
`acetate). The unreacted LAD was decomposed, and the
`mixture was filtered, dried over MgSO4, and evaporated to
`dryness. The crude product was purified by recrystallization
`and/or distillation, and purity was assessed by NMR and
`GC-MS.
`Para-Substituted Toluenes-R,R,R-2H3. p-Bromotoluene,
`p-chlorotoluene, and p-methylanisole were synthesized from
`the corresponding para-substituted benzyl-d2 alcohols ac-
`cording to the following method. Methane sulfonyl chloride
`(0.11 mol) was dissolved in a small amount of CH2Cl2 and
`added dropwise to a solution of the para-substituted benzyl-
`d2 alcohol (0.10 mol) and triethylamine (0.15 mol) in CH2-
`Cl2 that was cooled to -78 (cid:176)C in a dry ice -acetone bath.
`The reaction was stirred for another 30 min at -78 (cid:176)C.
`Because the mesylates are generally unstable, they were not
`isolated, and the procedure was continued by adding ice
`water, separating the layers, and extracting the aqueous layer
`with CH2Cl2. The organic layers were combined, washed
`with saturated sodium bicarbonate, and dried over MgSO4,
`and the solvent was removed under reduced pressure. The
`oil that remained was reconstituted in ether and reduced with
`LAD using the same procedure as described for the reduction
`of the benzoates. The products were purified by chroma-
`tography (silica gel: 100% CH2Cl2), distilled under vacuum
`using a Kugelrohr, and, when necessary, purified again with
`a neutral alumina column and 100% hexane to remove any
`residual benzyl alcohol. Any residual benzyl alcohol was
`detected by select ion monitoring using GC-MS as described
`below.
`Substituted Toluene-R-2H1 and -R-2H1-R¢ -2H1. Monodeu-
`terated analogues of p-bromotoluene, p-chlorotoluene, p-
`methylanisole, and toluene were synthesized by the mesyl-
`ation of the corresponding nondeuterated benzyl alcohols
`according to the same procedure described above for the
`trideuterated compounds, and p-xylene-R-2H1-R¢- 2H1 was
`synthesized with the same method from benzene dimethanol,
`except that the mole ratio of benzene dimethanol to the other
`reactants was 1:2.
`
`Higgins et al.
`For p-tolunitrile-R-2H1, the benzyl alcohol was first syn-
`thesized by the reduction of p-cyanobenzaldehyde using sod-
`ium borohydride. For this procedure, the sodium borohydride
`(0.05 mol) was mixed with about 100 mL of ethanol and
`stirred under N2. The aldehyde (0.025 mol) was dissolved
`in a small amount of ethanol and added dropwise to the
`NaBH4 solution. The reaction was monitored by TLC (silica
`gel: 50% ethyl acetate, 50% hexane). After about 1.5 h,
`the mixture was filtered and the solvent was evaporated. The
`oil that remained was mixed with water and then extracted
`with CH2Cl2. The organic layer was dried over MgSO4 and
`evaporated to yield a white solid, which was recrystallized
`with hexane and assessed for purity by NMR and GC-MS.
`Preparation of HepG2-Expressed CYP. Individual CYP
`isoforms (1A2, 2B1, 2E1, and 2C9) were obtained using the
`HepG2 vaccinia expression system described previously (9).
`Briefly, the recombinant vaccinia viruses containing cDNAs
`encoding a single CYP were propagated in human TK- 143
`(human embryoblast) cells, which were then used to infect
`confluent flasks of hepatoma HepG2 cells (American Type
`Culture Collection, Rockville, MD). For each isoform, cells
`were scraped from 50 flasks, pooled, divided into 1-mL
`aliquots, and stored at -78 (cid:176)C. Microsomal preparations
`from crude cell lysates were obtained immediately before
`use according to the following procedure: aliquots were
`thawed, mixed with 1 mL of 50 mM KPi buffer pH 7.4,
`sonicated with a Branson Sonifier Model 450, and centri-
`fuged for 15 min at 300000g in a cold (4 (cid:176)C) Beckman TL-
`100 ultracentrifuge; pellets were reconstituted with 2 mL of
`KPi buffer pH 7.4, pooled, and homogenized in a cold room.
`The total protein concentration of the final enzyme-buffer
`mixture was typically 3-4 mg mL-1 by Bradford assay (10)
`with bovine serum albumin as the standard. The typical low
`levels of P450 obtained using this expression system were
`undetectable by CO difference spectra.
`Cytochrome P450cam, Putidaredoxin, and Putidaredoxin
`Reductase. Construction of expression plasmids for P450cam,
`Pd, and PdR as well as protein expression and purification
`have been described previously (11).
`Incubation Conditions and Isolation of Products. All
`incubations using HepG2-expressed CYP were run for 20
`min in a shaking water bath at 37 (cid:176)C and contained the
`following: 300-400 (cid:237)g of protein suspension, depending
`on the isoform; 10 mM glucose-6-phosphate, 1 mM NADP+,
`and 1 U glucose-6-phosphate dehydrogenase as an NADPH
`generating system; and 1400 U catalase in a total volume of
`5 mL of 50 mM KPi buffer, pH 7.4, 2 mM MgCl2. For
`P450cam, the assays were carried out at 37 (cid:176)C for 20 min
`in a 1-mL reaction mixture in 50 mM KPi buffer, pH 7.4,
`200 mM KCl, 1.5 (cid:237)M P450cam, 3 (cid:237)M Pd, 1.5 (cid:237)M PdR,
`and 0.3 mM NADH. Reactions were initiated by the addition
`of 50 (cid:237)L (HepG2-expressed CYPs) or 10 (cid:237)L (P450cam) of
`a substrate solution in CH3CN, terminated upon addition of
`3 mL of CH2Cl2, and 10 (cid:237)L of internal standard in CH3CN
`was added,
`if applicable (see below). Samples were
`extracted three times with CH2Cl2; the organic fractions were
`pooled, dried with anhydrous MgSO4, and concentrated
`gently under a stream of N2 to a volume of approximately
`100 (cid:237)L. After reconstitution with 300 (cid:237)L (HepG2-expressed
`CYPs) or 100 (cid:237)L (P450cam) of CH3CN, 50 (cid:237)L of MTB-
`STFA + 1% t-BDMCS was added, and samples were heated
`overnight at 70 (cid:176)C.
`
`Apotex Ex. 1026
`
`

`

`P450 Mechanism, Kinetics, and Isotope Effects
`
`Biochemistry, Vol. 37, No. 19, 1998 7041
`
`Enzyme incubations with monodeuterated toluenes and
`p-xylene-R-2H1-R¢ -2H1 were performed in triplicate at a single
`substrate concentration of 0.25 mM. Determinations of Vmax
`and Km for D0 or D3 substrates were made using at least six
`samples, with six different substrate concentrations in a range
`between 0.5 and 100 (cid:237)M, and 10 nmol of the appropriate
`para-substituted benzyl alcohol as the internal standard.
`Product formation from D0 substrates was measured with
`the corresponding para-substituted benzyl alcohol dideuter-
`ated at the benzylic carbon as the internal standard; for D3
`substrates the internal standard was the corresponding
`nondeuterated benzyl alcohol. Competitive experiments
`were performed in triplicate with mixtures of D0 and D3
`substrates, each at a final concentration of 0.25 mM.
`Linearity studies using low and high concentrations of select
`D0 substrates showed linear product formation over 30 min.
`Two types of controls were run for each substrate with the
`identical procedure used for the complete samples except
`that in one case substrate was omitted, and in the other case,
`NADP+ and the NADPH-generating system or NADH was
`omitted. Controls for toluene-R-2H1 showed a small peak
`at the same retention time as product that originated from
`the derivatizing agent, which was quantified and subtracted
`from the observed amount of alcohol product. The kinetic
`parameters for the metabolism of the substituted toluenes
`were determined by GC-MS analysis of the M-(dimethyl-
`silyl) derivative of the product benzyl alcohols (see below).
`Control experiments using a mixture of deuterated and
`nondeuterated products showed that there was no isotope
`effect on product workup.
`Product Analysis by GC-MS. All samples were analyzed
`with an HP (Palo Alto, CA) GC/MS 5890/5972 system,
`equipped with an HP 7673A automatic injector. An HP-1
`capillary column (25 m, 0.2-mm ID, 0.5-(cid:237)m film thickness)
`was used for product analysis from all substrates except
`toluene and p-xylene, which were analyzed with an HPWax
`capillary column (30 m, 0.25-mm ID, 0.5-(cid:237)m film thickness).
`For the HP-1 column, the injector and detector temperatures
`were both set at 250 (cid:176) C; for the HPWax column, the injector
`and detector temperatures were 230 and 240 (cid:176) C, respectively.
`All samples were injected in the splitless mode. Oven
`conditions for the HP-1 column were 50 (cid:176)C for 0.5 min, 10
`(cid:176)C/min to 250 (cid:176)C, and 2 min at 250 (cid:176)C except when
`p-tolunitrile was used as the substrate, in which case a
`gradient of 15 (cid:176)C/min was used. For the HPWax column
`the oven conditions were 50 (cid:176)C for 0.5 min, 10 (cid:176)C/min to
`240 (cid:176)C, and 3 min at 240 (cid:176)C. The MS was operated at an
`ionizing voltage of -70 eV and a 20-ms dwell time. Scans
`of authentic standards of the derivatized benzyl alcohols were
`used to find peak retention times and to verify products
`formed from enzyme incubations. Selected ion recording
`of
`the M-(dimethylsilyl)
`ion of products and internal
`standards was used to determine the amount of metabolism.
`The ratio of the protio and deuterio products or of the product
`and internal standard was corrected for ion overlap and
`percent deuterium incorporation with Brauman’s least-
`squares approach (12). Deuterium incorporation of substrates
`was measured with an HP GC-MS 5890-5970 system with
`an HP-1 column using the same operating conditions as
`above except the MS ionizing voltage was adjusted to -10.0
`eV.
`
`Scheme 1
`
`DV )
`
`+ k5H
`k7
`+ k5H
`k7
`
`+ k5H
`k5H
`k5D
`k3
`1 + k5H
`k3
`D(V/K) ) 1
`
`(1)
`
`(2)
`
`Data Analysis. Kinetic parameters Vmax and Km were
`determined by nonlinear regression using GraphPad Prism
`v2.01 (San Diego, CA) data analysis software. Kinetic
`expressions were solved using the program Mathematica
`(Wolfram Research, Champain, IL).
`THEORY
`Linear Free-Energy Relationships and KIE. When LFERs
`are sought, experiments should be performed that definitively
`answer the following question:
`Is the rate of product
`formation masked (is product formation limited) by a rate-
`determining step in the catalytic cycle that is not the chemical
`step of interest? Fortunately, innovative methods for the
`investigation of masking in enzymatic systems were devel-
`oped by Cleland and Northrop (13, 14). With this approach,
`KIE are measured on the enzymatic parameters Vmax and V/K.
`Isotope effects on Vmax and V/K are expressions for the
`isotope effects on rates of product formation with the limiting
`conditions of high and low substrate concentrations, respec-
`tively. In this study we report deuterium isotope effects on
`Vmax and V/K which are designated DV and D(V/K). The
`expressions for both DV and D(V/K) contain the rate constant
`for the isotopically sensitive step, but they differ in the
`following ways: only DV contains the rate constant for the
`dissociation of product from enzyme, and only D(V/K)
`contains the rate constants for the binding of free substrate
`and enzyme, and dissociation of the first ES complex to free
`substrate and enzyme. The relative magnitudes of these steps
`as well as others to the isotopically sensitive step determine
`the degree to which the intrinsic KIE (the KIE associated
`solely with the bond breaking step, i.e., kH/kD) will be
`expressed in DV or D(V/K). Therefore, comparisons of DV
`and D(V/K) isotope effects to an intrinsic KIE can provide
`information regarding which steps, if any, are involved in
`masking the relative rates of product formation. These types
`of studies can reveal
`if LFERs can be used to make
`mechanistic conclusions. It should be noted that even if
`a linear relationship is seen between the natural log of
`the rates and the descriptors for a series of compounds,
`it does not mean that the relative rates are unmasked.
`Intermolecular Isotope Effects. Intermolecular deuterium
`isotope effects are often used for kinetic and mechanistic
`studies on enzyme systems. While this type of experiment
`usually does not provide significant mechanistic information
`since the intrinsic isotope effect is not observed, it can
`provide information about
`the kinetic mechanism. An
`intermolecular noncompetitive KIE experiment involves the
`
`Apotex Ex. 1026
`
`

`

`7042 Biochemistry, Vol. 37, No. 19, 1998
`
`independent determination of Vmax and V/K using protonated
`
`Scheme 2
`
`D(V/K)P1
`
`)
`
`+ k7Hk6
`k7H
`k7D
`k5k9
`1 + k7Hk6
`k5k9
`
`+ k7H
`k5
`+ k7H
`k5
`
`(3)
`
`DVP1
`
`)
`
`k7H
`k7D
`
`k5k9
`
`k5k9
`
`+[k7Hk6
`1 +[k7Hk6
`
`] +[2k3k7H
`] +[2k3
`
`k13
`
`+ k7H
`k5
`+ k7H
`k5
`
`+ 2k3k6k7H
`+ k3k7H
`+ k3k7H
`k11k5k9
`k13k5
`k11k5
`k13k7D
`+ 2k3k6k7H
`+ k3k7H
`+ k3k7H
`k11k5
`k13k5
`k11k5k9
`
`]
`
`]
`
`(4)
`
`and isotopically labeled substrates, for example, R1-CH3 and
`R1-CD3 (15). For the calculation of a KIE, the kinetic
`constant (Vmax or V/K) pertaining to the unlabeled substrate
`is divided by the corresponding constant for the labeled
`substrate. Scheme 1 shows a potential kinetic mechanism
`that describes an intermolecular noncompetitive KIE, and
`eqs 1 and 2 are the expressions for DV and D(V/K) obtained
`for this scheme. Substrate (SH or SD) combines with enzyme
`(E) reversibly to form the corresponding ES complex (ESH
`(Subscript “H” or “D” signifies the involvement
`or ESD).
`of a compound or a complex in the pathway for the
`abstraction of hydrogen or deuterium, respectively.) Forma-
`tion of an “activated” complex ESH* or ESD*, from ESH or
`ESD, respectively, is depicted here as an irreversible step
`that occurs prior to the isotopically sensitive step. This
`irreversible step is consistent with our understanding of the
`kinetic mechanism of CYP (16, 17).
`It
`is a kinetic
`representation of a collection of steps that
`include the
`following: the first electron reduction (Fe3+ f Fe2+); oxygen
`binding and reduction; the second electron reduction of
`oxygen; and the heterolytic cleavage of the peroxy anion, a
`postulated step which is presumed to be irreversible. Product
`is formed at the isotopically sensitive step, producing an
`enzyme-product complex (EPH or EPD), which is followed
`by the dissociation of the corresponding product from the
`enzyme. Rate constants k1, k2, k3, and k7 may each actually
`represent more than one step, and inclusion of these
`additional steps increases the complexity of the KIE equa-
`tions but does not modify this analysis. When there is no
`pathway for alternate product formation, which is typical for
`most enzymes, the value of DV depends on the relative
`magnitudes of k3 and k7 to k5H. As the rate constants k3 and
`k7 decrease relative to k5H, DV becomes masked (see eq 1).
`D(V/K) for the noncompetitive KIE with the kinetic mech-
`anism shown in Scheme 1, which has no branching pathway
`from ESH* or from the corresponding ESD* position, is
`completely masked (eq 2).
`
`Higgins et al.
`
`While the D(V/K) is completely masked given Scheme 1,
`a branched pathway has the potential to unmask both DV
`and D(V/K) (18). This is illustrated using Scheme 2, which
`represents the formation of two different products (Product
`1 ) P1, Product 2 ) P2) from one substrate, either deuterated
`or nondeuterated. Metabolism at an alternate, isotopically
`insensitive step, k9, is in competition with the isotopically
`sensitive step, k7. Rate constants k1 and k2 represent substrate
`binding to and debinding from the enzyme. Rate constants
`k3 and k3¢ are shown as irreversible steps that represent steps
`that occur after substrate binding, up to and including oxygen
`activation. These steps lead to the formation of “activated
`complexes” (ES1* and ES2*) between the enzyme and the
`substrate in the relative positions 1 and 2. Rate constants
`k3 and k3¢ are assumed to be equal for a given substrate. The
`active oxygen species is stable enough to permit translational
`and conformational changes of the substrate in the active
`site of the enzyme (19-21). Consequently, it is possible
`for an equilibrium to be established between ES1* and ES2*
`(see k5 and k6). This is the key to unmasking KIE. When
`deuterium is substituted for hydrogen at the site of metabo-
`lism, the rate constant for product formation is reduced. This
`can be expressed as k7D < k7H, where k7D represents the rate
`constant for oxidation at a deuterated position on a substrate,
`and k7H represents the rate constant for oxidation at the same
`position on a nondeuterated analogue of the substrate.
`Deuterium substitution will cause the concentration of ES1D*
`to increase relative to the concentration of ES1H* (which is
`the corresponding enzyme-substrate complex for a nondeu-
`terated substrate) unless ES1D* is in equilibrium with another
`complex, for example, ES2D*. (The corresponding equilib-
`rium would exist for the nondeuterated substrate as well.) If
`an alternate pathway does not exist, then [ES1D*] > [ES1H*],
`and an isotope effect is masked.
`Oxidation at an alternate, unlabeled position on a substrate
`with the formation of a second product (P2) can unmask a
`KIE. This is accomplished by the redirection of “excess”
`ES1D* toward ES2D* which allows [ES1D*] ) [ES1H*],
`which is necessary for the expression of a maximum, or
`unmasked, isotope effect. According to the convention of
`Northrop (15), a substrate with a low forward internal
`commitment to catalysis will branch from ES1 toward ES2
`(Scheme 2), and isotope effects on DV and D(V/K) will be
`unmasked, depending on the magnitude of the commitment
`terms. The extent to which D(V/K) is unmasked depends on
`the relative values of k5 and k9 to k7H (eq 3). As the rates of
`substrate rotation and branching increase relative to the rate
`of P1 formation, the amount of unmasking increases. The
`expression for a KIE on DV for Scheme 2 is quite complex
`(not shown) but can be simplified to eq 4 with the following
`two assumptions. If we assume that reduction of oxygen is
`rate-limiting (see prior discussion) and is significantly smaller
`than (i) the rotation of substrate in the active site (k3 , k5,
`k6) and (ii) the rates of product formation k7H and k9, the
`expression for DVP1 is simplified to eq 4.
`If the external
`reverse commitment to catalysis is small (k11, k13 . k3), that
`is, product release is not rate-limiting, then eq 4 reduces to
`eq 3 (DVP1 ) D(V/K)P1).
`Cytochrome P450 has the potential to function as an
`oxidase, whereby molecular oxygen undergoes a four-
`electron reduction, with the overall
`formation of
`two
`molecules of water (22). Formation of the second molecule
`
`Apotex Ex. 1026
`
`

`

`P450 Mechanism, Kinetics, and Isotope Effects
`
`Biochemistry, Vol. 37, No. 19, 1998 7043
`
`of water acts as a branch point leading from the ES complex
`in Scheme 2 and precludes substrate oxidation. As a
`consequence, if branching to water is fast relative to product
`formation, KIE (or rates of product formation) can become
`unmasked.
`Intramolecular Isotope Effects. To obtain better mecha-
`nistic information from enzyme kinetic studies, intramolecu-
`lar KIE can be used. While this type of experimental design
`cannot be used for all enzymes,
`in cases where it
`is
`appropriate it can be used to provide a closer estimate of
`the intrinsic isotope effect, since the potential for masking
`is decreased. For this experiment, a substrate is chosen that
`contains at least two symmetrical sites, with at least one site,
`but not all, isotopically labeled. An intramolecular KIE is
`a special case of branched kinetics in that the branch point
`occurs at the two symmetric sites. Numerous references
`provide discussions and schemes for intramolecular KIE and
`their relation to intrinsic KIE (19, 23, 24).
`An intramolecular KIE will always be greater than or equal
`to an intermolecular KIE. However, it has been shown that
`even intramolecular KIE can be masked (19, 25, 26),
`depending on the choice of substrate. Masking is signifi-
`cantly reduced when substrates are chosen that contain
`hydrogen and deuterium bonded to the same carbon atom
`(26).
`Isotope Effect Profiles. To determine if the mechanisms
`of two reactions are similar, IEPs are preferred over a single
`measurement since the trend in isotope effects provides
`additional information about the energetics of the reaction
`and makes fortuitous agreement between the two systems
`unlikely (27). An IEP consists of isotope effect measure-
`ments in a system (e.g., chemical or enzymatic) using a series
`of compounds that are structurally related. When intrinsic
`KIE are available, comparable trends in KIE profiles between
`different systems argue for similar mechanisms, while
`different trends are indicative of dissimilar mechanisms. The
`rationale for this inference is based on the fundamental idea
`of the Melander-Westheimer principle (28, 29) which is
`based on the Hammond postulate (30). According to the
`Melander-Westheimer principle, maximum isotope effects
`will be observed with reactions that have symmetrical
`reaction coordinates. Mechanisms with nonsymmetrical
`reaction coordinates will yield smaller isotope effects. From
`this information it can be inferred that intrinsic KIE that are
`measured for a series of compounds that differ in some
`systematic (e.g., chemical or physical) way can serve as a
`sensitive probe of the energetics of the isotopically sensitive
`step.
`RESULTS AND DISCUSSION
`Isotope Effect Profiles with Multiple CYPs. Isotope effect
`profiles for multiple CYP isoforms were measured in order
`to test the hypothesis that the active oxygen of multiple CYPs
`is conserved. The intramolecular isotope effects for six para-
`substituted toluenes (p-X-Ph-CH2D: X ) OCH3, CH2D, H,
`Cl, Br, CN; Ph ) phenyl) using expressed CYPs 1A2, 2B1,
`2E1, and 2C9 and purified P450cam are listed in Table 1.
`These experiments were performed with expressed CYP3A4
`also, but product formation was too low to accurately
`measure, most likely because these compounds are poor 3A4
`substrates.
`Isotope effects measured with substrates that
`contain hydrogen and deuterium bonded to the same carbon
`
`Table 1: Intramolecular Isotope Effects for Six Para-Substituted
`Toluenes Using Various CYP Isoforms, Where Isotopic Discrimin-
`ation Occurs at a Monodeuterated Benzylic Position (i.e., R-CH2D)a
`para substituent
`2E1
`2B1
`1A2
`2C9
`P450cam
`4.24 (1)
`3.69 (3)
`4.64 (4)
`4.3 (1)
`4.44 (1)
`OCH3
`CH2Db
`5.45 (5)
`6.23 (5)
`5.59 (1)
`5.9 (1)
`6.0 (2)
`ndc
`H
`6.1 (1)
`7 (1)
`6.1 (4)
`nd
`Cl
`6.75 (2)
`8.1 (1)
`7.06 (2)
`6.2 (1)
`6.5 (8)
`Br
`6.75 (3)
`8.02 (3)
`6.83 (8)
`6.9 (2)
`8.3 (1)
`CN
`10.1 (1)
`11.9 (3)
`10.1 (2)
`11.1 (6)
`11.6 (5)
`a CYPs 2E1, 2B1, 1A2, and 2C9 were expressed in HepG2 cells,
`and P450cam was purified from E. coli. Each result is the average of
`three determinations, and the numbers in parentheses indicate the
`standard deviation in the last significant digit of the mean isotope effect
`value. b p-Xylene-R-2H1-R¢- 2H1 was used as the substrate. c Not deter-
`mined due to low levels of product formation. For each substrate except
`toluene, a comparison of all the isotope effect values revealed that the
`means were statistically different from each other at the 0.05 signifi-
`cance level. This is explained by the fact that the standard deviations
`reflect the precision of the method only and do not take into account
`systematic and day-to-day errors.
`
`atom permit close measurement of the intrinsic isotope effects
`(see prior discussion of intramolecular isotope effects). Table
`1 shows that, for each isoform, the IEPs for the toluenes are
`practically identical: the values range from 4 to 11, and the
`magnitudes of the isotope effects have the same rank order.
`In addition, for each substrate the KIE for all five CYPs are
`in close agreement, which exemplifies the conserved nature
`of the active oxygen species.
`The IEPs presented in this paper provide the first evidence
`for a common reaction mechanism for aliphatic hydroxylation
`at the benzylic position in four mammalian CYP isoforms
`as well as one bacterial isoform. Inclusion of P450cam into
`our series of IEPs is the first extension of this application to
`a P450 outside of the Class II P450s. P450cam belongs to
`Class I, which consists of P450s that interact with two
`electron-transfer partners. Class II members differ in that
`they have only one protein as a reducing partner (31). The
`finding that the KIE profile for P450cam is not different from
`the CYPs that belong to Class II is significant because it
`indicates that the mechanism of hydrogen atom abstraction
`is independent of the steps that occur prior to the irreversible
`formation of an active oxygen species.
`Our results agree with other IEP data in the literature for
`different groups of compounds that are metabolized by CYP.
`Karki et al. measured isotope effects for a series of N,N-
`dimethylanilines using mammalian CYPs 2B1 (purified), 4B1
`and 1A2 (expressed), rat liver microsomes, and purified
`bacterial CYP102 and compared them to KIE in two different
`chemical systems (6). Practically identical IEPs for all of
`the P450s and the tert-butoxyl radical system provide strong
`evidence for a similar mechanism of N-dealkylation, that is,
`initial hydrogen atom abstraction. Furthermore, the mech-
`anism was conserved among all the CYPs that were tested,
`which includes both mammalian and bacterial members of
`the Class II P450s. Isotope effect profiles for various CYPs
`and for the tert-butoxy radical in several small molecules
`reported by Manchester et al. also support the mechanism
`of (initial) hydrogen atom abstraction for CYPs (7). This
`abundant amount of data for both mammalian and bacterial
`isoforms strongly supports a common mechanism for mul-
`tiple CYPs.
`LFER with Para-Substituted Toluenes and CYP2E1. Since
`the data above support the hypothesis that