JBC Ideal method for primary cell transfection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M411019200 on January 22, 2005

J. Biol. Chem., Vol. 280, Issue 13, 12279-12291, April 1, 2005
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
280/13/12279    most recent
M411019200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yun, C.-H.
Right arrow Articles by Guengerich, F. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yun, C.-H.
Right arrow Articles by Guengerich, F. P.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Kinetic Analysis of Oxidation of Coumarins by Human Cytochrome P450 2A6*{boxs}

Chul-Ho Yun{ddagger}, Keon-Hee Kim{ddagger}, M. Wade Calcutt§, and F. Peter Guengerich§

From the {ddagger}Hormone Research Center, School of Biological Sciences and Technology, Chonnam National University, Gwangju 500-757, Republic of Korea and the §Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146

Received for publication, September 24, 2004 , and in revised form, January 18, 2005.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Human cytochrome P450 (P450) 2A6 catalyzes 7-hydroxylation of coumarin, and the reaction rate is enhanced by cytochrome b5 (b5). 7-Alkoxycoumarins were O-dealkylated and also hydroxylated at the 3-position. Binding of coumarin and 7-hydroxycoumarin to ferric and ferrous P450 2A6 are fast reactions (kon ~ 106 M–1 s–1), and the koff rates range from 5.7 to 36 s–1 (at 23 °C). Reduction of ferric P450 2A6 is rapid (7.5 s–1) but only in the presence of coumarin. The reaction of the ferrous P450 2A6 substrate complex with O2 is rapid (k ≥ 106 M–1 s–1), and the putative Fe2+ ·O2 complex decayed at a rate of ~0.3 s–1 at 23 °C. Some 7-hydroxycoumarin was formed during the oxidation of the ferrous enzyme under these conditions, and the yield was enhanced by b5. Kinetic analyses showed that ~1/3 of the reduced b5 was rapidly oxidized in the presence of the Fe2+·O2 complex, implying some electron transfer. High intrinsic and competitive and non-competitive intermolecular kinetic deuterium isotope effects (values 6–10) were measured for O-dealkylation of 7-alkoxycoumarins, indicating the effect of C–H bond strength on rates of product formation. These results support a scheme with many rapid reaction steps, including electron transfers, substrate binding and release at multiple stages, and rapid product release even though the substrate is tightly bound in a small active site. The inherent difficulty of chemistry of substrate oxidation and the lack of proclivity toward a linear pathway leading to product formation explain the inefficiency of the enzyme relative to highly efficient bacterial P450s.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
P4501 enzymes are involved in the oxygenation of a variety of natural products and xenobiotic chemicals in microbial systems (3, 4). Much is known about the structure, function, and catalytic features of some of the P450s, particularly the more extensively studied of the bacterial P450s (4, 5). In mammalian systems P450s oxidize many drugs, steroids, carcinogens, fatty acids and eicosanoids, fat-soluble vitamins, and other endobiotic and xenobiotic chemicals (6). Less information is available about the biochemical details of most of the 57 human P450s (7). In particular, the basis of the inherently lower catalytic activities of these and other mammalian P450s relative to some of the microbial forms is not clear.

P450 2A6 is a low-to-medium abundance P450 in human liver (7–9) and is also expressed in some extrahepatic tissues (10). The history of this gene/protein goes back to Phillips et al. (11), who identified a human P450 cDNA as a relative of rat P450 2B1. The 7-hydroxylation of coumarin has long been used as an assay of P450 activity in animal and human liver microsomes (12, 13), and Yamano et al. (14) isolated a P450 2A6 cDNA (then termed 2A3) and first showed that the protein derived from heterologous expression had coumarin 7-hydroxylation activity. Miles et al. (15) also provided similar evidence for this particular sequence being associated with coumarin 7-hydroxylation. Our group purified a protein from human liver microsomes, identified it as P450 2A6, and showed it to be the major coumarin 7-hydroxylase in human liver (8). Subsequently P450 2A6 has been studied extensively, in large part because of its role in the metabolism of nicotine and carcinogenic N-alkylnitrosamines (16, 17). Genetic polymorphisms have been identified (18, 19) and may be of relevance to cancer risk; (i) impaired metabolism of nicotine has been proposed to reduce cigarette smoking in P450 2A6-deficient individuals (20); (ii) impaired metabolism can yield reduced levels of activation of the N-nitrosamines found in tobacco (21). Some drugs are oxidized by P450 2A6 (7, 22). P450 2A6 also catalyzes the oxidation of indoles (23, 24), and we have used P450 2A6 mutants to synthesize new indirubins with activity as protein kinase inhibitors (25).

Recently x-ray crystal structures have been reported for P450 2A6, including forms with the substrates coumarin and nicotine bound (26). These structures more than any of the other mammalian P450s solved to date have a small binding site akin to that of bacterial P450 101A1 (27). The space for the substrate coumarin is very restricted, and the coumarin-bound structure has the C-7 atom located near the heme iron (26). A major conformational change is required to open and close the enzyme and allow the substrate (coumarin) to enter and leave (26).

We have been studying aspects of catalysis of mammalian P450s, including the rate-limiting steps in reactions (28–31). P450 2A6 was of interest because of the recently reported structure, the useful fluorescence properties and common use of coumarins as P450 substrates, and our inherent interest in the catalytic properties of P450 2A6 (8, 23–25, 32). Rates of several steps were measured (Scheme 1). Studies of O-dealkylation of 7-OR coumarins showed high kinetic hydrogen isotope effects and demonstrate the kinetic difficulty of C–H bond breaking. The 7-OR coumarins showed extensive formation of 3-OH products as well as 7-OH coumarin. Together the results provide a picture of a very dynamic catalytic cycle with considerably more flexibility than apparent with the more efficient bacterial P450s, providing some potential insight into rate differences.



View larger version (12K):
[in this window]
[in a new window]
  		SCHEME 1.
Catalytic cycle of P450. Adapted from previous reviews (30, 33, 34). Step 5 can be divided into two parts, (a) Fe2+·O2–+ H+ {rightleftarrows} FeII-OOH, and (b) FeOOH -> FeO3+, although the dioxygen species have also been postulated to catalyze certain oxidations directly (35). Step 6 can also be divided into two parts, (a) FeO3+ + RH -> [FeOH3+ R·], and (b) [FeOH3+ R·] -> Fe3+ + ROH.

 

   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals—Coumarin, 7-OH coumarin, 7-OMe coumarin, and 7-OEt coumarin were purchased from Aldrich and recrystallized from C2H5OH/H2OorCH3OH/H2O before use. 5-Deazaflavin was a gift of the late V. Massey (University of Michigan, Ann Arbor, MI) (36).

Deuterated 7-OMe coumarins were prepared by reaction of 7-OH coumarin with deuterated methyl iodides (Cambridge Isotopes, Cambridge, MA) in the usual manner (29, 37, 38) and recrystallized from CH3OH/H2O. [1-Ethyl-d2]-7-OEt coumarin was prepared in the same way from [1-d2]-ethyl iodide (Cambridge). In the synthesis of [1-ethyl-d1]-7-OEt coumarin, CH3CHO was reduced with LiAlD4 in diethylene glycol diethyl ether to prepare [1-d1]-C2H5OH (39), and the distilled product was reacted with tosyl chloride in dry pyridine to form the tosylate (72% yield, m.p. 27–30 °C (literature (40), 21 °C). The tosylate was then reacted with 7-OH coumarin in acetone/K2CO3 (under reflux) in the same manner as used with the alkyl iodides to yield [1-ethyl-d1]-7-OEt coumarin, which was recrystallized from CH3OH, H2O (39% yield, m.p. 85–87 °C, literature (41), 88–90 °C); UV (CH3OH) {epsilon}323 1.23 x 104 M–1 cm–1; electrospray MS, m/z 191.1 (MH+); NMR (CDCl3) {delta} 1.37 (dd, 3H, CH3), 4.03 (m, 1H, CHD), 6.22 (d, 1H, H-3), 6.69–6.82 (m, 2H, H-6, H-8), 7.33 (d, 1H, H-5), 7.60 (d, 1H, H-4). All deuterated coumarin derivatives were >98% isotopically enriched at the site of modification as judged by MS and NMR spectroscopy.

The synthesis of 3-OH coumarins was done using the general procedure of Neubauer and Flatow (42), which involves condensation of salicylaldehyde or a 4-substituted salicylaldehyde with hippuric acid (N-benzoylglycine) to form the N-benzoyl enamine, which was hydrolyzed in 10 N NaOH (100 °C, 45 min) to give the desired coumarin. 4-OEt salicylaldehyde was prepared by BCl3 treatment of 2,4-(OEt)2 salicylaldehyde (Aldrich) in CH2Cl2 (80% yield) (43). The identities of the 3-OH coumarins were confirmed by their m.p. values and spectroscopy: 3-OH coumarin, m.p. 151–154 °C (literature (44), 154 °C), UV (CH3OH) {epsilon}307 1.22 x 104 M–1 cm–1, {epsilon}294 1.19 x 104 M–1 cm–1, {epsilon}235 4.91 x 103 M–1 cm–1, fluorescence (CH3OH) {lambda}excitation 310 nm, {lambda}emission 395 nm, electrospray MS, m/z 163.1 (MH+), NMR (CDCl3) {delta} 6.48 (bs, 1H, H-4), 7.46–7.67 (m, 4H, H-5,6,7,9); 3-OH,7-OMe coumarin, m.p. 179–182 °C (literature (45), 175.5–177.5 °C), UV (CH3OH) {epsilon}322 1.4 x 104 M–1 cm–1,{epsilon}235 5.6 x 103 M–1, {epsilon}215 8.4 x 10 M–1 cm–1, electrospray MS, m/z 193.1 (MH+), NMR (CDCl3) {delta} 3.79 (s, 3H, CH3), 6.88 (dd, 1H, H-6), 6.94 (d, 1H, H-8), 7.08 (s, 1H, H-4), 7.43 (d, 1H, H-5); 3-OH,7-OEt coumarin, m.p. 155–157 °C, UV (CH3OH) {epsilon}322 1.34 x 104 M–1 cm–1, {epsilon}235 5.6 x 103 M–1 cm–1, {epsilon}215 8.4 x 10 M–1 cm–1, electrospray MS, m/z 206.9 (MH+), NMR (CDCl3) {delta} 1.32 (t, 3H, CH3), 4.05 (q, 2H, CH2), 6.87 (dd, 1H, H-6), 6.89 (d, 1H, H-8), 7.08 (s, 1H, H-4), 7.42 (d, 1H, H-5).

Enzymes—P450 2A6 was expressed from a plasmid (originally obtained from P. Soucek, National Institute of Public Health, Prague) in Escherichia coli, except that a His5 tag was attached to the C terminus (24, 46). Rat NADPH-P450 reductase was expressed in E. coli and purified as described (47). Recombinant human b5 was expressed in E. coli JM109 cells from a plasmid (pSE420 (Amp)) kindly provided by Satoru Asahi (Takeda Pharmaceutical, Osaka, Japan). The protein was solubilized and purified to electrophoretic homogeneity using modifications of the DEAE-cellulose and hydroxylapatite chromatography methods described elsewhere (48).

Spectroscopy—NMR spectra were recorded using Bruker 300 and 400 MHz instruments in the Vanderbilt facility. UV-visible spectra were generally acquired using an OLIS/Cary 14 or a OLIS/Aminco DW2a instrument (OLIS, Bogart, GA). Mass spectra were recorded using HPLC-MS methods (octadecylsilane columns, positive ion-electrospray, or atmospheric pressure chemical ionization) in the Vanderbilt facility using a Thermo-Finnigan TSQ-7000 instrument (Thermo-Finnigan, Sunnyvale, CA). Fluorescence measurements were made using either an SPEX Fluoromax-3 instrument (SPEC/Jobin Yvon, Edison, NJ) or an OLIS RSM-1000 instrument (OLIS), operating in the stopped-flow mode.

Stopped-flow kinetic UV-visible measurements were made using an OLIS RSM-1000 instrument (slit width 1.24–3.16 nm for absorbance beam). Some kinetic traces were obtained in the single wavelength mode; the rapid scanning mode was used with a 16 x 1-mm scanning disk to obtain spectra (16–1000 scans s–1). In some cases the acquired spectra were used to derive kinetic traces at individual wavelengths, and fitting was done with the manufacturer's software.

Assays—Typical steady-state coumarin oxidation reactions included 50 pmol P450 2A6, 100 pmol of NADPH-P450 reductase, 50 pmol of b5 (when indicated), and 30 µg of di-12:0 GPC in 0.50 ml of 50 mM potassium phosphate buffer (pH 7.4) along with a specified amount of the coumarin substrate. An aliquot of an NADPH-generating system was used to start reactions (final concentrations, 10 mM glucose 6-phosphate, 0.5 mM NADP+, and 1 international unit of yeast glucose 6-phosphate ml–1 (49)). Stock coumarin solutions (5 mM) were made in H2O. 7-OMe and 7-OEt coumarin stocks (50 mM) were made in CH3CN and diluted into enzyme reactions, with final organic solvent concentrations <1% (v/v).

Incubations were generally done for 10 min at 37 °C, terminated with 0.10 ml of 17% HClO4, and centrifuged (103x g, 10 min). CH2Cl2 (1.0 ml) was added to the supernatant to extract the products followed by centrifugation at 103x g (process repeated one more time). The organic layers were combined, and the CH2Cl2 was removed under a N2 stream. The products were analyzed by HPLC using a Toso ODS-80TM octadecylsilane (C18) column (4.6 x 150 mm, 5 µm) with the mobile phase H2O:CH3CN (55:45, v/v) containing 10 mM HClO4, a flow rate of 1.0 ml min–1, and monitoring at A330. Kinetic parameters (Km and kcat) were determined using nonlinear regression analysis with Graph-Pad Prism software (Graph-Pad, San Diego, CA). In some cases (stopped-flow), 7-OH coumarin was monitored directly (F390/460).

Assays involving competitive deuterium isotope effects were done by HPLC-MS analysis of formaldehyde or acetaldehyde derived from [methyl-d3]-7-OMe coumarin or [1-ethyl-d2]-7-OEt coumarin, respectively, using the general approach described elsewhere (51, 52). With [1-ethyl-d1]-7-OEt and [1-ethyl-d2]-7-OEt coumarin as substrates, the mass spectra were complicated due to the presence of contaminants in some of the reagents, particularly the solvents and glycerol. To minimize complications arising from residual aldehyde contamination in reagents and solvents, the following changes were made to the procedure. Reconstituted enzyme solutions (P450 2A6, NADPH-P450 reductase, and b5) were dialyzed against glycerol-free 50 mM potassium phosphate buffer (pH 7.4) containing 0.2 mM EDTA and 0.1 mM dithiothreitol (two changes over 12 h at 4 °C) before the addition of di-12:0 GPC. Dissolution of 7-OMe and 7-OEt coumarin in H2O was achieved by sonication (Branson sonicator, microtip probe, 70% full power) (Branson, Danbury, CT) to avoid the introduction of organic solvents. Aqueous stock solutions (~0.5 mM) were then filtered and quantitated spectrophotometrically (see above). Hexanes and CH3CN were heated with and distilled from 2,4-dinitrophenylhydrazine. 2,4-Dinitrophenylhydrazine (for use as a derivatization reagent) was recrystallized twice from CH3OH/H2O, dried in vacuo, dissolved in 6 N HCl (0.1%, w/v), and washed multiple times with a hexane, CH2Cl2 mixture (7:3, v/v) to remove hydrazone impurities before use for derivatization. Deuterium incorporation was determined using HPLC/negative ion atmospheric pressure chemical ionization MS of the resulting 2,4-dinitrophenylhydrazone derivatives (source temperature 550 °C; heated capillary voltage 20 V; heated capillary temperature 180 °C; ionization current 5 µA; sheath gas (N2) pressure 70 p.s.i.; auxiliary gas (N2) pressure 10 p.s.i.) (53).

Anaerobic experiments involved the use of glass anaerobic cuvettes and tonometers with a gas train connected to a manifold, alternating between vacuum and Ar. The basic systems are described elsewhere (54–56), and recent further modifications have been described (31). The OLIS RSM-1000 stopped-flow spectrophotometer has stainless steel lines leading up to the observation cell instead of Teflon, reducing the diffusion of O2. As described earlier (31), the drive syringes were filled with anaerobic 0.10 M Na2S2O4 (in 0.2 M potassium phosphate buffer, pH 7.4) overnight before use to scrub O2 (31, 56). The drive syringes were then loaded with the contents of a tonometer containing 0.25 mM safranin T and 0.5 mM methyl viologen (photo-reduced) in anaerobic 0.10 M Tris·HCl buffer (pH 7.7) containing 10 mM EDTA. The lack of O2 in the system is indicated by blue color (methyl viologen radical cation), as opposed to red (oxidized safranin). Thus, the final displacement of the dye by the enzyme solution provides a reasonable check on the anaerobic nature of the system.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Oxidations Catalyzed by P450 2A6—Earlier work led to the characterization of P450 2A6 as the major coumarin 7-hydroxylase (8, 14, 15). Recently the structure of a P450 2A6 crystal has been reported (26), with coumarin bound and positioned with the C-7 atom near the iron atom.

Assays of P450 2A6-catalyzed coumarin oxidation commonly utilize a sensitive fluorescence assay that reports 7-hydroxylation (57, 58). HPLC-UV assays indicated the formation of the single product 7-OH coumarin by chromatographic and spectral comparison to standard material (Table I, also see Supplemental Fig. 1). 7-OMe and 7-OEt coumarins were O-dealkylated to form 7-OH coumarin, but both of these substrates also formed the 3-hydroxy products, as judged by comparisons with synthetic materials. The identification of 3-OH, 7-OEt coumarin as a product of oxidation of 7-OEt coumarin has been reported previously with human liver microsomes (59, 60). The 3-hydroxylation of coumarin has been reported with human liver microsomes (61, 62). Neither we nor Born et al. (62) detected conversion of coumarin to 3-OH coumarin by P450 2A6 systems.2


View this table:
[in this window]
[in a new window]
  		TABLE I
Steady-state parameters for coumarin oxidation catalyzed by P450 2A6

 

Subsequent analysis of steady-state kinetic parameters indicated that 3-hydroxylation was observed to a greater extent for 7-OEt coumarin than 7-OMe coumarin (Table I).3 The addition of b5 stimulated the formation of 7-hydroxylation of coumarin 2-fold, as reported by others (46, 63). However, the activities with 7-OR coumarins were not stimulated except for the decreased Km for the 3-hydroxylation of 7-OEt coumarin.

Spectral Properties of P450 2A6—The spectra of the ferric, ferric-coumarin, and ferrous-coumarin forms of P450 2A6 are shown in Fig. 1. The spectral properties were utilized in several subsequent experiments to measure rates of changes within the catalytic cycle (Scheme 1). Second-derivative analysis (64, 65) of the spectra (Fig. 1) yielded estimates of 98% low spin P450 iron in substrate-free ferric P450 2A6 and 88% high-spin iron with coumarin bound.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 1.
Spectra of P450 2A6 complexes. A, spectra were recorded in 50 mM potassium phosphate buffer (pH 7.4) with 5.2 µM P450 2A6 either with or without 50 µM coumarin. The ferrous form was produced by the addition of a few grains of Na2S2O4. The inset (B) shows an expansion of the {alpha},{beta} region. Solid lines, ferric P450 2A6 (Fe3+); dots and dashes, ferric P450 2A6·coumarin complex (Fe3+·S); dashes, ferrous P450 2A6·coumarin complex (Fe2+·S).

 
Substrate Binding (Step 1 of Scheme 1)—A large spectral difference is observed upon binding of the substrate coumarin (Fig. 2). Titration of P450 2A6 with coumarin yielded classic "Type I" difference spectra, with Ks = 0.38 (±0.03) µM (Fig. 2, A and B).



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 2.
Binding of coumarin to ferric (Fe3+) P450 2A6. A, difference spectra obtained by titration of a 2 µM P450 solution (100 mM potassium phosphate, pH 7.4) with increasing concentrations (spectra shown with arrows) of coumarin (to 7.5 µM). B, plot of data from A, fit to a quadratic expression Kd = ([Et–E·S][St–E·S])/ES, where Et = total P450 concentration, St = total coumarin concentration, and ES = P450-coumarin complex (66) yielded Kd = 0.38 (±0.03) µM. C, rate of {Delta}A390 for similar a binding experiment with 8 µM coumarin, fit to a single exponential of 17 s–1.D, kobs for traces obtained in C as a function of coumarin concentration. From kobs = kon [S] + koff (67), k = 2.7 x 106 M–1 s–1, and koff = 5.7 s–1 (k0ff/kon = 2.1 µM).

 
The rates of spectral changes could be monitored using stopped-flow spectroscopy (Fig. 2C). These traces were fit to single-exponential plots. Analysis of rates as a function of coumarin concentration (Fig. 2D) fit a relationship describing a two-state system (Fe3+ + S {rightleftarrows} Fe3+·S), i.e. kobs = kon[S] + koff (67) and yielding kon = 2.7 x 10 M–1 s–1 and koff = 5.7 s–1, with Kd = koff/kon = 2.1 µM, in reasonable agreement with the Kd values estimated by titration (Fig. 2B).

Product Release (Step 7 of Scheme 1)—Titration of ferric P450 2A6 with 7-OH coumarin yielded a "Type II" difference spectrum, which is probably indicative of ligation of the phenolic oxygen to the iron atom (Fig. 3A). The titration indicated binding as tight as for the substrate, Kd = 0.82 ± 0.05 µM (Fig. 3B). The rate of binding could be measured using stopped-flow spectroscopy (Fig. 3C). Analysis of the rate as a function of 7-OH coumarin concentration (Fig. 3D) yielded kon = 2.0 x 106 M–1 s–1 and koff = 6.8 s–1 (koff/kon ~ 3.4 µM).



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 3.
Binding of 7-OH coumarin to ferric (Fe3+) P450 2A6 (product release). A, difference spectra obtained by titration of a 2 µM P450 2A6 solution (100 mM potassium phosphate, pH 7.4) with increasing concentrations (spectra shown with arrows) of 7-OH coumarin (to 8 µM). B, plot of data from A, fit to a quadratic expression yielded Kd = 0.82 (±0.05) µM. C, rate of {Delta}A230 for a similar binding experiment with 16 µM 7-OH coumarin, fit to a single exponential of 18 s–1. D, kobs for traces obtained in C as a function of 7-OH coumarin concentration. From kobs = kon [S]+ koff (67), kon = 2.0 x 106 M–1 s–1 and koff = 6.8 s–1 (koff/kon ~ 3.4 µM).

 
P450 2A6-catalyzed oxidations of coumarin, 7-OMe coumarin, and 7-OEt coumarin were examined and did not show kinetic bursts (Fig. 4). On the basis of these results and the rates measured in the experiment of Fig. 4, steps after product formation are not rate-limiting in the formation of 7-OH coumarin from any of these substrates.



View larger version (43K):
[in this window]
[in a new window]
  		FIG. 4.
Lack of kinetic bursts in oxidations of coumarins to form 7-OH coumarin. In each case a reaction was done at 22 °C in the stopped-flow spectrofluorimeter (OLIS RSM-1000) by adding 22 µM NADPH and an NADPH-generating system (49) from one syringe to a typical mixture (final concentrations indicated) of P450 2A6 (0.5 µM), NADPH-P450 reductase (1.0 µM), b5 (0.5 µM), di-12:0 GPC (45 µM), and the substrate. Fluorescence excitation was at 390 nm, and emission was monitored at 420 nm. The substrate concentrations were 50, 200, and 200 µM for coumarin, 7-OMe coumarin, and 7-OEt coumarin, respectively. The fluorescence corresponding to 0.5 µM 7-OH coumarin (indicative of one enzyme cycle) was determined by comparison with a conventional mixing experiment done with these preparations and coumarin in an OLIS DM45 spectrofluorimeter. The vertical line at t = 0 is the mixing artifact. The origins of the four traces are not intended to be identical due to variations in the background fluorescence (as a result of both the coumarins and the NADPH).

 
Reduction of Ferric P450 2A6 (Step 2 of Scheme 1)—Rates of reduction of ferric P450 2A6 were measured in an anaerobic CO environment, with ferrous P450 trapped as the CO complex (Fig. 5). The rate of binding of ferrous P450 and CO is much faster than reduction.



View larger version (17K):
[in this window]
[in a new window]
  		FIG. 5.
Reduction of ferric (Fe3+) P450 2A6. Experiments were done anaerobically under a CO atmosphere in an OLIS RSM-1000 stopped-flow spectrophotometer. One syringe contained NADPH (300 µM) in 100 mM potassium phosphate buffer (pH 7.4) The other tonometer contained a mixture of P450 2A6 (1.0 µM), NADPH-P450 reductase (2.0 µM), and di-12:0 GPC (45 µM) in 100 mM potassium phosphate buffer (pH 7.4). Coumarin (40 µM) was also included when indicated. The inset shows an expansion of the plot with coumarin present. Rates were fit to exponential plots using the manufacturer's software; without coumarin, 0.13 s–1 (no correction for the apparent lag); with coumarin, bi-exponential fit of 7.5 s–1 and 0.13 s–1.

 
The rate of reduction of ferric P450 was slow in the absence of coumarin, with a fit to a single exponential of 0.13 s–1 (Fig. 5). In the presence of coumarin, a bi-exponential fit yielded k1 = 7.5 s–1 and k2 = 0.13 s–1 (Fig. 5). About 50% of the P450 was reduced in the fast phase.

Dissociation of Substrate from Ferrous P450 2A6 (Step 2a of Scheme 1)—Although the binding of a substrate (coumarin) to ferric P450 2A6 produces a major spectral change, the changes observed upon the addition of coumarin to ferrous P450 2A6 are much weaker. A difference spectrum was observed with a trough at 438 nm and a peak at 460 nm (Fig. 6A). The rate of binding to NADPH-P450 reductase-reduced P450 2A6 under anaerobic conditions yielded kon = 1.5 x 106 M–1 s–1 and koff = 36 s–1 (koff/kon = 24 µM) (Figs. 6, B and C).



View larger version (14K):
[in this window]
[in a new window]
  		FIG. 6.
Binding of coumarin to ferrous (Fe2+) P450 2A6. A, difference spectrum obtained with 2.6 µM P450 2A6 with and without 50 µM coumarin (in 50 mM potassium phosphate buffer, pH 7.4). Both cuvettes were reduced with solid Na2S2O4. B, kinetics of {Delta}A438 observed upon the addition of 15 µM coumarin to ferrous P450 2A6. The experiment was done in a stopped-flow spectrophotometer. One syringe contained 100 µM coumarin (in 50 mM potassium phosphate buffer, pH 7.4), and the other contained P450 2A6 (4 µM), NADPH-P450 reductase (4 µM), di-12:0 GPC (45 µM), potassium phosphate buffer (50 mM potassium phosphate buffer, pH 7.4), and an NADPH-generating system composed of 10 mM glucose 6-phosphate, 0.5 mM NADPH, and 1 international unit of yeast glucose-6-phosphate dehydrogenase ml–1 (49). Both syringes were under an anaerobic environment (Ar). The data fit to a single exponential of 57 s–1. C, kobs for traces obtained in B as a function of coumarin concentration. Fitting to the expression kobs = kon [S] + koff yielded values of kon = 1.5 x 106 M–1 s–1 and koff= 36 s–1 (Kd = Koff/Kon = 24 µM).

 
Formation (Step 3 of Scheme 1) and Decomposition of Ferrous P450 2A6·O2 Complex—Ferrous P450 2A6 (in the presence of 50 µM coumarin) was introduced anaerobically into one syringe of the stopped-flow spectrophotometer and mixed with aerobic buffer (200 µM O2). A rapid spectral change was observed, and the early spectra are shown in Fig. 7A. The rate of this reaction, as estimated from {Delta}A390, was ~75 s–1. This complex was not very stable and decayed to yield ferric P450 2A6 (Fig. 7B), with an estimated rate of 0.3 s–1 (t1/2 = 2.3 s) at 23 °C (Fig. 7C).



View larger version (20K):
[in this window]
[in a new window]
  		FIG. 7.
Reaction of ferrous P450 2A6 with O2. P450 2A6 (4.9 µM), in an anaerobic mixture of 50 mM potassium phosphate, 100 mM Tris-HCl, 10 mM EDTA buffer (pH 7.4) containing 50 µM coumarin, 75 µM di-12:0 GPC, and 2 µM 5-deazaflavin (36), was photo-reduced and loaded into one syringe of the stopped-flow spectrophotometer. The other syringe contained air-saturated 50 mM potassium phosphate buffer. A, traces (rapid scanning) of the reaction. The first trace shown was recorded 21 ms after mixing, and the subsequent spectra were obtained every 25 ms. A390 decreased, and the broad absorbance at higher wavelength increased and then decreased. B, further spectral changes after mixing with O2. The first trace was recorded 64 ms after mixing, with subsequent spectra recorded every 1.6 s, with the direction shown by the arrow. C, kinetics of reaction from B ({Delta}A390), fit to a single exponential of 0.30 s–1.

 
Formation of Product in Limited Turnover Experiments— The oxidation of coumarin was measured under conditions in which the electron input was limited to what should be a single cycle, e.g. as in Fig. 7. P450 2A6 was either photo-reduced with 5-deazaflavin or reduced (by NADPH-P450 reductase) with a limited amount of NADPH (enough to fully reduce NADPH-P450 reductase and P450 once).4

The product 7-OH coumarin could be measured in all cases (Fig. 8), although the yields were low (Table II). The formation of product could be detected with only P450 present (Fig. 8A), which apparently indicates that two FeO2+2 complexes can interact to provide the second electron for product formation (e.g. 2FeO2+2 -> FeO2+2 (->-> FeO3+) + Fe3+). Product formation was more efficient when electrons were delivered from the reductase (Table II), although we do not know exactly how many electrons are delivered, i.e. it is possible that multiple 1-electron transfers could occur that are not relevant during steady-state catalysis.



View larger version (11K):
[in this window]
[in a new window]
  		FIG. 8.
Oxidation of coumarin to 7-OH coumarin in limited cycle experiments. A, P450 2A6 (5 nmol) was photo-reduced under an Ar atmosphere (in the presence of 50 µM coumarin, 1.0 µM 5-deazaflavin, and 1.0 µM safranin T in 3.0 ml of buffer as described under Fig. 7). Reduction was monitored (e.g. Fig. 2) in the OLIS/Cary 14 spectrophotometer and was complete after 3 min of irradiation (36). The sample was mixed with air and quenched by the addition of 0.3 ml of 43% H3PO4 after 3 min. Products were extracted 3 times with 3.0 ml of CH2Cl2, and the combined organic phase was concentrated to dryness under an N2 stream and analyzed by HPLC using the fluorescence method of Soucek (58). The scale used in A is expanded 8-fold compared with the other parts. B, the experiment was as in A, except that 5 nmol of b5 was included (and photo-reduced before oxygenation). C, the experiment was as in A, except that 5 nmol of NADPH-P450 reductase was included, and 12.5 nmol of NADPH was added anaerobically to achieve reduction (instead of photoreduction). D, the experiment was as in C except that 5 nmol of b5 were also present, and 15 nmol of NADPH were used for reduction.

 


View this table:
[in this window]
[in a new window]
  		TABLE II
Yields of products formed from coumarin in limited cycle experiments h{nu}, photochemical reduction.

 
In the cases of both photo-reduced P450 2A6 and P450 2A6 that was reduced with NADPH/reductase, the presence of reduced b5 raised the yield of product (Fig. 8, Table II). This result suggests that ferrous b5 might be able to transfer electrons to the P450 2A6 FeO22+-substrate complex.

Kinetic Deuterium Isotope Effects (Step 6 of Scheme 1)—The experiments presented thus far have dealt with the binding of ligands and the activation of O2 by P450. The relative rate of the step in which chemical transformation of the substrate occurs had not been addressed. This is a difficult question in the case of an aromatic or olefinic hydroxylation, unless an enzyme intermediate can be isolated and reacted directly to yield product. Such is not the case with P450s, so an alternate approach was used, that of measurement of kinetic deuterium isotope effects for reactions involving C–H bond cleavage of simple analogs. The work described in Table I showed that 7-OMe and 7-OEt coumarin O-dealkylations were catalyzed by P450 2A6 with catalytic efficiencies approaching that for coumarin 7-hydroxylation.

Several types of kinetic isotope effect experiments are possible (see Scheme 3) and reveal different information. A so-called non-competitive intramolecular experiment approximates the intrinsic kinetic deuterium isotope effect, the isotope effect on the C–H bond breaking step itself (1, 2). In the case of a C–H bond breaking step with a methyl group, the rapid rotation does not provide a kinetic barrier to attenuate the reaction, and the only limitation to the interpretation of this experiment is the potential contribution of a smaller geminal secondary kinetic isotope effect. P450 2A6-catalyzed 7-OMe coumarin O-demethylation showed a non-competitive intramolecular isotope effect of 9.8 (Table III). A somewhat lower isotope effect was measured for the O-deethylation of 7-OEt coumarin (6.0) (Table III), although the point must be made that the O-deethylation involves a prochiral substrate, 7-OEt coumarin, which can preclude interpretation about intrinsic isotope effects.



View larger version (10K):
[in this window]
[in a new window]
  		SCHEME 3.
Outline of kinetic deuterium isotope effect experiments. A, non-competitive intramolecular experiment for approximating the intrinsic isotope effect (with a methyl; pro-chirality is a potential confounder with a methylene). B, competitive intermolecular experiment. C, non-competitive intermolecular experiment (provides information about how rate-limiting the C–H bond breaking step is in the overall steady-state reaction. This experiment provides information about the relative rates of exchange of substrates, i.e. if attenuation of the intrinsic isotope effect is observed, then exchange rates are partly rate-limiting (1, 2).

 


View this table:
[in this window]
[in a new window]
  		TABLE III
Kinetic deuterium isotope effects for 7-alkoxycoumarin O-dealkylation reactions catalyzed by P450 2A6

 
We also measured competitive intermolecular isotope effects (see Scheme 3B). Attenuation of the intrinsic isotope effect in such an experiment is suggestive of a relatively slow exchange of substrate. The values for 7-OMe and 7-OEt coumarin in this experiment were still high (9.0 and 7.0, respectively) (Table III). The results are consistent with the rapid koff rate for the substrate coumarin (Fig. 2D).

The third type of experiment used was a non-competitive intermolecular system (see Scheme 3C). This type of experiment can provide some information about the extent to which the C–H bond-breaking step is rate-limiting (in the steady state) (1, 2). The value can be compared with the intrinsic isotope effect (see Scheme 3A, Table III). High kinetic isotope effects (both DV and D(V/K)) were observed in these experiments (Fig. 9, Table IV) (regardless of whether b5 was added or not; results not presented). These experiments provide strong evidence that the oxidation of the substrate by the active hypervalent oxygen species is difficult and can limit the rate of formation of these products, at least in the case of the O-dealkylation reactions. In the case of 7-OMe coumarin considerable "switching" to 3-hydroxylation was observed (Table IV, Fig. 9B). This result implies that the rotation of the substrate can occur to generate this product (68). Alternatively, the FeO3+-deuterated substrate complex could decompose, and the cycle can begin again with a protiated substrate (31). The switching effect (to 3-hydroxylation) was not significant in the case of 7-OEt coumarin (Fig. 9D, Table IV).



View larger version (18K):
[in this window]
[in a new window]
  		FIG. 9.
Effect of deuterium substitution on oxidations of 7-OMe and 7-OEt coumarin by P450 2A6. Typical steady-state reaction conditions were used with d0 (A) or [methyl-d3]-7-OMe-coumarin (B) or d0 (C), or [1-ethyl-d2]-7-OEt-coumarin (D), and the products were analyzed by HPLC (100 µM substrate in each case).

 


View this table:
[in this window]
[in a new window]
  		TABLE IV
Intermolecular non-competitive kinetic deuterium isotope effects for 7-alkoxycoumarin O-dealkylation reactions catalyzed by P450 2A6

 
Stoichiometry of NADPH Utilization—Rates of NADPH oxidation, formation of products, and H2O2 formation were measured, and H2O formation was calculated by difference (69) (Table V). The results indicate that formation of coumarin products is a relatively inefficient process, with the bulk of the NADPH used to reduce O2. As in the report of Tan et al. (63) with a baculovirus-based system, the addition of b5 to P450 2A6 suppressed the formation of H2O2. The suppression of H2O2 formation was also noted in the presence of the 7-OR coumarin substrates, although formation of the coumarin product was not enhanced (except for 7-OEt coumarin 3-hydroxylation).


View this table:
[in this window]
[in a new window]
  		TABLE V
NADPH oxidation, product formation, and H2O2 formation by P450 2A6

 
Steady-state Spectral Measurements—Steady-state spectra of the coumarin 7-hydroxylation reaction were recorded with the prospect of identifying an accumulating intermediate. An aerobic mixture of P450 2A6, NADPH-P450 reductase, b5, di-12:0 GPC, coumarin, and an NADPH-regenerating system was mixed rapidly with NADPH, and spectra were recorded. The first phase occurred within 1 s (Fig. 10A). The reduction of the flavins of the reductase was observed (e.g. 440–500 nm). Several steps occur during this time, and a relatively stable spectrum accumulated by 1 s. The {lambda}max was ~415 nm and probably represents some iron-oxygen complex (or a mixture of more than one). It is not the ferric nor ferrous P450 2A6 (Fig. 1) and probably not the FeO22+ complex (Fig. 7).



View larger version (21K):
[in this window]
[in a new window]
  		FIG. 10.
Spectra of P450 2A6 coumarin-oxidizing reaction. One syringe of the stopped-flow instrument contained P450 2A6 (10 µM), NADPH-P450 reductase (20 µM), b5 (10 µM), di-12:0 GPC (50 µM), coumarin (100 µM), glucose 6-phosphate (10 mM), yeast glucose 6-phosphate (1 international unit ml–1), and potassium phosphate buffer (100 mM, pH 7.4). The other syringe contained 100 µM NADPH in 100 mM potassium phosphate buffer (pH 7.4). The contents of the syringes were mixed, and spectra were recorded (23 °C). A, the indicated spectrum was observed at 58 ms, and the subsequent traces were recorded every 100 ms, to an elapsed time of 858 ms. The decrease in A390 was fit to a single exponential of 20 s–1. B, the indicated spectra were recorded every 3 s after mixing and correspond to a slow reduction of the b5 in the mixture. The decrease in A409 was fit to a single exponential of 0.14 s–1, and the increase in A424 was fit to 0.17 s–1.

 
This spectrum persisted for several seconds, and then the reduction of b5 was observed (Fig. 10B). The reduction appeared to follow first-order kinetics (as judged by {Delta}A409 or {Delta}A424), with k = 0.15 s–1. In other experiments with no P450 present, we measured the reduction of b5 to be a bi-exponential process, with a first step (stoichiometric with 1 b5 per NADPH-P450 reductase) of 18 s–1 and a subsequent rate of 1.7 s–1 (results not shown).

Oxidation of Ferrous b5—One hypothesis for the enhancement of the catalytic activity of P450 2A6 by b5 is the transfer of an electron from ferrous b5 to the P450 FeO22+-substrate complex (70, 71). This hypothesis has some support in the limited turnover experiments (Table II), where yields were enhanced in the presence of b5. However, alternate hypotheses could be valid (e.g. improvement of P450 efficiency through protein-protein interactions).

b5 was photo-reduced and mixed with air-saturated buffer (Fig. 11A). The first-order rate of oxidation was 0.13 s–1 (Fig. 11B). The experiment was repeated in the presence of P450 2A6 and coumarin. Thus, ferrous (photo-reduced) P450 2A6 reacts rapidly with O2 to form an oxygenated complex (Fig. 7A); if reduced b5 is present, it might transfer electrons to this FeO22+ complex before the complex decomposes. In this experiment (Fig. 11C), reduction was faster than in the absence of P450 2A6. About 1/3 of the b5 was oxidized rapidly (as judged by {Delta}A410 or {Delta}A424 measurements, and the data were fit to a bi-exponential plot with k1 = 3.6 s–1 and k2 = 0.04 s–1 (Fig. 11D). We conclude that b5 can participate in transferring electrons to the P450 2A6 FeO22+-substrate complex, although this does not appear to be a well coupled process.



View larger version (23K):
[in this window]
[in a new window]
  		FIG. 11.
Oxidation of reduced b5 by ferrous-oxy-P450 2A6 (FeO22+). b5 (3.0 µM), in 50 mM potassium phosphate buffer (pH 7.4) containing 75 µM di-12:0 GPC, 100 mM Tris ·HCl, 20 mM EDTA, 50 µM coumarin, and 1 µM 5-deazaflavin (36), was photo-reduced and introduced into one side of the stopped-flow spectrophotometer. The sample was mixed with air-saturated 50 mM potassium phosphate buffer (pH 7.4) with a nominal O2 concentration of 200 µM. A, spectral traces of the reaction as described, with the first spectrum shown at 160 ms. Spectra were recorded every 3 s and progressed in the directions shown with the arrows. B, the data from A were fit to a single exponential with a rate of 0.13 s–1. C, the experiment of A was repeated in the presence of 3.0 µM P450 2A6. The first spectrum shown was recorded at 800 ms, and the subsequent spectra were recorded every 3 s. D, data from C ({Delta}A410) were fit to a bi-exponential equation with k1 = 3.6 s–1 and k2 = 0.04 s–1.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The purpose of this work was to characterize individual steps in catalysis by a mammalian P450 enzyme, human P450 2A6. The enzyme has a small active site (260 Å3), which is filled by the substrate coumarin (26) and binds the substrate reasonably tightly (Kd ~ 1 µM, Fig. 2B). However, catalysis is still relatively slow and inefficient in the context of the use of electrons for substrate oxidation (Tables I and V). Our results are interpreted in the context of a model with rapid exchange of ligands at multiple steps (requiring major conformational changes inferred from the x-ray crystallography work (26)), an unstable FeO22+ complex, and a relatively difficult step for the chemistry of oxidation by the FeO3+ complex.

Most of the work with P450 2A6 and coumarin substrates has been focused on the 7-hydroxylation reaction, which can be readily observed because of the strong fluorescence of the product 7-OH coumarin at neutral or alkaline pH (57, 58). HPLC-UV analysis also indicated the oxidation of a second site with the 7-OR coumarins (but not coumarin), which was identified as the 3-position by chromatographic and spectral comparison (Scheme 2, Fig. 9). Coumarin 3,4-epoxide has been reported not to convert to 3-OH coumarin (62), and therefore, we presume that the chemical mechanism is one involving formation of a bond between (Fe)O and the C-3 atom of the coumarin (4CH+-3CH-O-Fe), which collapses to form the enol product (favored over the keto tautomer). O-Dealkylation (at C-7) is a favorable reaction for 7-OR coumarins, presumably because the positioning approximates that observed in the P450 2A6-coumarin complex (26). The "shift" to 3-hydroxylation with 7-OEt coumarin was more extensive than for 7-OMe coumarin (Tables I and IV). Apparently the steric restriction imposed by the larger alkyl group shifts the equilibrium to an alternate form with the lactone carbonyl and the C-3 atom near the heme iron.



View larger version (9K):
[in this window]
[in a new window]
  		SCHEME 2.
Oxidations of coumarins catalyzed by P450 2A6.

 
The molecular basis for the difference between oxidation at the 7 and 3 positions awaits the availability of the coordinates of the P450 2A6-coumarin crystal structure (26). The balance would appear to be sensitive not only to the bulk at the 7-position (Table I) but also the ease of oxidation of the C-7 substituent (Table IV). Thus, the explanation for the specificity will probably be both steric and dynamic. How difficult the prediction of regioselectivity and rates of oxidation will be for new putative ligands of P450 2A6 is unknown. In a general sense, the relatively low rates of coumarin oxidations by P450 2A6 (~10 min–1, Table I) would not have been expected if only a tight fit of the substrate and H2O exclusion from the active site (as evidenced by a strong shift to high spin iron) (Figs. 1 and 2) are the major factors involved in predicting catalysis, i.e. the properties of the P450 2A6-coumarin complex resemble those of P450 101A1-camphor, a system that turns over ~100-fold faster (5, 26).

The binding and dissociation rates were estimated for ferric P450 2A6 and two ligands, coumarin and the 7-OH product (Figs. 2 and 3). The results can be fit reasonably well to a simple 2-state model, and a more complex system may not be justified, at least at this point. The kon rates are near those reported for bacterial P450 101A1 (5, 72). Whether or not these are really a diffusion-limited rates is unclear, in that the second-order rates are still lower than for many enzymes, and theoretical calculations predict rates of ~109 M–1 s–1 (73, 74). A rapid rate of binding could be obscured by a slower, reversible transformation of the low spin to the high spin iron, yielding an artificially low apparent Kd (67). In principle, such a phenomenon should yield a hyperbolic plot of the apparent rate of binding (versus substrate concentration) instead of a linear form (Figs. 2D and 3D), but more analysis is needed to address this possibility.

The koff rates were 5.7 and 6.8 s–1 for the substrate coumarin and the product 7-OH coumarin, respectively (for ferric P450 2A6). The equation koff/kon = Kd yields parameters close to those estimated by steady-state spectral analysis (Figs. 2 and 3). The rapid off-rates are consistent with other results presented here, including the lack of a burst of 7-OH coumarin formation (Fig. 4) and the lack of attenuation of the kinetic deuterium isotope effect in the competitive intermolecular experiment (Table III). The rapid koff for coumarin (5.7 s–1 at 23 °C) is realistic in the context of the other parameters and is also competitive with the rate of reduction (Fig. 5).

Rapid reduction of ferric P450 2A6 (Fig. 5) was highly dependent upon the presence of the substrate coumarin, although only approximately one-half of the P450 2A6 was reduced in the fast phase (7.5 s–1 at 23 °C) even with a 2-fold excess of NADPH-P450 reductase. This rate is certainly much faster than overall catalysis (Table I) and should not be rate-limiting, even if fast reduction is only partial (Fig. 5). Our previous experience with (purified) microsomal P450s has been that about one-half of them show rapid reduction in the absence of substrate, and the other half requires substrate (75). Other P450s often show biphasic kinetics (75), probably due to spatial issues with reductase in the complexes (76). Although the iron of coumarin-saturated P450 2A6 is high spin (Fig. 1), a conclusion that only high spin P450 is rapidly reduced is unwarranted (75). Second-derivative analysis of the ferric Soret spectrum (Fig. 1) indicated ~88% high spin iron, but only 50% of the P450 2A6 was reduced rapidly (Fig. 5). We have previously presented evidence against a general linkage of substrate binding, low to high spin iron conversion, rapid reduction, and more positive redox potentials (Em,7) in P450s (65, 75, 77), unlike the situation with bacterial P450 101A1 (78). Although we have not directly estimated the Em,7 of P450 2A6 with and without substrate, consideration of the estimated Kd values of the Fe3+ and Fe2+ enzymes (Figs. 2 and 6) and the Nernst equation would suggest that the Em,7 of the Fe3+/Fe2+ couple would become somewhat more negative in the presence of coumarin, not more positive, applying a "thermodynamic box" analysis of substrate binding and reduction (73) and considering that the substrate is bound more tightly to the oxidized form of P450 2A6 (74) (Figs. 2, 6). Very recently Ost et al. (79) reported that the ligand 4-cyanopyridine is bound 3 orders of magnitude more tightly to ferrous than ferric P450 102A1, and accordingly, the presence of this ligand lowers the Em,7.

As discussed above, the koff rate for coumarin from ferrous P450 2A6 is considerable (36 s–1) (Fig. 6C) and may be an issue in the functionality of the Fe2+-substrate complex. That is, some fraction might not be competent in that it could dissociate, bind O2, and then decompose. The possibility also exists that substrate might dissociate from the FeO22+-substrate complex (or Fe-O complexes further in the catalytic cycle (Scheme 1)), although we do not have any measurements. In the case of rabbit P450 1A2, we recently presented evidence that dissociation of substrate from the FeO3+ complex itself (due to prevention of C–H bond breaking by deuterium substitution) was only accompanied by decomposition of the complex (31).

Reaction of the Fe2+ P450 2A6-coumarin complex with O2 produced two sets of spectral changes (Fig. 7). Although we did not examine the effect of varying O2 concentration on the rates, our view is that the rapid initial changes (Fig. 7A, e.g. A390 decrease) represent the formation of an FeO22+ (-substrate) complex and the succeeding changes (Fig. 7B, e.g. A390 increase) represent the decay of the complex to regenerate Fe3+ P450 2A6. The estimated rate of the first reaction was ~75 s–1, and the decay was 0.3 s–1 (18 min–1) at 23 °C. This complex appears to be much less stable than the Fe2+-O2 complexes reported for bacterial P450 101A1 (72, 80) and P450 108A1 (81) but has a stability similar to rabbit P450 1A2 (31, 82) and the heme domain of P450 102A1 (81); it is probably more stable than the complexes of rabbit P450 2B4 (83) and bacterial P450 119A1 (84). Analysis of 7-OH coumarin indicated low (but finite) yields in "limited cycle" experiments, e.g. when the kinetic reaction described above was analyzed for product formation (Fig. 8, Table II). Product formation apparently involves dismutation of two Fe2+·O2 complexes to achieve the requisite 2-electron stoichiometry (31). The amount of product formation in these experiments was increased when reduced NADPH-P450 reductase was present, indicating that electron transfer from reduced NADPH-P450 reductase to the P450 2A6 FeO22+-substrate complex occurs under these conditions, although the efficiency was low (Table II).

As reported previously (46, 63, 85), the presence of b5 enhanced the steady-state kcat for coumarin 7-hydroxylation (Table I). One general hypothesis for the enhancement of P450 catalytic activities by b5 is transfer of an electron from ferrous b5 to the P450 FeO2+-substrate complex (70, 71). The enhancement of yields of products derived from FeO22+-coumarin by ferrous b5 (Fig. 8, Table II) is consistent with this hypothesis but does not necessarily prove electron transfer. Another set of experiments was done in which a mixture of Fe2+ P450 2A6-coumarin-ferrous b5 was mixed with O2 to form a FeO22+-substrate-ferrous b5 complex, and the rate of oxidation of the ferrous b5 was monitored (at 410 or 424 nm) (Fig. 11). About one-third of the b5 was oxidized rapidly (3.6 s–1), whereas only the second, slower phase was observed for b5 oxidation in the absence of P450 2A6 (Fig. 11B). This result is consistent with transfer of an electron from reduced b5 to the P450 2A6 FeO22+-substrate complex, although the process appears to be less than quantitative. The situation is further complicated by the results of the experiment presented in Fig. 10, where a mixture of oxidized P450 2A6, NADPH-P450 reductase, and b5 (plus coumarin and di-12:0 GPC) was mixed with NADPH. In the first part of the reaction, the b5 was not reduced (up to ~1 s) and then became reduced at a rate of ~0.15 s–1 (at 23 °C). The b5 then stayed reduced for at least several minutes. This apparent rate was similar to the oxidation rate of ferrous b5 (Fig. 11B). For comparison, an experiment with only NADPH-P450 reductase and ferric b5 yielded a bi-exponential fit for reduction, with k1 = 18 s–1 and k2 = 1.7 s–1. The b5 experiments are complex and difficult to interpret, but our overall conclusion is that some electron transfer from ferrous b5 to the P450 2A6 FeO22+-coumarin complex occurs but that this does not seem to be a particularly efficient process. In other work we have shown that P450 2A6-catalyzed coumarin 7-hydroxylation can be enhanced by apo-b5 devoid of heme and precluding electron transfer (85). We conclude that part of the enhancement may be attributable to the electron transfer (Figs. 8 and 11 and Tables I and II) but that b5 probably has another effect, probably some type of conformational effect on the P450.5

Although the focus of this investigation was coumarin hydroxylation, the observed reaction (7-hydroxylation) was not amenable to the application of kinetic deuterium isotope effect studies (Scheme 2). However, 7-OMe and 7-OEt coumarin appear to be reasonable surrogates in that the P450 2A6 heme iron atom is apparently positioned in sites that yield O-dealkylation (C-7) or 3-hydroxylation. The apparent intrinsic deuterium isotope effect for 7-OMe coumarin was high (9