Analysis of Coumarin 7-Hydroxylation Activity of Cytochrome P450 2A6 using Random Mutagenesis*

Cytochrome P450 (P450) 2A6 is an important human enzyme involved in the metabolism of many xenobiotic chemicals including coumarin, indole, nicotine, and carcinogenic nitrosamines. A combination of random mutagenesis and high-throughput screening was used in the analysis of P450 2A6, utilizing a fluorescent coumarin 7-hydroxylation assay. The steady-state kinetic parameters (kcat and Km) for coumarin 7-hydroxylation by wild-type P450 2A6 and 35 selected mutants were measured and indicated that mutants throughout the coding region can have effects on activity. Five mutants showing decreased catalytic efficiency (kcat/Km) were further analyzed for substrate selectivity and binding affinities and showed reduced catalytic activities for 7-methoxycoumarin O-demethylation, tert-butyl methyl ether O-demethylation, and indole 3-hydroxylation. All mutants except one (K476E) showed decreased coumarin binding affinities (and also higher Km values), indicating that this is a major basis for the decreased enzymatic activities. A recent x-ray crystal structure of P450 2A6 bound to coumarin (Yano, J. K., Hsu, M. H., Griffin, K. J., Stout, C. D., and Johnson, E. F. (2005) Nat. Struct. Mol. Biol. 12, 822-823) indicates that the recovered A481T and N297S mutations appear to be close to coumarin, suggesting direct perturbation of substrate interaction. The decreased enzymatic activity of the K476E mutant was associated with decreases both in NADPH oxidation and the reduction rate of the ferric P450 2A6-coumarin complex. The attenuation is caused in part to lower binding affinity for NADPH-P450 reductase, but the K476E mutant did not achieve the wild-type coumarin 7-hydroxylation activity even at high reductase concentrations.

Recently x-ray crystal structures have been reported for P450 2A6 bound with coumarin and methoxsalen (23). The active site of P450 2A6 is six times smaller than that of human P450 2C8 (23,24), even smaller than that of bacterial P450 101A1 (25). The decrease in size of the active site in P450 2A6 relative to P450 2C8 is caused by repositioning of helix BЈ and helices F to H toward the active site. Large aromatic residues in the active site of P450 2A6 also reduce the volume of the substrate binding site.
In recent years, random mutagenesis and high-throughput screening methods have been applied to investigate structure-function relationships of proteins and to develop enzymes with desirable activities or altered catalytic specificity (26 -29). P450s are a class of enzymes that have attracted particular interest for their potential applications in biotransformation and the analysis of structure-function relationships. Random mutagenesis approaches have been used when a P450 enzyme can be functionally expressed and a suitable screen is available (20, 29 -36).
In this study, a random mutagenesis approach was applied to study structure-function relationships with P450 2A6. Mutations throughout the coding region have some effects on activity. Several of the mutants have altered substrate affinity and can be rationalized by the available structure, but an unexpected finding was that the mutant with the lowest activity is related to attenuated reduction by NADPH-P450 reductase.

EXPERIMENTAL PROCEDURES
Chemicals-Coumarin, 7-hydroxycoumarin, and 7-methoxycoumarin were purchased from Aldrich Chemical Co. (Milwaukee, WI) and recrystallized from C 2 H 5 OH/H 2 O or CH 3 OH/H 2 O before use. The synthesis of 3-hydroxy-7-methoxy coumarin is described elsewhere (37,38). Other chemicals were of the highest grade commercially available.
Construction of P450 2A6 Libraries-The general approach has been described previously (for P450 1A2) (36). Briefly, the P450 open reading frame region (1.5 kb) of a bicistronic plasmid (pCW2A6bc) expressing both human P450 2A6 and NADPH-P450 reductase was randomly mutated using low-fidelity PCR (Mutazyme, Stratagene, La Jolla, CA). The amplified PCR library fragment was purified and cloned into the pCW2A6bc vector using the NdeI and XbaI restriction sites.
Screening of P450 2A6 Libraries-P450 2A6/NADPH-P450 reductase bicistronic library DNA was transformed into Escherichia coli DH5␣FЈIQ and then plated on Luria-Bertani medium (with ampicillin, 100 g/ml) Q-tray plates (25 ϫ 25 cm; Genetix, Hampshire, UK). Individual colonies were picked and transferred to single wells of 96-well fluorescence plates filled with 200 l of TB expression medium, using a Genetix colony picking robot system (Genetix) (Fig. 1). Approximately 10 4 colonies were picked. The 96-well plates were then incubated at 30°C for 24 h. For the measurement of activity of mutants, coumarin (2 M) in minimal medium (including 0.04% glucose, w/v) was added into 96-well plates after the TB medium was removed. After 30 min of incubation at 37°C followed by the addition of alkaline solution (10 mM NaOH, containing 1 M NaCl), formation of the product 7-hydroxycoumarin was determined fluorometrically with a fluorescence plate reader (BD Bioscience, Palo Alto, CA). The mutants possessing low activity were selected and verified with duplicate assays for coumarin 7-hydroxylation in whole cells.
Enzymes-Bacterial inner membrane fractions containing mutant P450 2A6 and NADPH-P450 reductase were isolated and prepared from 1 liter of TB (with ampicillin, 100 g/ml) expression cultures of E. coli DH5␣FЈIQ, as described previously (39). Purification of P450 2A6 enzymes using a Ni 2ϩ -nitrilotriacetate column was as previously described (40,41). Recombinant rat NADPH-P450 reductase was expressed in E. coli and purified to apparent homogeneity as described elsewhere (42).
Steady-state Kinetic Assays for Coumarin 7-Hydroxylation and 7-Methoxycoumarin O-Demethylation and 3-Hydroxylation-Steadystate coumarin and 7-methoxycoumarin metabolism reactions included 20 pmol of P450 2A6 (wild type or mutants, in bicistronic membranes), in 0.50 ml of 100 mM potassium phosphate buffer (pH 7.4), along with a specified amount of the substrate. An NADPH-generating system (43) was used to start reactions. Incubations were generally done for 5 min at 37°C and terminated with 0.05 ml of 2 M HCl. Two volumes of CHCl 3 were added to extract the products, followed by vortex mixing and centrifugation. The organic layer (lower) was transferred to a clean tube and an equal volume of 0.1 M sodium borate buffer (pH 9.3) was added. The upper layer, containing the phenolic product, was transferred to fluorescence microtiter plates, and the 7-hydroxycoumarin was measured using a microtiter plate fluorescence plate reader (43). For the analysis of 3-hydroxy-7-methoxycoumarin, the reaction products were analyzed by HPLC using a Phenomenex Prodigy ODS (C 18 ) column (4.6 mm ϫ 150 mm, 3 m, Phenomenex, Torrance, CA) with the mobile phase H 2 O:CH 3 CN (70:30, v/v (solvent A); and 55:45, v/v (solvent B)) containing 10 mM HClO 4 at a flow rate of 1.0 ml Ϫ1 , monitoring A 322 (38). The percentage of solvent A was decreased linearly (to 0% solvent A,100% solvent B) over 10 min and held for 3 min; the percentage of solvent B was then decreased linearly (to 100% solvent A,0% solvent B) over 2 min and held for 3 min.
Other Catalytic Assays-Assays of tert-butyl methyl ether O-demethylation, an established P450 2A6 reaction (44,45), were done with bacterial membranes prepared from E. coli DH5␣-expressing P450 2A6 (or a mutant) and NADPH-P450 reductase. Membranes were dialyzed twice versus 25 volumes of 200 mM potassium phosphate buffer (pH 7.4) containing 0.1 mM EDTA at 4°C, first for 90 min and then 14 h, to remove glycerol, which contains traces of HCHO and reduces the assay sensitivity. 3 Membranes, containing 0.3 nmol of P450, were incubated with 100 M potassium phosphate buffer (pH 7.4) containing 20 -800 M tert-butyl methyl ether 4 in a 0.75-ml volume in 4-ml glass vials fitted with Teflon-lined septa (screw caps). The vials were preincubated for 5 min at 37°C, and reactions were initiated by the addition of 0.10 ml of an NADPH generating system (43). After 20 min at 37°C, the reactions were terminated by the addition of 0.25 ml of 17% HClO 4 and chilling on ice. Following centrifugation ((3 ϫ 10 3 ) ϫ g, for 10 min), aliquots (0.80 ml) of the supernatant from each incubation were transferred to clean glass vials and mixed with 0.32 ml of double-strength Nash reagent (43,47). The tubes were heated at 60 -70°C for 20 min and then centrifuged again at (3 ϫ 10 3 ) ϫ g for 5 min to remove any turbid material; the A 412 readings were used to estimate HCHO production by comparison with a reference curve prepared with a standard HCHO solution in the same way.
The P450 2A6-catalyzed 3-hydroxylation of indole yields the blue dye indigo (19). A continuous spectral assay (20) has a complication in that only the initial event, 3-hydroxylation, is catalytic. 3-Hydroxyindole (indoxyl) undergoes rapid, base-catalyzed oxidation in the presence of O 2 to yield the blue dimer indigo (48). An assay has been developed in which the oxidation of indole (or a phenyl ring-substituted indole) is carried out and stopped by the addition of base after a fixed time, with some additional time allowed for development of the indigo, which is stable. Further details are provided elsewhere (49). These assays were done with P450 2A6 (or mutants) and NADPH-P450 reductase expressed in an E. coli trnA Ϫ strain (22), devoid of tryptophanase required to produce endogenous indole and the subsequent blue color in the membranes (expression at 28°C for 14 h). Membranes containing 0.20 nmol of P450 2A6 (or a mutant) in 1.0 ml of 100 mM potassium phosphate buffer (pH 7.4) were incubated with 0 to 400 M concentrations of indole for 20 min at 37°C and quenched by the addition of 0.1 ml of 0.5 M NaOH; the formation of indigo was estimated colorimetrically by comparison with standard indigo (⑀ 600 ϭ 3.9 ϫ 10 3 M Ϫ1 cm Ϫ1 under these conditions (31)). The assumptions were made that all indoxyl was coupled to indigo formation and that the formation of a single indigo molecule represents the formation of two molecules of indoxyl.
NADPH oxidation rates for P450 2A6 enzymes were determined using P450/NADPH-P450 reductase/phospholipid reconstituted systems for some steady-state kinetic experiments (43). Reconstituted enzymes were preincubated for 5 min at 37°C in the presence or absence of coumarin (20 M). Reactions were initiated with the addition of 15 l of 10 mM NADPH, and the decrease of A 340 was monitored. Rates were calculated using the value ⌬⑀ 340 ϭ 6.22 mM Ϫ1 cm Ϫ1 .
Spectral Binding Titrations-Purified P450 2A6 and mutants were diluted to 1 M in 100 mM potassium phosphate buffer (pH 7.4) and divided between two glass cuvettes. Spectra (350 -500 nm) were recorded with subsequent additions of coumarin (from an aqueous stock, Ref. 38) or DTT (also aqueous stock) using an Aminco DW-2a/ OLIS spectrophotometer (On-Line Instrument Systems, Bogart, GA). The difference in absorbance between the difference wavelength maximum (385 nm) and minimum (418 nm) was plotted versus the substrate concentration (50) to estimate K s . 5 Anaerobic Reduction Kinetics-The modified systems for anaerobic stopped-flow experiments have been previously described (38,51). The instrument drive syringes were flushed with anaerobic buffer, filled with a solution of safranine T in 0.25 M Tris-HCl/25 mM EDTA buffer (pH 7.5) (38), and illuminated (with a lamp) to scrub the system. Enzyme mixtures (of purified wild-type 2A6 and the K476E mutant) and NADPH (150 M) were prepared separately and made anaerobic by alternative cycles of vacuum and argon, using a gas manifold. The dye solution was flushed from the stopped-flow instrument with anaerobic buffer, and then anaerobic samples were introduced from the tonometers. Reduction kinetics was monitored at 448 nm using an RSM-1000 stopped-flow apparatus (On-Line Instrument Systems).

RESULTS
Selection of Low Activity Mutants-More than 10 4 mutant clones were screened for coumarin 7-hydroxylation activity, and the mutants with apparently reduced activity were verified in duplicate cultures ( Fig.  1). Some of the mutants were found to have low activity because of very low P450 expression upon subsequent analysis and were discarded. Thirty-five mutants showing expression of P450 holoenzyme were isolated, and the nucleotide sequences of the DNAs were analyzed. Fiftyone amino acid residues were changed in the 35 mutants, and 12 of the changed residues were located in SRS regions (Fig. 2).
Catalytic Activities of Selected Mutants-Characterization of mutants was initially done with membrane preparations. Fe 2ϩ -CO versus Fe 2ϩ difference spectroscopy indicated that the whole cell expression levels of P450 2A6 varied among the mutants, ranging from 50 to 350 nmol per liter culture (see supplemental data Table  S1). The rates of coumarin 7-hydroxylation by P450 2A6 and the mutants were measured and normalized for P450 content (TABLE ONE, Figs. 3 and 4). Steady-state kinetic analysis of coumarin 7-hydroxylation indicated lower catalytic efficiencies (k cat /K m ) than wildtype P450 2A6 for most mutants (12-fold decreases in catalytic efficiency for mutants C004 and C332).
Wild-type P450 2A6 and five low activity mutants (k cat /K m Ͻ 1.5 min Ϫ1 M Ϫ1 ; mutants C004, C065, C188, C215, and C332) were selected for further analysis, because these represent a variety of levels of attenuation and also of sites of amino acid substitution. These mutants were used for analysis of four other catalytic activities (TABLES TWO and THREE) and were also purified to establish that the altered (coumarin 7-hydroxylation) activity was actually caused by the modifications of the P450 2A6 component (Fig. 4B).  All of the mutants displayed decreased catalytic efficiencies of 7-methoxycoumarin O-demethylation, mainly because of the increase of K m , but mutant C332 showed a highly decreased k cat value and decreased catalytic efficiency (TABLE TWO). Catalytic activities of P450 2A6 were also examined with two other, rather different substrates of P450 2A6, tert-butyl methyl ether (44,45) and indole (19,49). The same five selected mutants yielded patterns generally similar (TABLE THREE) to the results presented for the coumarin substrates (TABLES ONE and TWO), with some differences.
Kinetic Parameters of Site-directed Mutants-Three site-directed mutants related to the C332 mutant (K476E) were constructed (K476A, K476D, and K476R). Steady-state kinetic analysis of coumarin 7-hydroxylation activity (with the bacterial membranes) indicated that all of these three mutants had lower catalytic efficiency (ϳ20% that of the wild-type enzyme), but not as low as for K476E. These results suggest that the mutation Lys to Glu selectively induced the change in enzymatic activity.
Substrate Binding Affinities-Titration of purified wild-type P450 2A6 and the mutants showed classic "Type I" spectral changes on binding, suggesting the loss of coordination of H 2 O to the P450 heme. The absorbance spectra were characterized by an increase in absorbance at 385 nm and a decrease at 418 nm. The binding of coumarin to wild-type P450 2A6 yielded a K s of 1.1 M (TABLE FOUR). All of the mutants except C332 showed highly decreased coumarin binding affinities (increased K s values) (TABLE FOUR). These results suggest that the decreased enzymatic activities by the mutants might be attributable in part to the changes in binding affinity. Mutant C332 did not exhibit significantly altered binding affinity for coumarin (TABLE FOUR). However, the titration spectra of mutant C332 showed different absorbance characteristics (e.g. an increase at 392 nm and a decrease at 460 nm) from the wild-type P450 2A6 and the rest of the purified mutants (C004, C065, C188, and C215). This change was found to be caused by the presence of an unusual hyperporphyrin absolute spectrum for the ferric enzyme (split Soret peaks at ϳ360 and ϳ460 nm), which has been observed in some other hemoproteins (52, 53) (Fig. 5). To establish whether a covalently attached heme was responsible for the changes of activity and spectra in mutant C332, wild-type P450 2A6 and mutant C332 were resolved by SDS-polyacrylamide gel electrophoresis and stained using 3,3Ј,5,5Ј-tetramethylbenzidine (54 -56). In contrast to cytochrome c (positive control), there was no staining of wild-type P450 2A6 or the C332 mutant, which would have indicated the covalent binding of heme (see supplemental data Fig. S1).
The addition of other sulfur donors such as 2-mercaptoethanol or DTT generated very similar hyperporphyrin spectra in the purified wild-type P450 2A6 enzyme (results not shown). Strong binding affinity of DTT for the C332 mutant (K d of 5 M, TABLE FIVE) and the shift of the C332 mutant to a normal heme spectrum following dialysis to remove DTT (results not shown) suggests that the observed unusual spectra observed with the C332 mutant is caused by DTT ligation to heme. However, the shift of heme to the high spin state upon binding of coumarin to C332 mutant (same K s for coumarin) indicated that DTT binding was not responsible for its low enzymatic activity. The P450 2A6 C065 mutant (N297S) and human P450s 1A2, 2E1, and 3A4 displayed very weak or no binding affinity for DTT (TABLE FIVE). Thus, the high affinity of P450 2A6 for DTT is not general among all P450s and varies with the P450 2A6 mutants.
NADPH Oxidation and Reduction Kinetics-P450 2A6 oxidizes NADPH at a high rate in the presence of substrate (TABLE SIX). The C332 mutant (K476E) showed a highly decreased rate of NADPH oxidation (ϳ10%) both in the absence and presence of substrate (TABLE   TABLE TWO     SIX). This result suggests that the low enzymatic activity in C332 may be caused by a decrease in utilization of electrons. Wild-type P450 2A6 was reduced rapidly in the presence of coumarin, with a fit to a biexponential plot with k 1 ϭ 8.0 s Ϫ1 and k 2 ϭ 0.9 s Ϫ1 (Fig. 6), similar to that reported previously (38). In contrast, the rate of reduction of mutant C332 was very slow, with a zero-order fit of ϳ0.012 s Ϫ1 (Fig. 6). These results indicated that the C332 mutant has highly decreased enzymatic activities for oxidation of coumarin and 7-methoxycoumarin because of perturbation of electron transfer to the P450.

Kinetic parameters for tert-butyl methyl ether O-demethylation and indole 3-hydroxylation by P450 2A6 mutants measured in bacterial membrane fractions
Binding of P450 2A6 Enzymes to NADPH-P450 Reductase-Titration of P450 2A6 enzymes with NADPH-P450 reductase showed that the enzymatic activity of the wild-type enzyme was fully saturated at a low reductase concentration (molar ratio Յ5) but the C332 mutant did not reach saturation even at high concentrations (100-fold molar ratio) (Fig.  7). Difference spectra for the complexation of wild-type P450 2A6 enzyme with the reductase yielded a Type I-like binding spectrum (i.e. partial shift to high spin state) whereas mutant C332 did not (Fig. 8).
(We have demonstrated elsewhere that substrate is not needed for wildtype P450 2A6 to bind the reductase (57).) These results suggest that lower affinity and lack of productive binding of reductase may be responsible for the low NADPH oxidation and reduction rates observed with the K476E mutant (mutant C332).

DISCUSSION
Among the 35 P450 2A6 mutants from random libraries found to have decreased coumarin 7-hydroxylation activity, mutations were found not only in substrate recognition sites but also throughout the whole open reading frame of P450 2A6 (Fig. 2), with the most inhibitory mutation was found to have an effect different than substrate binding.
The positions of these mutants are placed in a structure of a P450 2A6coumarin complex (Fig. 9) that has been reported recently (23).
Recently He et al. (58) reported that residues Val 117 and Arg 372 of P450 2A6 are critical for coumarin 7-hydroxylation, based on an alignment study with P450 2A13, with catalytic efficiencies (k cat /K m ) of the mutants V117A and R372H reduced 8-and 5-fold, respectively. Interestingly, mutant R372H from that study (corresponding to our mutant C353) was found in the current random mutagenesis study and similar results were observed (mainly affecting the k cat value) (TABLE ONE).
Fitting to the homology model of He et al. (58) suggests that Val 117 and Arg 372 may play an indirect functional role in coumarin 7-hydroxylation through their interaction with two conserved residues (Arg 437 and Leu 370 ) that have the potential to influence the rate of coumarin 7-hydroxylation. The x-ray crystal structure (Fig. 9) does not show Val 117 or Arg 372 in contact with the substrate. A point that should be emphasized    here is that the boundaries of the "consensus" SRS regions may differ among P450s, depending on the size of their active sites. The K476E mutation (mutant C332) affected mainly the k cat for coumarin 7-hydroxylation rather than K m . The change of Lys (basic) to Glu (acidic) suggests that this residue might play a role through an ionic interaction, involving recognition of substrate or a conformational change in the protein. Replacement of Lys 476 by Ala or Asn (neutral) yielded lower catalytic efficiency than observed for the wild-type enzyme, mainly by affecting the coumarin K m value, not k cat . Interestingly, substitution with another acidic residue, Asp, did not yield catalytic efficiency as low as with the Glu (K476E). These results may exclude the possibility that the K476E mutant decreased the activity through perturbation of the ionic interaction in the structural moiety or with substrate, although the distance for ionic interaction by Glu may be an issue.
We previously reported that the reduction rate of ferric P450 2A6 (450 min Ϫ1 at 23°C) is much faster than overall catalysis (38). In the current study, wild-type P450 2A6 also showed a similar rapid reduction rate in the fast phase (480 min Ϫ1 at 23°C), which is not rate-limiting. However, the K476E mutant enzyme displayed a very slow reduction rate (ϳ0.16 min Ϫ1 at 23°C) in repeated experiments. In addition, the NADPH oxidation rate was highly decreased in the K476E mutant. These results suggest a role for Lys 476 in electron transfer to P450 2A6. Previous studies of rat P450 1A2 and neuronal nitric-oxide synthase by Shimizu and co-workers (59,60) have implicated similar roles of Lys residues in heme electron transfer pathways. Their work proposes that the interactions between Lys residues in the proximal heme site (of nitric-oxide synthase and P450 1A2) with the reductase domain may help to form a reductase to heme electron transfer pathway (59,60). In the x-ray crystal structure of P450 2A6 (23), Lys 476 appears to be located at the ␤4 hairpin region and exposed to the solvent (Fig. 9A).
Because intermolecular electron transfer across the interfacial surface between the oxygenase and adjacent reductase was found to involve Lys/ Arg residues of the proximal surface that interact with Asp/Glu residues of reductase surface, the K476E mutation may reduce catalytic activity by changing the intermolecular electron transfer process by forming ionic interactions between P450 and the reductase or by orienting geometry for the appropriate electron transfer on the interfacial surface between two proteins (59). Titration of the K476E mutant with reductase did not yield the level of enzymatic activity of wild-type P450 2A6, even at high concentrations (Fig. 7), indicating the role of Lys 476 in the interaction with the reductase.
Studies with additional substrates, which may have binding different from coumarin, also suggest that the K476E mutation perturbs the electron transfer to result in an inefficient enzyme rather than a change in the recognition of substrate (TABLES TWO and THREE). On the other hand, the binding affinities of the other mutants have much higher dissociation constants (K s ) than wild-type P450 2A6 (TABLE FOUR), suggesting that any structural changes elicited by these mutations may affect substrate binding to yield attenuated enzymatic activities. The increased K m values for coumarin oxidation in these mutants may reflect, in part, the weaker binding affinities to the substrate, although   the meaning of K m in the steady-state kinetics is not clear even in this case (38). The mutation N297S is located in the I-helix and appears to directly affect the substrate in the x-ray crystal structure (Fig. 9). The x-ray crystal structure shows hydrogen bonding interaction between Asn 297 and an oxygen atom of coumarin (23). This interaction presumably helps orient coumarin for hydroxylation. The mutation N297Q was previously identified in a random mutagenesis screen and found to increase catalytic activity toward indole (20). This view is consistent with the results that the K m and K s values are mainly affected for all of the substrates with these mutants.
The mutation A481T appears to be located in the formal active site (Fig. 9). The change of residue size and polarity in the active site may change the interaction of the mutated Thr residue to be direct to the substrate and affect its binding and orientation. Also, the structural change imposed by the A481T substitution may have an effect on the functional role of Phe 480 , which is known to be an important residue participating in the compact, hydrophobic active site architecture (23). The effect of the A481T mutation was more significant in the case of the substrate tert-butyl methyl ether (TABLE THREE) than the other substrates (TABLES TWO and THREE). Phe 480 is in some proximity to this substrate in a model (supplementary data, Fig. S2), although not obviously more than in the cases of coumarin (Fig. 9) and indole (49). The mutations F286L, D93H, and D169N are located in the ␤5 region, B-helix, and D-/E-helix ␤-turn regions, respectively, and appear to be outside of the active site (Fig. 9A). The effects of these substitutions suggest that structural elements outside of the active site also may play roles in changing the catalytic activity through a variety of changes proposed in the previous studies; e.g. blocking of substrate or product access channels, reductase binding, effects on the motion of the protein in conformational changes that occur in the catalytic cycle, and global changes in distant regions that affect residues in the substrate binding region to influence the distribution of non-productive binding modes (29,36). The effects of the D93H and D169N mutations were, for whatever reason, more pronounced with indole 3-hydroxylation (TABLE THREE).
The absolute spectrum of the K476E mutant was different from that of wild-type P450 2A6 (Fig. 5). The change of the heme spectrum was attributed to ligation with DTT and did not seem to affect the binding affinity to coumarin (and is probably not responsible for the decreased activity of K476E mutant) (TABLE FOUR). However, caution is suggested in the indiscriminate use of DTT and other thiols in the handling of P450 enzymes (TABLE FIVE), in the absence of a documented need to add these reducing agents.
In conclusion, we have applied random mutagenesis and fluorometric high-throughput screening approaches to study structure-function relationships with P450 2A6. Mutations at residues throughout the coding sequence can have at least small effects on catalytic activity. Some important residues involved in the enzymatic mechanism, especially substrate binding and electron transfer process, have been identified and explained in more detail with structural information from the x-ray crystal structure. One unexpected result was the role of a Lys distant from the active site in delivery of electrons from NADPH-P450 reductase.