The substrate specificity of cytochrome P450(cam)

Article abstract of DOI:10.1016/S0968-0896(98)00091-1

The catalytic turnover of xenobiotics by cytochrome P450(cam) results in both the formation of organic metabolites and the uncoupled production of H₂O₂, and H₂O. Previous studies have shown that a receptor-constrained thre

Full text of DOI:10.1016/S0968-0896(98)00091-1

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 1501±1508
The Substrate Speci®city of Cytochrome P450cam
Zhoupeng Zhang,a Ole Sibbesen,a,* Roy A. Johnsonb
and Paul R. Ortiz de Montellanoa,y
aDepartment of Pharmaceutical Chemistry, University of California, San Francisco, CA 94143-0446, USA
bResearch Laboratories, Pharmacia and Upjohn, Inc., Kalamazoo, MI 49001, USA
Received 31 December 1997; accepted 3 March 1998
AbstractÐThe catalytic turnover of xenobiotics by cytochrome P450cam results in both the formation of organic
metabolites and the uncoupled production of H2O2, and H2O. Previous studies have shown that a receptor-constrained
three-dimensional screening program (DOCK) can be used to identify potential ligands (ergo substrates) for the
enzyme (De Voss, J. J.; Sibbesen, O.; Zhang, Z.; Ortiz de Montellano, P. R. J. Am. Chem. Soc. 1997, 119, 5489). A new
set of 10 compounds has now been examined to further test the substrate speci®city of P450cam and the ability of
DOCK to identify substrates for this enzyme. The results expand the known speci®city of P450cam and de®ne limita-
tions in the use of DOCK to predict its substrate speci®city. # 1998 Elsevier Science Ltd. All rights reserved.
Introduction
species and to bind and orient the substrate in its
proximity. The chemo- and regiospeci®city of substrate
oxidation, however, are largely determined by the
intrinsic reactivities of the substrate sites that are acces-
sible to the ferryl species in the enzyme±substrate com-
plex. In addition to coupled turnover, which results in
oxidation of the organic substrate, cytochrome P450
enzymes undergo uncoupled turnover to produce O2-,
H2O2, and H2O.2,3 The uncoupling of NAD(P)H con-
sumption from organic substrate oxidation is due to
dissociation of superoxide from the ferrous dioxy com-
plex, dissociation of H2O2 from the ferric peroxide
complex, and/or reduction of the ®nal ferryl species to
water by electrons provided by NAD(P)H (Fig. 1). A
low degree of uncoupled turnover is normally observed
with the resting enzyme, but the extent of uncoupled
oxygen reduction is greatly increased by the binding of
some ligands or substrates. Uncoupled turnover is thus
a substrate-dependent catalytic process that produces
reduced oxygen rather than organic metabolites.
Cytochrome P450 enzymes catalyze a range of oxidative
transformations, including carbon hydroxylation,
heteroatom oxidation, p bond oxidation, and hydro-
carbon desaturation.1 Despite di€erences in the phylo-
genetic origin, cellular localization, electron donor
partners, and substrate speci®city of P450 enzymes,
their active sites and catalytic mechanisms appear to be
remarkably similar. The catalytic cycle of the known
P450 enzymes consists of: (a) substrate binding with a
concomitant decrease in the reduction potential of the
heme iron atom; (b) reduction of the iron to the ferrous
state; (c) formation of the ferrous dioxy (Fe+2-O2)
complex; (d) reduction of the ferrous dioxy complex to
an intermediate formally equivalent to a ferric peroxide
(Fe+3 -OOH); (e) cleavage of the dioxygen bond to give
a ferryl (formally Fe+5=O) species; (f) oxidation of the
substrate by the ferryl species, yielding an enzyme±
product complex and (g) dissociation of the product
(Fig. 1).1 The primary function of the enzyme in this
catalytic cycle is to promote formation of the ferryl
The broad substrate speci®city of individual P450
enzymes,4 their preference for lipophilic substrates,5,6
and the hydrophobic nature of the crystallographically-
de®ned P450 active sites7±10 suggest that the binding of
P450 substrates is governed by hydrophobic and steric
factors. Control of substrate binding by speci®c hydro-
gen bonding or ionic interactions is relatively infrequent.
Little is known, however, about the detailed control of
Key words: Cytochrome P450; DOCK; substrate speci®city;
P450 uncoupling; molecular modeling.
*Current address: Danisco Biotechnology, Langebrogade 1,
DK-1001 Copenhagen K, Denmark.
yCorresponding author. Tel: (415) 476-2903; Fax: (415) 502-
4728; E-mail: ortiz@cgl.ucsf.edu
0968-0896/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00091-1

1502
Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
allowed between substrate and active site atoms.12±14
One important ®nding of the previous study was that a
certain degree of mobility within the active site, or of
active site deformability that allows substrate reorienta-
tion, is required for a compound to be a substrate.
Varying the minimum allowed distance between sub-
strate and protein atoms, which are arti®cially held in a
rigid matrix, provides a formal construct for represent-
ing active site deformability. Decreasing the value of
this parameter increases the extent to which substrate
and protein atoms are allowed to overlap, which is for-
mally equivalent to allowing the protein atoms a certain
degree of freedom to move out of the way. To test the
generality of the parameters derived earlier, we have
used the same DOCK parameters to examine an inde-
pendent library of compounds as potential substrates
for P450cam. The library consists of compounds pre-
viously investigated at Upjohn in studies of the fungal
metabolism of organic compounds.15±21 This relatively
small library of compounds was chosen because; (a) it
consists of small to medium size non-ionic compounds
with a workable degree of water solubility; (b) it
includes a related but diverse set of structural frame-
works; and (c) metabolite identi®cation is greatly sim-
pli®ed due to the availability of authentic standards for
many of the potential metabolites.
Figure 1. Catalytic cycle of cytochrome P450, showing the
individual steps at which uncoupling can occur. The iron of the
heme group is indicated as an iron atom in brackets.
substrate speci®city even though a better understanding
of P450 speci®city would facilitate the identi®cation of
substrates for individual P450 enzymes, the design of
P450 enzymes with tailored speci®cities, and the predic-
tion of potentially toxic interactions.
Computational approaches to the identi®cation of P450
substrates are attractive because the question of speci®-
city can be dissected into distinct steps, nonpolar inter-
actions primarily determine speci®city, and the protein
does not participate actively in the reaction of the ferryl
species with the substrate. We recently described the use
Results
P450cam substrate predictions
of a receptor-constrained docking program (DOCK) to
11
The receptor-constrained docking program DOCK was
recently used to identify potential cytochrome P450cam
substrates from a data base.12±14 To further test the
validity of the system we have carried out a series of
studies with a di€erent library of potential substrate
structures. For the present studies, the structure data
base was assembled from compounds that had been
previously investigated at Upjohn as substrates for
microbial metabolism.15±21 This approach was taken to
obtain a data base of compounds that would be rea-
sonably diverse, would have favorable physical proper-
ties (e.g. water solubility), and for which many of the
metabolites would be available to simplify product
identi®cation. As reported earlier,12±14 the three-dimen-
sional structures of the compounds in the present library
of structures were constructed using Sybyl energy mini-
mization. The list of structures in the data base is avail-
able in a supplement to this manuscript.
identify potential substrates for P450cam
,
a soluble
bacterial enzyme for which a crystal structure is avail-
able.12±14 The study examined 16 compounds selected
from a list of compounds identi®ed by DOCK from a
library of 20,000 compounds as either potential P450cam
substrates or non-substrates. It was assumed in this
study that binding of a compound, as predicted by the
DOCK analysis of its ability to ®t within the P450cam
active site, is sucient to trigger catalytic turnover of
the enzyme. Catalytic turnover was de®ned as NADH-
and substrate-dependent formation of either organic or
reduced oxygen metabolites. In e€ect, after empirical
parameter optimization,
a correlation was found
between the predicted ability of a compound to ®t
within the P450cam active site and whether it was a sub-
strate or not. Some of the compounds identi®ed as sub-
strates were oxidized with moderate eciency to organic
products, but others primarily triggered uncoupled
turnover.12±14
Analysis of the in¯uence of the variable parameters of
DOCK on the accuracy of the predictions in the earlier
study suggested that a key parameter was the minimum
distance allowed between substrate and protein atoms.
Empirically, the best predictions were obtained in the
The set of compounds examined in the previous study
was used to de®ne the values for the variable parameters
in DOCK, particularly the minimum contact distance

Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
1503
earlier study with a value of 2.9 A for this distance.14
The value of this parameter that gives the best predic-
tions has been similarly determined in the present study.
Although the value of the parameter was initially set at
di€erent values for polar and nonpolar atoms, in the
studies reported here a single value was used for both
polar and nonpolar atoms. Thus, the binding of Sybyl-
minimized structures to wild-type P450cam was exam-
ined by DOCK while gradually increasing the minimum
contact distance from 2.6 to 3.2 A (Fig. 2). Eight of the
10 structures were correctly predicted to be substrates
for the P450cam model when the minimum contact dis-
tance was set to 2.7 A for both polar and nonpolar
atoms, but the two non-substrates were not correctly
predicted.
to 16 mM for 2, are to be compared with Ks=1.1 mM for
the natural substrate camphor.
Saturation of the P450cam active site with camphor
results in complete conversion of the enzyme from the
low to the high spin state. Of the new compounds
examined, the two with the lowest Ks values,
6
(Ks=0.4 mM) and 10 (Ks=0.6 mM), convert 59% and
78%, respectively, of the enzyme to the high spin state.
However, compound 2, which has the highest Ks value
(16 mM), causes a comparable maximum spin state
change (69%). As reported earlier, there is little corre-
lation between the binding anity and the extent to
which the enzyme is converted to the high spin state by
a saturating concentration of the substrate.14
Binding of compounds to P450cam
Oxygen consumption and uncoupling
The binding of a potential substrate to the cytochrome
P450cam active site is a prerequisite for catalytic turn-
over. The spectroscopic dissociation constants (Ks),
determined from the relationship between substrate
concentrations and the spectroscopically measured low-
to-high spin transitions for wild-type P450cam, are given
in Table 1. The Ks values, which vary from 0.4 mM for 6
NADH consumption provides a direct measure of the
extent of catalytic turnover engendered by the binding
of a substrate to P450cam, with catalytic turnover encom-
passing both coupled turnover (substrate oxidation) and
uncoupled turnover (reduction of O2 to H2O2 and H2O).
In the presence of the natural substrate camphor,
P450cam consumes NADH at the rate of
262 nmol min 1 nmol 1, ꢀ98% of which is coupled to
.
the 5-exo-hydroxylation of camphor. None of the
unnatural substrates achieves a level of NADH con-
sumption comparable to that of camphor (Table 1). The
NADH consumption rates decrease in the order
10>2>3>7>6>1>9>5>4. Compound
8 pre-
cipitates from the incubation solution under the experi-
mental conditions, making its activity as a substrate
dicult to evaluate. The proportion of the uncoupled
formation of H2O2 increases in the order
1<10<3<7<6<2<9<4, 5.
P450cam metabolites
Most of the metabolites produced by P450cam were
identi®ed by gas±liquid chromatographic and mass
spectrometric comparison with authentic standards. In
the case of compound 3, one of the metabolites was
characterized by spectroscopic methods after isolation
from a large scale incubation of the substrate with a
PdR-Pd-P450cam fusion protein.22 The metabolite
formed from compound 5 was determined by compar-
ison with an authentic standard synthesized by bromin-
ation of 5 followed by basic hydrolysis.
The oxidation of 1-benzoylcycloheptamethylenimine (1)
by P450cam yields two hydroxylated metabolites in a 6:1
ratio. The major metabolite is identical to an authentic
standard, N-benzoylhexahydro-5(2H)-azocin-5-ol, and
the minor metabolite, for which no standard is avail-
able, has the mass spectrometric molecular ion expected
Figure 2. Structures of the compounds tested as potential sub-
strates for P450cam and of the metabolites obtained from them.

1504
Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
Table 1. Parameters for the interaction of DOCK-predicted substrates with wild-type P450cam
Compd
Ks Spin state change
NADH Useda
O2 Used
H2O2 Formed Organic products Product formationb
1
1
.
mM
Camphor 1.1
%
100
59
nmol min 1 nmol
.
nmol min 1 nmol
%
%
96
55
31
94
<1
1
262
26
45
41
6
250
32
40
33
9
2
+
+
+
+
±
1
2
2.8
16.0
3.0
18
65
69
57
3
38
4
NA
1.4
<4
32
100
100
58
5
8
8
+
+
+
ND
+
+
6
0.4
2.1
59
36
32
37
NDc
14
49
28
26
ND
8
37
4
7
39
8
0.9
4.5
24
33
ND
72
ND
1
9
10
0.6
78
31
25
20
a
1
Background NADH consumption is 4±6 nmol min 1 nmol
.
.
bThe turnover is the ratio of the area of the organic product peak divided by the area of the starting material plus organic product
after both have been normalized versus the internal standard peak times Â100.
cCompound 8 precipitated at the substrate concentration (1 mM) utilized to measure catalytic turnover. No product was detected
when lower concentrations were used.
for a monohydroxylated product. The chromatographic
and mass spectrometric properties of the latter product
suggest the ring system is intact, so the hydroxyl is at
positions 3 or 4 rather than at position 2. The oxidation
Discussion
Of the 10 compounds examined in this study as poten-
tial substrates for cytochrome P450cam, eight are oxi-
dized by the wild-type enzyme to organic products, in all
cases with the concomitant formation of H2O2
(Table 1). The other two compounds (8 and 4) are not
detectable substrates. None of the compounds is turned
of compound
2 produces two monohydroxylated
metabolites in a 3:1 ratio. The major product is identi-
®ed by comparison with authentic standards as N-ben-
zoyl-2-methylpiperidin-3-ol and the minor product as
N-benzoyl-2-methylpiperidin-4-ol. Oxidation of N-ben-
zoyl-2,6-dimethyl-piperidine (3) produces N-benzoyl-
2,6-dimethylpiperidin-3-ol as the sole metabolite. a,a-
Dimethyl-3-(1-methylethyl)benzene-methanol is the sole
metabolite produced by P450cam from 1,3-diisopropyl-
benzene (5). The oxidation of compound 6 results, as
indicated by the mass spectrometric and chromato-
graphic properties of the single metabolite, in mono-
hydroxylation of the bicyclic ring system. Although the
exact position and stereochemistry of hydroxylation
remain unde®ned, the site of the hydroxylation is not
the bridgehead carbon as this would lead to formation
of a ring opened structure. The metabolism of com-
pound 7 by P450cam produces two products in a 2:1
ratio. The major metabolite is the endo-3-hydroxylated
product, and the minor metabolite, which is shown by
its mass spectrometric molecular ion at m/z 233.1 to be
monohydroxylated, is the exo-3-hydroxylated product.
Oxidation of compound 9 produces two metabolites in a
1:1 ratio. The two metabolites are the 3-exo- and 3-endo-
hydroxyl derivatives of 9. 1-Azidoadamantane (10) is
cleanly oxidized to the bridgehead tertiary alcohol.
Under the experimental conditions, compound 8 pre-
cipitates from the incubation solution and no products
were detected even though 8 causes a 24% low to high
spin shift with a binding constant (Ks) of 0.9 mM. Com-
pound 4 is not a detectable P450cam substrate.
over at
a rate comparable to that of camphor
(262 nmol min 1 nmol 1) but six (1, 2, 3, 6, 7, and 10)
are oxidized at rates at least ®ve times higher than the
background level of NADH consumption. Among the
compounds that give organic products, coupled turn-
over is best for compounds 1 and 10 and worst for
compound 5. In all cases, the organic products are
obtained by monohydroxylation of a saturated carbon
atom.
.
The hydroxylation of compound 10 occurs exclusively
at a bridgehead position to give the tertiary alcohol
(Fig. 2). The hydroxylation of 10 is thus similar to the
previously reported hydroxylation of adamantane by
P450cam, which also occurs exclusively at the bridgehead
carbon.23 The crystal structure of adamantane bound in
the active site of P450cam has been determined and evi-
dence has been obtained that adamantane has a fair
degree of mobility within the active site.24 It is there-
fore not surprising that the oxidation occurs primarily
at the carbon with the weakest C-H bond. The azido
group in 10 is unlikely to suppress active site mobility
and therefore also gives the product in which the weak-
est (bridgehead) C-H bond is oxidized.
The hydroxylation regiospeci®cities for the other sub-
strates, most of which have several distinct but equally

Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
1505
reactive sites, are less readily rationalized and are likely
to be in¯uenced by protein-controlled orientation of the
substrate within the active site. The oxidation regio-
speci®cities in situations where protein interactions play
a strong role are dicult to rationalize without recourse
to molecular dynamics calculations. Nevertheless, two
observations are warranted. Compounds 2 and 3 are
closely related but compound 3 is both more symmetric
and conformationally more rigid than 2. These di€er-
ences may explain the tighter binding of 3 than 2 (Ks=3
and 16 mM, respectively) and the fact that 3 gives one
whereas 2 gives two metabolites. A second notable point
is that aromatic ring hydroxylation is not observed,
either because the aromatic rings are bound away from
the oxidizing species or because the intrinsic reactivity
of the aromatic ring is lower than that of the secondary
or tertiary carbons hydroxylated in the substrates.
The use of DOCK to distinguish between substrates and
non-substrates was found in an earlier study of 16 com-
pounds to be most accurate if the minimum distance
allowed between substrate and active site atoms was set
at 2.9 A.14 However, the activities as substrates of the
present library of structures are most accurately ratio-
nalized when the DOCK minimum contact value is set
at 2.7 rather than 2.9 A (Fig. 3). This disparity con®rms
that the accuracy of DOCK in identifying P450cam sub-
strates depends signi®cantly on the choice of the mini-
mum contact distance. The ®nding that an optimum
value for the minimum contact distance may be dicult
to de®ne suggests that the primary utility of DOCK in
the prediction of substrate speci®city for P450cam may
be in selecting substrate candidates from large data
bases rather than in discriminating between closely
related structures.
Figure 3. Relationship between the minimum distance allowed
to separate a substrate and a protein atom and the prediction
by DOCK that the compound is a substrate. The solid bars
indicate compounds shown experimentally to be substrates,
and the open bars compounds shown not to be substrates. The
bars end at the minimum allowed distance beyond which the
compounds are predicted by DOCK to not be substrates. The
enzyme used in this study was wild-type ferric P450cam
.
sulfonyl-7-azabicyclo[2.2.1]heptane (8), 7-tert-butyl-
oxycarbonyl-7-azabicyclo[2.2.1]heptane (9), and 1-azido-
adamantane (10) were obtained from Upjohn. 1,4-Di-
isopropylbenzene (4) and 1,3-diisopropylbenzene (5)
were purchased from Aldrich. Of the standards used in
metabolite identi®cation, two were synthesized (see
below) and the following were obtained from Upjohn:
N-benzoylhexahydro-5(2H)-azocin-5-ol,18 1-benzoyl-2-
methylpiperidin-3-ol,19 1-benzoyl-2-methylpiperidin-4-
In sum, studies with a new library of compounds extend
the range of known substrates for P450cam and con®rm
that the enzyme accepts a wide range of xenobiotics,
although catalytic turnover is most highly coupled and
ecient with the natural substrate camphor. The results
also identify a limitation in the use of DOCK to predict
ol,19
2-endo-7-phenyloxycarbonyl-7-azabicyclo[2.2.1]-
heptan-2-ol,15 2-endo-7-tert-butyl-oxycarbonyl-7-azabi-
cyclo[2.2.1]heptan-2-ol,15 2-exo-7-tert-butyloxycarbonyl-
7-azabicyclo-[2.2.1]heptane,15 and 3-hydroxy-1-azido-
adamantane.16
substrates for P450cam
.
1-Benzoyl-cis-2,6-dimethylpiperidin-3-ol. The title com-
pound was enzymatically produced by adding a 6 mg
sample of 1-benzoyl-cis-2,6-dimethyl-piperidine (3) in
0.1 mL of ethanol to 200 mL of E. coli culture
(OD600ꢀ0.8) expressing the PdR-Pd-P450cam fusion
proteind22 (approximately 80 nmol). The mixture was
aerated at 25 ꢁC for 2 h. The cell culture was extracted
with CHCl3 (40 mLÂ3) and the combined extracts were
then washed with water and dried over anhydrous
Na2SO4. The residue obtained on solvent removal was
chromatographed on a silica gel column with CH2Cl2/
Experimental
Materials and general methods
Cytochrome P450cam, putidaredoxin reductase, and puti-
daredoxin were expressed heterologously in Escherichia
coli and were puri®ed as previously reported.14 The
substrates 1-benzoylheptamethylenimine (1), 1-benzoyl-2-
methylpiperidine (2), 1-benzoyl-cis-2,6-dimethylpiperidine
(3), benzoyl-7-azabicyclo[2.2.1]heptane (6), 7-phenyl-
oxycarbonyl-7-azabicyclo[2.2.1]heptane (7), 7-p-toluene-

1506
Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
MeOH (98:2) as the eluent. Solvent removal from the
appropriate fractions gave 4.2 mg (68%) of the title
compound: MS (EI): 233.2 (M+); 1H NMR (300 HZ,
CDCl3): d 1.31 (d, 3H), 1.45 (d, 3H), 1.62 (m, 4H), 1.92
(m, 1H), 2.06 (m, 1H), 4.13 (m, 1H), and 7.28ꢀ7.38 ppm
(m, 5H). An attached proton test (APT) was performed
to distinguish the di€erent carbons: (CDCl3,d ppm):
22.1, 23.1, 36.6, 39.0, 47.7, 60.3, 64.4, 125.5, 128.4,
128.5, 136.8, 171.7.
Gilson Oxy5 Oxygraph electrode were connected to the
cell. The reaction mixture contained, in a ®nal volume
of 1.1 mL, 200 mM potassium phosphate bu€er (pH
7.4), 1 mM camphor-free P450cam, 2 mM PdR, 4 mM Pd,
250 mM NADH, and the substrate added in 10 mL of
ethanol. Substrates were present at a concentration 10-
fold higher than their Ks values or just below their
solubility limits. The NADH was added last to initiate
the reaction. After 5 min incubation, a 0.3 mL aliquot of
the incubation mixture was added to 0.2 mL of 0.3 M
H2SO4 and the mixture was stored on ice until the H2O2
assay was carried out essentially as described previously
(see below).28 b-Mercaptoethanol was removed from
the Pd prior to use by passage through a G-25 gel ®l-
tration column (100 mM TRIS, pH 7.4) because the
thiol interferes with the peroxide assay and, at higher
concentrations, raises the background NADH and O2
consumption values. The kinetic rate for the NADH
consumption was obtained from the resulting data by
ꢀ,ꢀ-Dimethyl-3-(1-methylethyl)benzenemethanol.25. The
title compound was synthesized by slowly adding bro-
mine (80 mg, 0.5 mmol) in CCl4 (1 mL) to 1,3-diisopro-
pylbenzene (96%, 423 mg, 2.5 mmol) at room
temperature, followed by stirring of the reaction mixture
for an additional 15 min. After removal of CCl4 under
reduced pressure, the residue was poured into 10 mL of
10% aqueous NaOH and the solution was re¯uxed for
12 h. The reaction mixture was extracted with ether
(3Â10 mL). The combined extract was washed with 10%
NaHCO3 solution (10 mL) and water (10 mLÂ2). The
solvent was removed under reduced pressure and the
residue was chromatographed through a small silica gel
column with CH2Cl2 as the eluent. The appropriate
fractions were collected and the solvent was removed in
vacuo and a colorless oil was isolated (54 mg, 61%). EI-
a
zero-order calculation. The background NADH
consumption values were recorded when ethanol
(instead of substrate solution) was added to the mix-
ture. The O2 electrode was calibrated in distilled water
at 25 ꢁC, in which the O2 concentration was taken to
be 261 mM.
1
HRMS: m/e 178.1350 (178.1358 calcd for C12H18O); H
NMR (300 Hz, CDCl3): d 1.25 (d, 6H), 1.58 (s, 6H), 2.16
(s, 1H), 2.92 (m, 1H), 7.11ꢀ7.37 (m, 4H).
To quantitate H2O2, 0.125 mL of 6 mM ferrous ammo-
nium sulfate and 0.125 mL of 6 M ammonium thio-
cyanate were added to the acidi®ed aliquot of the
reaction mixture taken above. The mixture was cen-
trifuged at 12,000 RPM for 2 min to pellet precipitated
protein and the absorbance was measured at 480 nm on a
Hewlett±Packard 8452A diode array spectrophotometer.
The amount of H2O2 formed was calculated from a
standard curve made from a stock H2O2 solution.
Binding of compounds to P450cam. The test compound in
ethanol was added in increasing amounts to the sample
cuvette, and an equal amount of ethanol was added to a
matched reference cuvette. Both cuvettes contained
1 mM camphor-free P450cam in 500 mL of 200 mM
potassium phosphate bu€er (pH 7.4). The ethanol con-
centration did not exceed 1%. The solutions were mixed
after each addition of substrate and/or ethanol and the
di€erence spectrum was recorded. The value of Ks was
determined by plotting 1/ÁA versus 1/[S], the x-inter-
cept of which yields -1/Ks from the relationship 1/
[S]=[E]Áe/KsÁA-1/Ks, where [S] is the concentration of
Metabolite analysis and characterization
A
1 mL solution containing 1 mM camphor-free
P450cam, 2 mM putidaredoxin reductase, 8 mM putidare-
doxin, 1 mM catalase (to remove any H2O2 formed in the
reaction), 1 mM test compound (added in 10 mL etha-
nol), and 5 mM NADH was incubated at 25 ꢁC for 1.5 h.
A suitable amount of organic solvent (e.g. chloroform
or diethyl ether) was added to the above reaction solu-
tion with 1-bromoadamantane as the internal standard.
The extracted material was characterized by comparison
with authentic standards by gas chromatography and
tandem gas chromatography-mass spectrometry. Gas
chromatography was performed on a Hewlett±Packard
5890 instrument equipped with a 30 m DB-17 capillary
column under the following conditions: 60 ꢁC for 1 min;
the test compound, [E] is the concentration of P450cam
,
Áe is the di€erence in molar absorptivity of the free and
bound P450cam is the spectroscopic binding constant of
the compound, and ÁA is the peak-to-trough absor-
bance in the di€erence spectrum.26,27
Oxygen consumption, NADH consumption, and H2O2
production
The reaction was carried out in a custom made quartz
cell thermostated to 25 ꢁC in which the O2 concentration
and the decrease in the NADH absorption maximum at
340 nm can be simultaneously monitored. A Hewlett±
Packard 8452A diode array spectrophotometer and a
gradient of 10 ꢁC min from 60±250 ꢁC; ®nally 300 ꢁC
1
for 9 min. Tandem gas chromatography-mass spectro-
metry was carried out on a Hewlett±Packard model
5890 gas chromatograph interfaced with a VG-70 mass

Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
1507
spectrometer. The column and gas chromatographic
program were as described above.
2. Goeptar, A. R.; Scheerens, H.; Vermeulen, N. P. E. Crit.
Rev. Toxicol. 1995, 25, 25.
3. Archakov, A. I.; Bachmanova, G. I. Cytochrome P-450 and
Active Oxygen; Taylor & Francis: New York, 1990; pp 129±
138.
Molecular modeling
4. Guengerich, F. P. In Cytochrome P450: Structure, Mechan-
ism, and Biochemistry, 2nd Edn; Ortiz de Montellano, P. R.,
Ed.; Plenum: New York, 1995; pp 473±535.
The coordinates of camphor-bound cytochrome
P450cam were obtained from the Brookhaven National
Laboratory's public data bank (PDB access code 2cpp) .
The principles and algorithms that subserve the pro-
gram DOCK 3.5 are described, and a manual can be
found, online at http://www.cmpharm.ucsf.edu/kuntz/
dock.html.29 To use the program, it is necessary to: (a)
generate a molecular surface of the enzyme active site;
(b) generate a set of overlapping spheres of varying sizes
that ®ll the cavity de®ned by the generated surface (the
centers of the spheres serve as potential ligand atom
positions); (c) generate a gridmap that allows calcula-
tion of the minimum distance between ligand and
receptor (enzyme) atoms; (d) generate a 3-D ligand
database in DOCK 3.5 readable format; and (e) run
DOCK 3.5. We have chosen to run the program using
contact (as opposed to force-®eld) scoring,30 and to run
the program repeatedly while varying the minimum
allowed distance between ligand and enzyme atoms to
®nd the optimal minimum distance to be used for dis-
tinguishing substrates from nonsubstrates. The mini-
mum allowed distance was incrementally increased from
2.6 to 3.2 A using the same distance for polar and non-
polar interactions. The 3-D structures of the test com-
pounds were generated with the software package Sybyl
6.1 and were minimized by Sybyl energy minimization
methods. Those structures were then written into a
DOCK 3.5 readable database, which was searched by
DOCK 3.5.
5. White, R. E.; McCarthy, M. Arch. Biochem. Biophys. 1986,
246, 19.
6. Duquette, P. H.; Erickson, R. R.; Holtzman, J. L. J. Med.
Chem. 1983, 26, 1343.
7. Poulos, T. L.; Finzel, B. C.; Howard, A. J. J. Mol. Biol.
1987, 195, 687.
8. Ravichandran, K. G.; Boddupalli, S. S.; Hasemann, C. A.;
Peterson, J. A.; Deisenhofer, J. Science 1993, 261, 731.
9. Hasemann, C. A.; Ravichandran, K. G.; Peterson, J. A.;
Deisenhofer, J. J. Mol. Biol. 1994, 236, 1169.
10. Cupp-Vickery, J. R.; Poulos, T. L. Nature Struct. Biol.
1995, 2, 144.
11. Nelson, D.; Koymans, L.; Kamataki, T.; Stegeman, J.;
Feyereisen, R.; Waxman, D.; Waterman, M.; Gotoh, O.; Coon,
M.; Estabrook, R.; Gunsalus, I.; Nebert, D. Pharmacogenetics
1996, 6, 1.
12. De Voss, J. J.; Ortiz de Montellano, P. R. J. Am. Chem.
Soc. 1995, 117, 4185.
13. De Voss, J. J.; Ortiz de Montellano, P. R. Meth. Enzymol.
1996, 272, 336.
14. De Voss, J. J.; Sibbesen, O.; Zhang, Z.; Ortiz de Mon-
tellano, P. R. J. Am. Chem. Soc. 1997, 119, 5489.
15. Davis, C. R.; Johnson, R. A.; Cialdella, J. I.; Liggett, W.
F.; Mizsak, S. A.; Marshall, V. P. J. Org. Chem. 1997, 62, 2244.
16. Davis, C. R.; Johnson, R. A.; Cialdella, J. I.; Liggett, W.
F.; Mizsak, S. A.; Han, F.; Marshall, V. P. J. Org. Chem. 1997,
62, 2252.
17. Johnson, R. A. In Oxidation in Organic Chemistry; Traha-
novsky, W. S., Ed.; Academic: New York, 1978; Part C, pp
131±210.
Acknowledgements
18. Johnson, R. A.; Herr, M. E.; Murray, H. C.; Fonken, G. S.
J. Org. Chem., 1968, 33, 3187.
The authors thank Dr. I. D. Kuntz for a number of
suggestions concerning the manuscript. This work was
supported by grant GM25515 from the National Insti-
tutes of Health and by a fellowship to Ole Sibbesen
from the Danish Veterinary and Agricultural Research
Council. Mass spectrometry was carried out in the Bio-
medical Bioorganic Mass Spectrometry Facility (A.
Burlingame, director) of the University of California at
San Francisco supported by National Institutes of
Health Grants RR 01614 and Liver Core Center 5 P30
DK26743.
19. Johnson, R. A.; Murray, H. C.; Reinecke, L. M.; Fonken,
G. S. J. Org. Chem. 1969, 34, 2279.
20. Johnson, R. A.; Hall, C. M.; Krueger, W. C.; Murray, H.
C. Bioorg. Chem. 1973, 2, 99.
21. Johnson, R. A.; Herr, M. E.; Murray, H. C.; Reinecke, L.
M.; Fonken, G. S. J. Org. Chem. 1968, 33, 3195.
22. Sibbesen, O.; De Voss, J. J.; Ortiz de Montellano, P. R. J.
Biol. Chem. 1996, 271, 22462.
23. White, R. E.; McCarthy, M.-B.; Egeberg, K. D.; Sligar, S.
G. Arch. Biochem. Biophys. 1984, 228, 493.
24. Raag, R.; Poulos, T. L. Biochemistry 1991, 30, 2674.
25. Iwamuro, H.; Kanehiro, M.; Matsubara, Y. Yukagaku,
1982, 31, 222.
References
1. Ortiz de Montellano, P. R. In Cytochrome P450: Structure,
Mechanism, and Biochemistry, 2nd Ed; Ortiz de Montellano, P.
R., Ed.; Plenum: New York, 1995; pp 245±304.
26. Estabrook, R. W.; Peterson, J.; Baron, J.; Hildebrandt, A.
In Methods in Pharmacology; Chignell, C.F., Ed.; Appleton-
Century-Crofts: New York, 1972; Vol. 2, pp 303±350.

1508
Z. Zhang et al./Bioorg. Med. Chem. 6 (1998) 1501±1508
27. Rao, S.; Wilks, A.; Hamberg, M.; Ortiz de Montellano, P.
R. J. Biol. Chem. 1994, 269, 7210.
29. Gschwend, D. A.; Good, A. C.; Kuntz, I. D. J. Molec.
Recog. 1996, 9, 175.
28. Fruetel, J.; Chang, Y. T.; Collins, J.; Loew, G.; Ortiz de
Montellano, P. R. J. Am. Chem. Soc. 1994, 116, 11643.
30. Meng, E. C.; Shoichet, B. K.; Kuntz, I. D. J. Comput.
Chem. 1992, 13, 505.