Kinetic analysis of hydroxylation of saturated fatty acids
by recombinant P450foxy produced by an
Escherichia coli
expression
system
7
Tatsuya
7
Kitazume
1
, Akinori Tanaka
1
, Naoki Takaya
1
, Akira Nakamura
1
, Shigeru Matsuyama
1
,
Takahisa Suzuki
1
and Hirofumi Shoun
2
1
Institute of Applied Biochemistry, University of Tsukuba, Tsukuba, Ibaraki, Japan;
2
Department of Biotechnology, Graduate School
of Agricultural and Life Sciences, The University of Tokyo, Japan
Cytochrome P450foxy (P450foxy, CYP505) is a fused pro-
tein of cytochrome P450 (P450) and its reductase isolated
from the fungus Fusarium oxysporum, which catalyzes the

apparent with P450BM3, a bacterial counterpart of
P450foxy, which was the first obvious difference in their
catalytic properties to be identified. Kinetic data were
consistent with the inhibition due to binding of the second
substrate. We discuss the inhibition mechanism based on
differences between P450foxy and P450BM3 in key amino
acid residues for substrate binding.
Keywords: fatty acid hydroxylase; cytochrome P450;
P450foxy; dodecanoic acid; Fusarium oxysporum.
Cytochrome P450 (P450) is a group of heme proteins that
are widespread in nature [1–3]. It is generally accepted that
all P450s originated from the same, ancient gene (P450
superfamily), which has acquired unparalleled molecular
and functional diversity during evolution [1,4]. Most of the
P450 enzymes function as monooxygenases that act on
various lipophilic compounds, whereas others catalyze a
variety of reactions [2]. P450s can be classified into several
classes according to their redox partners [5,6]. Bacterial and
mitochondrial P450 systems are of class I; they receive
electrons from NAD(P)H via ferredoxin reductase and
ferredoxin coupling. Eukaryotic microsomal P450 systems
are of class II; they receive electrons from P450 reductase,
which contains both FAD and FMN. These two classes
comprise typical, multicomponent P450 monooxygenase
systems, whereas the functions of P450 are most diversified
in other classes. Class III P450s are not monooxygenases
but catalyze isomerization [7] or dehydration [8], and
require neither external redox equivalents nor any redox
partners. P450nor is the only class IV P450, and catalyzes
the reduction of nitric oxide (NO) to nitrous oxide (N

Abbreviations: GC, gas chromatography; GC-EIMS, gas chromato-
graphy-electron impact mass chromatography; P450, cytochrome
P450, rP450foxy, recombinant P450foxy.
(Received 19 November 2001, revised 22 February 2002, accepted 25
February 2002)
Eur. J. Biochem. 269, 2075–2082 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02855.x
like other eukaryotic P450s [14], whereas P450BM3 is a
soluble protein like other bacterial P450s.
Several P450 functions are of interest with respect to
potential industrial applications [16–18] and for basic studies.
The most inconvenient properties of the P450 system when
considering industrial applications would be the complexity
of its electron transport. In vitro P450 function can only be
exhibited in a reconstituted mixture of many components,
yet the activity is usually very low under such conditions. One
way to overcome this obstacle would be to use a fused
protein consisting of P450 and its reductase. Okawa et al.
originally constructed an artificial fused protein using the
yeast expression system and applied the system to the
bioconversion of fine chemicals [19]. Thereafter, P450BM3
[20] and P450foxy [13,14] were identified as naturally fused
proteins. The catalytic turnover of both P450s is exception-
ally high, possibly because they are catalytically self-
sufficient due to the fusion of two domains. The naturally
fused protein would be more useful for such applications and
attempts have made using P450BM3 [21,22].
We have produced recombinant P450foxy (rP450foxy) in
the host-vector system of Saccharomyces cerevisiae [14].
However, a more efficient production system is required for
advancing both basic and application studies of P450foxy.

C
GCCGGGTTATCCGCTT-3¢ and 5¢-TGTTTGCTTG
ATCTCCAAAGCGTAGTT-3¢ (mutated residues are
underlined), which have homology to the nucleotide
sequences of the 5¢ and the 3¢ ends of the P450foxy
cDNA, respectively [14]. PCR products were purified,
digested by NdeIandBamHI, then ligated to the plasmid
vector pCWori+ [23] that had been digested with the same
restriction enzymes. The resulting plasmid was designated
pCWfoxy. Standard DNA techniques proceeded according
to Sambrook et al. [24].
Preparation of recombinant P450foxy
E. coli DH5a was transformed with pCWfoxy, cultured
in LBA overnight, transferred to 2 L of TBA in a 5-L
volume flask, and rotated at 120 r.p.m.
2
at 30 °C.
When D
600
¼ 0.5, 1 m
M
isopropyl thio-b-
D
-galactoside,
0.5 m
M
5-aminolevulinic acid and 1 lgÆmL
)1
chloram-
phenicol (final concentration) were added to the medium

Pharmacia Biotech) equilibrated with buffer A. The sam-
ple after this elution step was used as purified rP450foxy.
Spectroscopy
Optical and fluorescence spectra were measured using a
Beckman DU 7500 spectrophotometer and a Hitachi
F-3010 fluorescence spectrophotometer, respectively. Heme
was identified and determined by the pyridine ferrohemo-
chromogen method using the molar absorption coefficient
(e) of the chromogen of the protoheme as 34.4 m
M
)1
Æcm
)1
at 557 nm [25]. FAD and FMN were identified and
quantified as described by Faeder & Siegel [26] using e at
450 nm ¼ 11.5 m
M
)1
Æcm
)1
. The P450 content was deter-
mined using an extinction coefficient of 91 m
M
)1
Æcm
)1
for
the difference in the carbon monoxide (CO) difference
spectrum between 450 nm and 490 nm [13]. The ratio in the
high/low spin states in the bound heme of P450 was

Data analysis
Steady state kinetic analyses for P450foxy were examined
with varying concentrations of fatty acid and a saturating
2076 T. Kitazume et al. (Eur. J. Biochem. 269) Ó FEBS 2002
concentration of the second substrate (NADPH, 125 l
M
).
Substrate inhibition was observed for longer fatty acid
substrates. Assuming that the inhibition depended on the
second binding of fatty acid substrate at its higher concen-
trations, the data were fitted to Eqn (1) using
ORIGIN
Software, in which v, e, and [S] represent the experimentally
determined initial velocity, and the enzyme and the substrate
(fatty acid) concentrations, respectively. K
s
represents the
dissociation constant for the second binding of S.
v=e ¼ k
cat
½S=fK
m
þ½Sð1 þ½S=K
S
Þg ð1Þ
Determination of the reaction products
The structures of the reaction products (hydroxyl fatty
acids) of rP450foxy were determined fundamentally as
described previously [15] by gas chromatography (GC) and
GC-electron impact-mass spectrometry (GC-EIMS). Prior

in contrast to native P450foxy, which is cofractionated with
themembranefractionofF. oxysporum [13]. The produc-
tion of rP450foxy in the E. coli system was further
confirmed by Western blotting that gave a specific signal
with the predicted M
r
of 118 000 (data not shown) and
dodecanoic acid-dependent NADPH oxidase activity. None
of these properties characteristic of P450foxy were detected
in the extract of E. coli cells that harbored only the vector.
The fraction containing P450 was purified to homogeneity
from the soluble fraction at a yield of 26%. The M
r
estimated by SDS/PAGE and gel filtration
5
was 118 000 and
132 000, respectively, indicating that rP450foxy exists as a
monomer-like native protein and rP450foxy produced by
yeast [14].
Spectral properties
The absorption spectra of purified rP450foxy (Fig. 1) in
its resting oxidized, dithionite-reduced ferrous, and car-
bon monoxide (CO)-ligated forms are identical to the
corresponding spectra of native P450foxy purified from
F. oxysporum [13]. The CO-difference spectrum with a
peak at 448 nm and a trough at 407 nm (Fig. 1, inset)
was also identical. The calculated heme content was
0.4 mol of protoheme per mol of protein. In contrast to
these spectral characteristics due to the bound heme, the
absorbance around 450 nm (characteristic of flavin) was

spectrally (above results) indistinguishable from the native
protein.
Fig. 1. Absorption spectra of rP450foxy. Solid line, resting (oxidized);
dotted line, dithionite-reduced; dashed line, dithionite-reduced + CO.
Inset, CO-difference spectrum (CO-bound minus dithionite-reduced);
7.2 l
M
purified rP450foxy in 100 m
M
sodium phosphate buffer
(pH 7.3), 10% glycerol, at room temperature.
Ó FEBS 2002 Fungal P450foxy catalyzing fatty acid hydroxylation (Eur. J. Biochem. 269) 2077
We determined the substrate specificity of rP450foxy
against saturated fatty acids. Figure 2 shows that rP450foxy
was active against fatty acids with a chain length of C9 (nine
carbon atoms, nonanoic acid) to C18 (18 carbon atoms,
octadecanoic acid), with the activity on tridecanoic acid
(C13) being the highest. These results are similar to those
obtained using the native protein [13], but rP450foxy can
also efficiently use the shorter substrate, nonanoic acid (C9).
Stoichiometry between the consumption of NADPH and
O
2
was 1.3 : 1 with all of the enzymatic reactions on
substrates with chains of C10 to C15 in length (data not
shown), consistent with the theoretical value for 2 electron
reduction coupling to monooxygenation of the substrates.
Interaction of the resting rP450foxy with the substrate
fatty acids
Spectral changes were observed upon mixing the resting

3m
M
. These results showed that the heme in rP450foxy is in
equilibrium between high and low spin states when a fatty
acid substrate binds and that the chain length of the
substrate affects the equilibrium.
Steady-state kinetics
Apparent K
m
and k
cat
values for fatty acid substrates were
determined (Table 2). The K
m
value was in a similar range
(8–36 l
M
) between the substrates from C12 to C16, and
each value approximately agreed with the respective K
d
value for the same substrate. In contrast, the K
m
value
increased with decreasing chain length of the substrate
Table 1. Catalytic activities of native and recombinant P450foxy. Data are mean values of three experiments.
Enzyme
Activity (nmol NADPHÆmin
)1
Ænmol P450
)1

M
Mes
(pH 6.5), 1 l
M
FAD, 1 l
M
FMN, 10% glycerol, at 30 °C] in the
presence of 0, 2.5, 5, 7.5, 10, 15, 20, 30, 40, 50 l
M
pentadecanoic acid,
respectively (lines 1–10). (B) Difference spectra. Each difference spec-
trum was obtained by subtracting line 1 from each of lines 2–10 in A.
2078 T. Kitazume et al. (Eur. J. Biochem. 269) Ó FEBS 2002
below C12 (dodecanoic acid), and became larger than the
K
d
value for the same substrate. The kinetic constants for
fatty acids over C16 could not be obtained because of the
substrate inhibition. The k
cat
value was of the same
magnitude for all substrates tested (1200–1800 min
)1
).
Substrate inhibition occurred for fatty acids with a chain
length of C13 or longer (Fig. 4). The inhibition was
apparent at higher substrate concentration and was eluci-
dated by the mechanism described in Eqn (1), in which a
second substrate binds to form an abortive enzyme-
substrate complex. Such inhibition was not evident when

M
or
lower. The K
m
for NADH was 74 l
M
. These results are
same as those obtained with the native enzyme [13].
Determination of the reaction products
We identified the metabolites (reaction products) of fatty
acids due to the enzymatic reaction of rP450foxy by GC-
EIMS. The results using dodecanoic acid (C12) are shown in
Fig. 5. The derivatives of products were separated into three
peaks on GC (Fig. 5A), each of which gave a fragment
pattern typical of TMSlated alcohol on electron impact mass
chromatography (EIMS)
6
(Fig. 5B–D). The mass number of
each fragment identified these metabolites as x-1, x-2 and
x-3 hydroxy derivatives of dodecanoic acid, respectively.
The ratio of these metabolites was estimated from the signal
intensity on GC (Fig. 6). These results using a C12 fatty acid
are similar to those we obtained using cell-free extracts of
F. oxysporum [15]. However, the present study is the first to
identify the reaction products of purified P450foxy. We also
confirmed that the time-dependent decrease of the substrate
during the enzymatic reaction approximately agreed with
the accompanying increase in the sum of the products (data
not shown). The same sets of experiments were replicated
for, decanoic and undecanoic acids as substrates, and the

M
)
K
s
(l
M
)
k
cat
(min
)1
)
k
cat
/K
m
(min
)1
Æl
M
)1
)
Spin state
high spin(%)
Nonanoic acid 170.0 3200 > 2000 1500 0.5 25
Decanoic acid 8.7 260 > 2000 1200 4.6 23
Undecanoic acid 14.0 160 > 2000 1900 11.5 30
Dodecanoic acid 9.4 30 > 2000 1500 49.1 33
Tridecanoic acid 9.0 36 > 2000 1800 62.7 46
Tetradecanoic acid 2.8 19 1100 1300 68.4 43

m
and K
d
values were similar for longer
fatty acids whereas K
m
became much larger than K
d
for
fatty acids shorter than C12. As K
m
¼ (k
cat
+ k
off
)/k
on
and K
d
¼ k
off
/k
on
,wherek
on
and k
off
are, respectively, the
rate constants for association and dissociation of the
substrate, and k

entrance, the hydrophobic residues, Leu75, Phe87, Leu181,
Ile263, and Leu437, form a hydrophobic stretch in
P450BM3 that allow access to the aliphatic chains of fatty
acids. All of these residues are conserved in P450foxy.
These alignments indicate that all of the key amino-acid
residues at the entrance of the active site pocket of
P450BM3 (Phe42, Arg47, and Tyr51) are replaced by
others (Leu43, Lys48, and Phe52) in P450foxy although all
of other key residues inside the pocket are conserved. The
positive charge essential for fixing the carboxylate of fatty
acids is maintained by replacing Arg with Lys. However,
hydrophobicity at the entrance to P450foxy should be
significantly increased as the result of the substitutions from
Phe to Leu and from Tyr to Phe. Why substrate inhibition is
observed only with P450foxy may be explained by these
substitutions. The first interaction of P450BM3 with fatty
acids may occur between the carboxylate anion of the
Fig. 5. Gas chromatographic separation (A)
and EIMS-spectra of TMSlated and methyla-
ted derivatives of the reaction products (B–D)
from dodecanoic acid. Mass spectra of deriva-
tives 1–3 (A) are shown in B–D, respectively.
Relative abundance of ions in panel D was
expanded by threefold for fragments with high
m/z-values (indicated by horizontal arrow).
Fig. 6. Regio-specificity of reaction products determined with the
substrates, decanoic (C10), undecanoic (C11), and dodecanoic (C12)
acids. Relative amount of each product was determined by the extent
to which corresponding peaks separated on GC. Closed, open,
and striped bars represent x-1, x-2, and x-3 hydroxy fatty acids,

Basic Research Activities for Innovative Biosciences), SBPB (Structural
Biology Sakabe Project) of FAIS (Foundation for Advancement of
International Science), and Grant-in-Aid for Scientific Research from
Ministry of Education, Science, Culture and Sports of Japan.
REFERENCES
1. Omura, T. (1999) Forty years of cytochrome P450. Biochem.
Biophys. Res. Commun. 266, 690–698.
2. Sono, M., Roach, M.P., Coulter, E.D. & Dawson, J.H. (1996)
Heme-containing oxygenases. Chem. Rev. 96, 2841–2887.
3. Nelson, D.R., Koymans, L., Kamataki, T., Stegeman, J.J.,
Feyereisen, R., Waxman, D.J., Waterman, M.R., Gotoh, O.,
Coon, M.J., Estabrook, R.W. et al. (1996) P450 superfamily:
update on new sequences, gene mapping, accession numbers and
nomenclature. Pharmacogenetics 6, 1–42.
4. Nelson, D.R. (1999) Cytochrome P450 and the individuality of
species. Arch. Biochem. Biophys. 369, 1–10.
5. Boddupalli, S.S., Hasemann, A.C., Ravichandran, G.K., Lu, Y.J.,
Goldsmith, J.E., Deisenhofer, J. & Peterson, A.J. (1992) Crystal-
lization and preliminary X-ray diffraction analysis of P450terp and
the hemoprotein domain of P450BM-3, enzymes belonging to two
distinct classes of the cytochrome P450 superfamily. Proc. Natl
Acad.Sci.USA89, 5567–5571.
6. Shimizu, H., Park, Y.S., Lee, S.D., Shoun, H. & Shiro, Y. (2000)
Crystal structures of cytochrome P450nor and its mutants
(Ser286 fi Val, Thr) in the ferric resting state at cryogenic tem-
perature: a comparative analysis with monooxygenase cyto-
chrome P450s. J. Inorgan. Biochem. 81, 191–205.
7. Haurand, M. & Ullrich, V. (1985) Isolation and characterization
of thromboxane synthase from human platelets as a cytochrome
P-450 enzyme. J. Biol. Chem. 260, 15059–15067.

763.
16. Tsujita,Y.,Ikeda,T.,Serizawa,N.,Hishimura,A.&Komai,T.
(1997) Pravastatin sodium: new clinical aspects and recent pro-
gress. Ann. Report Sankyo Res. Lab. 49, 1–61.
17. Shiota, N., Nagasawa, A., Sakaki, T., Yabusaki, Y. & Ohkawa,
H. (1994) Herbicide-resistant tobacco plants expressing the fused
enzyme between rat cytochrome P4501A1 (CYP1A1) and yeast
NADPH-cytochrome P450 oxidoreductase. Plant Physiol. 106,
17–23.
18. Carmichael, A.B. & Wong, L.L. (2001) Protein engineering of
Bacillus megaterium CYP102. The oxidation of polycyclic aro-
matic hydrocarbons. Eur. J. Biochem. 268, 3117–3125.
19. Murakami, H., Yabusaki, Y., Sakaki, T., Shibata, M. & Ohkawa,
H. (1987) A genetically engineered P450 monooxygenase: con-
struction of the functional fused enzyme between rat cytochrome
P450c and NADPH-cytochrome P450 reductase. DNA 6, 189–
197.
20. Narhi, L.O. & Fulco, A.J. (1986) Characterization of a catalyti-
cally self-sufficient 119,000-dalton cytochrome P-450 mono-
oxygenase induced by barbiturates in Bacillus megaterium. J. Biol.
Chem. 261, 7160–7169.
21. Schneider, S., Wubbolts, M.G., Sanglard, D. & Witholt, B. (1998)
Biocatalyst engineering by assembly of fatty acid transport and
oxidation activities for In vivo application of cytochrome
P-450BM-3 monooxygenase. Appl. Environ. Microbiol. 64, 3784–
3790.
22. Schneider, S., Wubbolts, M.G., Oesterhelt, G., Sanglard, D. &
Witholt, B. (1999) Controlled regioselectivity of fatty acid oxida-
tion by whole cells producing cytochrome P450BM-3 mono-
oxygenase under varied dissolved oxygen concentrations.

32. Noble, M.A., Miles, C.S., Chapman, S.K., Lysek, D.A., MacKay,
A.C.,Reid,G.A.,Hanzlik,R.P.&Munro,A.W.(1999)Rolesof
key active-site residues in flavocytochrome P450 BM3. Biochem.
J. 339, 371–379.
2082 T. Kitazume et al. (Eur. J. Biochem. 269) Ó FEBS 2002