An in vivo cytochrome P450cin (CYP176A1) catalytic system for metabolite production

Article abstract of DOI:10.1016/j.molcatb.2012.03.007

Cytochrome P450_cin (CYP176A1) is a bacterial P450 isolated from Citrobacter braakii that catalyses the hydroxylation of 1,8-cineole to (1R)-6β-hydroxycineole. P450_cin uses two redox partners in vitro for catalysis: cindoxin, its physiological FMN-containing redox partner, and Escherichia coli flavodoxin reductase. Here we report the construction of a tricistronic plasmid that expresses P450_cin, cindoxin and E. coli flavodoxin reductase and a bicistronic plasmid that encodes only P450 _cin and cindoxin. E. coli transformed with the bicistronic vector effectively catalysed the oxidation of 1,8-cineole, with the endogenous E. coli flavodoxin reductase presumably acting as the terminal electron transfer protein. This in vivo system was capable of producing enantiomerically pure (1R)-6β-hydroxycineole in yields of ～1 g/L culture, thus providing a simple, one-step synthesis of this compound. In addition, the metabolism of (1R)- and (1S)-camphor, structural homologues of 1,8-cineole was also evaluated in order to investigate the ability of this in vivo system to produce compounds for mechanistic studies. Significant quantities of five of the six possible secondary alcohols arising from methylene oxidation of both (1R)- and (1S)-camphor were isolated and structurally characterised. The similarity of the (1R)- and (1S)-camphor product profiles highlight the importance of the inherent reactivity of the substrate in determining the regiochemistry of oxidation in the absence of any specific enzyme-substrate binding interactions.

Full text of DOI:10.1016/j.molcatb.2012.03.007

Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
An in vivo cytochrome P450cin (CYP176A1) catalytic system for metabolite
production
Kate E. Slessor, David B. Hawkes, Anthony Farlow, Andrew G. Pearson, Jeanette E. Stok∗,
James J. De Voss∗∗
Department of Chemistry, School of Chemistry and Molecular Biosciences, University of Queensland, St. Lucia, Brisbane QLD 4072, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
Cytochrome P450cin (CYP176A1) is a bacterial P450 isolated from Citrobacter braakii that catalyses the
hydroxylation of 1,8-cineole to (1R)-6␤-hydroxycineole. P450cin uses two redox partners in vitro for catal-
ysis: cindoxin, its physiological FMN-containing redox partner, and Escherichia coli flavodoxin reductase.
Here we report the construction of a tricistronic plasmid that expresses P450cin, cindoxin and E. coli flavo-
doxin reductase and a bicistronic plasmid that encodes only P450cin and cindoxin. E. coli transformed
with the bicistronic vector effectively catalysed the oxidation of 1,8-cineole, with the endogenous E. coli
flavodoxin reductase presumably acting as the terminal electron transfer protein. This in vivo system
was capable of producing enantiomerically pure (1R)-6␤-hydroxycineole in yields of ∼1 g/L culture, thus
providing a simple, one-step synthesis of this compound. In addition, the metabolism of (1R)- and (1S)-
camphor, structural homologues of 1,8-cineole was also evaluated in order to investigate the ability of
this in vivo system to produce compounds for mechanistic studies. Significant quantities of five of the
six possible secondary alcohols arising from methylene oxidation of both (1R)- and (1S)-camphor were
isolated and structurally characterised. The similarity of the (1R)- and (1S)-camphor product profiles
highlight the importance of the inherent reactivity of the substrate in determining the regiochemistry of
oxidation in the absence of any specific enzyme–substrate binding interactions.
Received 19 January 2012
Received in revised form 8 March 2012
Accepted 16 March 2012
Keywords:
Cytochrome P450
In vivo
Escherichia coli
Cineole
Camphor
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
in whole cells can be employed to overcome some of these
barriers.
Cytochromes P450 enzymes (P450s) form a super family of
versatile catalytic hemoproteins that can activate molecular oxy-
gen, allowing them to catalyse a remarkable array of interesting
oxidative transformations. These include reactions such as the
hydroxylation of non-activated hydrocarbons, epoxidation, het-
eroatom oxidation and carbon–carbon bond cleavage [1]. The
ability of P450s to catalyse the often regio-, stereo- and enantiospe-
cific hydroxylation of non-activated hydrocarbons, a significant
challenge for traditional synthetic means, makes them attrac-
tive targets both as biocatalysts for organic synthesis and for
mechanistic investigations [2–4]. Despite their oxidative poten-
tial, the application of P450s as biocatalysts is limited for a
number of reasons. These include: substrate specificity, reaction
rates, the required multi-component electron transfer proteins
and the stoichiometric amounts of expensive reducing cofactors
(NAD(P)H) needed when using purified or semi-purified enzymes.
Heterologous coexpression of the P450 and its redox partners
Over-expression of both the P450 and its reductase partners
in whole cells has been extensively explored with a variety of
P450s from a number of different organisms [5–13]. Two bac-
terial examples that investigate the bio-catalytic potential of
these whole cell systems include P450cam [5,8,12] and P450BM-3
[10,11,14,15]. Cytochrome P450cam catalyses the hydroxylation
of (1R)-camphor to 5-exo-hydroxycamphor. It has been demon-
strated that the expression of P450cam, putidaredoxin (Pdx) and
putidaredoxin reductase (PdR) in whole cells either individually
(tricistronic plasmid) [5,8] or as a fusion-protein [12] can effec-
tively catalyse camphor oxidation. This P450cam/Pd/PdR whole-cell
system has been investigated with non-natural substrates includ-
ing limonene [8]. P450BM-3, a natural fusion of a P450 and its
reductase components, has also been demonstrated to convert a
variety of long-chain saturated fatty acids to ␻-1, ␻-2 and ␻-
3 hydroxy fatty acids in whole cells [10,11]. Such systems have
also been used to produce minor metabolites for structure elu-
cidation, leading to important insights into the mechanism of
P450 catalysed oxidations [16]. In vivo systems of both P450cam
and P450BM-3 have also been engineered to enhance oxidation of
non-natural substrates by introducing specific mutations [5,16].
For example, the P450cam Y96F-V247L mutant in vivo system
∗
Corresponding author. Tel.: +61 7 3365 3795; fax: +61 7 3365 4273.
Corresponding author. Tel.: +61 7 3365 3825; fax: +61 7 3365 4273.
∗∗
E-mail addresses: j.stok1@uq.edu.au (J.E. Stok), j.devoss@uq.edu.au (J.J. De Voss).
1381-1177/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2012.03.007

16
K.E. Slessor et al. / Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
treated with DNA polymerase I (Klenow fragment) to create blunt
ends and re-ligated to form pCW-P450cin/Cdx.
O
O
O
HO
2.3. Construction of pCW-P450cin/Fpr/Cdx
O
1
4
2
pET11a-Fpr [22] was digested with XbaI and BamHI to excise
the Fpr gene. The fragment was treated with DNA polymerase I and
ligated into pCW-P450cin/Cdx which had been digested with BstBI
and similarly treated with DNA polymerase I. Correct orientation
of the Fpr gene within the construct was achieved using HindIII
digests to give pCW-P450cin/Fpr/Cdx.
Scheme 1. P450cin oxidation of cineole 1 to produce (1R)-6␤-hydroxycineole 2.
Under certain conditions, over-oxidation by P450cin can result in the production
of ketocineole 4.
O
3
O
6
5
2.4. Evaluation of polycistronic constructs: expression
(1R)-3
(1S)-3
A polycistronic plasmid (either pCW-P450cin/CdR/Cdx, pCW-
P450cin/Cdx or pCW-P450cin/Fpr/Cdx) was transformed into E. coli
DH5␣FꢀIQ and used to inoculate Terrific broth (1 L in 2.8 L
Fernbach flask) containing ampillicin (50 ␮g/mL). This culture
was incubated at 37 ◦C (approx. 180 rpm; Innova 4000, New
Brunswick Scientific) until an OD600 of approximately 0.6 was
attained and protein expression was induced with isopropyl ␤-
D-1-thiogalactopyranoside (1 mM). The culture was incubated at
27.5 ◦C for a further 18 h and the cells harvested by centrifugation
(4000 × g). To determine the relative amounts of each protein a
crude purification was performed. The cells were lysed by soni-
cation (Branson sonifier 450; 6 × 1 min; 50% output) in Buffer A
(50 mM Tris·HCl pH 7.4, 50 mM KCl, 1 mM EDTA and 0.5 mM DL-
dithiothreitol) containing 0.1 mM phenylmethylsulfonyl fluoride
and the cellular debris removed (15,000 × g). The supernatant was
loaded onto a DEAE ion-exchange column, washed with Buffer A
and each protein eluted during a 50–500 mM KCl gradient in Buffer
A. The individual proteins were pooled and the amount of protein
Fig. 1. The two enantiomers of camphor (1R) and (1S)-3 used in this study.
was observed to successfully catalyse the turnover of styrene and
ethylbenzene [5].
P450cin (CYP176A1) is a soluble, microbial P450 isolated from
Citrobacter braakii that catalyses the hydroxylation of cineole 1 to
(1R)-6␤-hydroxycineole 2 (Scheme 1) [17]. This hydroxylation step
is believed to be the first in the metabolism of cineole 1 [18] by
C. braakii allowing it to survive on cineole as its sole carbon and
energy source [17]. Reconstitution of P450cin activity in vitro has
been achieved with cindoxin (Cdx), its native redox partner, and
Escherichia coli flavodoxin reductase (Fpr) [19]. We report here the
construction of polycistronic plasmids encoding the genes of this
P450cin system. These constructs enabled the preparative scale bio-
production of enantiopure (1R)-6␤-hydroxycineole 2 in a single
oxidative step from cineole 1, mimicking the natural transforma-
tion catalysed by P450cin. We also demonstrate that this system
has potential for the in vivo transformation of compounds that are
much poorer substrates for P450cin than cineole, which should be
of significant value for mechanistic investigation. The in vivo oxida-
tion of the (1R)- and (1S)-camphor 3 (Fig. 1) was demonstrated to
yield sufficient quantities of metabolites for structural elucidation,
even in cases where the non-natural substrates produced complex
metabolite mixtures. Development of this in vivo P450cin system
may also be a useful tool in producing antimicrobial, bactericidal
agents [20], or fragrances in the perfume industry [21].
estimated by UV absorption. P450cin 415 nm (ε = 150 cm−1 mM−1
)
[17], Cdx 456 nm (ε = 10,825 cm−1 M−1) [19] and Fpr at 456 nm
(ε = 7100 cm−1 M−1) [23].
2.5. Evaluation of polycistronic constructs: cineole oxidation
Protein expression was performed as outlined above with the
following modifications. The cell culture (1 L in 2.8 L Fernbach
flask) was supplemented with cineole 1 (0.5 mL/L) concurrently
to induction with isopropyl ␤-d-1-thiogalactopyranoside (1 mM).
The cultures were incubated at 27.5 ◦C for 43 h (approx. 180 rpm;
Innova 4000, New Brunswick Scientific). The cells were removed
by centifugation, the supernatant extracted with ethyl acetate
and dried over MgSO4. This was followed by GC/MS analysis
as has previously been described [17]. ␣-Terpineol was used as
an internal standard to examine the progress of the oxidation
over time.
2. Experimental
2.1. Construction of pCW-P450cin/CdR/Cdx
pUC19-Cdx [19] was digested with SphI and BamHI, effec-
tively removing one PstI site from the multiple cloning site, and
retaining the second PstI site found within the Cdx gene. This plas-
mid was treated with DNA polymerase I (Klenow fragment) to
create blunt ends and re-ligated. A PstI–PstI 4 kb fragment gener-
ated from pC1 [17] (containing P450cin, CdR and 98% of the Cdx
gene) was ligated into the altered PstI cut pUC19-Cdx to form
pUC19-C3. The PstI fragment was oriented using PpuMI/HindIII
which removes a 3 kb fragment if inserted correctly. This 3 kb
PpuMI/HindIII fragment was then inserted into pUCPKS-cinA [17]
which was similarly digested with PpuMI/HindIII to create pUC-
PHC3. pUC-PHC3 was then digested with NdeI/HindIII and the
fragment cloned into similarly cut pCW to yield the tricistronic
plasmid pCW-P450cin/CdR/Cdx.
2.6. In vitro catalytic turnover
A solution of P450cin (0.5 mM), Cdx (4 mM) and Fpr (1 mM) in
Tris·HCl buffer (50 mM, pH 7.4) was prepared and the substrate
was added in excess (5 mM in ethanol). Turnover was initiated
by the addition of NADPH (0.2–5 mM) and the reaction incubated
at room temperature with stirring for 30–60 min. The solution
was then extracted into ethyl acetate, dried over MgSO4 and ana-
lysed by GC/MS [17]. Calculation of the dissociation constant (Kd),
NADPH consumption and coupling have all been outlined else-
where [17,19,24]. In this study they were all performed in 100 mM
potassium phosphate, pH 7.4. The percentage spin state change
was calculated from the maximum absorbance shift observed in
the presence of substrate (5 mM final concentration; 417–392 nm)
and standardised against P450cin with cineole.
2.2. Construction of pCW-P450cin/Cdx
pCW-P450cin/CdR/Cdx was digested with BstBI and the 800 bp
fragment containing the CdR gene discarded. The vector was

K.E. Slessor et al. / Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
17
2.7. Large scale oxidation of cineole 1 (pCW-P450cin/Cdx)
93 (42), 70 (15), 55 (36), 41 (100). This is consistent with literature
data [27].
P450cin and Cdx were expressed as outlined above with the plas-
mid pCW-P450cin/Cdx and the following modifications. Cineole 1
(0.6 mL/L broth) was added to the culture (1 L in 2.8 L Fernbach
flask) approximately every 12 h (approx. 180 rpm; Innova 40, New
Brunswick Scientific). After three days of expression, a total of
11.6 mL of 1 had been added to 3.5 L of culture. The cells were pel-
leted at 4000 × g for 30 min and the supernatant collected. NaCl
was added to the supernatant, extracted with ethyl acetate, filtered
through celite and dried over MgSO4. The solvent was removed
in vacuo and the crude material purified by flash chromatography
(30% ethyl acetate in hexane) to yield (1R)-6␤-hydroxycineole 2
(3.1 g, 26%). Spectral data matched the literature [24]. m.p. 91–92 ◦C
(Lit. [24] m.p. 90–91 ◦C) [␣]D +29.1 ◦C (Lit. [24] [␣]D −26.0 ◦C for
enantiomer).
6-endo-Hydroxycamphor 7: 1H NMR (500 MHz, CDCl3) ı 0.80
(3H, s), 0.95 (6H, s), 4.15 (1H, dd, J = 9.5 Hz, 2.5 Hz). 13C NMR
(125 MHz, CDCl3) ı 63.9. MS (70 eV) m/z (%) 168 (M+, 5), 153 (29),
135 (6), 125 (5), 108 (33), 107 (33), 93 (29), 81 (14), 69 (24), 55 (55),
41 (100). This is consistent with the literature [28].
5-exo-Hydroxycamphor 8: 1H NMR (500 MHz, CDCl3) ı 0.81 (s,
3H), 0.89 (s, 3H), 1.22 (s, 3H), 3.98 (1H, dd, J = 3.5 Hz, 7.5 Hz). 13C
NMR (125 MHz, CDCl3) ı 74.3. MS (70 eV) m/z (%) 168 (M+, 2), 153
(7), 125 (18), 111 (49), 83 (20), 69 (33), 55 (56), 41 (100). This is
consistent with the literature [27].
3. Results and discussion
Three polycistronic plasmids were constructed that contained
combinations of the P450cin, Cdx, cindoxin reductase (CdR) and
E. coli Fpr genes (Table 1). The pCW vector was chosen to con-
tain the polycistronic constructs as it has previously been shown to
be effective for heterologous P450 expression in E. coli. In particu-
lar, it proved effective for the expression of both P450cin and Cdx
[17,19]. The first construct investigated, pCW-P450cin/CdR/Cdx,
was arranged identically to the native arrangement of the operon
in C. braakii with expression of the encoded proteins inducible by
the addition of IPTG [17]. It was hoped that this arrangement, using
the natural upstream ribosome binding site and in the presence of
P450cin and Cdx, may enable the expression of active CdR. CdR is
the proposed natural terminal redox partner which has yet to be
expressed in an active form despite significant effort [19]. How-
ever, SDS-PAGE analysis of crude cell lysates of E. coli transformed
with this plasmid and induced with IPTG, indicated expression of
the CdR gene only resulted in polypeptide that accumulated in the
insoluble fraction. Despite this, the construct was able to oxidise
cineole 1 to (1R)-6␤-hydroxycineole 2 effectively as determined by
GC/MS analysis of culture extracts (Table 1). We hypothesise that
this oxidation is supported by endogenous Fpr in E. coli, mimicking
the in vitro system we had previously reported [19].
To establish that a small amount of CdR was not being expressed
in active form and hence supporting the observed P450cin activ-
ity, a second plasmid, pCW-P450cin/Cdx, was constructed in which
approximately 75% of the CdR gene had been deleted. Expression
in E. coli produced a system that was again capable of cineole
oxidation and with a similar efficiency to that observed with the
pCW-P450cin/CdR/Cdx tricistronic construct. Thus, it appeared that
cineole 1 oxidation observed in vivo is supported by the endogenous
Fpr from the E. coli host.
2.8. Large scale oxidation of (1R) and (1S)-camphor 3
(pCW-P450cin/Cdx)
Oxidation of (1R) and (1S)-3 were performed in an identical fash-
ion. Camphor 3 (0.3 g/L broth) was added to the culture (1 L in 2.8 L
Fernbach flask) approximately every 12 h (approx. 180 rpm; Innova
40, New Brunswick Scientific). After three days of expression, a
total of 1 g/L of 3 had been added to 3.5 L of culture. (1R)-camphor:
The crude material from ethyl acetate extraction of the culture was
purified by flash chromatography in sequential stages to yield the
hydroxycamphors. Firstly, a gradient of 5–100% ethyl acetate in
hexane was used to fractionate the crude mixture. Two pools were
collected: Pool 1 contained (1R)-3 (1.1 g, 31% recovery of start-
ing material); Pool 2 contained several hydroxycamphors. These
were further purified by flash chromatography (30% ethyl acetate
in hexane). Three mixed fractions containing hydroxycamphors
were isolated and the structures of their components identified
by GC/MS, 1H and 13C NMR. Fraction 1 (70 mg) contained a 4:1
mixture of 3-exo-hydroxycamphor 5 to 3-endo-hydroxycamphor
6. Fraction 2 (2 mg) was identified as 6-endo-hydroxycamphor 7.
Fraction 3 (85 mg) contained 5-exo-hydroxycamphor 8, 5-endo-
hydroxycamphor 9 and 6-endo-hydroxycamphor 7 in a 3:3:1 ratio,
respectively. The yield of hydroxycamphors was approximately 6%
based on recovered starting material.
2.8.1. Key data for metabolite identification
3-exo-Hydroxycamphor 5: 1H NMR (500 MHz, CDCl3) ı 0.93 (3H,
s), 0.95 (3H, s), 0.99 (3H, s), 3.74 (1H, s). This is consistent with
the literature [25]. 3-endo-Hydroxycamphor: 1H NMR (500 MHz,
CDCl3) ı 0.88 (3H, s), 0.94 (3H, s), 1.01 (3H, s), 4.20 (1H, dm,
J = 5.0 Hz). This is consistent with the literature [26]. 5 and 6 co-
eluted by GC. MS (70 eV) m/z (%) 168 (M+, 1.5), 153 (1.3), 140 (2.9),
125 (21), 100 (10), 84 (61), 83 (70), 55 (88), 41 (100).
With this result in mind, a third plasmid, pCW-P450cin/Fpr/Cdx,
was constructed in order to supplement native Fpr expression in
E. coli and ensure that P450 activity was not limited by low lev-
els of endogenous Fpr within the cell. Surprisingly, however, poor
oxidation of 1 was observed in this whole-cell system (Table 1).
Employing both SDS-PAGE and estimating protein content by UV
it was observed that over-expression of Fpr had occurred. How-
ever, it effectively diminished Cdx expression (Table 1), presumably
in part by competition for flavin cofactors. Interestingly, P450cin
5-endo-Hydroxycamphor 9: 1H NMR (500 MHz, CDCl3) ı 0.82
(3H, s), 0.83 (3H, s), 0.97 (3H, s), 4.60 (1H, apparent dtd, J = 10 Hz,
4 Hz, 2 Hz). 13C NMR (125 MHz, CDCl3) ı 69.3. MS (70 eV) m/z (%)
168 (M+, 2), 153 (37), 150 (0.5), 135 (6), 125 (2), 108 (41), 107 (39),
Table 1
Evaluation of polycistronic constructs.
Construct
P450cin (nmol/La)
CdR/Fpr (nmol/L)
Cdx (nmol/L)
Cineole oxidation
pCW-P450cin/CdR/Cdx
pCW-P450cin/Cdx
pCW-P450cin/Fpr/Cdx
560
510
380
NMb
NAc
4000
1700
2000
NM
Yes
Yes
Poor
a
nmol per litre of culture expressed.
Not measurable.
Not applicable.
b
c

18
K.E. Slessor et al. / Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
H
O
O
O
H
HO
OH
H
OH
(1R)-6
(1R)-5
(1R)-7
O
O
H
OH
H
(1R)-8
OH
(1R)-9
Fig. 3. Hydroxycamphors ((1R)-5–9) identified from P450cin in vivo turnover of (1R)-
camphor 3.
Fig. 2. Monitoring (1R)-6␤-hydroxycineole 2 production by GC/MS using the in vivo
system expressing both P450cin and Cdx.
structural homologues of the natural substrate cineole 1, as model
systems. We had previously reported that the enantioselectivity of
cineole oxidation was largely controlled by a hydrogen bond from
the ethereal oxygen of 1 to asparagine 242 of P450cin [29,30]. The
regiochemistry of oxidation of (1R)-camphor 3 by P450cam is simi-
larly controlled by a hydrogen bond between the camphor carbonyl
and a protein residue, tyrosine 96 [31]. This similarity in mecha-
nism of regiocontrol led us propose that there may be interesting
differential selectivity in the oxidation of (1R)- and (1S)-camphor
concentration was also reduced during expression using the pCW-
P450cin/Fpr/Cdx construct. Hence, the native Fpr expression in
E. coli must be produced at the correct concentration to adequately
support efficient in vivo oxidation.
The bicistronic pCW-P450cin/Cdx construct was then used to
evaluate the biocatalytic viability of this system to produce useful
quantities of oxidised metabolites. Naturally, the oxidation of cine-
ole 1 to (1R)-6␤-hydroxycineole 2 was investigated initially. Due to
the poor aqueous solubility and volatility of 1, aliquots were added
at approximately 12 h intervals to an E. coli culture expressing both
P450cin and cindoxin from pCW-P450cin/Cdx to ensure that an ade-
quate concentration of substrate was maintained. The production
of 6␤-hydroxycineole 2 was monitored by GC/MS analysis of cul-
ture extracts, employing ␣-terpineol as an internal standard (Fig. 2).
This revealed that the bicistronic system was actively oxidising sub-
strate for approximately 40 h, after which oxidation ceased. The
6␤-hydroxycineole 2 was isolated and purified by extraction and
subsequent column chromatography to yield approximately 1 g per
litre of bacterial culture (26% based upon added 1). The 1H and
13C NMR data for 6␤-hydroxycineole 2 generated enzymatically
matched that of the authentic standard synthesised previously [24].
Enantioselective GC analysis revealed that only the (1R)-2 enan-
tiomer had been formed. A small amount of (1R)-ketocineole 4
was also observed during the in vivo turnover of 1, which pre-
sumably results from the oxidation of 2 (Scheme 1). This type of
over-oxidation has also been observed in an analogous in vivo sys-
tem using P450cam/PdR/Pdx where 5-oxocamphor was produced
ultimately from camphor (1R)-3 [5].
(1R)-6␤-hydroxycineole 2 can be synthesised chemically by
the treatment of ␣-(−)-terpineol with m-CPBA followed by treat-
ment with an acid catalyst [24]. However, the somewhat variable
enantiomeric purity of the starting material and the formation of
unwanted side products make large scale production via this route
unattractive. The use of P450cin as a biocatalyst shows significant
potential as a new synthetic tool to generate enantiomerically pure
(1R)-6␤-hydroxycineole 2. Although cineole 1 itself is a meso com-
pound, hydroxylation by P450cin occurs only at the pro-R carbon
generating enantiomerically pure (1R)-6␤-hydroxycineole 2 with
three new stereogenic centres.
by P450cin
.
We first evaluated the interaction of (1R)-and (1S)-camphor 3
with P450cin in vitro and showed that they had the ideal character-
istics for further work in vivo. Both (1R)- and (1S)-camphor 3 are
able to access the active site of P450cin with reduced affinity for
the enzyme as compared to cineole 1 (over 100 fold difference in
Kd; Table 2). (1R)- and (1S)-3 were also found to partially displace
the water molecule normally coordinated to the heme in the rest-
ing active site of P450cin as monitored by the change in the Soret
absorbance (from 417 nm to 392 nm); 27 2% ((1R)-3) or 32 1%
((1S)-3) efficient as cineole 1. Catalytic turnover was investigated
and it was shown that the rate of NADPH consumption by P450cin
in the presence of (1R)- and (1S)-3 was approximately a third of
that observed with cineole 1 (Table 2). Additionally, only a quarter
to a third of the NADPH consumed during (1R)- or (1S)-3 oxidation
by P450cin was directed to the production of a somewhat surpris-
ingly complex mixture of hydroxycamphors. In sum, both (1R)- and
(1S)-3 was clearly a much poorer substrate for P450cin than cineole
1. Based on these observations the catalytic turnover of camphor
3 appeared ideal for investigation with the in vivo P450cin sys-
tem both to test whether it could produce metabolites in sufficient
quantities for structure elucidation and to provide information as
to the selectivity of camphor oxidation.
Oxidation of camphor 3 via a typical P450 catalysed oxidation
reaction, oxygen insertion into a C-H bond, has the potential to
generate ten different hydroxyl camphors. There are three methyl
groups, three methylene groups with six diastereomeric positions
and a single methine position that can be hydroxylated. Oxidation
of (1R)-camphor 3 by P450cin using the whole cell oxidation sys-
tem described above generated at least seven different oxidised
(1R)-camphor 3 metabolites (Data not shown). The GC/MS trace
of the crude extract revealed six different metabolites with m/z
168, which is expected for a hydroxycamphor as well as a peak
with m/z 166, required for an oxocamphor; the latter would result
from a second oxidation of a hydroxycamphor. The products of
in vivo transformation were extracted from the culture medium
into ethyl acetate and partially purified by column chromatogra-
phy. This resulted in the isolation of the five major (1R)-camphor 3
metabolites (Fig. 3, 5–9) in three different pools in sufficient quan-
tities to be unambiguously identified by a combination of GC/MS,
GC and NMR.
3.1. Oxidation of unnatural substrates: (1R)- and (1S)-camphor
It was also important to determine if this in vivo P450cin cat-
alytic system could be useful for compounds that were much
poorer substrates than cineole 1. Although not valuable synthet-
ically, this would be useful for metabolite production for structural
elucidation in the investigation of P450 mechanism and struc-
tural determinants of oxidation. This possibility was evaluated
by exploring the in vivo oxidation of (1R)- and (1S)-camphor 3,

K.E. Slessor et al. / Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
19
Table 2
In vitro comparison of catalytic turnover of cineole and (1R)- and (1S)-camphor by P450cin
a
.
Spin changeb (%)
Kd (␮M)
NADPH consumption (%)
100
Couplingc (%)
Cineole
(1R)-camphor
(1S)-camphor
100
0.7 [17]
80
35
25
2
2
3
27
32
2
1
101
81
4
5
33
29
1
1
a
See Section 2 for condition details.
b
Both the percentage spin state and the rate of NADPH consumption were standardised against cineole with P450cin (assuming P450cin:cineole is 100%).
Ratio of the amount of NADPH consumed compared to the amount of product formed.
c
GC/MS analysis of Pool 1 revealed only one peak with an m/z 168.
However, by 1H NMR it was apparent that two different hydrox-
ycamphors were present in a 4:1 ratio: there were six signals
corresponding to methyl groups and two downfield signals (ı 4.17
and 3.70 ppm) for hydroxymethine positions. The simplicity of the
downfield signals suggested that hydroxylation had occurred on
the parent camphor at C3, as this has few available vicinal coupling
partners. The major isomer was 3-exo-hydroxycamphor 5, provid-
ing the resonance at ı 3.70 ppm (s, H3); the absence of coupling is in
accord with the dihedral angle of almost 90◦ to H4. The minor iso-
mer was identified as 3-endo-hydroxycamphor 6 and corresponded
to the resonance at ı 4.17 ppm (dd J 5.2, 1 Hz, H3). These assign-
ments were confirmed by comparison of full NMR data with the
literature [32,33] (ESI). The epimeric 3-hydroxycamphors 5 and 6
were shown to be resolved when analysed by enantioselective GC
(␤-cyclodextrin), with 5 eluting first.
Pool 2 was found to contain only 6-endo-hydroxycamphor 7,
characteristically displaying a hydroxmethine at ıH 4.14 ppm (ddd J
9.5, 2.5, 1 Hz, H6), the smallest coupling corresponding to long
range coupling to H4. This assignment was confirmed by the cor-
respondence of 1H and 13C data with literature data (ESI) [28,34].
Three different hydroxycamphors were found in Pool 3, includ-
ing a portion of 7 identified above. The remaining two compounds
were readily identified by their diagnostic NMR data (ESI) as 5-exo-
hydroxycamphor, 8 and 5-endo- hydroxycamphor, 9. In particular,
8 exhibits a hydroxymethine resonance at ıH 3. 96 (dd J 7.2, 4.4 Hz)
whilst 9 is characterised by a complex signal at d 4.58 (dddd J 9.5,
3.5, 3.5,1.5 Hz). Full NMR data of each compound determined on
the mixture was in good agreement with that reported previously
(ESI) [27].
Fig. 4. GC/MS trace comparing the oxidation products of (1R)- and (1S)-camphor 3
of catalytic turnover by P450cin
.
O
O
42%
39%
4% 5%
14%
32%
(1R)-camphor
19%
7%
21% 17%
With these results in hand, we investigated the P450cin
-
(1S)-camphor
catalysed oxidation of (1S)-camphor 3. The hydroxycamphors 5–9
isolated above would of course provide standards for (1S)-camphor
metabolites analysed under non-enantioselective conditions. (1S)-
3 was oxidised under in vitro and in vivo conditions as described
above and the products analysed by GC/MS. Unexpectedly, the
product profile was remarkably similar to that observed with (1R)-
3 (Fig. 4). The major site for hydroxylation is at C3 in both (1R)-
and (1S)-camphor 3 (approximately 50% of total oxidation) with
the 3-exo-product ((1R)- and (1S)-5) the dominant isomer in both
instances. The remainder of the oxidation products were mainly 5-
and 6-endo-hydroxycamphors ((1R)- and (1S)-7 and 9) with a small
amount of 5-exo-hydroxycamphor ((1R)- and (1S)-8) (Fig. 5).
The similarity in the P450cin metabolite profiles from (1R)- and
(1S)-camphor and the large number of products formed strongly
suggest that these substrates are highly mobile in the active site
of the enzyme. Clearly, the H-bond formed to the ethereal oxy-
gen of cineole bound to P450cin cannot be replicated in a similar
interaction with the carbonyl of camphor, despite such an interac-
tion controlling the regiochemstry of (1R)-camphor oxidation by
Fig. 5. Oxidation profile of both (1R)- and (1S)-camphor by the in vivo P450cin
expression system.
from hydrogen atom abstraction. The major site of hydroxylation of
both enantiomers of camphor is C3 and one rationalisation for this
lies in the stabilities of the radicals produced: hydrogen abstrac-
tion at C3 is expected to generate the most stable carbon radical
due to delocalisation of the unpaired electron into the neighbour-
ing carbonyl group [35]. In contrast, the hydroxylation of (1R)- and
(1S)-norcamphor hydroxylation by P450cam [36] primarily occurs
at C5 in the (1R)-isomer and at C6 in the (1S)-isomer. This is pro-
posed to be due to a hydrogen bond between the substrate ketone
and Y96 of P450cam, controlling the substrate orientation in the
enzyme–substrate complex. The apparent absence of such an ori-
enting H-bond in P450cin means the reactivity of the substrate and
stability of the intermediates become more important in determin-
ing the regiochemical outcome of the reaction.
P450cam
.
4. Conclusion
The regiochemical outcome of P450 catalysed oxidation is
dependent not only on the substrate’s orientation in the active site
but also on the inherent reactivity of the substrate itself (essentially
Several vectors for the polycistronic coexpression of P450cin
and cindoxin with or without a terminal redox partner were
constructed. The bicistronic plasmid containing only P450cin and
C
H bond strength), or, the relative stabilities of the radicals formed

20
K.E. Slessor et al. / Journal of Molecular Catalysis B: Enzymatic 79 (2012) 15–20
cindoxin proved to be the most effective in the whole cell oxidation
of cineole. When expressed in E. coli this plasmid was capable of oxi-
dising cineole 1 to (1R)-6␤-hydroxycineole 2 presumably utilising
E. coli’s native flavodoxin reductase as the terminal redox protein.
This system could produce approximately 1 g/L culture over a 48 h
incubation and thus provides a simple and convenient way to make
enantiomerically pure (1R)-6␤-hydroxycineole 2 on a preparative
scale. The oxidation of unnatural substrates was also investigated
with a view to developing this in vivo P450cin system as a biocata-
lyst and a tool for investigating P450 mechanism. P450cin oxidised
both (1R)- and (1S)-camphor with little regiospecificity but suffi-
cient material was isolated from in vivo transformation to allow
structural characterisation of the products. The metabolite profiles
from the oxidation of (1R)- and (1S)-camphor were surprisingly
similar. This highlights the importance of the inherent reactivity of
the molecule in determining the regiochemical outcome of P450
catalysed oxidation when substrate orientation within the active
site is not controlled.
In the future, construction of a bicistronic vector with the
recently described P450cin-N242A mutant [37], which introduces
oxygen at the pro-S carbon (90% ee), may enable the prepara-
tive scale production of the enantiomeric (1S)-6␤-hydroxycineole
2. The ease with which preparative amounts of enantiopure
hydroxycineole from the original P450cin bicistronic system can be
generated provides easy access to a large range of cineole deriva-
tives via functionalisation of the newly added hydroxyl moiety. We
plan to make use of such derivatives as probes for studying the
mechanism of P450s but they may also be useful as antimicrobial
or bactericidal agents [20], or as fragrances in the perfume industry
[21].
[2] G. Grogan, Curr. Opin. Chem. Biol. 15 (2011) 241–248.
[3] V.B. Urlacher, S. Eiben, Trends Biotechnol. 24 (2006) 324–330.
[4] V.B. Urlacher, M. Girhard, Trends Biotechnol. 30 (2012) 26–36.
[5] S.G. Bell, C.F. Harford-Cross, L.L. Wong, Protein Eng. 14 (2001) 797–802.
[6] J.A.R. Blake, M. Pritchard, S.H. Ding, G.C.M. Smith, B. Burchell, C.R. Wolf, T.
Friedberg, FEBS Lett. 397 (1996) 210–214.
[7] E.M.J. Gillam, R.M. Wunsch, Y.F. Ueng, T. Shimada, P.E.B. Reilly, T. Kamataki, F.P.
Guengerich, Arch. Biochem. Biophys. 346 (1997) 81–90.
[8] D. Kim, P.R. Ortiz de Montellano, Biotechnol. Lett. 31 (2009) 1427–1431.
[9] A. Parikh, E.M.J. Gillam, F.P. Guengerich, Nat. Biotechnol. 15 (1997)
784–788.
[10] S. Schneider, M.G. Wubbolts, D. Sanglard, B. Witholt, Appl. Environ. Microbiol.
64 (1998) 3784–3790.
[11] S. Schneider, M.G. Wubbolts, D. Sanglard, B. Witholt, Tetrahedron: Asymmetry
9 (1998) 2833–2844.
[12] O. Sibbesen, J.J. DeVoss, P.R. Ortiz de Montellano, J. Biol. Chem. 271 (1996)
22462–22469.
[13] S.G. Bell, A. Dale, N.H. Rees, L. Wong, Appl. Microbiol. Biotechnol. 86 (2010)
163–175.
[14] R. Fasan, N.C. Crook, M.W. Peters, P. Meinhold, T. Buelter, M. Landwehr, P.C.
Cirino, F.H. Arnold, Biotechnol. Bioeng. 108 (2011) 500–510.
[15] H. Schewe, D. Holtmann, J. Schrader, Appl. Microbiol. Biotechnol. 83 (2009)
849–857.
[16] C. Whitehouse, N. Rees, S. Bell, L. Wong, Chem. Eur. J. 17 (2011)
6862–6868.
[17] D.B. Hawkes, G.W. Adams, A.L. Burlingame, P.R. Ortiz de Montellano, J.J. De Voss,
J. Biol. Chem. 277 (2002) 27725–27732.
[18] D.R. Williams, P.W. Trudgill, D.G. Taylor, J. Gen. Microbiol. 135 (1989)
1957–1967.
[19] D.B. Hawkes, K.E. Slessor, P.V. Bernhardt, J.J. De Voss, ChemBioChem 11 (2010)
1107–1114.
[20] M. Miyazawa, Y. Hashimoto, J. Agric. Food Chem. 50 (2002) 3522–3526.
[21] E. Mariani, M. Bruno, V. Villa, C. Neuhoff, Int. J. Cosmetic Sci. 22 (2000)
227–234.
[22] C.M. Jenkins, M.R. Waterman, Biochemistry 37 (1998) 6106–6113.
[23] K. Fujii, F. Huenneke, J. Biol. Chem. 249 (1974) 6745–6753.
[24] R.M. Carman, M.T. Fletcher, Aust. J. Chem. 37 (1984) 1117–1122.
[25] T. Utsukihara, O. Misumi, N. Kato, T. Kuroiwa, C. Horiuchi, Tetrahedron: Asym-
metry 17 (2006) 1179–1185.
[26] Q. Tan, D. Li, H. Bao, Y. Wang, J. Wen, T. You, Synth. Commun. 34 (2004)
2945–2950.
[27] M. Miyazawa, Y. Miyamoto, J. Mol. Catal. B: Enzym. 27 (2004) 83–89.
[28] J. Alberto Marco, J.F. SanzCervera, V. GarciaLliso, N. Batlle, Phytochemistry 45
(1997) 755–763.
[29] Y.T. Meharenna, H. Li, D.B. Hawkes, A.G. Pearson, J. De Voss, T.L. Poulos, Bio-
chemistry 43 (2004) 9487–9494.
[30] K.E. Slessor, A.J. Farlow, S.M. Cavaignac, J.E. Stok, J.J. De Voss, Arch. Biochem.
Biophys. 507 (2011) 154–162.
Acknowledgements
This work was supported by Australian Research Council Dis-
covery Grants (DP0881116; DP110104455) to JDV and Australian
Postgraduate Research Awards to KES, DBH and AF.
[31] T.L. Poulos, B.C. Finzel, I.C. Gunsalus, G.C. Wagner, J. Kraut, J. Biol. Chem. 260
(1985) 6122–6130.
Appendix A. Supplementary data
[32] W. Chai, H. Hamada, J. Suhara, C.A. Horiuchi, Phytochemistry 57 (2001)
669–673.
[33] H. Hata, S. Shimizu, S. Hattori, H. Yamada, J. Org. Chem. 55 (1990)
4377–4380.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcatb.2012.03.007.
[34] F.S. Sariasiani, L.R. McGee, M.K. Trower, F.G. Kitson, Biochem. Biophys. Res.
Commun. 170 (1990) 456–461.
[35] J.R. Collins, G.H. Loew, J. Biol. Chem. 263 (1988) 3164–3170.
[36] P.J. Loida, S.G. Sligar, M.D. Paulsen, G.E. Arnold, R.L. Ornstein, J. Biol. Chem. 270
(1995) 5326–5330.
References
[1] P.R. Ortiz de Montellano, J.J. De Voss, in: P.R. Ortiz de Montellano (Ed.),
Cytochrome P450: Structure, Mechanism, and Biochemistry, Kluwer Aca-
demic/Plenum Publishers, New York, 2005, pp. 183–245.
[37] Y.T. Meharenna, K.E. Slessor, S.M. Cavaignac, T.L. Poulos, J.J. De Voss, J. Biol.
Chem. 283 (2008) 10804–10812.