A diversified library of bacterial and fungal bifunctional cytochrome P450 enzymes for drug metabolite synthesis

Article abstract of DOI:10.1002/adsc.200900190

Innovative biohydroxylation catalysts for the preparation of drug metabolites were developed from scratch. A set of bacterial and fungal sequences of putative and already known bifunctional P450 enzymes was identified by protein sequence alignments, expre

Full text of DOI:10.1002/adsc.200900190

FULL PAPERS
DOI: 10.1002/adsc.200900190
A Diversified Library of Bacterial and Fungal Bifunctional
Cytochrome P450 Enzymes for Drug Metabolite Synthesis
a,b
a
a
c
Roland Weis, Margit Winkler, Matthias Schittmayer, Spiros Kambourakis,
c
c
a,
Mandy Vink, J. David Rozzell, and Anton Glieder *
a
Research Centre Applied Biocatalysis, c/o Institute of Molecular Biotechnology, TU Graz, Petersgasse 14, 8010 Graz,
Austria
Fax : (+43)-316-873-9302; phone: (+43)-316-873-4074; e-mail: a.glieder@tugraz.at
Current address: VTU Technology, Parkring 18, 8074 Grambach, Austria
Codexis Inc., 200 Penobscot Drive, Redwood City, CA 94063, USA
b
c
Received: March 24, 2009; Published online: September 8, 2009; Accepted: July 8, 2009
th
Dedicated to Kalle Hult on the occasion of his 65 birthday.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/adsc.200900190.
Abstract: Innovative biohydroxylation catalysts for levels prominent single mutations were introduced.
the preparation of drug metabolites were developed Selected improved variants were then used as lyophi-
from scratch. A set of bacterial and fungal sequences lised bacterial lysates for the synthesis of 4’-hydroxy-
of putative and already known bifunctional P450 en- diclofenac and 6-hydroxychlorzoxazone, the two me-
zymes was identified by protein sequence alignments, tabolites of active pharmaceutical compounds diclo-
expressed in Escherichia coli and characterised. No- fenac and chlorzoxazone representing the same me-
tably, a fungal self-sufficient cytochrome P450 (CYP) tabolites as generated by human P450s.
from Aspergillus fumigatus turned out to be especial-
ly stable during catalyst preparation and application
and also in presence of organic co-solvents. To en- Keywords: active pharmaceutical ingredients (APIs);
hance the catalytic activity and broaden the substrate cytochrome P450 (CYP); hydroxylation; metabo-
specificity of those variants with high expression lites; mutagenesis
[
3]
Introduction
form. Parallel heterologous expression on the other
hand must provide exact ratios of components to
[
2g,4]
Cytochrome P450 (CYP) monooxygenases catalyse ensure maximised catalytic activity.
The same
the regio- and often stereoselective introduction of holds true for a soluble bacterial representative of a
oxygen into a vast range of substrates. In the presence multi-component P450 enzyme. Pseudomonas putida
of the cofactor NAD(P)H and a corresponding elec- P450cam (CYP101A1) activity is bound to the co-ex-
tron transfer system or alternatively with hydrogen pression of the electron transfer proteins putidaredox-
peroxide using the peroxy-shunt pathway, CÀH acti- in and putidaredoxin reductase, along with the P450
vation by hydroxylation, epoxidation of CÀC double domain. Notwithstanding, the relatively cheap and
bonds and aromatic hydroxylation represent reactions stable cofactor NADH renders P450cam a valuable
of cytochrome P450 enzymes which are hardly feasi- system. A triple-fusion enzyme containing all three
[
1]
ble by conventional chemical methods.
Industrial processes using microbial strains have rates and hence depreciates its use.
overcome some limitations but, generally, poor sub- Microbial bifunctional P450 monooxygenase sys-
necessary domains showed diminished conversion
[5]
strate uptake, product degradation, product toxicity tems, consisting of a single polypeptide chain includ-
or significant by-product generation pose severe ing both the heme domain and the electron transfer
[2]
drawbacks. As membrane-bound two-enzyme sys- (reductase) domain, alleviate some of the above-men-
tems, mammalian P450 enzymes are notoriously diffi- tioned problems: frequently, they are soluble en-
cult to express and hard to maintain in an active zymes, relatively stable and show high enzymatic ac-
2140
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2140 – 2146

A Diversified Library of Bacterial and Fungal Bifunctional Cytochrome P450 Enzymes
FULL PAPERS
[6]
tivity. The first cloned and most extensively engi- non-natural substrates were expected, as was also
[
8a]
neered and studied example of this enzyme sub-group shown for CYP102A2/3. Herein we present the het-
(
class III) is cytochrome P450-BM3 from Bacillus erologous expression and characterisation of bacterial
[7]
megaterium (CYP102A1).
and fungal bifunctional cytochrome P450 enzymes as
Currently, CYP102A1 (B. megaterium), CYP102A2 well as further diversification of this enzyme library
[
8]
and CYP102A3 (B. subtilis) and, as the only stable by prominent single mutations. Selected examples for
and soluble eukaryotic representative, CYP505A1 their utilisation as biocatalysts for the semi-prepara-
[
9]
(
fungus Fusarium oxysporum) are the predominant- tive scale synthesis of high-value drug metabolites are
ly used and characterised self-sufficient P450 mono- demonstrated.
[
10]
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
oxygenases. Recently, CYP102A5 (B. cereus)
and
[
11]
CYP102A7 (B. licheniformis)
were heterologously
expressed and characterised, representing efficient Results and Discussion
and regiospecific enzymes, which were in parallel also
investigated in this study.
To identify novel bifunctional cytochrome P450s we
Biotechnological application of (bifunctional) cyto- carried out
a
local alignment search (http://
chrome P450 enzymes using regio- and stereoselective www.ncbi.nlm.nih.gov/BLAST, then blastp option) on
hydroxylation includes, e.g., synthesis of drug lead the basis of the CYP102A1 (B. megaterium) sequence.
compounds or generation of value-added chemicals This enzyme was also chosen to benchmark expres-
[
1a,2e]
such as active pharmaceutical ingredients (APIs).
sion, activity and stability pattern analysis of 16 target
Mandatory studies of API metabolism in the course enzymes of both prokaryotic (Bacilli, Ralstonia) and,
of ADME analyses (administration, distribution, me- more importantly, also eukaryotic (Aspergilli, Fusari-
tabolism and excretion) require a good access to the um, Gibberellae, Magnaporthe, Neurospora) origin,
respective metabolites. Human CYPs have been ap- which were selected for recombinant expression
[
2g,4,12]
plied for this purpose
stable bifunctional P450 variants in a simple over-ex-
pression system with broad substrate/product profiles genomic DNA of the respective strains, total RNA or
but highly active and more (Table 1).
The genes of interest were amplified either from
[
13]
enable improved and scalable processes.
synthetic genes, inserted into proper expression vec-
Although a large number of CYP102 relatives was tors (see Experimental Section) and expressed in E.
[
6]
recognised by in silico nucleotide similarity searches,
coli BL21. To enhance correct protein folding and
the number of readily available self-sufficient CYPs is therefore protein yield, we created co-transformants
still limited. On the other hand, there is no obvious for the co-expression of the GroEL/ES chaperon
[
17]
reason why the first cloned bifunctional CYP should system.
Particularly for all prokaryotic CYPs as
be the best variant from nature. In addition to well well as for CYP505A1, co-expression of GroEL/ES
studied CYP102A1/A2/A3 and CYP505A1, only two greatly improved the P450 concentration in the solu-
other bifunctional microbial P450s were expressed ble protein fraction. Compared to previously pub-
and characterised and both of them are of bacterial lished protein amounts for the expression of soluble
[
10,11,14]
[18]
origin.
The self-sufficient P450 enzymes from CYP505A1 in E. coli,
we obtained a 15-fold in-
[6a,15]
[6b]
Rhodococcus sp.
and Ralstonia metallidurans,
crease with this system. Notably, cytochrome P450 en-
which were recently expressed and characterised, zymes from A. oryzae (native as well as the optimised
were not classified as CYP102 homologues due to the synthetic sequence) and the protein CYPX from G.
diverse characteristics of their reductase domains. A moniliformis were located in the insoluble protein
worthwhile strategy to generate novel self-sufficient fraction, irrespective of co-expression of GroEL/ES.
CYPs is the creation of chimeras by fusion of the
Based on their expression efficiency, nine bifunc-
heme domain of interest with a reductase domain of tional P450s (biCYPs) were chosen to be character-
[
16]
the same or another strain.
ised further. All these enzymes were active towards
A survey for relatives of CYP102A1 in public data- oxidation of the typical CYP102 reporter substrates
bases (NCBI BLAST, BRENDA, drnelson.utmem. lauric acid (see Table 2) and also ethyl dodecanoate
edu/CytochromeP450.html) revealed a broad number (data not shown). The CYP102 enzymes from B. meg-
of microbial self-sufficient P450 enzymes with reason- aterium, B. cereus and A. fumigatus showed the high-
able similarity to CYP102A1. Other approaches to est product formation although the coupling efficien-
get hold of novel bifunctional P450s have been pub- cy under our conditions was relatively low. As expect-
[
3,10b]
lished recently.
To overcome the limited availabil- ed, the enzymes from B. megaterium and B. cereus
[
10]
ity of bifunctional P450s for industrial biotechnology, were especially fast with these substrates. Enzyme
we selected a substantial number of both pro- and eu- stability against shear forces as well as organic co-sol-
karyotic representatives of self-sufficient P450s. De- vents was assayed by residual activity for lauric acid
spite apparent similarities of those targets with hydroxylation and was generally satisfying (see Sup-
CYP102A1, varying enzymatic properties towards porting Information). Remarkably, CYPX from A. fu-
Adv. Synth. Catal. 2009, 351, 2140 – 2146
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2141

Roland Weis et al.
FULL PAPERS
À1
Table 1. Amino acid sequence similarity of selected bifunc- expressed (1 mgmL ), this enzyme is highly stable
[a]
tional P450 s.
and to the best of our knowledge the wild-type bi-
functional CYP with the highest co-solvent stability
reported so far.
Native host
CYP No.
Acc. #
Id
S
[%] [%]
Subsequent to expression and stability studies, an
HPLC-MS based screening of all bifunctional CYPs
against API hydroxylation (testosterone, diclofenac,
dextromethorphan, chlorzoxazone, norbornylamine
benzamide, tolbutamide, harmine and phenacetin)
was carried out to estimate the usefulness of this
enzyme collection for the preparation of high-value
drug metabolites. The performance of these wild-type
biCYPs, however, was low. With a few exceptions the
conversions of APIs were on average less than 10%
on the screening scale. For details see Supporting In-
formation.
Prokaryotic
B. megaterium DSM 32
[
b]
102A1
P14779
102A5
AAP10153
102A7
YP 079990
X
YP 037304
102E1
100 100
[
b]
B. cereus ATCC14579
B. licheniformis ATTC14580
B. thuringiensis serovar konkukian
60
59
60
40
76
76
76
57
[b]
[b]
str.97-27
[
b]
R. metallidurans CH34
YP 585608
Eukaryotic
A. fumigatus Af293
[
d]
505X
EAL92660
505A8
EAA58234
505A3
Q2U4F1
X
Q2UNA2
505A1
Q9Y8G7
505A1
Q9Y8G7
X
AAG27132
505A7
EAA67736
505C2
EAA77183
505A5
XP 365223
505 A2
XP 961848
36
37
35
33
36
36
32
36
34
36
38
56
55
54
54
54
54
51
55
56
57
57
P450-BM3 (CYP102A1) – including countless mu-
tants thereof – is one of the most extensively studied
[
d]
A. nidulans FGSC A4
[20]
enzymes to date. The most commonly catalysed re-
[
[
c]
action involves the hydroxylation or epoxidation of
fat-soluble substrates such as fatty acids. The wild-
type CYP102A1 enzyme, moreover, is able to metab-
A. oryzae ATCC42149
A. oryzae ATCC42149
c]
[21]
olise certain xenobiotics including chlorzoxazone.
[
b]
F. oxysporum
F. oxysporum
The introduction of specific mutations was reported
to additionally broaden the substrate specificity of
CYP102A1. Especially the single mutation F87V
[d]
[
22]
proved to be important. Several groups independ-
ently engineered CYP102A1 towards the hydroxyl-
ation of unnatural substrates such as polycyclic aro-
matic hydrocarbons and found double, triple and mul-
tiple mutations to increase the enzymeꢁs performance
significantly, e.g., F87V/L188Q and A74G/F87V/
[d]
G. moniliformis
[
c]
G. zeae PH1
G. zeae PH1
[
b]
[
23]
[24]
[d]
L188Q
or
R47L/Y51F/A264G.
Recently,
M. grisea 70-15 syn
CYP102A1 variants with four mutations were created
[
d]
[25]
N. crassa OR74 A
mainly for O- or N-dealkylation purposes.
Based
on this knowledge as well as structure- and sequence-
based analysis we designed and introduced specific
mutations for the bifunctional CYPs of this study for
an upgrade of the basic bifunctional CYP library by
diversification, employing prominent site-specific mu-
tations. A summary of the investigated single muta-
tions is given in Table 3 and a brief experimental de-
scription in the supporting information. To the best of
[
a]
Id: identity, S: similarity, Acc. #: amino acid sequence ac-
cession number.
From genomic DNA.
From total RNA.
From synthetic DNA.
[
[
[
b]
c]
d]
migatus showed excellent stability in the presence of our knowledge, this is the first report of muteins from
0 and 20% of DMSO maintaining >95% of initial these particular strains, especially for self-sufficient
activity towards lauric acid hydroxylation after 6 h. fungal P450s.
For comparison, wild type CYP102A1 retained The mutein CYP102A7 A266G showed 24% hy-
95% activity at 10% DMSO as well but activity droxylation of chlorzoxazone after 30 min incubation
1
>
dropped to 70% when exposed to 20% v/v DMSO. time in presence of the co-factor recycling system glu-
Directed evolution towards solvent stability of cose/GDH (>60% conversion after overnight incuba-
CYP102A1 was reported to lead to improved variants tion). In comparison, wild type CYP102A1 gave only
[
19]
containing multiple mutations. In 10 or 20% v/v of 2.6% of compound 2 under the same conditions al-
[
21]
ethanol, activity levels of wild-type BM3 CYP102A1 though Di Nardo et al. reported that the bacterial
[
19]
dropped to 40% and complete inactivation whereas CYP102A1 is already about 10 times faster than the
those of the wild-type A. fumigatus CYP under com- human CYP2E1. However, no comparisons on a
parable conditions retained 80–90%, respectively. preparative scale were described. Likewise, the fungal
Hence, apart from being highly active and very well CYP505X F90V and the similar bacterial CYP102A5
2142
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2140 – 2146

A Diversified Library of Bacterial and Fungal Bifunctional Cytochrome P450 Enzymes
FULL PAPERS
[a]
Table 2. Conversion of lauric acid.
[b]
À1
[c]
À1
biCYP
NADPH depletion [nmolmin ] Lauric acid conversion [nmolmin ] Coupling efficiency [%]
Aspergillus fumigatus
Giberellum zeae
Bacillus megaterium
Bacillus cereus
760
160
26
21
2.5
38
21
7
2
2
10
2
1020
1320
4260
2000
500
890
140
0.8
15
Bacillus licheniformis
Ralstonia metallidurans 40
[
d]
Fusarium oxysporum[
770
190
470
e]
Fusarium oxysporum
Nꢀurosporum crassa
19
9
[
[
[
[
[
a]
b]
c]
d]
e]
2
mM lauric acid, 200 mM of NADPH in a final volume of 1 mL of 50 mM KP , pH 8.0.
i
Determined by NADPH depletion assay.
Determined by HPLC-MS.
Subcloned from gDNA.
From synthetic DNA.
Table 3. Investigated single mutations.
[a]
biCYP
Mutations
A. fum.
W50L, L78R, I81A, F90A, F90V, Y209S,
Y209L, A268G and F279S
505X
G. zeae
L75R, V78A, F87A, F87V, N188Q, F206S,
F206L, A265G and F276S
B. cer.
D49L, L77R, V80A, F89A, F89V, L190Q,
F207S, F207L, A267G and F278S
A48L, L76R, V79A, F88A, F88V, T189Q,
F206S, F206L, A266G and F277S
L82R, F94A and F94V
1
02A5
B. lich.
02A7
1
R. met.
F. oxy.
L77R, V78A, L86I, F87A and F87V
[
a]
The positions refer to the real positions in the respective
proteins not the comparable position in CYP102 A1
which, on average, differ by few amino acid residues.
Scheme 1. Preparative scale biohydroxylations with self-suf-
ficient CYP mutants.
F89V mutant (the positions mimick the CYP102A1
F87V mutation, respectively) both catalysed the 4’-hy-
For the production of diclofenac and chlorzoxazone
droxylation of diclofenac in moderate conversions metabolites (Scheme 1) on the preparative scale, we
3.0% and 0.9% opposed to no hydroxylation product used the respective muteins as lyophilised lysates in
(
at all generated by the wild-type CYP102A1). A very stirred reactors with cofactor recycling and under pH
recent paper describes the formation of 4’-hydroxydi- control (see Experimental Section). Due to the poor
clofenac catalysed by a CYP102A1 variant with six solubility, especially of chlorzoxazone, in the reaction
mutations, including F87V, which might well be the buffer, 1 v/v % of DMSO was applied as co-solvent.
key for this activity, although the specific effect of The products and unreacted substrates were isolated
[
26]
each mutation was not rationalised by the authors.
from the reaction mixtures, purified and characterised
Human CYPs heterologouosly expressed, for exam- (see Table 4). The desired metabolites were obtained
ple, in yeasts have also been screened for diclofenac
[
2g,12b]
Table 4. Preparative scale biohydroxylation of active phar-
maceuticals.
hydroxylation but with rare exceptions
are limited to the analytical scale.
reports
[
4,27]
The wild-type
CYP102A1 as well as a number of mutants thereof
have been evaluated towards API hydroxylation to
find activities analogous to those of human
[a]
[b]
biCYP
[S]
Product Conv.
[%]
Yield
[%]
À1
[
gL ]
[13a,21,26,28]
CYP505X F90V 1.69
CYP102A7
A266G
1
2
15
44
8
12
CYPs.
Although the scaleability of those bio-
1.36
hydroxylations on the screening scale was predicted
repeatedly, literature reports of preparative metabo-
lite syntheses using bifuncional CYPs are practically
non-existent.
[
[
a]
b]
Determined by HPLC-MS.
Isolated yield after purification.
Adv. Synth. Catal. 2009, 351, 2140 – 2146
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2143

Roland Weis et al.
FULL PAPERS
in milligram quantities which is usually the amount suspended pellets (in buffer A) were tested for enzymatic
activity and P450 occurrence (see next section). Protein con-
required as reference compounds for metabolic stud-
centrations were determined using the BCA method (Nova-
ies. The reactions were certainly not optimised sys-
gen). CO difference spectra were recorded by diluting the
tematically in this study but the conditions offer a
supernatant fraction in potassium phosphate buffer (50 mM,
good starting point for further improvement.
pH 7.4) to a final volume of 1 mL. Sodium dithionite was
used as reducing agent, and half of this sample was used as
blank reference for the reduced, non-CO coupled status.
After bubbling CO for 30 sec to the other half with 1–3 bub-
Conclusions
bles per second, the absorption spectrum from 410 to
In total, four bacterial and seven fungal cytochrome 500 nm was recorded and the protein content calculated.
P450 enzymes have been heterologously expressed in Catalytic characterisation of the enzymes is briefly described
in the Supporting Information.
E. coli and characterised, giving rise to an expanded
portfolio of highly interesting biocatalysts. Upon in-
troduction of specific mutations, catalytic activities to-
wards API hydroxylation were significantly increased Culturing and Work-Up of Self-Sufficient CYP
which was exemplified by the preparative scale bio- Mutants
transformations of diclofenac and chlorzoxazone with
E. coli BL21 (DE3) strains carrying both the mutated P450
CYP102A5/CYP505X and CYP102A5 variants, re-
expression plasmid as well as the plasmid with GroEL/ES,
spectively.
were cultured in 50 mL TB medium supplemented with
À1
À1
5
0 mgmL
kanamycin and 33 mgmL
chloramphenicol
À1
(100 mgmL EtOH), and grown at 378C in non-baffled
Experimental Section
flasks with constant shaking at 180 rpm until an OD600 of 0.6
was reached. The cultures were supplemented with 0.5 mM
g-aminolevulinic acid and induced with 0.5 mM IPTG and
then incubated for 16 h at 258C at 130 rpm. Samples were
centrifuged for 10 min at 4,000 rpm and 48C, pellets were
resuspended in buffer A (20 mL, 50 mM MOPS, 10% glycer-
ol, 1 mM DTT, 0.1 mM EDTA, pH 7.4), and disrupted by
sonification for 8 min. After ultracentrifugation at
40,000 rpm for 30 min at 48C the supernatants (lysates)
were used for further experiments.
Unless otherwise stated, all chemicals were purchased from
Carl Roth GmbH (Karlsruhe, Germany) and Sigma–Aldrich
(
St. Louis, MO, USA). Standard molecular biology proce-
dures were performed according to Current Protocols in
Molecular Biology 2007. Cloning, site-specific mutagenesis,
enzyme activity and enzyme stability determinations are de-
scribed in the Supporting Information. High performance
liquid chromatography/mass spectrometry measurements
were done on an Agilent 1200 with a UV detector and an
Agilent G1956B mass detector in series. For separation, a
Merck LiChroCART Purospher RP18 endcapped column
Activity Screening of Wild-Type Strains and Mutants
(
250 mmꢂ4.6 mmꢂ5 mm) was employed. The mass detector
settings were optimised for product detection with a single
ion mode (SIM) at the expected m/z of the product in the
preferred ionisation mode (determined experimentally with
an authentic standard). Additionally, scans (100–1000 m/z)
were recorded to monitor substrate depletion and possible
by-products.
Typically, 300 mL of appropriately diluted lysate were added
to 100 mL of a solution of NADPH (4 mM), glucose
À1
(
40 mM), glucose dehydrogenase (4 mgmL ) and FAD
(
0.4 mM) in deep-well plates. The substrates (3 mL, 100 mM
in DMSO) were added and the deep-well plates agitated on
a Titramax for 15 min. The reaction proceeded for 18 h at
3
08C. Subsequently, the proteins were precipitated by addi-
tion of acetonitrile (400 mL) and removed by centrifugation
4,000 rpm, 15 min). The supernatants were analysed by
Culturing, Work-Up and Enzyme Characterisation of
Wild-Type Self-Sufficient CYPs
(
high performance liquid chromatography/mass spectrometry
(HPLC-MS) on an Agilent 1200 with a UV detector and an
Agilent G1956B mass detector in series. For separation, a
Merck LiChroCART Purospher RP18 endcapped column
(250 mmꢂ4.6 mmꢂ5 mm) was employed. The mobile phase
composition and flow rate were optimised for each substrate
(typically acetonitrile/0.1% aqueous acetic acid or formic
E. coli BL21 (DE3) strains carrying both the P450 expres-
sion plasmid as well as the plasmid with GroEL/ES, were
cultured in 50 mL TB medium supplemented with
À1
À1
5
0 mgmL kanamycin and 33 mgmL chloramphenicol, and
grown at 288C with constant shaking at 180 rpm until an
OD600 of 0.5 was reached. The cultures were supplemented
with 0.25 mM or 0.5 mM g-aminolevulinic acid (as heme
precursor), and after 1 h induced with IPTG in a test range
of 0.1 to 0.5 mM in 0.1 increments, and incubated for 16 h at
À1
acid was used at flow rates between 0.5 and 1.0 mLmin ).
The mass detector settings were optimised for product de-
tection with a single ion mode (SIM) at the expected m/z of
the product in the preferred ionisation mode (determined
experimentally with an authentic standard). Additionally,
scans (100–1,000 m/z) were recorded to monitor substrate
depletion and possible by-products. The product detection
limit depended on the respective substrate but was in every
case <1% of the substrate concentration.
208C, 25 8C and 288C at 130 rpm.
Samples were centrifuged for 10 min at 4,000 rpm and
8C, pellets were resuspended in buffer A (20 mL, 50 mM
4
MOPS, 10% glycerol, 1 mM DTT, 0.1 mM EDTA, pH 7.4),
and disrupted by sonification for 8 min. After ultracentrifu-
gation at 40,000 rpm for 30 min at 48C, supernatants and re-
2144
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2140 – 2146

A Diversified Library of Bacterial and Fungal Bifunctional Cytochrome P450 Enzymes
FULL PAPERS
Example for Preparative Scale Biotransformation of
Diclofenac
[2] a) K. Petzoldt, K. Annen, H. Laurent, R. Wiechert,
982, U.S. Patent 4,353,985, 1982; b) S. Picataggio, T.
1
Rohrer, K. Deanda, D. Lanning, R. Reynolds, J. Mie-
lenz, L. D. Eirich, Biotechnology 1992, 10, 894–898;
c) A. Trefzer, V. Jungmann, I. Molnꢃr, A. Botejue, D.
Buckel, G. Frey, D. S. Hill, M. Jçrg, J. M. Ligon, D.
Mason, D. Moore, J. P. Pachlatko, T. H. Richardson, P.
Spangenberg, M. A. Wall, R. Zirkle, J. T. Stege, Appl.
Environ. Microbiol. 2007, 4317–4325; d) A. Weiss, in:
Modern Biooxidation, 1st edn., (Eds.: R. D. Schmid,
V. B. Urlacher), Wiley-VCH, Weinheim, 2007, pp 193–
To a solution of 1.48 g of lyophilised enzyme preparation
À1
(
2 nmol CYP505X F90Vmg ) in potassium phosphate
buffer (300 mL, 100 mM, pH 7.0) NADPH (1 mM), glucose
(
(
50 mM), FAD (0.3 mM) and glucose dehydrogenase
0.5 mgmL ) were added. The reaction was started by addi-
À1
tion of 507 mg of sodium diclofenac and proceeded in a
stirred Minifermenter (DAS-GIP) at 308C with O supply.
2
Due to strong foaming, the addition of a few droplets of
Acepol anti foaming agent was necessary. The pH was auto-
matically adjusted by addition of NaOH (0.5M). The reac-
tion was stopped after overnight incubation. The conversion
was 15% as determined by HPLC-MS analysis (0.1% acetic
2
10; e) O. Ghisalba, M. Kittelmann, in: Modern Biooxi-
dation, 1st edn., (Eds.: R. D. Schmid, V. B. Urlacher),
Wiley-VCH, Weinheim, 2007, pp 211–232; f) S. P.
Hanlon, T. Friedberg, C. R. Wolf, O. Ghisalba, M. Kit-
telmann, in: Modern Biooxidation, 1st edn., (Eds.:
R. D. Schmid, V. B. Urlacher), Wiley-VCH, Weinheim,
À1
acid/acetonitrile 30/70, flow: 1.0 mLmin , detection: UV
230 nm and APCI in negative mode). The aqueous reaction
mixture was evaporated to dryness and the product was pu-
rified by preparative reversed phase HPLC. 4’-Hydroxydi-
clofenac {2-[2-(2,6-dichloro-4-hydroxyphenylamino)phenyl]-
acetic acid} was thus obtained in 8% yield (40 mg).
2
007, pp 233–252; g) R. B. Vail, M. J. Homann, I.
Hanna, A. Zaks, J. Ind. Microbiol. Biotechnol. 2005, 32,
7–74.
6
[
[
[
[
3] S. Eiben, L. Kaysser, S. Maurer, K. Kꢄhnel, V. B. Ur-
lacher, R. D. Schmid, J. Biotechnol. 2006, 124, 662–669.
4] A. Parikh, E. M. J. Gillam, F. P. Guengerich, Nat. Bio-
technol. 1997, 15, 784–788.
Example for Preparative Scale Biotransformation of
Chlorzoxazone
5] O. Sibbesen, J. J. De Voss, P. R. Ortiz de Montellano, J.
To a solution of 255 mg of lyophilised enzyme preparation
À1
Biol. Chem. 1996, 271, 22462–22469.
(
2 nmol CYP102A7 A266G mg ) in potassium phosphate
6] a) R. de Mot, A. H. A. Parret, TRENDS Microbiol.
2002, 10, 502–508; b) A. W. Munro, H. M. Girvan, K. J.
McLean, Biochim. Biophys. Acta 2007, 1770, 345–359.
7] a) L. O. Narhi, A. J. Fulco, J. Biol. Chem. 1986, 261,
buffer (50 mL, 100 mM, pH 7.4) NADPH (0.9 mM), glucose
À1
(
60 mM) and glucose dehydrogenase (0.9 mgmL ) was
added. The reaction was started by addition of 68 mg of
chlorzoxazone in DMSO (500 mL) and proceeded in a
stirred pH stat (Mettler Toledo) at 308C. The pH was auto-
matically adjusted by addition of NaOH (0.5M). After 14 h
the reaction was stopped. The conversion was 44% as deter-
mined by HPLC-MS analysis (0.1% acetic acid/acetonitrile
[
7
160–7169; b) R. T. Ruettinger, L. P. Wen, A. J. Fulco,
J. Biol. Chem. 1989, 264, 10987–10995; c) H. Li, T. L.
Poulos, Nat. Struct. Biol. 1997, 4, 140–146; d) A. J.
Warman, O. Roitel, T. Neeli, H. M. Girvan, H. E.
Seward, S. A. Murray, K. J. McLean, M. G. Joyce, H.
Toogood, R. A. Holt, D. Leys, N. S. Scrutton, A. W.
Munro, Biochem. Soc. Trans. 2005, 33, 747–753.
3
3
2
0/70 for 1 min, linear gradient to 77.5% acetonitrile within
min, hold for 0.5 min, flow: 1.0 mLmin , detection: UV
95 nm and APCI in negative mode). The product and un-
À1
[
8] a) M. C. U. Gustafsson, O. Roitel, K. R. Marshall,
M. A. Noble, S. K. Chapman, A. Pessegueiro, A. J.
Fulco, M. R. Cheesman, C. von Wachenfeldt, A. W.
Munro, Biochemistry 2004, 43, 5474–5487; b) M.
Budde, S. C. Maurer, R. D. Schmid, V. B. Urlacher,
Appl. Microbiol. Biotechnol. 2005, 66, 180–186.
reacted starting material were isolated by extraction of the
aqueous reaction mixture with ethyl acetate. After drying
with Na SO and evaporation, a silica gel column with cyclo-
2
4
hexane/ethyl acetate as eluent was used for purification. 6-
Hydroxychlorzoxazone {5-chloro-6-hydroxybenzo[d]oxazol-
2(3H)-one} was thus obtained in 12% yield (8 mg).
[
9] a) N. Nakayama, A. Takemae, H. Shoun, J. Biochem.
1
996, 119, 435–440; b) T. Kitazume, Y. Yamazaki, S.
Matsuyama, H. Shoun, N. Takaya, Appl. Microbiol.
Biotechnol. 2008, 79, 981–988.
Acknowledgements
[
[
10] a) P. K. Chowdhary, M. Alemseghed, D. C. Haines,
Arch. Biochem. Biophys. 2007, 468, 32–43; b) B. S.
Kim, S. Y. Kim, J. Park, W. Park, K. Y. Hwang, Y. J.
Yoon, W. K. Oh, B. Y. Kim, J. S. Ahn, J. Appl. Micro-
biol. 2007, 102, 1392–1400.
This research was supported by Codexis, the FFG, the Prov-
ince of Styria, the SFG and the City of Graz. Beate Pscheidt
is gratefully acknowledged for assistance in mutation design
and Hannelore Mandl for technical support.
11] M. Diethrich, S. Eiben, C. Asta, T. A. Do, J. Pleiss,
V. B. Urlacher, Appl. Microbiol. Biotechnol. 2008, 79,
9
31–940.
References
[12] a) J. A. R. Blake, M. Pritchard, S. Ding, G. C. M. Smith,
B. Burchell, C. R. Wolf, T. Friedberg, FEBS Lett. 1996,
397, 210–214; b) T. H. Rushmore, P. J. Reider, D.
Slaughter, C. Assang, M. Shou, Metab. Eng. 2000, 2,
115–125.
[1] a) V. B. Urlacher, S. Eiben, TRENDS Biotechnol. 2006,
2
1
4, 324–330; b) R. Bernhardt, J. Biotechnol. 2006, 124,
28–145; c) H. M. Bolt, P. H. Roos, in: Cytochromes
P450, 1st edn., (Ed.: C. Ioannides), RSC Publishing,
Cambridge, 2008, pp 47–86.
[13] C. R. Otey, G. Bandara, J. Lalonde, K. Takahashi, F. H.
Arnold, Biotechnol. Bioeng. 2006, 93, 494–499.
Adv. Synth. Catal. 2009, 351, 2140 – 2146
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2145

Roland Weis et al.
FULL PAPERS
[
14] N.-S. Park, H.-J. Park, K. Han, E.-S. Kim, J. Microbiol.
Biotechnol. 2006, 16, 295–298.
[20] a) J. D. Bloom, S. T. Labthavikul, C. R. Otey, F. H.
Arnold, Proc. Natl. Acad. Sci. USA 2006, 103, 5869–
5874; b) V. B. Urlacher, S. G. Bell, L.-L. Wong, in:
Modern Biooxidation, 1st edn., (Eds.: R. D. Schmid,
V. B. Urlacher), Wiley-VCH, Weinheim, 2007, pp 99–
122.
[
15] a) G. A. Roberts, G. G. Grogan, A. Greter, S. L. Flitsch,
N. J. Turner, J. Bacteriol. 2002, 184, 3898–3908;
b) G. A. Roberts, A. Celik, D. J. B. Hunter, T. B. W.
Ost, J. H. White, S. K. Chapman, N. J. Turner, S. L.
Flitsch, J. Biol. Chem. 2003, 278, 48914–48920; c) L.
Liu, R. D. Schmid, V. B. Urlacher, Appl. Microbiol.
Biotechnol. 2006, 72, 876–882.
[
21] G. Di Nardo, A. Fantuzzi, A. Sideri, P. Panicco, C. Sas-
sone, C. Giunta, G. Gilardi, J. Biol. Inorg. Chem. 2007,
1
2, 313–323.
[
22] a) S. E. Graham-Lorence, G. Truan, J. A. Peterson,
J. R. Falck, S. Wei, C. Helvig, J. H. Capdevila, J. Biol.
Chem. 1997, 272, 1127–1135; b) W. T. Sulistyaningdyah,
J. Ogawa, Q.-S. Li, C. Maeda, Y. Yano, R. D. Schmid,
S. Shimizu, Appl. Microbiol. Biotechnol. 2005, 67, 556–
[
16] a) Y. Y. Deeni, M. J. I. Paine, A. D. Ayrton, S. E.
Clarke, R. Chenery, C. R. Wolf, Arch. Biochem. Bio-
phys. 2001, 396, 16–24; b) M. Nodate, M. Kubota, N.
Misawa, Appl. Microbiol. Biotechnol. 2006, 71, 455–
4
62; c) E. Leonard, M. A. G. Koffas, Appl. Environ.
5
62.
Microbiol. 2007, 73, 7246–7251; d) S. Eiben, H. Bartel-
maes, V. B. Urlacher, Appl. Microbiol. Biotechnol.
[
[
[
23] Q.-S. Li, J. Ogawa, R. D. Schmid, S. Shimizu, Appl. En-
viron. Microbiol. 2001, 67, 5735–5739.
24] A. B. Carmichael, L.-L. Wong, Eur. J. Biochem. 2001,
2
007, 75, 1055–1061; e) S. Li, L. M. Podust, D. H. Sher-
man, J. Am. Chem. Soc. 2007, 129, 12940–12941.
17] M. Mayhew, A. C. da Silva, J. Martin, H. Erdjument-
Bromage, P. Tempst, F. U. Hartl, Nature 1996, 379,
2
68, 3117–3125.
[
[
[
25] B. M. A. van Vugt-Lussenburg, E. Stjernschantz, J.
Lastdrager, C. Oostenbrink, N. P. E. Vermeulen,
J. N. M. Commandeur, J. Med. Chem. 2007, 50(3), 455–
4
20–426.
18] T. Kitazume, A. Tanaka, N. Takaya, A. Nakamura, S.
4
61.
Matsuyama, T. Suzuki, H. Shoun, Eur. J. Biochem.
[
26] M. C. Damsten, B. M. A. van Vugt-Lussenburg, T. Zel-
denthuis, J. S. B. de Vlieger, J. N. M. Commandeur,
N. P. E. Vermeulen, Chem.-Biol. Interact. 2008, 171, 96–
2
002, 269, 2075–2082.
19] T. S. Wong, F. H. Arnold, U. Schwaneberg, Biotechnol.
Bioeng. 2004, 85, 351–358. In the cited paper the re-
ported relative activities were determined using cell ly-
sates monitoring p-nitrophenoxydodecanoic acid
1
07.
[
[
27] a) A. Mancy, M. Antignac, C. Minoletti, S. Dijols, V.
Mouries, N.-T. H. Duong, P. Battioni, P. M. Dansette,
D. Mansuy, Biochemistry 1999, 38, 14264–14270.
28] B. M. A. van Vugt-Lussenburg, M. C. Damsten, D. M.
Maasdijk, N. P. E. Vermeulen, J. N. M. Commandeur,
Biochem. Biophys. Res. Commun. 2006, 346, 810–818.
hydroxy ACHTUNGTRENNUNGl ation spectrophotometrically. In our study, rel-
ative activities were determined using lyophilised
enzyme preparations employing lauric acid hydroxyl-
ation in combination with HPLC-MS monitoring.
2146
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 2140 – 2146