Stereospecific biocatalytic epoxidation: The first example of direct regeneration of a FAD-dependent monooxygenase for catalysis

Article abstract of DOI:10.1021/ja034119u

Catalysis for chemical synthesis by cell-free monooxygenases necessitates an efficient and robust in situ regeneration system to supply the enzyme with reducing equivalents. We report on a novel approach to directly regenerate flavin-dependent monooxygena

Full text of DOI:10.1021/ja034119u

Published on Web 06/12/2003
Stereospecific Biocatalytic Epoxidation: The First Example of
Direct Regeneration of a FAD-Dependent Monooxygenase for
Catalysis
Frank Hollmann, Po-Chi Lin, Bernard Witholt, and Andreas Schmid*
Contribution from the Institute of Biotechnology, Swiss Federal Institute of Technology,
CH-8093, Zurich, Switzerland
Received January 10, 2003; E-mail: andreas.schmid@biotech.biol.ethz.ch
Abstract: Catalysis for chemical synthesis by cell-free monooxygenases necessitates an efficient and robust
in situ regeneration system to supply the enzyme with reducing equivalents. We report on a novel approach
to directly regenerate flavin-dependent monooxygenases. The organometallic complex [Cp*Rh(bpy)(H2O)]2+
catalyzes the transhydrogenation reaction between formate and isoalloxazine-based cofactors such as
FAD and FMN. Coupling this FADH2 regeneration reaction to the FADH2-dependent styrene monooxygenase
(StyA) resulted in a chemoenzymatic epoxidation reaction where the organometallic compound substitutes
for the native reductase (StyB), the nicotinamide coenzyme (NAD), and an artificial NADH regeneration
system such as formate dehydrogenase. Various styrene derivatives were converted into the essentially
optically pure (S)-epoxides (ee > 98%). In addition, StyA was shown to be capable of performing sulfoxidation
reactions. The productivity of the chemoenzymatic epoxidation reaction using 6.5 µM StyA reached up to
6.4 mM/h, corresponding to approximately 70% of a comparable fully enzymatic reaction using StyB, NADH,
and formate dehydrogenase for regeneration. The coupling efficiency of the nonenzymatic regeneration
reaction to enzymatic epoxidation was examined in detail, leading to an optimized reaction setup with
minimized quenching of the electron supply for the epoxidation reaction. Thus, up to 60% of the reducing
equivalents provided via [Cp*Rh(bpy)(H2O)]2+ could be channeled into epoxide rather than hydrogen peroxide
formation, allowing selective synthesis with high yields.
Introduction
reaction conditions; and the in situ activation of molecular
oxygen.7,8 In principle, cell extracts containing a suitable
The versatility of the oxirane function as starting point for
chemical synthesis has driven considerable research effort for
the preparation of enantiopure epoxides. Diverse approaches
such as the use of chiral matrixes,1 chiral dioxiranes,2 and metal
oxoporphyrins as low-molecular weight mimics for P450
enzyme systems3 have been reported. Since the breakthrough
of the catalytic asymmetric epoxidation by the Katsuki-
Sharpless method,4 great advances were achieved, especially
with the (salen) manganese(III) complexes reported by Jacobsen
and co-workers.5 Despite the increasing number of examples
for highly specific catalysts,3 selectivities achieved so far with
terminal and trans-olefins are less satisfactory.6 Other limitations
of this method include the laborious design and synthesis of
suitable chiral ligands. Furthermore, activated oxygen must be
supplied stoichiometrically, generally in the form of hydrogen
peroxide, hypochlorite, or iodosylbenzene, causing undesired
side reactions.
biocatalyst (e.g. from recombinant microbial strains) can be used
as chemical reagents. The major challenge in using cell-free
monooxygenases as biocatalysts is to supply the monooxygenase
with the reducing equivalents needed for the reductive activation
of molecular oxygen.9 In their natural environment, monooxy-
genases derive these reducing equivalents from reduced nico-
tinamide cofactors that are supplied by cellular metabolism.
Therefore, one widespread approach is to mimic this natural
regeneration and to use a dehydrogenase for the in situ
regeneration of NAD(P)H from its oxidized form (NAD(P)+).
Most commonly, formate dehydrogenase is used for this
purpose.10-13 However, these cofactor regeneration processes
are expensive and require considerable optimization efforts to
function adequately.14
(7) May, S. W. Curr. Opin. Biotechnol. 1999, 10, 370-375.
(8) Li, Z.; van Beilen, J. B.; Duetz, W. A.; Schmid, A.; de Raadt, A.; Griengl,
H.; Witholt, B. Curr. Opin. Chem. Biol. 2002, 6, 136-144.
(9) Vilker, V. L.; Reipa, V.; Mayhew, M. P.; Holden, M. J. J. Am. Oil Chem.
Soc. 1999, 76, 1283-1289.
In contrast, enzymatic routes offer excellent chemo-, regio-,
and enantiospecificities; high catalytic performance under mild
(1) Lauret, C.; Roberts, S. M. Aldrichimica Acta 2002, 35, 47-51.
(2) Frohn, M.; Shi, Y. Synthesis 2000, 14, 1979-2000.
(3) Katsuki, T. Coord. Chem. ReV. 1995, 140, 189-214.
(4) Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102, 5974-5976.
(5) Zhang, W.; Loebach, J. L.; Wilson, S. R.; Jacobsen, E. N. J. Am. Chem.
Soc. 1990, 112, 2801-2803.
(10) Schmid, A.; Vereyken, I.; Held, M.; Witholt, B. J. Mol. Catal. B: Enzym.
2001, 11, 455-462.
(11) Zambianchi, F.; Pasta, P.; Carrea, G.; Colonna, S.; Gaggero, N.; Woodley,
J. M. Biotechnol. Bioeng. 2002, 78, 489-496.
(12) Rissom, S.; Schwarz-Linek, U.; Vogel, M.; Tishkov, V. I.; Kragl, U.
Tetrahedron: Asymmetry 1997, 8, 2523-2526.
(6) Archelas, A.; Furstoss, R. In Topics in Current Chemistry; Springer: Berlin,
1999; pp 160-181.
(13) Lutz, J.; Mozhaev, V. V.; Khmelnitsky, Y. L.; Witholt, B.; Schmid, A. J.
Mol. Catal. B: Enzym. 2002, 19-20, 177-187.
9
10.1021/ja034119u CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 8209-8217
8209

A R T I C L E S
Hollmann et al.
An interesting alternative is to directly regenerate the catalyti-
cally active monooxygenase component. Direct regeneration of
monooxygenases has mostly been described for cytochrome
P450 enzymes. Direct reduction with dithionite15 and indirect
methods using cobalt(III) sepulchrate as mediator between the
enzyme and an electrode16,17 or elementary zinc18 have been
reported. By cathodic reduction of the iron-sulfur protein
putidaredoxin, Vilker and co-workers were able to omit the
reductase component and NADPH in cell-free hydroxylation
and epoxidation reactions.19,20 However, direct regeneration of
a FAD (or FMN)-dependent monooxygenase has not been
reported so far. Typically, such enzymes show turnover numbers
of about 10-20 s-1 21,22 for product formation and are therefore
very interesting catalysts.
In the present study we show that a FADH2-dependent
monooxygenase (in this case StyA) can be regenerated directly
by means of non-native redox catalysts such as [Cp*Rh(bpy)-
(H2O)]2+ (Scheme 1, lower panel). We show that this cell-free
chemoenzymatic approach can be used for the production of
enantiopure epoxides via asymmetric synthesis. In particular,
we compare the non-native redox catalyst to the native reductase
with respect to conversion rate, substrate range, and optical
purity of the resulting products. Finally, we identify the factors
that influence the extent to which the catalytic regeneration
reaction is coupled to the biocatalytic epoxidation reaction,
provide a quantitative measure for the coupling efficiency, and
compare it with the “native”, fully enzymatic approach.
Experimental Section
Organometallic complexes such as (2,2′-bipyridyl)(penta-
methylcyclopentadienyl) rhodium ([Cp*Rh(bpy)(H2O)]2+) have
been described as potent redox catalysts for the specific
regeneration of nicotinamide cofactors (NAD(P)H) using either
formate or a cathode as source of reducing equivalents.23-26
Recently, we reported that [Cp*Rh(bpy)(H2O)]2+ can also
reduce alloxazine derivatives such as FAD or FMN at rates
comparable to those obtained for NAD(P)H regeneration.27
Therefore, [Cp*Rh(bpy)(H2O)]2+ appears to be an ideal can-
didate to catalyze the direct regeneration of FAD-dependent
monooxygenases for synthetic purposes.
Styrene monooxygenase (StyAB) from Pseudomonas sp.
VLB12028 catalyzes the specific (S)-epoxidation of a broad range
of m- and p-, as well as R- and â-, substituted styrene
derivatives.29 The enzyme is composed of a FAD-dependent
monooxygenase component (StyA) that catalyzes the epoxida-
tion reaction and a NADH-dependent reductase component
(StyB) that delivers the reducing equivalents from NADH to
StyA via FADH2.30 StyAB may be classified as a member of
the “two-component flavin-diffusible monooxygenase” fam-
ily.30,31 Partially purified StyAB from recombinant Escherichia
coli has been used for cell-free biocatalytic epoxidation reactions
using formate dehydrogenase for the in situ regeneration of
NADH (Scheme 1, upper panel).32
Chemicals. Unless indicated otherwise, all chemicals were purchased
from Fluka (Buchs, Switzerland) in the highest quality available and
used without further purification. Appropriate safety precautions were
taken where suitable, especially when working with rhodium, 3-chlo-
roperoxobenzoic acid, and organic solvents.
[Cp*Rh(bpy)(H2O)]2+ was synthesized according to literature meth-
ods33 by addition of 2 equiv of 2,2′-bipyridine to a suspension of
[{Cp*RhCl(µ-Cl)2}] (Aldrich, Buchs, Switzerland) in methanol. [Cp*Rh-
(bpy)Cl]Cl was precipitated upon addition of diethyl ether. Aqueous
stock solutions were stored at room temperature for several months
without detectable loss of activity. Racemic epoxides were produced
from the corresponding styrene derivatives with 3-chlorobenzoic acid
according to literature procedures.34
StyA (95% pure) was obtained from recombinant E. coli JM101
(pSPZ10)35 as described.32 StyB was purified from recombinant E. coli
JM109 (pTEZ302) as reported previously.36
Methods. Determination of the Catalytic FAD Reduction by
[Cp*Rh(bpy)H]+ and StyB. The rates of the formate-driven, [Cp*Rh-
(bpy)(H2O)]2+-catalyzed reduction of FAD and FMN were measured
in oxygen-free media by monitoring the depletion of oxidized FAD
(FMN) at λ ) 450 nm (ꢀ ≈ 10 000 M-1 cm-1). Alternatively, the fast
reoxidation of reduced isoalloxazine in O2-containing media leading
to hydrogen peroxide was quantified on the basis of either the depletion
of O2 using a Clark electrode (Strathkelvin Instruments Ltd., Glasgow,
UK) or the formation of hydrogen peroxide, determined using the
colorimetric assay of Tanaka and co-workers.37 Each method gave
comparable values for the catalytic performance of [Cp*Rh(bpy)-
(H2O)]2+ [e.g. at 37 °C and c(formate) ) 150 mM turnover frequencies
for [Cp*Rh(bpy)(H2O)]2+ of 74.3, 69.9, and 67.6 h-1 were determined
using FAD depletion, H2O2 formation, or O2 consumption, respectively].
The activity of StyB was determined on the basis of the hydrogen
peroxide formation rate.
General Reaction Conditions for the Chemoenzymatic Epoxi-
dation Reactions. Unless indicated otherwise, the chemoenzymatic
epoxidations were performed in 2.0-mL Eppendorf tubes placed in an
Eppendorf 5436 thermomixer. The reactions were performed in a 50
mM potassium phosphate buffer containing 150 mM sodium formate,
pH 7.6. After supplementation with StyA and FAD, the reaction mixture
was maintained at the indicated reaction temperature. The reactions
were started by the simultaneous addition of [Cp*Rh(bpy)(H2O)]2+ and
(14) Drauz, K.-H.; Waldmann, H. Enzyme Catalysis in Organic Synthesis, 2nd
ed.; Wiley-VCH: Weinheim, 2002.
(15) Fang, X.; Halpert, J. R. Drug Metab. Dispos. 1996, 24, 1282-1285.
(16) Estabrook, R. W.; Shet, M. S.; Fisher, C. W.; Jenkins, G.; Waterman, M.
R. Arch. Biochem. Biophys. 1996, 333, 308-315.
(17) Faulkner, K. M.; Shet, M. S.; Fisher, C. W.; Estabrook, R. W. Proc. Natl.
Acad. Sci. U.S.A. 1995, 92, 7705-7709.
(18) Schwaneberg, U.; Appel, D.; Schmitt, J.; Schmid, R. D. J. Biotech. 2000,
84, 249-257.
(19) Reipa, V.; Mayhew, M. P.; Vilker, V. L. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 13554-13558.
(20) Mayhew, M. P.; Reipa, V.; Holden, M. J.; Vilker, V. L. Biotechnol. Prog.
2000, 16, 610-616.
(21) Walsh, C. T.; Chen, Y.-C. J. Angew. Chem. 1988, 100, 342-352.
(22) Palfey, A.; Massey, V. In ComprehensiVe Biological Catalysis; Sinnott,
M., Ed.; Academic Press: San Diego, London, 1998; Vol. III, pp 83-154.
(23) Ruppert, R.; Herrmann, S.; Steckhan, E. Tetrahedron Lett. 1987, 28, 6583-
6586.
(24) Westerhausen, D.; Herrmann, S.; Hummel, W.; Steckhan, E. Angew. Chem.,
Int. Ed. Engl. 1992, 31, 1529-1531.
(25) Lo, H. C.; Buriez, O.; Kerr, J. B.; Fish, R. H. Angew. Chem., Int. Ed. Engl.
1999, 38, 1429-1432.
(32) Schmid, A. In DiVision of Biochemical Technology at the 224TH ACS
National Meeting; Boston, 2002, p BIOT 128.
(33) Ko¨lle, U.; Gra¨tzel, M. Angew. Chem. 1987, 99, 572-574.
(34) Schwetlick, K. In Organikum: Organisch-chemisches Grundpraktikum;
Barth Verlagsgesellschaft: Leipzig, 1993; Vol. 19, pp 272-274.
(35) Panke, S.; Wubbolts, M. G.; Schmid, A.; Witholt, B. Biotech. Bioeng. 2000,
69, 91-100.
(36) Otto, K.; Hofstetter, K.; Witholt, B.; Schmid, A. In FlaVins and FlaVopro-
teins; Chapman, S. K., Perham, R. N., Scrutton, N. S., Eds.; Rudolf
Weber: Berlin, 2002; pp 1027-1033.
(37) Saito, Y.; Mifune, M.; Nakashima, S.; Odo, J.; Tanaka, Y. Talanta 1987,
34, 667-669.
(26) Hollmann, F.; Schmid, A.; Steckhan, E. Angew. Chem., Int. Ed. Engl. 2001,
40, 169-171.
(27) Hollmann, F.; Witholt, B.; Schmid, A. J. Mol. Catal. B: Enzym. 2002,
19-20, 167-176.
(28) Panke, S.; Witholt, B.; Schmid, A.; Wubbolts, M. G. Appl. EnVirom.
Microbiol. 1998, 64, 2032-2043.
(29) Schmid, A.; Hofstetter, K.; Feiten, H.-J.; Hollmann, F.; Witholt, B. AdV.
Synth. Catal. 2001, 343, 732-737.
(30) Schmid, A. Unpublished results.
(31) Gala´n, B.; D´ıaz, E.; Prieto, M. A.; Garc´ıa, J. L. J. Bacteriol. 2000, 183,
627-636.
9
8210 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Stereospecific Biocatalytic Epoxidation
A R T I C L E S
Scheme 1. Comparison of the “Traditional” Cell-Free Biocatalytic Epoxidation Reaction Using the Reductase Unit StyB and Formate
Dehydrogenase (FDH)-Catalyzed NADH Regeneration (upper) and the Novel Chemoenzymatic Approach with [Cp*Rh(bpy)(H2O)]2+
Replacing FDH, NAD, and StyB (lower)
The temperature dependence of the [Cp*Rh(bpy)(H2O)]2+
reactant. Unless indicated otherwise, the reactions were terminated after
-
catalyzed reduction of FAD followed an Arrhenius-like curve.
The turnover rate (number of catalytic cycles per time) at 30
and 37 °C was 28.9 ( 0.5 and 69.9 ( 2.5 h-1, respectively.
The FAD reduction rate was linearly dependent on the
concentration of [Cp*Rh(bpy)(H2O)]2+, whereas the concentra-
tion of FAD had no significant influence on the initial reduction
rates. However, the total number of catalytic cycles of [Cp*Rh-
(bpy)(H2O)]2+ showed a distinct dependency on the molar ratio
FAD/[Cp*Rh(bpy)(H2O)]2+. Figure 1 shows the catalytic per-
formance of [Cp*Rh(bpy)(H2O)]2+ over time at varying FAD/
[Cp*Rh(bpy)(H2O)]2+ ratios (measured via the formation of
hydrogen peroxide). Above a 2.5-fold molar excess of FAD
over [Cp*Rh(bpy)(H2O)]2+, no decrease of H2O2 formation was
observed throughout the experiment. However, below this ratio,
the catalytic activity ceased with time. In the absence of FAD,
(when hydrogen peroxide is generated directly by the oxidation
of the reduced hydrido rhodium complex)26 [Cp*Rh(bpy)-
(H2O)]2+ performs only about 11 catalytic cycles.
15 min by the addition of 500 µL of ice-cold acetonitrile and
centrifugation of the precipitated proteins at 10300g for 15 min at 4
°C. The supernatant was analyzed by high performance liquid chro-
matography (HPLC).
HPLC Analysis. HPLC analyses were performed on a Merck/Hitachi
system (LaChrom series) consisting of a L-7200 autosampler, a D-7100
gradient-pump, a L-7360 column oven, and a L-7455 diode array
detector. Chromatograms were recorded at 210 nm. For achiral HPLC,
a CC250/4 Nucleosil 100-5 C18HD column with acetonitrile/water (60:
40 isocratic, 1 mL/min) was used. Product identification was done by
comparison of the retention times with authentic samples. Optical
purities of the prepared epoxides were determined by HPLC using a
CC 200/4 column with Nucleodex R-PM as chiral stationary phase and
methanol/0.1% TEAA buffer (40:60 isocratic, 0.7 mL/min) as eluent.
Commercially available enantiopure epoxides (for styrene oxide and
3-chlorostyrene oxide) or the chemically prepared racemic epoxides
were used as standards. Optical purities of indene oxide, trans-â-
methylstyrene oxide, and the sulfoxides were determined on a DIACEL
OB-H chiral column using hexane/2-propanol (98:2 isocratic at 0.5 mL/
min).
Thus, given appropriate reaction conditions (especially an
optimized ratio of FAD to reduction catalyst), [Cp*Rh(bpy)-
(H2O)]2+ appears to be a suitable catalyst for the continuous
regeneration of reduced flavin cofactors.
The Influence of StyB on the Chemoenzymatic Reaction.
With the aim of elucidating whether the reductase component
(StyB) is crucial for the epoxidation mechanism (e.g. by exerting
an activating influence on StyA or by facilitating product
release), chemoenzymatic epoxidation reactions in the presence
Results
Characterization of the Regeneration System. The catalytic
performance of [Cp*Rh(bpy)(H2O)]2+ for the formate-driven
regeneration of FADH2 was examined as a function of pH,
temperature, and concentration of reactants. In accordance with
our prior findings for the [Cp*Rh(bpy)(H2O)]2+-catalyzed
regeneration of NAD(P)H,27 we found no influence of the pH
on the rate of FAD reduction between at least pH 6 and 8 (data
not shown).
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8211

A R T I C L E S
Hollmann et al.
Table 1. Comparison of the Fully Enzymatic and
Chemoenzymatic Epoxidation of Indene at Different FAD
Concentrationsa
(1S,2R)-indene oxide formation rate (mM/h)
[FAD] (µM)
enzymaticb
chemoenzymaticc
relative activity (%)d
50
100
300
600
4.48
3.46
1.65
0.90
3.13
2.43
0.81
0.34
70
70
49
38
a General conditions: 50 mM sodium phosphate buffer (pH 7.6), c(StyA)
) 167 µg mL-1 (3.6 µM), c(catalase) ) 650 U mL-1, c(NaHCO2) ) 150
mM, c(indene) ) 2 mM, T ) 30 °C, 800 rpm, reaction time ) 15 min.
b c(StyB) ) 1.6 µg ml-1 (89 nM, 0.16 U mL-1); c(FDH) ) 2.6 U mL-1
;
c(NADH) ) 2 mM. c c([Cp*Rh(bpy)(H2O)]2+) ) 0.2 mM. d ratio of
productivities.
Figure 1. Time course of hydrogen peroxide formation at different FAD/
[Cp*Rh(bpy)(H2O)]2+ ratios. c([Cp*Rh(bpy)(H2O)]2+) ) 19 µM, c(NaH-
CO2) ) 0.15 M, T ) 37 °C, c(FAD) ) 0 µM (b), 10 µM (O), 20 µM ([),
50 µM (]), 100 µM (2), 200 µM (4).
life of approximately 45 min was found for StyA. The
inactivation rate is increased about 2-fold when the monooxy-
genase was exposed to shear stress (caused by rotation of the
reaction vessel). However, at 30 °C, StyA retained more than
90% of its initial activity after 45 min, even under shear stress
conditions.
A maximal epoxidation rate was observed at pH 7.6 with
more than 80% of maximum activity between pH 7.4 and 7.8.
Considering the pH independence of the mediator activity within
this pH range, the optimum observed can be attributed entirely
to the monooxygenase.
Comparison of [Cp*Rh(bpy)(H2O)]2+- and StyB-Cata-
lyzed Regeneration of FADH2. One measure for the efficiency
of the chemoenzymatic process is its productivity (in terms of
product formation rate) compared to the fully enzymatic (native)
regeneration. To obtain comparable reaction conditions, equiva-
lent “reductase activities” of both StyB and [Cp*Rh(bpy)-
(H2O)]2+ were coupled with StyA. Thus, a reaction mixture
containing 0.26 U/mL formate dehydrogenase, 2 mM NADH,
and 0.16 U/mL StyB (approximately 100 U/mg) under otherwise
identical conditions (Table 1) was found to exhibit the same
FAD reduction rate as 0.2 mM [Cp*Rh(bpy)(H2O)]2+ (one unit
is defined as the catalyst amount that produces one µmol of
product per minute). Under these conditions, the rate of the
chemoenzymatic epoxidation of styrene to (S)-styrene oxide was
up to 76 ( 1.6% of the styrene epoxidation rate of the fully
enzymatic reaction (data not shown). Table 1 compares the
initial productivities for the epoxidation of indene to (1S,2R)-
indene oxide catalyzed by StyA using the chemoenzymatic
approach with the initial productivities of the fully enzymatic
regeneration at variable FAD concentrations.
Interestingly, the epoxidation rate in both setups depends
inversely on the FAD concentration in the reaction medium.
The effect however is less pronounced for the fully enzymatic
approach. This effect was examined in more detail.
Uncoupling of the Regeneration Reaction from the Ep-
oxidation Reaction. Styrene oxide was not formed in control
experiments without StyA [50 mM potassium phosphate buffer
(pH 7.6), c(NaHCO2) ) 150 mM, c([Cp*Rh(bpy)(H2O)]2+) )
0.2 mM, c(styrene) ) 2 mM, T ) 30 °C, 800 rpm, t ) 15 min].
We therefore conclude that the enantiospecific epoxidation
reaction strictly depends on the presence of StyA and a putative
4R-hydroperoxoflavin in the active site of StyA.
Figure 2. Temperature dependence of the chemoenzymatic epoxidation
reaction ([, left ordinate) and of the [Cp*Rh(bpy)(H2O)]2+-catalyzed
regeneration reaction (4, right ordinate). Conditions for the styrene oxide
formation rate: 50 mM potassium phosphate buffer (pH 7.6), c([Cp*Rh-
(bpy)(H2O)])2+ ) 0.2 mM, c(NaHCO2) ) 150 mM, c(FAD) ) 50 µM,
c(StyA) ) 1.97 µM, c(catalase) ) 1.3 kU mL-1, c(styrene) ) 4 mM, u )
1500 rpm. Data for FADH2 regeneration are based on previous results
[c([Cp*Rh(bpy)(H2O)]2+) ) 25 µM, c(NaHCO2) ) 0.5 M, c(FAD) ) 0.25
mM]27 and were confirmed experimentally.
and absence of StyB were performed. No significant difference
in the styrene oxide productivity was observed if 0.42 equiv of
StyB (relative to StyA) was applied under otherwise identical
conditions. Furthermore, optical purities of the products obtained
from the chemoenzymatic reaction were as high as reported for
whole-cell biotransformations.29 Apparently, the reductase
component does not influence the catalytic efficiency of the
monooxygenase component (StyA) and is not necessary if
FADH2 is regenerated by [Cp*Rh(bpy)H]+, the catalytically
active reduced complex. Therefore, all subsequent chemoen-
zymatic reactions were performed in the absence of StyB.
Optimal pH and Temperature Conditions. The effect of
temperature on the activity of the chemoenzymatic epoxidation
setup was examined over the range of 25-50 °C. The styrene
oxide productivity increased linearly from 25 to 40 °C. Above
this value, a distinct decrease was observed with only 28% of
the maximum activity at 50 °C (Figure 2).
Since [Cp*Rh(bpy)(H2O)]2+ was shown to exhibit increasing
catalytic performance toward the reduction of FAD within this
temperature range,27 the stability of the biocatalyst with increas-
ing temperature was examined in more detail. At 37 °C, a half-
Figure 3 shows that significant molar surpluses of FAD over
StyA are necessary in order to obtain high productivities. As a
9
8212 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Stereospecific Biocatalytic Epoxidation
A R T I C L E S
Figure 3. Styrene oxide formation rates (left ordinate) and the correspond-
ing electron-transfer yields (ETYs, right ordinate) at varying FAD concen-
trations. General conditions: 50 mM potassium phosphate buffer (pH 7.6),
c([Cp*Rh(bpy)(H2O)]2+) ) 0.2 mM, c(NaHCO2) ) 150 mM, c(catalase)
) 1.3 kU mL-1, c(styrene) ) 4 mM, reaction time ) 15 min, T ) 37 °C,
u ) 1400 rpm. (4) c(StyA) ) 1.25 µM, (9) c(StyA) ) 2.23 µM.
Figure 4. Styrene oxide productivity ([) and specific StyA activity (4)
at varying StyA concentrations. General conditions: 50 mM potassium
phosphate buffer (pH 7.0, T ) 37 °C, t ) 15 min, u ) 1400 rpm), c([Cp*Rh-
(bpy)(H2O)]2+) ) 0.4 mM, c(FAD)) 50 µM, c(NaHCO2) ) 150 mM,
c(catalase) ) 1.3 kU mL-1, c(styrene) ) 4 mM.
consequence, large amounts of freely diffusible FAD are present
in the reaction medium. Reduction of these FAD molecules
results in an uncoupling of the [Cp*Rh(bpy)(H2O)]2+-catalyzed
regeneration reaction from the StyA-catalyzed epoxidation
reaction, because FADH2 is rapidly reoxidized in the presence
of molecular oxygen.38,39 As a quantitative measure for the
uncoupling, also including uncoupling at the active site of StyA,
we defined an electron-transfer yield (ETY) as the ratio between
the styrene oxide formation rate and the FAD reduction rate.
The latter is based on the hydrogen peroxide formation rate
determined in the absence of the vinyl aromatic substrate and
catalase under otherwise identical conditions. The ETY corre-
sponds to the coupling efficiency of the regeneration reaction
and the epoxidation reaction.
The increasing specific StyA activity with increasing StyA
concentration can be explained by the more complete utilization
of nascent FADH2, which leads to lower uncoupling. The styrene
oxide formation rate showed saturation-like behavior with
respect to StyA, and a maximum ETY of 23% was calculated.
This suggests that even high concentrations of StyA cannot
compete with the fast uncoupling reaction and that an ETY
around 23% is maximal under these reaction conditions.
Overall, the uncoupling of the regeneration reaction from the
epoxidation reaction can be identified as the factor limiting the
efficiency of the chemoenzymatic reaction.
Product Spectrum of the StyA/[Cp*Rh(bpy)(H2O)]2+
System. From a mechanistic point-of-view, the isoalloxazine-
based cofactors FAD and FMN (as well as their reduced forms)
should be rather similar. The specificity of StyA for FAD as
cofactor was examined by substituting FAD with FMN. The in
situ FMN concentration was varied from 50 to 600 µM;
however, no formation of styrene oxide or indene oxide was
detectable. Since [Cp*Rh(bpy)(H2O)]2+ exhibits identical reduc-
ing activity for FAD and FMN,27 this lack of epoxidation activity
is attributed to a high specificity of StyA for FAD.
To determine the influence of the total FAD concentration
on the ETY, reactions at varying FAD and fixed catalyst (StyA
and [Cp*Rh(bpy)(H2O)]2+) concentrations were performed for
two biocatalyst concentrations (Figure 3).
We found a strong dependence of the overall styrene oxide
formation rate on the amount of FAD applied. For both enzyme
concentrations maximum styrene oxide formation rates were
found at FAD concentrations between 20 and 50 µM with
maximum specific StyA activities of 0.49 ( 0.02 U mg-1
,
During in vivo biotransformations, styrene monooxygenase
has been shown to catalyze to greater than 98% enantiospecific
epoxidation of a broad range of vinyl aromatic compounds.29
To elucidate the synthetic potential of the chemoenzymatic
approach, various substrates were tested (Table 2).
whereas the optimal ETYs were 13% and 25% at StyA
concentrations of 1.25 and 2.23 µM, respectively (Figure 3).
The sharp increase of product formation rate and ETY with
increasing FAD concentration up to approximately 20 µM can
partially be attributed to a decreasing inactivation of the
regeneration catalyst over time with increasing FAD concentra-
tion. Exceeding this FAD concentration, the FADH2-regenera-
tion rate remains unchanged (Figure 1), indicating that the
decreasing styrene oxide productivity is either due to inhibition
of StyA or due to the increasing uncoupling of the regeneration
reaction from the epoxidation reaction.
To further investigate the chemoenzymatic reaction, a series
of experiments at fixed FAD and [Cp*Rh(bpy)(H2O)]2+ and
varying StyA concentrations were performed (Figure 4).
Similar to the findings given in Figure 3, a maximum for the
specific enzyme activity of 0.48 ( 0.02 U mg-1 was obtained.
Apart from trace amounts of diols (products from spontaneous
degradation of the corresponding epoxides), only epoxides were
detected as products formed from the vinyl aromatic substrates.
The enantiomeric purities of the products were generally greater
than 98% (with the exception of p-bromostyrene oxide, ee )
95.6%) and thereby comparable to the ee values obtained in
the in vivo reactions.29 For styrene oxide and 3-chlorostyrene
oxide the (S)-configuration was confirmed using commercially
available enantiopure standards. For (1S,2R)-indene oxide the
absolute stereochemistry was confirmed by its chromatographic
behavior.40 In contrast to whole-cell reactions,29 all substrates
tested (with the exception of 3-chlorostyrene) were converted
at higher initial rates than styrene. No obvious correlation
between the initial epoxidation rate and electron-donating or
(38) Kemal, C.; Chan, T. W.; Bruice, T. C. J. Am. Chem. Soc. 1977, 99, 7272-
7286.
(39) Massey, V. J. Biol. Chem. 1994, 269, 22459-22462.
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8213

A R T I C L E S
Hollmann et al.
Table 2. Chemoenzymatic Conversion of Various Vinyl Aromatic Compoundsa,d
a Values in parentheses were taken from whole-cell biotransformations;29 n.d. ) not determined. b Determined after 15 min: TF (turnover frequency) )
c(product)15min/(c([Cp*Rh(bpy)(H2O)]2+) × reaction time). c Determined after 2 h: Yield ) 100 × (c(product)final/(c(product)final + c(substrate)final). Product
ratio ) m(Product)/m(StyA + [Cp*Rh(bpy)(H2O)]2+). TTN (total turnover number of catalyst) ) c(product)final/c(catalyst). d General conditions: 50 mM
potassium phosphate buffer (pH 7.6), c(NaHCO2) ) 150 mM, c(catalase) ) 650 U mL-1, c(StyA) ) 3.3 µM, c([Cp*Rh(bpy)(H2O)]2+) ) 0.2 mM, c(FAD)
) 50 µM, T ) 30 °C, u ) 800 rpm, substrate was added above its solubility (approximately 2 mM).
-withdrawing effects of the substituents could be detected.
Qualitatively, substitution of the double bond increased the initial
reaction rate; this effect was more distinct for cis- than for trans-
substitution (trans-â-methylstyrene vs indene and dihydronaph-
thalene). Methyl substitution at the aromatic ring increased the
reactivity, irrespective of the position of the methyl group,
whereas only p-halogenation at the aromatic ring increased the
reaction rate as compared to styrene (m- vs p-chlorostyrene).
Obviously, steric effects influencing substrate binding interfere
with electronic effects activating the double bond.
Among the vinyl aromatic compounds that were not converted
by the chemoenzymatic reaction, two groups can be distin-
guished. The first group contains bulky substrates such as p-tert-
butylstyrene as well as cis- and trans-stilbene or vinylferrocene.
These substrates probably exhibit too much steric hindrance to
efficiently bind to the active site of StyA. The second group of
nonsubstrates such as 4-aminostyrene and indole as well as
pyridines inhibits the formate driven formation of the catalyti-
(40) Jinyoua, Z.; Reddya, J.; Robergea, C.; Senanayakea, C.; Greashama, R.;
Chartrain, M. J. Ferm. Bioeng. 1995, 80, 244-246.
9
8214 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Stereospecific Biocatalytic Epoxidation
A R T I C L E S
cally active hydridorhodium complex (data not shown). Here,
the only exception was 2-vinylpyridine, which was converted
to a yet undefined more polar compound. Overall, a broad range
of chiral epoxides is accessible via the chemoenzymatic
epoxidation procedure.
of 0.15 M, [Cp*Rh(bpy)H]+ is not regenerated at maximal rates,
limiting the rate of the overall reaction.
The extent to which the reducing equivalents provided by
the regeneration system are coupled to the epoxidation reaction
influences the efficiency of the presented chemoenzymatic
approach. The correct amount of FAD present in the reaction
medium is crucial for high efficiency. Excess of FAD over
[Cp*Rh(bpy)(H2O)]2+ is desirable in order to minimize the
direct oxidation of the hydridorhodium complex to a so far
undefined complex species that is catalytically inactive under
the reaction conditions used. Maximizing the specific biocatalyst
activity necessitates a defined surplus of FAD with respect to
StyA of approximately 10-20-fold. Exceeding this value leads
to increasing redirection of the reducing equivalents into
unproductive and hazardous formation of hydrogen peroxide.
The occurrence of this uncoupling reaction is independent of
the regeneration catalyst ([Cp*Rh(bpy)(H2O)]2+ or StyB) used
(Table 1). The nature of this uncoupling at high FAD concentra-
tions is not fully understood yet. One possible explanation is
based on the assumption that only the reduction of StyA-bound
FAD can productively be channeled into epoxide formation (this
assumption is supported by the low stability of reduced FADH2
in the presence of O2). Above a certain FAD concentration, StyA
would be saturated with FAD and further increase of c(FAD)
would exclusively increase the concentration of free FAD. As
a consequence, the (constant) reducing activity of [Cp*Rh(bpy)-
(H2O)]2+ would then increasingly be channeled into unproduc-
tive reduction of free FAD at the expense of the reduction rate
of StyA-bound FAD, leading to a decreased epoxidation rate.
Another explanation is the assumption of a competition of FAD
and FADH2 (irrespective, whether it was formed at the active
site of StyA, or not) for the flavin binding site on StyA. In this
case, high overall concentrations of FAD would shift the binding
equilibrium toward StyA-bound FAD and freely diffusing
FADH2, which in the presence of O2 is reoxidized to FAD and
H2O2. Alternatively, a combination of both mechanisms is
possible. Interestingly, the fully enzymatic reaction appears to
be less sensitive to high FAD concentrations (Table 1). Further
investigations to clarify this phenomenon are underway.
The maximization of the ETY necessitates high biocatalyst
concentrations relative to regeneration catalyst and FAD con-
centration. Under the conditions outlined in Figure 4, a maximal
ETY of 23% for the epoxidation of styrene was found at a molar
ratio of StyA to [Cp*Rh(bpy)(H2O)]2+ above 1:40. This can be
explained by the more complete utilization of nascent FADH2,
which leads to lower uncoupling.
Most flavoprotein monooxygenases utilize 4R-hydroperoxo-
flavin as oxygenating reagent.22 This reactive intermediate is
chemically similar to peracids that are of broad utility for
synthetic chemistry, not only for epoxidation reactions but also
for Baeyer-Villiger oxidations or the oxygenation of hetero-
atoms. To elucidate the applicability of StyA for such reactions,
we chose structural analogues of styrene such as benzyl alcohol
(oxidation of alcohols), benzaldehyde (oxidation of aldehydes
and/or Baeyer-Villiger oxidation), and acetophenone (Baeyer-
Villiger oxidation), as well as thiophenol ethers (sulfoxidation)
as test substrates. However, only organic sulfides such as methyl
and ethyl phenyl sulfide (but not the bulky diphenyl sulfide)
were converted into the corresponding sulfoxides (data not
shown). Sulfoxides were not formed in the absence of StyA
under otherwise identical conditions (H2O2, which may oxidize
the S-atom, is dismutated by catalase). The sulfoxidation rates
were in the same range as the epoxidation rates of styrene (e.g.
thioanisol was converted at 2.1 mM h-1 under the conditions
indicated in Table 2). Further oxidation of the sulfoxide group
to sulfones was not observed. Quite interestingly, the enan-
tiospecificity of the sulfoxidation reaction was rather poor with
only 26% ee for methyl phenyl sulfoxide and 13% ee for the
ethyl phenyl sulfoxide.
Therefore, we conclude that the putative 4R-hydroperoxofla-
vin in the active site of StyA is activated to oxygenate at least
sulfur atoms in addition to double bonds.
Discussion
One of the major limitations for cell-free biocatalysis using
monooxygenases hitherto has been the complexity of the
regeneration chain supplying the enzymes with the reducing
equivalents that are needed for catalysis. Most approaches focus
on the regeneration of costly nicotinamide cofactors (NAD(P)H).
In the present study, we report on the simplification of such
electron transport chains while preserving the benefits of the
native cycles such as high reaction rates and enantiospecificity.
By means of the organometallic complex [Cp*Rh(bpy)(H2O)]2+
,
we directly regenerated the flavin-dependent styrene monooxy-
genase and thereby circumvented the need for a reductase, the
nicotinamide coenzyme, and an enzymatic regeneration system.
In these experiments we used formate as a source of reducing
equivalents for the in situ regeneration of the catalytically active
hydridorhodium complex. However, the use of electrodes as a
source of reducing equivalents is possible as well.26
[Cp*Rh(bpy)(H2O)]2+ was found to be active for the reduc-
tion of FAD over a broad range of pH and temperature,
exhibiting a catalytic performance with formate as electron
donor of approximately 1.2 turnovers per minute at 37 °C,
corresponding to a specific activity of 2.8 U mg-1. The apparent
zero-order kinetics of the regeneration reaction with respect to
FAD suggests that under the conditions applied, the formation
of the catalytically active hydridorhodium complex is the rate-
limiting step. This becomes evident considering the Michaelis-
Menten-like kinetics of its formation41 and the apparent KM
value of 78 mM for formate.27 Thus, at a formate concentration
Table 2 indicates that the ETY also strongly depends on the
substrate used. For example, the ETY obtained with dihy-
dronaphthalene as substrate is almost twice as high as for
styrene. These results may be explained by assuming that
substrate-binding (or electronic activation of the double bond)
limits the rate of the chemoenzymatic epoxidation. Furthermore,
the rate-limiting step seems to be less temperature sensitive than
the substrate-independent formation of hydrogen peroxide. The
optimal ETY obtained for the epoxidation of styrene at 37 °C
is 23%, whereas it is more than 30% at 30 °C (Table 2).
Apparently, high temperatures accelerate the unproductive
formation of H2O2 more than the epoxidation reaction and thus
(41) Steckhan, E.; Herrmann, S.; Ruppert, R.; Dietz, E.; Frede, M.; Spika, E.
Organometallics 1991, 10, 1568-1577.
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8215

A R T I C L E S
Hollmann et al.
Table 3. Comparison of Chemoenzymatic Reactions with Respect to Total Turnover Number and Activity
mediator
source of reducing
a
enzyme
substrate/product
equivalents/mediator
c (mM)
TTN
relative activity
P450 BM3f
12-pNPDb
decanoic acid
camphor
styrene
Zn/Co(sep)c
0.5
1
1.03
0.5
0.2
0.15
0.015
0.51
0.06
21.8
59.1
3.5
P450rFP4504A1g
P450camh
cathode/Co(sep)
cathode/Pdxd
cathode/Pdxd
2.9e
76
StyA
styrene
formate/[Cp*Rh(bpy)(H2O)]2+
10.9
a Ratio of initial productivities of the mediated setup and the “native” reaction cycle. b 12-pNCA ) p-nitrophenoxydodecanoic acid. c Co(sep) ) cobalt(III)
sepulchrate. d Pdx ) putidaredoxin. e The regeneration of reduced Pdx was achieved by Pdx reductase. f see ref 18. g see ref 17. h see refs 9, 19, 20. Values
were recalculated from data in the references.
result in lower ETYs. An alternative approach for the optimiza-
tion of the ETY might be to control the concentration of
dissolved oxygen. Provided the KM of StyA for O2 is suitably
low, a lower O2 concentration would minimize direct reaction
of FADH2 with O2.
The relative epoxidation rates of StyA for different substrates
observed in the cell-free chemoenzymatic reactions differ
significantly from the results obtained from whole-cell biotrans-
formations,29 where the highest epoxide formation rate was
achieved for styrene. This is most likely due to the lack of the
transport limitations across microbial membranes and other
influences of complex whole cell systems.
and emphasizes the potential of this concept for preparative
application. This concept has now been proven for StyA, which
seems to use FAD as a substrate rather than as prosthetic group.
Scaling up for preparative purposes is straightforward, since
the process setup may consist of an aerated stirred tank reactor.
On the basis of the numbers given in Table 2, it can be estimated
that the production of 1g of any epoxide using the present setup
requires a biocatalyst amount of approximately 500 mg. This
amount of enzyme can be easily produced by simple fed-batch
fermentation followed by chromatographic purification on a
technical scale.32 The present reaction setup has so far not been
optimized for preparative synthesis. It may be possible to
improve the StyA catalyst by common reaction engineering
procedures such as immobilization onto a solid support to
enhance thermal and mechanical stability,42,43 and/or the use
of a second organic phase to withdraw the highly reactive
products from the enzyme,44 and by protein engineering
procedures.45-47
Overall, a broad range of synthetically interesting enantiopure
epoxides is accessible with the presented chemoenzymatic
approach. The unsatisfying low optical purities of the sulfoxides
might be due to structural reasons. So far, only thiophenol
derivatives were tested, which might not present the sulfur in
the most suitable position for the oxygenation reaction. Benzyl
thioethers might deliver the sulfur closer to the putative reactive
4R-hydroperoxoflavin, which could enhance the stereospeci-
ficity. This structural assumption might also explain why no
Baeyer-Villiger oxidation was observed with the substrates
investigated so far (carbonyl carbon atom adjacent to the
benzene ring). Therefore, â-keto aromatics might be the
substrates of choice for StyA-catalyzed Baeyer-Villiger oxida-
tions.
Table 3 lists different direct regeneration systems that were
used in combination with monooxygenases in order to replace
NAD(P)H. Cobalt(III) sepulchrate and the redox protein puti-
daredoxin were used as shuttle systems to transfer electrons from
the source of reducing equivalents (cathode or elementary zinc)
to the monooxygenase. Multiple turnovers (up to 4000) were
reported for the monooxygenases. However, the mediators were
applied in concentrations up to 60 times higher than the maximal
product concentration with the result that the TTNs of the
mediators were between 0.015 and 0.51. The authors state that
the uncoupling of the regeneration reaction from the enzymatic
reaction is a limiting factor in the mediated reaction setups.
However, no quantification for the impact of quenching of the
electron supply chain by molecular oxygen is given. The high
excesses of mediators used suggest that there is a major impact
on the overall reaction, necessitating high concentrations of the
reduced mediators. We achieve excellent catalytic performance
of the chemoenzymatic reaction with up to 60% coupling of
the redox equivalents provided by the regeneration reaction to
the enzymatic epoxidation reaction. The epoxidation rates of
the chemoenzymatic reactions are in the range of 70% of the
comparable fully enzymatic reaction. Thus, the described
[Cp*Rh(bpy)(H2O)]2+-based regeneration setup represents sub-
stantial progress in the direct regeneration of monooxygenases
Compared to chemical catalysis, e.g. using (salen) manganese-
(III) complexes (Jacobsen catalysts), the biocatalytic, chemo-
enzymatic approach features some important advantages. The
enantiospecificity of the enzymatic epoxidation reaction resulted
in product ee’s of more than 98%. For the Jacobsen catalysts,
comparable stereospecificities have been reported for various
substrates, but specificity and activity tend to be low for terminal
and trans-configured double bonds.6 The catalytic performances
of such transition metal catalysts (that are used in one to several
mole percent) are typically in the range of a few hundred
catalytic cycles at rates in the range of several turnovers per
hour.3 These chemical catalyst values are poorer than those for
the biocatalyst, whereas they correspond to the numbers
calculated for the regeneration catalyst [Cp*Rh(bpy)(H2O)]2+
.
In recent years the preparative usefulness of synthetic flavin
hydroperoxides has been recognized and exploited for catalytic
oxidation reactions. In addition, asymmetric variants are in the
development stage.48 However, since these reactions do not
imply in situ regeneration of the reduced flavins, they depend
on the presence of stoichiometric amounts of peroxides to drive
the oxidation reaction.49,50
(42) Tischer, W.; Wedekind, F. In Topics in Current Chemistry; Springer: Berlin,
Heidelberg, 2000; Vol. 200, pp 95-126.
(43) Katchalski-Katzir, E.; Kraemer, D. M. J. Mol. Catal. B: Enzym. 2000, 10,
157-176.
(44) Held, M.; Schmid, A.; Kohler, H.-P. E.; Suske, W. A.; Witholt, B.;
Wubbolts, M. G. Biotechnol. Bioeng. 1999, 62, 641-648.
(45) Meyer, A.; Schmid, A.; Held, M.; Westphal, A. H.; Ro¨thlisberger, M.;
Kohler, H.-P. E.; van Berkel, W. J. H.; Witholt, B. J. Biol. Chem. 2002,
277, 5575-5582.
(46) Meyer, A.; Wu¨rsten, M.; Schmid, A.; Kohler, H.-P. E.; Witholt, B. J. Biol.
Chem. 2002, 277, 34161-34167.
(47) Arnold, F. H. Nature 2001, 409, 253-257.
(48) Murahashi, S.-I.; Ono, S.; Imada, Y. Angew. Chem., Int. Ed. Engl. 2002,
114, 2472-2474.
9
8216 J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003

Stereospecific Biocatalytic Epoxidation
A R T I C L E S
In this context, the polypeptide StyA can be considered as a
chiral, macromolecular ligand for 4R-hydroperoxoflavin. This
ligand not only provides an asymmetric matrix for the reaction
but also exerts an activating effect on the 4R-hydroperoxoflavin
to perform epoxidation reactions. Numerous other chiral ligands
(flavin-dependent monooxygenases) supporting asymmetric
epoxidation, hydroxylation, Baeyer-Villiger oxidation, het-
eroatom oxygenation, and halogenation reactions have been
epoxides and have addressed the major limitation of the in vitro
application of monooxygenases. By using [Cp*Rh(bpy)(H2O)]2+
as a regeneration catalyst, we eliminated the need for a
complicated enzymatic electron delivery system for the mo-
nooxygenase and the nicotinamide cofactor. Factors influencing
the coupling efficiency of the regeneration reaction to the
epoxidation reaction were identified and partially optimized.
Despite the early stage of development, the chemoenzymatic
epoxidation reaction is complementing traditional chemical
routes. Overall, we have shown that biocatalytic epoxidation
reactions using an isolated oxygenase can easily be applied as
tools for chemical synthesis.
described.22,51 The concept of the [Cp*Rh(bpy)(H2O)]2+
-
catalyzed regeneration of FADH2 (FMNH2) can now be
extended to these interesting catalysts.
Thus, the chemoenzymatic reaction may be considered as
complementary to “traditional” homogeneous catalysts, extend-
ing the toolbox of chemical oxidation catalysts.
Acknowledgment. The authors express their gratitude to
Karin Hofstetter and Martina Ro¨thlisberger for preparing styrene
monooxygenase and to Katja Otto for helpful discussions and
preparation of StyB. Antonio Togni and Bruno Bu¨hler are
gratefully acknowledged for critical reading of the manuscript.
This work was financially supported by the ETH research grant
TH 24/99-3.
Conclusion
In the present study we have demonstrated the feasibility of
the cell-free biocatalytic assymmetric synthesis of enantiopure
(49) Minidis, A. B. E.; Ba¨ckvall, J.-E. Chem. Eur. J. 2001, 7, 297-302.
(50) Jonsson, S. Y.; Fa¨rnegårdh, K.; Ba¨ckvall, J.-E. J. Am. Chem. Soc. 2001,
123, 1365-1371.
(51) Moonen, M. J. H.; Fraaije, M. W.; Rietjens, I. M. C. M.; Laane, C.; van
Berkel, W. J. H. AdV. Synth. Catal. 2002, 344, 1023-1035.
JA034119U
9
J. AM. CHEM. SOC. VOL. 125, NO. 27, 2003 8217