[Cp*Rh(bpy)(H2O)]2+ as a coenzyme substitute in enzymatic oxidations catalyzed by Baeyer-Villiger monooxygenases

Article abstract of DOI:10.1039/b504921k

[Cp*Rh(bpy)(H₂O)]²⁺ was applied as a flavin regenerating reagent in BVMO catalyzed oxidations of organic sulfides to chiral sulfoxides. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b504921k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
[Cp*Rh(bpy)(H2O)]2+ as a coenzyme substitute in enzymatic oxidations
catalyzed by Baeyer–Villiger monooxygenases
Gonzalo de Gonzalo,a Gianluca Ottolina,*a Giacomo Carreaa and Marco W. Fraaijeb
Received (in Cambridge, UK) 8th April 2005, Accepted 24th May 2005
First published as an Advance Article on the web 16th June 2005
DOI: 10.1039/b504921k
[Cp*Rh(bpy)(H2O)]2+ was applied as a flavin regenerating
have been classified as Baeyer–Villiger monooxygenases (BVMOs,
flavin containing enzymes),9 were chosen on the basis of their
reagent in BVMO catalyzed oxidations of organic sulfides to
sequence distance in order to cover an ample volume of sequence
space for this class of enzymes9 and thus maximize possible
differences in their catalytic properties. These enzymes are able to
catalyze the nucleophilic oxidation of ketones (Baeyer–Villiger
reaction) as well as the electrophilic oxidation of various
heteroatoms such as that of organic sulfides to the corresponding
optical active sulfoxides,10 valuable compounds for medicinal and
fine chemistry.11
chiral sulfoxides.
In organic synthesis, the application of isolated enzymes depending
on expensive coenzymes [e.g. NADP+] is often hindered by the
consumption of the respective coenzymes in stoichiometric
amounts. The most likely solution to avoid this drawback is the
coupling of the main reaction with an ancillary (enzymatic)
reaction able to recycle the coenzyme.
Recently, it has been reported that certain organorhodium
complexes can reduce some biologically active structures, such as
flavins,1 porphyrins1 and nicotinamides.1,2 Treatment of
[Cp*Rh(bpy)(H2O)]2+, 1 (Scheme 1)3 with a formate salt leads to
[Cp*Rh(bpy)H]+, 1a. This compound (1a) is able to act as an
electron donor capable of reducing the FAD cofactor. By this, the
need for NADPH can be avoided. This methodology has been
employed in enzymatic epoxidations catalyzed by an FAD-
dependent styrene monooxygenase.4 Another attractive feature is
that, also, a cost-effective electrochemical method can be used for
the regeneration of 1a.2,3
Before testing the performance of 1 as a regenerating agent,
NADP+ and the glucose-6-phosphate/glucose-6-phosphate dehy-
drogenase recycling system were employed and both the conver-
sion and enantiomeric excess so obtained were used as benchmarks
for the model substrates tested. Using the enzymatic regeneration
system, at first, enzymatic sulfoxidation was performed with
PAMO and benzyl ethyl sulfide, 2. It was found that 2 was
oxidized to (S)-benzyl ethyl sulfoxide [(S)-3] with high enantio-
meric excess (ee 98%, Table 1, entry 1). Next, 1, synthesized
according to the literature,3 was employed at two concentrations
(1.0 and 10.0 mM; Table 1, entries 2 and 3), in the presence of
sodium formate, in the PAMO oxidation of 2. In these conditions,
both conversion and, even more so, enantiomeric excess were low
but measurable. This indicates that the rhodium-based regenera-
tion system is able to replace coenzyme regeneration.{
In the present report, we describe the application of 1 in the
enzymatic oxidation of organic sulfides and sulfoxides catalyzed by
four isolated flavin-monooxygenases, namely: phenylacetone
monooxygenase (PAMO)5 from Thermobifida fusca, 4-hydroxya-
cetophenone monooxygenase (HAPMO)6 from Pseudomonas
fluorescens ACB, cyclohexanone monooxygenase (CHMO)7 from
Acinetobacter NCIMB 9871 and ethionamide monooxygenase
(EtaA)8 from Mycobacterium tuberculosis. These enzymes, which
A schematic representation of the reactions that were likely
involved is reported in Scheme 2, part a. The oxidation of the
sulfide to the corresponding sulfoxide is obtained at the expense of
formate ion and oxygen without the involvement of NADPH; in
this case, 1a directly reduces the FAD bound to the enzyme to
Table 1 PAMO oxidation of benzyl ethyl sulfide, 2, and benzyl ethyl
sulfoxide, (¡)-3, using different recycling systemsa
Entry Substrateb [1]c
[NADP+]c Conversion (%)d Ee (%)e
1f
2
3
4
5
6f
7
8
9
a
2
2
2
2
2
—
1.0
10.0
10.0
1.0
0.26
—
—
0.26
1.3
0.26
—
0.26
1.3
36
22g
17
21
27
49
10
14
24
¢98
7
9
97
97
93
,5
6
(¡)-3
(¡)-3
(¡)-3
(¡)-3
—
10.0
10.0
1.0
Scheme 1 Structures of organorhodium complexes involved in the flavin
regeneration system.
22
Reactions stopped after 6 h for 2 and 4 h for (¡)-3. b 16 mmol L21
.
Measured by HPLC
aIstituto di Chimica del Riconoscimento Molecolare, CNR, via Mario
Bianco 9, 20131, Milano, Italy. E-mail: gianluca.ottolina@icrm.cnr.it;
Fax: +39 02 2890 1239; Tel: +39 02 2850 0021
c
d
Concentrations expressed in mmol L21
using acetanilide as internal standard. Determined by chiral
.
e
f
HPLC. Reactions carried on using the glucose-6-phosphate/
bLaboratory of Biochemistry, Groningen Biomolecular Sciences and
Biotechnology Institute, University of Groningen, Nijenborgh 4,
9747 AG Groningen, The Netherlands
g
glucose-6-phosphate dehydrogenase system.
after 24 h.
Reaction stopped
3724 | Chem. Commun., 2005, 3724–3726
This journal is ß The Royal Society of Chemistry 2005

View Article Online
(Table 2, entry 1).10 When 1 was used at 1.0 and 10.0 mM
concentrations the racemic sulfoxide was recovered in low yields
(Table 2, entries 2 and 3). No significant improvement in the
biocatalyst properties was found in the presence of NADP+, when
10 mM 1 was utilized (Table 2, entries 4 and 5). Instead, at a lower
concentration of 1 (1 mM), (S)-5 was obtained with higher
conversion and good optical purity (ee 81%) (Table 2, entry 7).
Thioanisole sulfoxide 5 was also oxidized by HAPMO to the
corresponding sulfone, but no selectivity was found for any of
the recycling systems tested (Table 2, entries 8–11). However the
combined use of 1 and NADP+ gave a conversion close to that of
the control reaction.
CHMO is the best known flavin containing monooxygenase for
synthetic applications10 and has been successfully employed in the
selective oxidation of several sulfides.12 The oxidation of 4 led to
the enantiopure (R)-5 with high conversion (Table 3, entry 1). The
oxidation of 4 carried out in the presence of 10 mM of 1 yielded
the sulfoxide 5 with low enantiomeric excess and conversion
(Table 3, entry 2). The presence of NADP+ increased both the
conversion and the enantiomeric excess of (R)-5. A decrease in the
concentration of 1 (1.0 mM), and an increase in that of NADP+,
induced a slight improvement in enantioselectivity and a higher
degree of conversion (Table 3, entry 4). No reaction was observed
when the oxidation of (¡)-5 to the corresponding sulfone was
attempted.
Scheme 2 Enzymatic oxidation of organic sulfides catalyzed by BVMOs
using (a) complex 1/formate and (b) adding NADP+ to the mixture
1/formate.
FADH2. When the reaction was carried out in the presence of
NADP+ (0.26 mM) and 1 (Table 1, entry 4), PAMO selectivity was
completely recovered, and (S)-3 was obtained with almost the
same enantiomeric excess as that of the control reaction, even
though with lower conversion. In addition, when the concentration
of 1 was decreased to 1.0 mM and that of NADP+ was increased
to 1.3 mM, a small improvement in the conversion was observed
(Table 1, entry 5). In this case the scenario should be as depicted in
Scheme 2, part b. Compound 1a reduces NADP+ to the
enzymatically active NADPH isomer which, in turn, reduces
FAD to FADH2.1,2
Finally, EtaA was investigated. In the control reaction, 2 was
oxidized to the corresponding S-sulfoxide with good selectivity
(ee 83%) (Table 4, entry 1). When 1 was employed at 1.0 or
10.0 mM concentration, 3 was obtained in almost racemic form
(Table 4, entries 2 and 3). The same result was observed using a
low concentration of NADP+ (Table 4, entry 4). Only when the
concentration of 1 was reduced to 1.0 mM and that of NADP+
was increased to 1.3 mM, was (S)-2 formed with an enantiomeric
excess of 28% and a conversion close to that achieved in the
Racemic benzyl ethyl sulfoxide, (¡)-3, was also studied as a
PAMO substrate. Using the enzymatic reaction as an NADPH
recycling system, the (R)-3 enantiomer was oxidized to the
corresponding sulfone, leaving the remaining (S)-3 with good
enantiomeric excess (Table 1, entry 6). With 1, the biocatalyst lost
its selectivity, yielding racemic (¡)-3 as the remaining compound.
Although the reactions performed with two different mixtures of
NADP+/ 1/ formate showed improved enzyme selectivity, still very
low enantiomeric excesses for the remaining substrate and low
degrees of conversion were obtained (Table 1, entries 8 and 9).
Enzymatic oxidation of thioanisole (methyl phenyl sulfide, 4)
with HAPMO and enzymatic coenzyme recycling, leads to
enantiopure (S)-thioanisole sulfoxide, (S)-5, with high yield
Table 3 CHMO oxidation of thioanisole, 4, using different recycling
systems
Entry Substratea [1]b
[NADP]b Conversion (%)c Ee (%)d
1e
2
3
4
a
4
4
4
4
—
0.26
—
0.26
1.3
88
5
31
55
¢98
9
10.0
10.0
1.0
91
94
c
16 mmol L21
.
Concentrations expressed in mmol L21
.
After
b
Table 2 HAPMO oxidation of thioanisole, 4, and methyl phenyl
sulfoxide, (¡)-5, using different recycling systems
24 h and measured by HPLC using acetanilide as internal standard.
Determined by chiral HPLC. Reactions carried using the glucose-
6-phosphate/glucose-6-phosphate dehydrogenase system.
d
e
Entry Substratea [1]b [NADP]b t (h) Conversion (%)c Ee (%)d
1e
2
3
4
5
6
4
4
4
4
4
—
1.0
10.0
10.0 0.26
10.0 1.3
5.0 0.72
1.0 1.3
0.26
—
—
24
48
24
24
24
24
24
20
40
20
20
96
4
¢98
,5
,5
6
11
46
Table 4 EtaA oxidation of benzyl ethyl sulfide, 2, using different
recycling systems
10
11
18
21
29
30
14
20
29
Entry Substratea [1]b [NADP]b t (h) Conversion (%)c Ee (%)d
4
4
1e
2
3
4
5
a
2
2
2
2
2
—
1.0
10.0
10.0 0.26
1.0 1.3
b
0.26
—
—
16
32
16
16
16
24
6
10
19
26
83
,5
,5
,5
28
7
81
8e
9
(¡)-5
(¡)-5
(¡)-5
(¡)-5
—
0.26
—
—
,5
,5
,5
,5
1.0
10.0
10.0 0.26
10
11
16 mmol L21
Measured by HPLC using acetanilide as internal standard.
.
Concentrations expressed in mmol L21
.
a
b
16 mmol L21
.
Concentrations expressed in mmol L21
.
c
c
Measured by HPLC using acetanilide as internal standard.
Determined by chiral HPLC. Reactions carried using the glucose-
6-phosphate/glucose-6-phosphate dehydrogenase system.
d
e
Determined by chiral HPLC. Reactions carried on using the
glucose-6-phosphate/glucose-6-phosphate dehydrogenase system.
d
e
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3724–3726 | 3725

View Article Online
control reaction. No oxidation was observed for (¡)-5 with any of
the conditions tested.
reduce the enzyme bound FAD. Unfortunately, the enantioselec-
tivity displayed by the enzymes in these conditions was
disappointingly low. The presence of NADP+ in the reaction
medium restored selectivity. This suggests that the role of NADPH
in the catalytic process is not only that of a mere reducing agent: it
enhances the enantioselectivity by properly shaping the active site
of these enzymes.
On the whole, the results indicate that the enantioselectivity
displayed by the four studied BVMOs in the oxidation of sulfides
is very low, when using only 1 as a coenzyme substitute. Similar
results (ee 26%) were previously obtained with another mono-
oxygenase, styrene monooxygenase (StyA),13 in the oxidation of 4
to 5.4 The lack of enantioselectivity might be ascribed to i)
formation of hydrogen peroxide obtained through oxygen
reduction2 by 1 and consequent abiotic oxidation of the sulfide
to the racemic sulfoxide or ii) spontaneous sulfide oxidation by air.
However, these hypotheses were ruled out by the fact that in
control experiments, conducted in the presence of the various
reagents but without the monooxygenases, the formation of
sulfoxides did not occur. Thus, sulfoxides are produced by genuine
enzyme catalysis and, more importantly, 1 is able to reduce the
FAD associated with the enzymes.
We thank CERC3 for funding and G. d. G. thanks FICYT
(Principado de Asturias, Spain) for a postdoctoral fellowship.
COST D25/0005/03 is gratefully acknowledged.
Notes and references
{ General method for enzymatic oxidations using the rhodium complex as
coenzyme: the organic sulfide or sulfoxide (1.9–2.7 mg; 16 mmol L21) was
added to potassium phosphate buffer (50 mM, pH 9.0, 0.9 mL) containing
sodium formate (6.8 mg; 0.1 M), an aqueous solution of 1 (100 mM,
0.1 mL), 1.0 U of the flavin containing monooxygenase and acetanilide
(0.02 mg) as an internal standard. When necessary, different amounts of
NADP+ were added. The system was shaken at 250 rpm and 25 uC. The
reaction was then stopped, worked up and analyzed by chiral HPLC
Although styrene monooxygenase and BVMOs are closely
related enzymes, there are differences in the role of FAD. In
styrene monooxygenase the cofactor is loosely bound to the
enzyme and can diffuse in the medium,13 whereas in the studied
BVMOs the flavin cofactor is tightly bound to the enzyme, which
implies that 1a has to enter the enzyme active site in order to
reduce the cofactor. This latter hypothesis is also supported by the
finding that it is possible to accommodate 1a, by manual docking,
in the active site of PAMO (data not shown). Because of the
bulkiness of 1a compared to that of NADPH, we must also
assume a certain degree of enzyme flexibility to make it possible to
accommodate 1a in enzyme active site.
When NADP+ is present in the medium, it is likely that the
reaction mechanism is not as simple as illustrated in Scheme 2,
part b. A more reliable representation might be that both 1a and
NADPH compete for the same active site. This view is also
supported by literature kinetic data demonstrating that 1a is more
efficient in reducing NADP+ to NADPH than FAD to FADH2.2
Therefore a decrease in concentration of 1 (e.g. from 10 to 1 mM)
and an increase in that of NADP+ (e.g. from 0 to 1.3 mM) will
decrease the fraction of FAD directly reduced by 1a (Scheme 2,
part a) and increase that reduced through NADPH (Scheme 2,
part b). This partial switch of mechanism was also reflected in the
general increase of both conversion and enantiomeric excess. For
example, with PAMO the recovery of the enantioselectivity was
full when supplementing the system with NADP+ (Table 1, entries
4 and 5). The different influence of 1a and NADPH on enzyme
enantioselectivity is quite intriguing. It has been shown previously
that, in the catalytic cycle of BVMOs, NADPH is bound to the
enzyme during the substrate oxidation step.14 We hypothesize that,
as NADP+ is bound during the oxidation reaction, it is also part of
the active site cavity, thereby tuning the (enantio)selectivity of the
enzyme. Instead 1a, even if able to reduce FAD, will not show any
real affinity for the enzyme and will diffuse from the active site
before oxidation takes place. Therefore, in this case, the enzyme
active site will be less defined, resulting in lowered selectivity.
In conclusion, we have demonstrated that 1/formate can be used
as a coenzyme substitute in BVMO oxidations as 1a is able to
(Chiracel OD, Daicel, chiral column; l254 nm; flow rate 1 mL min21
;
eluent: light petroleum ether, i-PrOH).12
1 J. H. van Esch, M. A. M. Hoffmann and R. J. M. Nolte, J. Org. Chem.,
1995, 60, 1599.
2 H. Christine Lo, O. Buriez, J. B. Kerr and R. H. Fish, Angew. Chem.
Int. Ed., 1999, 38, 1429; F. Hollmann, A. Schmid and E. Steckhan,
Angew. Chem. Int. Ed., 2001, 40, 169; F. Hollmann, B. Witholt and
A. Schmid, J. Mol. Catal. B: Enzym., 2003, 19–20, 167; J. Lutz,
F. Hollmann, T. V. Ho, A. Schnyder, R. H. Fish and A. Schmid,
J. Organomet. Chem., 2004, 689, 4783.
3 U. Ko¨lle and M. Gra¨tzel, Angew. Chem., 1987, 99, 572.
4 F. Hollmann, P.-C. Lin, B. Witholt and A. Schmid, J. Am. Chem. Soc.,
2003, 125, 8209.
5 E. Malito, A. Alfieri, M. W. Fraaije and A. Mattevi, Proc. Natl. Acad.
Sci. U. S. A., 2004, 101, 13157; M. W. Fraaije, J. Wu, D. P. H.
M. Heuts, E. W. van Hellemond, J. H. Lutje Spelberg and D. B. Janssen,
Appl. Microbiol. Biotechnol., 2005, 66, 393.
6 N. M. Kamerbeek, M. J. H. Moonen, J. G. M. van der Ven, W. J.
H. van Berkel, M. W. Fraaije and D. B. Janssen, Eur. J. Biochem., 2001,
268, 2547; N. M. Kamerbeek, A. J. J. Olsthoorn, M. W. Fraaije and
D. B. Janssen, Appl. Environ. Microbiol., 2003, 419.
7 N. A. Donoghue, D. B. Norris and P. W. Trudgill, Eur. J. Biochem.,
1976, 63, 175; S. D. Doig, L. M. O’Sullivan, S. Patel, J. M. Ward and
J. M. Woodley, Enzyme Microb. Technol., 2001, 28, 265.
8 M. W. Fraaije, N. M. Kamerbeek, A. J. Heidekamp, R. Fortin and
D. B. Janssen, J. Biol. Chem., 2004, 5, 279.
9 M. W. Fraaije, N. M. Kamerbeek, W. J. H. van Berkel and
D. B. Janssen, FEBS Lett., 2002, 518, 43.
10 S. M. Roberts and P. W. H. Wan, J. Mol. Catal. B: Enzym., 1998, 4,
111; V. Alphand, G. Carrea, R. Wohlgemuth, R. Furstoss and
J. M. Woodley, Trends Biotechnol., 2003, 21, 318; N. M. Kamerbeek,
D. B. Janssen, W. J. H. Berkel and M. W. Fraaije, Adv. Synth. Catal.,
2003, 345, 667; M. D. Mihovilovic, F. Rudroff and B. Gro¨tzl, Curr. Org.
Chem., 2004, 8, 1057.
11 I. Ferna´ndez and N. Khiar, Chem. Rev., 2003, 103, 3651; J. Legros,
J. R. Dehli and C. Bolm, Adv. Synth. Catal., 2005, 347, 19.
12 S. Colonna, N. Gaggero, P. Pasta and G. Ottolina, Chem. Commun.,
1996, 2303; S. Colonna, S. Del Sordo, N. Gaggero, G. Carrea and
P. Pasta, Heteroat. Chem., 2002, 13, 467; F. Zambianchi, S. Raimondi,
P. Pasta, G. Carrea, N. Gaggero and J. A. Woodley, J. Mol. Catal. B:
Enzym., 2004, 31, 165.
13 K. Otto, K. Hofstetter, M. Ro¨thlisberger, B. Witholt and A. Schmid,
J. Bacteriol., 2004, 186, 5292.
14 D. Sheng, D. P. Ballou and V. Massey, Biochemistry, 2001, 40, 11156.
3726 | Chem. Commun., 2005, 3724–3726
This journal is ß The Royal Society of Chemistry 2005