Enantioselective oxidation of thioanisole with an alcohol oxidase/peroxidase bienzymatic system

Article abstract of DOI:10.1016/j.tetasy.2007.03.010

Optimization of a new bienzymatic couple alcohol oxidase/peroxidase for the asymmetric oxidation of thioanisole was realized. The main advantages in the application of this system is the use of inexpensive and easily available enzymes and substrates, perm

Full text of DOI:10.1016/j.tetasy.2007.03.010

Tetrahedron: Asymmetry 18 (2007) 701–704
Enantioselective oxidation of thioanisole with an alcohol
oxidase/peroxidase bienzymatic system
Fabio Pezzotti and Michel Therisod*
ECBB, ICMMO, Univ Paris-Sud, UMR 8182, F-91405 Orsay, France
Received 21 February 2007; accepted 9 March 2007
Abstract—Optimization of a new bienzymatic couple alcohol oxidase/peroxidase for the asymmetric oxidation of thioanisole was
realized. The main advantages in the application of this system is the use of inexpensive and easily available enzymes and substrates,
permitting gram scale enantioselective syntheses. Methanol, as a substrate for alcohol oxidase, also facilitates the solubilization of
organic substrates in the reaction medium.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
Asymmetric sulfoxides have proven to be useful tools in
organic synthesis, where they are used as chiral auxiliaries.1
They are also important compounds per se: Esomeprazol,
(sold as Losecꢁ, an anti-ulcerous compound) is one of
the highest selling drugs.2 They also have tentatively been
used as prodrugs.3 The global importance of this class of
Scheme 1. Synthesis of chiral aryl-methyl sulfoxides catalyzed by an
oxidase/peroxidase bienzymatic system.
products appears in many publications, where hundreds
of preparations are reported. Among them, many use
biological and enzymatic methods. The most efficient bio-
chemical systems, in terms of yields and enantioselectivity,
advantages, as compared to the direct addition of hydrogen
make use of chloroperoxidase (CPO)4 or cyclohexanone
peroxide to the reaction medium are:
monoxygenase (CMO).5 However, CPO is rare and expen-
sive, while CMO uses a cofactor which must be recycled
– An improved operational stability of the peroxidase
during the operation. Hence, these two systems are only
(this class of enzymes is inactivated by hydrogen perox-
suitable for small scale syntheses of chiral sulfoxides.
ide, acting as a ‘suicide substrate’).
– A very significantly increased ee of the products (hydro-
We have previously reported the elaboration of two bienzy-
gen peroxide directly oxidizes the sulfide, leading to
matic systems, making use of an inexpensive, robust indus-
racemic sulfoxide).
trial peroxidase (CiP, EC: 1.11.1.7, from Coprinus cinereus,
expressed in Aspergillus oryzae, marketed by Novozymes)
We previously reported the use of glucose oxidase and
and an oxidase producing hydrogen peroxide in situ.6–8
D-aminoacid oxidase as parts of these systems. These two
The general principle of these systems is summarized in
enzymes offer different advantages: glucose oxidase from
Scheme 1. The method was proven to efficiently permit
Aspergillus niger (GOD; EC. 1.1.3.4) is a well known and
the oxidation of a variety of aryl- and heteroaryl-methyl-
robust enzyme, produced on an industrial scale and at
sulfides into the corresponding (S)-sulfoxides, with high
low cost. D-Glucose, its natural substrate, is also relatively
yields and enantiomeric excesses (up to >99%). Their main
inexpensive. D-Aminoacid oxidase (DAOx; EC. 1.4.3.3)
from Trygonopsis variabilis is available from Fluka in an
*
immobilized and easily recoverable form. In our hands, it
gave sulfoxides with the highest ee. However they also have
Corresponding author. Fax: +33 1 69 157 281; e-mail: therisod@icmo.
u-psud.fr
0957-4166/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2007.03.010

702
F. Pezzotti, M. Therisod / Tetrahedron: Asymmetry 18 (2007) 701–704
Table 1. Influence of the ratio of alcohol oxidase/peroxidase used for
oxidation of thioanisole (0.25 mmol in 5 mL aqueous buffer; 1 M
methanol)
several drawbacks: (1) D-Alanine, the substrate of DAOx,
is too expensive for large scale synthesis. (2) The produc-
tion of water soluble gluconic acid by GOD implies that
a pH-stat must be used to maintain a constant pH. (3) In
some cases, the isolation of oxidized products of peroxi-
dase might be complicated by the presence of 1 equiv of
the water soluble products of oxidases (gluconate and
pyruvate, respectively).
Entry
AOX (U)
CiP (lmol)
Yield
ee
1
2
3
4
5
6
7
8
0.5
1
2
4
8
1
4
8
0.25
0.25
0.25
0.25
0.25
0.5
43
60
58
48
26
60
90
55
78
77
75
76
74
77
76
70
Thus we reasoned that alcohol oxidase (AOX; EC.
1.1.3.13) from the yeast Pichia pastoris could be an interest-
ing alternative to these previously described systems, with
the following improvements:
0.5
0.5
(1) Its best substrate is methanol, which is very inexpensive,
readily available and miscible with water in all proportions.
(2) Methanol can act as a useful cosolvent to assist the dis-
solution of hydrophobic substrates in the mainly aqueous
reaction medium. (3) The product of oxidation (formalde-
hyde) is very volatile. (4) Although the commercial enzyme
is relatively expensive (compared to glucose oxidase), the
cultivation of the yeast and the preparation of cell-free
extracts can be conducted on a medium scale (several litres)
in most laboratories equipped for bioorganic chemistry.9
Alternatively, frozen cells can be obtained from a labora-
tory equipped for scant microbiology, stored in a freezer
and used when needed.
ditions found above (4 U AOX, 0.5 lmol peroxidase), we
varied the concentration of methanol in the reaction med-
ium from 0.25 M to 4 M. As shown in Figure 1, the initial
rate of oxidation increased with the concentration up to
1 M. After this, the rate was somewhat insensitive to the
concentration, but the final yield increased to a maximum
of 100% for a 4 M concentration of methanol. The exact
nature of this effect was not determined, and might be com-
plex: the final concentration of alcohol is far above the KM
of alcohol oxidase for methanol (1–2 mM). Thus, AOX
works at its maximal velocity in all cases. We determined
the maximal solubility of thioanisole in pure buffer as
0.5 mM. With 8% methanol (4 M), the solubility was
roughly the same (0.6 M), indicating that the observed
effect is not due to an increased solubility of the substrate.
On an other hand, the aspect of the suspension was very
different: large droplets of thioanisole were formed in pure
buffer, while a fine milky suspension was observed in 4 M
methanol: the reaction catalyzed by CiP is probably
favoured by a better dispersion of the substrate in the aque-
ous medium. Under our conditions, the sulfoxide formed
was completely dissolved in 4 M methanol, which may pre-
vent an inhibition of peroxidase by the product.
Alcohol oxidase itself has been shown to be an interesting
enzyme in organic synthesis.10,11 The coupled system alco-
hol oxidase/peroxidase has been widely used for analytical
applications.12 Our goal was to devise a similar system
applicable to preparative asymmetric synthesis.
2. Results and discussion
We chose the classical oxidation of thioanisole into chiral
methyl-phenyl-sulfoxide as a model reaction by using a
limited amount of alcohol oxidase (purified enzyme, from
Sigma and from Applied Enzyme Technology Ltd) and a
large excess of peroxidase, in order to limit the concentra-
tion of free hydrogen peroxide in the reaction medium.
The first step was to determine the minimal amount of
AOX and the suitable AOX/CiP ratio to be used for the
oxidation of a given amount of thioanisole (0.25 mmol).
The best results were obtained with a mixture containing
0.5 lmol of peroxidase and 4 U of AOX (Table 1, entry
7) resulting in a 90% yield in sulfoxide with 76% ee.
It is clear from these results that a large excess of peroxi-
dase is needed for an optimal oxidation. The ee was essen-
tially constant at 75% for the (S)-sulfoxide, which indicates
that all the hydrogen peroxide transits through the peroxi-
dase. This ee value is also in accordance with those re-
ported with other systems using the same peroxidase,6,7,13
and is probably representative of the enantioselectivity of
this enzyme for this substrate.
Figure 1. Effect of the concentration of methanol in the reaction medium
on the oxidation of thioanisole (0.25 mmol in 5 mL aqueous buffer;
0.5 lmol of peroxidase; 4 U of AOX).
Another important feature of this reaction to be optimized
is the final concentration of methanol. Using the best con-

F. Pezzotti, M. Therisod / Tetrahedron: Asymmetry 18 (2007) 701–704
703
Since commercial purified AOX is somewhat expensive and
4.2. P. pastoris production
sold in minimal quantities, we attempted to use a ‘home
made’ crude extract of P. pastoris as a source of alcohol
oxidase. The yeast was grown following standard proce-
dures in a liquid medium containing methanol as a sole
source of carbon. After centrifugation, the cells were kept
deep-frozen at ꢀ18 ꢂC. They were later resuspended in
dilute buffer solution and disrupted with a French press
immediately before use.14 We measured an average activity
of 19 lmol of methanol oxidized/min/mL (19 U/mL) of
crude extract, which could probably be increased, if
needed, by optimization of the culture conditions. The
use of 2–4 U of AOX allowed us to obtain the sulfoxide
with 100% yield and 75% ee.
A preculture was realized by inoculation of 100 mL of
medium (phosphate buffer pH 6; 0.7% yeast nitrogen base
(DIFCO); 0.5% glycerol) with P. pastoris. The yeast was
grown at 30 ꢂC for two days with orbital agitation. This
preculture was then poured in 900 mL of medium (0.5%
methanol instead of glycerol for induction of alcohol oxi-
dase production) at 30 ꢂC under agitation (150 rpm) for
three days. The culture was then centrifuged (8000 rpm;
0 ꢂC). The cells were recovered, freeze-dried and stored at
ꢀ20 ꢂC. Crude extracts were obtained by cell disruption
of lyophilized cells (previously suspended in 0.1 M,
pH 7.5 phosphate buffer) in a French press at 1100 psi.
Centrifugation (8000 rpm; 0 ꢂC) allowed us to obtain a
crude extract solution with 0.664 mg of proteins/mL as
determined by Bradford test, with an activity of 19.6 lmol
of oxidized methanol/min/mL (115.8 U/g dried cells).
Finally, in order to prove the usefulness of the method in
preparative asymmetric synthesis, we used 10 mmol of
thioanisole for a gram scale preparation of sulfoxide. After
treatment of the reaction medium (see Experimental) and
careful evaporation of the solvent, we obtained 1.1 g of
pure sulfoxide (72% yield) as indicated by NMR, with
75% ee. The total turnover number (>700) is quite modest
when compared to other peroxidases (6 · 104 for CPO15),
but is favourably counterbalanced by the availability of
CiP. Moreover, CPO gives the (R)-sulfoxide, while CiP
gives the (S)-enantiomer. Thus, the two enzymes have com-
plementary enantioselectivities.
4.3. Thioanisole oxidation
4.3.1. Small scale oxidation. Thioanisole (0.25 mmol) was
suspended in aqueous buffer (20 mL, phosphate buffer,
0.1 M, pH 7.5) followed by the addition of methanol
(0.5–4 M), alcohol oxidase and peroxidase. At different
time intervals, aliquots were extracted with ethyl acetate
and the organic phase evaporated. The degree of conver-
sion was determined by GC. Enantiomeric excess was
determined by HPLC on a Chiralcel OD-H column (hex-
1
ane/i-propanol 95:5, flow 0.7 mL/min) and by H NMR
3. Conclusion
with the addition of (S)-(+)-N-(3,5-dinitrobenzoyl)-a-
methylbenzylamine as the chiral shifting agent.
In conclusion, an enantioselective oxidation of thioanisole
was realized with a bienzymatic couple alcohol oxidase/
peroxidase. Different parameters including the number of
units of alcohol oxidase, the ratio alcohol oxidase/peroxi-
dase and the concentration of methanol were optimized
leading to the (S)-methyl-phenyl-sulfoxide with high yield
and 75% ee. Similar results were obtained via the use of
crude extracts of alcohol oxidase prepared from culture
of P. pastoris. The method was shown to be efficient on a
gram scale preparation. There is no doubt that the condi-
tions optimized on the model compound thioanisole can
be applied to other aryl-methyl-sulfides, as was demon-
strated with the previously described bienzymatic systems.
4.3.2. Gram scale preparation. The initially turbid reac-
tion medium contained 10 mmol of thioanisole (1.24 g),
30 U of AOX as freeze-dried crude cell-free extract (equiv-
alent to 1.5 mL of extract) and 3.5 mL of CiP solution
(7 lmol) in 150 mL of phosphate buffer and 12 mL of
methanol (2 M). The mixture was vigorously stirred in an
open vessel at 25 ꢂC. After 24 h, the same amount of
enzymes and 6 mL of methanol were added, and the mix-
ture was stirred for another 24 h. After this additional time,
the reaction solution became completely limpid. This aque-
ous solution was subjected to a continuous extraction with
ethyl acetate for 48 h. The solvent was dried and carefully
evaporated to give 1.1 g of the sulfoxide (72%; 75% ee).
Acknowledgements
4. Experimental
4.1. Enzymes
We are grateful to Dr. Schneider (Novozymes) for kindly
supplying C. cinereus peroxidase and to Dr. Drago
(Applied Enzyme Technology Ltd) for supplying Pichia
pastoris alcohol oxidase. This work was supported by
scholarship to F.P. from the Mexican Government
CONACyT.
Purified alcohol oxidase from P. pastoris was obtained
from Applied Enzyme Technology Ltd (Leeds, UK) with
an activity of 1780 U/mL and from Sigma. Coprinus cine-
reus peroxidase was obtained from Novozymes as an aque-
ous solution. After dialysis against 0.5 M NaCl, the
peroxidase concentration was estimated to be mM from a
measure of absorbance at 404 nm (e = 108 mMꢀ1 cmꢀ1).
Specific activity: 208 mmol of ABTS (2,20-azinobis(3-ethyl-
benzothiazoline-6-sulfonate)) transformed/min/mL of
solution (2 mmol of ABTS transformed/mg of protein).
References
1. Recent reviews (a) Pellissier, H. Tetrahedron 2006, 62, 5559–
5601; (b) Fernandez, I.; Khiar, N. Chem. Rev. 2003, 103,
´
3651–3705; (c) Solladie, G. Heteroat. Chem. 2002, 13, 443–
452.

704
F. Pezzotti, M. Therisod / Tetrahedron: Asymmetry 18 (2007) 701–704
2. Lindberg, P.; Bra¨nstrom, A.; Wallmark, B.; Mattson, H.;
10. Dienys, G.; Jarmalavicius, S.; Budriene, S.; Citavicius, D.;
Sereikaite, J. J. Mol. Cat. B: Enzym. 2003, 21, 47–49.
11. Borzeix, F.; Monot, F.; Vandecasteele, J.-P. Enzyme Microb.
Technol. 1995, 17, 615–622.
12. Vijayakumar, A. R.; Csoregi, E.; Heller, A.; Gorton, L. Anal.
Chim. Acta 1996, 327, 223–234.
Rikner, L.; Hoffman, K. J. Med. Chem. Rev. 1990, 10, 1–54.
3. Caturla, F.; Amat, M.; Reinoso, R. F.; Calf, E.; Warrellow,
G. Bioorg. Med. Chem. Lett. 2006, 16, 3605–3608.
4. Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P.
Tetrahedron: Asymmetry 1992, 3, 95–106.
5. Carrea, G.; Redigolo, B.; Riva, S.; Colonna, S.; Gaggero, N.;
Battistel, E.; Bianchi, D. Tetrahedron: Asymmetry 1992, 3,
1063–1068.
13. Tuynman, A.; Vink, M. K. S.; Dekker, H. L.; Schoemaker,
H. E.; Wever, R. Eur. J. Biochem. 1998, 258, 906–913,
Oxidation of thioanisole is described on a lmol scale, with a
controlled slow addition of hydrogen peroxide (80% yield,
73% ee). Following the same method on a mmol scale
however, we were not able to obtain ee >45%.
´
6. (a) Okrasa, K.; Guibe-Jampel, E.; Therisod, M. J. Chem.
´
Soc., Perkin Trans. 1 2000, 1077–1079; (b) Okrasa, K.; Guibe-
Jampel, E.; Therisod, M. Tetrahedron: Asymmetry 2003, 14,
2487–2490.
14. In our hands, frozen cells were found to be much more stable
than frozen crude extracts; the former kept their activity for
several months, while extracts could only be stored for a few
days. Consequently, the cells were disrupted into crude
extracts when needed.
´
7. Okrasa, K.; Falcimaigne, A.; Guibe-Jampel, E.; Therisod, M.
Tetrahedron: Asymmetry 2002, 13, 519–522.
8. Pezzotti, F.; Okrasa, K.; Therisod, M. Tetrahedron: Asym-
metry 2005, 16, 2681–2683.
9. Duff, S. J. B.; Murray, W. D. Biotechnol. Bioeng. 1988, 31,
44–49.
15. Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P.
Tetrahedron: Asymmetry 1992, 3, 95–106.