Enantioselective synthesis of sulfoxides catalysed by an oxidase-peroxidase bienzymatic system

Article abstract of DOI:10.1016/S0957-4166(02)00142-8

We have devised a new bienzymatic oxidase-peroxidase system in which hydrogen peroxide is produced from air by a D-amino acid oxidase. The generated hydrogen peroxide is instantly used for asymmetric oxygenation of sulphides by a peroxidase. In the absenc

Full text of DOI:10.1016/S0957-4166(02)00142-8

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 13 (2002) 519–522
Enantioselective synthesis of sulfoxides catalysed by an
oxidase–peroxidase bienzymatic system
Krzysztof Okrasa,a,b Aude Falcimaigne,b Eryka Guibe´-Jampelb and Michel Therisodb,*
aZTiBSL, Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-662 Warsaw, Poland
bLab. Chimie Bioorganique et Bioinorganique, I. C. M. O.-Bat. 420, Universite´ Paris-Sud, F-91405 Orsay cedex, France
Received 22 February 2002; accepted 15 March 2002
Abstract—We have devised a new bienzymatic oxidase–peroxidase system in which hydrogen peroxide is produced from air by a
D
-amino acid oxidase. The generated hydrogen peroxide is instantly used for asymmetric oxygenation of sulphides by a
peroxidase. In the absence of peroxidase no spontaneous oxidation is observed, indicating that hydrogen peroxide production by
the -amino acid oxidase is fully controlled. Arylmethyl sulfoxides can be synthesised on a preparative scale with high yields and
D
good enantiomeric excess. © 2002 Published by Elsevier Science Ltd.
1. Introduction
Bienzymatic systems coupling an oxidase producing
H2O2 and a peroxidase can allow ‘in vivo like’ oxida-
tions by atmospheric oxygen.
There is an ever increasing demand for more selective
and clean oxidation catalysts, especially for the produc-
tion of enantiomerically pure compounds. Peroxidases
are widely distributed enzymes which have a consider-
able scope for undertaking potentially useful transfor-
mations in organic synthesis and their applications have
been extensively reviewed.1–3 However, there is no
example of their use as catalysts in industrial organic
synthesis. Limited commercial availability, high price
and low operational stability have hampered their
application. Nowadays, new ‘ready to use’ inexpensive
industrial peroxidases have appeared on the market.
Nevertheless, one major problem is the inhibition of
these heme-containing enzymes by an excess of hydro-
gen peroxide, their substrate. In addition to oxidation
of usual substrates (phenols, aromatic amines) involv-
ing single electron transfers, peroxidases can also
catalyse the enantioselective transfer of oxygen from
hydrogen peroxide (sulfoxidation, epoxidation).1–3
Improvement of both operational stability of perox-
idase and enantioselectivity of the oxygen transfer by
maintaining concentration of hydrogen peroxide at a
low level can be achieved either by step-wise or contin-
uous feed on demand, or in situ generation of hydrogen
peroxide.1
In our previous work,4 we devised a bienzymatic system
in which hydrogen peroxide is progressively generated
in situ at the expense of glucose and oxygen (air) by
glucose oxidase to effect the preparative enantioselec-
tive oxidation of thioanisoles by a peroxidase from
Coprinus cinereus. This procedure has also been
attempted with chlorperoxidase.5,6
In those systems, the pH of the reaction medium has to
be kept close to the optimum pH of the enzymes (5.5–7
for glucose oxidase); the use of a pH-stat is therefore
required, since the production of hydrogen peroxide is
concomitant to the synthesis of gluconic acid. Glucose
oxidase (GOD, specific activity: 10 U/mg) is hardly
inhibited by H2O2. As a result, an excess of hydrogen
peroxide can be produced if the peroxidase becomes
deactivated during the process, and non-enzymatic sub-
strate oxidation may take place.
In principle, other oxidases can be used in this
approach which allow a fine tuning of the oxidase–per-
oxidase combination in terms of substrate, optimum
pH and controlled H2O2 production. Therefore, we
investigated the in situ H2O2 production by
D-amino
acid oxidase (DAO; EC 1.4.3.3), a flavin-containing
enzyme which catalyses the oxidative deamination of
D
-amino acids to the corresponding alpha-ketoacids.
* Corresponding author. Tel.: +33-169-156311; fax: +33-169-157281;
e-mail: therisod@icmo.u-psud.fr
Reoxidation of the flavin coenzyme by molecular oxy-
0957-4166/02/$ - see front matter © 2002 Published by Elsevier Science Ltd.
PII: S0957-4166(02)00142-8

520
K. Okrasa et al. / Tetrahedron: Asymmetry 13 (2002) 519–522
gen is accompanied by the release of H2O2. DAO can
catalyse the deamination of a wide spectrum of
amino acids. -Alanine is a good substrate, but inex-
pensive DL-alanine can be used as well, leading to
-alanine and ammonium pyruvate as side products.
ity, and to be inhibited by the substrates and products
of peroxidase. DAO from Trigonopsis variabilis has its
principal industrial application in the preparation of
semisynthetic cephalosporin antibiotics.10 It has also
been used for the kinetic resolution of DL-amino acids,
the production of a-ketoacids7 and in analytical chem-
istry.11 This enzyme is commercially available from
Fluka in an immobilised form on polymeric support.
Fungal peroxidase from Coprinus cinereus (CiP; EC.
1.11.1.7) is a low cost industrial enzyme produced by
Novozyme as an additive for laundry detergent (liquid
preparation; specific activity: 231 kU/ml). CiP has
excellent thermostability and a broad pH optimum (5
to 9). It has been investigated as a catalyst in sulfoxida-
tion of thioanisoles4,12,13 and epoxidation of styrene
derivatives (on a microscale).14
D
-
D
L
Data from the literature indicate deactivation of DAO
occurs in the presence of hydrogen peroxide, but this
can be avoided by catalase addition.7–9 Our preliminary
experiments have shown that only limited amount of
H2O2 can be produced by DAO in the absence of
peroxidase. Thus, low-cost substrate, constancy of reac-
tion medium pH and limitation of hydrogen peroxide
overproduction make DAO an interesting alternative
for a new bienzymatic oxidase–peroxidase system.
Moreover, DAO optimum pH is 7 to 9. In terms of pH,
the DAO–peroxidase system is then complementary to
the glucose oxidase–peroxidase couple previously
described. For our study, the well documented oxida-
tion of thioanisoles was adopted as a model (Scheme 1).
The oxidation rate of thioanisole was sensitive to the
amount of both TvDAO and CiP used (Fig. 1). In the
DAO/Catalase-mediated resolution of DL-phenylala-
nine authors have recommended an enzyme ratio of 1
U/2.6 kU.8 Similarly, in our experiments, satisfactory
rates of sulfoxidation could be obtained only when a
large amount of peroxidase was used. Optimisation of
the reaction led us to use a TvDAO/CiP ratio of 1 U/66
kU. This value was chosen for all our investigations.
Under these conditions only a small excess of
D-amino
acid was necessary. The yield of sulfoxide relative to the
starting alanine was greater than 65%, indicating a low
catalase activity in the reaction medium. The flux of
hydrogen peroxide produced by the DAO-mediated
oxidation of DL-alanine could be evaluated by titration
of the pyruvate formed (Fig. 2). The kinetics of pyru-
vate formation was monitored spectrophotometrically
Scheme 1.
using an
L-lactate dehydrogenase-catalysed oxidation of
2. Results and discussion
NADH.15
Two DAOs were tested: from Porcine Kidney
(PKDAO; Sigma, 0.15 U/mg) and from yeast Trigonop-
sis variabilis (TvDAO; Fluka, 28 U/g). Unfortunately,
PKDAO was found to have too low operational stabil-
Figure 2. Formation of pyruvate from DL-alanine (2.8 mmol)
and of methylphenyl sulfoxide from thianisole (1 mmol)
catalysed by the TvDAO/CiP system ꢀ: pyruvate in the
presence of CiP; ꢁ: sulfoxide in the presence of CiP; ꢂ:
pyruvate in the absence of CiP ꢃ: sulfoxide formation when
CiP was added after 120 min.
Figure 1. Optimisation of DAO/CiP ratio in sulfoxidation of
1a (1 mmol) in H2O pH 8.5 at rt. Reaction time 4 h. ꢀ
TvDAO in the presence of 231 kU of CiP, ꢂ CiP in the
presence of 3.5 U of TvDAO.

K. Okrasa et al. / Tetrahedron: Asymmetry 13 (2002) 519–522
521
reaction with glycine can be interesting for difficult
peroxidase-catalysed oxygenation like epoxidation,
where a lower flux of H2O2 is required.14
In the absence of CiP, formation of pyruvate from
DL-alanine (2.8 mmol) and conversion of sulphide was
less than 7%, and conversion of sulphide (1 mmol) to
less than 2%. This is probably a consequence of retroin-
hibition of DAO by H2O2 and/or hydrogen peroxide
oxidation of pyruvate.16 The normal rate of sulfoxida-
tion can be restored even after 2 h by addition of CiP
(Fig. 2, Table 1).
Three arylmethyl sulphides were tested as substrates of
the peroxidase in various conditions (Table 1). TvDAO
was added at once to thioanisole dispersed in aqueous
solution of CiP and amino acid at pH 8.5. The mixture
was mechanically stirred under atmospheric air or in a
closed vessel under oxygen.17 Reaction medium pH is
constant during the process, thus avoiding any buffer
utilisation. All obtained chiral arylmethyl sulfoxides
have the S absolute configuration.4,12
At optimum TvDAO/CiP ratio, the rate of sulfoxide
formation can be influenced by temperature and choice
of amino acid. (Fig. 3)
The best result was obtained with
strate of TvDAO. Only a slightly lower rate was
observed with the more economical DL-alanine. Slower
D-alanine as a sub-
Compared with the GOD-mediated oxidation of 1a
(entry 1), where 43% of racemic 2a is formed in the
absence of peroxidase, the reaction catalysed by
TvDAO in the same conditions (entry 3) leads to less
than 2% of 2a. Under the optimal oxygenation condi-
tions (entry 9–12) conversion is >90% and enantiomeric
excesses of sulfoxides are the best reported for CiP-
catalysed sulfoxidation. The enantiomeric purities of
sulfoxides 2b and 2c are higher than those previously
obtained in tandem glucose oxidase/CiP oxidation.4
Moreover, the enantiomeric excesses stay constant dur-
ing the reaction (entry 8) and no further oxidation to
the sulfone was observed. CiP can also be replaced by
microperoxidase-11 (entry 13), but the reaction leads to
racemic sulfoxide.19
The reaction can be easily scaled up and carried out on
a multigram scale without decrease of yield or enan-
tiomeric excess (entry 14). The immobilised DAO could
be reused several times without significant loss of activ-
ity.20 The total turnover number of CiP in this system
and for oxidation of thioanisole is superior to 500
(versus 100–200 for the same reaction catalysed by the
same enzyme fed directly with hydroperoxide12,13).
Figure 3. Formation of methylphenyl sulfoxide using various
substrates of TvDAO ": DL-alanine at 40°C, ꢀ:
D
-alanine at
25°C, ꢄ: DL-alanine at 25°C, ꢃ:
ꢂ: glycine at 25°C.
D-phenylalanine at 25°C,
Table 1. Asymmetric oxidation of arylmethyl sulfides 1a-c catalysed by oxidase/CiP systems
Entry
Sulphide
Oxidase substrate
(mmol)
Oxidase (mg)
Peroxidase (ml) Temp. (°C)
Time (h)
Conversion (%)
(yieldd)
Ee (%)
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1a
1a
D
D
-Glucose, 2.00
-Glucose, 2.00
GOD, 4
GOD, 4
–
1
–
1
1
1
1
1
Rt
Rt
Rt
Rt
Rt
40
Rt
Rt
6
6
6
6
6
24
6
2
6
10
6
43
90
2
60
48
40
89
38
79 (70)
93 (85)
97
82
88 (76)
56
0
72
0
DL-Ala, 5.60
DL-Ala, 5.60
TvDAO, 125
PKDAO, 23
TvDAO, 125
TvDAO, 125
TvDAO, 125
TvDAO, 125
77
56
76
75
75
76
79
73
96
97
0
D
-Phe, 2.80
Gly, 2.00
-Ala, 2.80
D
DL-Ala, 5.60
9
1a
1a
1b
1c
1a
1a
DL-Ala, 2.80
DL-Ala, 2.80
DL-Ala, 2.80
DL-Ala, 5.60
DL-Ala, 0.28
DL-Ala, 42.00
TvDAO, 125
TvDAO, 125
TvDAO, 125
TvDAO, 125
TvDAO, 3
1
1
1
1
Rt
40
Rt
40
Rt
40
10
11
12
13a
14b
8
6
10
10
0.5 mg
TvDAO, 1875 15
93 (86)
75c
Reaction conditions: sulfide 1 (1 mmol), water (20 mL), pH 8.5
a Microperoxidase-11 (Sigma), water (2 mL), sulfide 1a (0.1 mmol).
b Sulfide 1a (15 mmol); water (200 mL).
c [h]=−110 (c=3, EtOH; lit.:18 [h]=+146 for the (R)-enantiomer)
d Yield of pure isolated product.

522
K. Okrasa et al. / Tetrahedron: Asymmetry 13 (2002) 519–522
4. Okrasa, K.; Guibe´-Jampel, E.; Therisod, M. J. Chem.
3. Experimental
Soc., Perkin Trans. 1 2000, 1077–1079.
5. Neidleman, S. L.; Amon, W. P., Jr.; Geigert J.; USP 4
284 723 (Chem. Abstr. 1981, 95, 167173).
6. van de Velde, F.; Lourenco, N. D.; Bakker, M.; van
Rantwijk, F.; Sheldon, R. A. Biotech. Bioeng. 2000, 69,
286–291.
7. Trost, E. M.; Baskir, I.; Fisher, L. 5th International
Symposium on Biocatalysis and Biotransformation,
Darmstadt, September 2001
8. Fernandez-Lafuente, R.; Rodrigez, V.; Guisan, J. M.
Enzym. Microb. Technol. 1998, 23, 28–33.
9. Gabler, M.; Hensel, M.; Fisher, L. Enzym. Microb. Tech-
nol. 2000, 27, 605–611.
3.1. Typical procedure
The sulphide (1 mmol) and immobilised TvDAO (125
mg) were suspended in a mixture of aqueous DL-alanine
solution (20 mL, 2.8 mmol) and CiP (1 mL, previously
dialysed against 0.5 M NaCl). The pH of the mixture
was adjusted to 8.5 and the mixture was mechanically
stirred either at rt or 40°C until completion of the
reaction (Table 1). The formed sulfoxide was then
extracted with ethyl acetate, and purified if necessary
on a short silica gel column. The enantiomeric excess
was determined by HPLC on a Chiracel OD-H column
(iso-hexane/propan-2-ol 95:5). The degree of conversion
and chemical purity of the sulfoxides was determined
by GC and NMR.
10. Dey, E. S.; Miller, J. R.; Kovacevic, S.; Kosbach, K.
Biochem. Mol. Biol. Int. 1990, 20, 1169–1178.
11. Xu, D. K.; Lin, H.; Chen, H. Y. Anal. Chim. Acta 1997,
349, 215–219.
12. Tuynman, A.; Vink, M. K. S.; Dekker, H. L.; Schoe-
maker, H. E.; Wever, R. Eur. J. Biochem. 1998, 256,
906–913.
13. Adam, W.; Mock-Knoblanch, C.; Saha-Moller, C. R. J.
Org. Chem. 1999, 64, 4834–4839.
14. Tuynman, A.; Luftje-Spelberg, J.; Kooter, I. M.; Schoe-
maker, H. E.; Wever, R. J. Biol. Chem. 2000, 275,
3025–3030.
15. von Korff, R. W. In Methods in Enzymology; Lowen-
stein, J. M., Ed.; Academic Press: New York, 1969; Vol.
XIII, pp. 520–521.
16. Greenstein, J. P.; Winitz, M. Chemistry of the Amino
Acids; John Wiley and Sons: New York, 1961; Vol. 2, pp.
1782–1783.
17. In order to prevent loss of volatile substrates, use of a
closed vessel under an atmosphere of oxygen is the best
choice.
4. Conclusions
The devised TvDAO/CiP system mediates the prepara-
tive oxidation of sulphides by oxygen, leading to the
corresponding sulfoxides in high yield and with good
enantiomeric excess. Immobilised oxidase and its low
cost substrates ensure a ‘cheap and green’ synthetic
entry to chiral sulfoxides. Product isolation is easy and
background non-selective oxidation, often observed
with slow reacting peroxidase, is suppressed by self-con-
trolled hydrogen peroxide production.
Acknowledgements
We are grateful to Dr. Wohlgemuth and Fluka for the
supply of TvDAO, and to Dr. Schneider and
Novozyme for the supply of CiP.
18. Folli, U.; Iarossi, D.; Montanari, F.; Torre, G. J. Chem.
Soc. C 1968, 1317–1322.
19. Colonna, S.; Gaggero, N.; Carrea, G.; Pasta, P. Tetra-
hedron Lett. 1994, 48, 9103–9104; the authors report an
enantiomeric excess of 20% in oxidation of thioanisole
catalysed by MP11 but with a very high MP-11/substrate
ratio of 1/6.
20. Immobilised commercial TvDAO loses 30% activity after
the first washing with water, probably as an indication of
the presence of some non-immobilised enzyme adsorbed
on the support. No further loss of activity of recycled
TvDAO was observed even after three runs.
References
1. van de Velde, F.; van Rantwijk, F.; Sheldon, R. A.
Trend. Biotechnol. 2001, 19, 73–80.
2. Colonna, S.; Gaggero, N.; Richelni, C.; Pasta, P. Trend.
Biotechnol. 1999, 17, 163–168.
3. van Deurzen, M. P. J.; van Rantwijk, F.; Scheldon, R. A.
Tetrahedron 1997, 53, 13183–13220.