Synthesis of S-licarbazepine by asymmetric reduction of oxcarbazepine with Saccharomyces cerevisiae CGMCC No. 2266

Article abstract of DOI:10.1016/j.molcatb.2011.07.004

S-licarbazepine was synthesized by asymmetric reduction of oxcarbazepine with CGMCC No. 2266. The optimum batch reduction conditions were found to consist of a reaction time of 36 h, temperature of 30 °C, and initial pH value of 7.0. The optimum concentration of the glucose co-substrate was found to be 0.3 mol L^-1. The addition of glucose contributed to in situ regeneration of NADPH in cells and improved conversion. Conversion increased with the addition of more biomass and with a decrease in the initial concentration of substrate. Within the membrane reactor, a continuous reduction process was used to improve production efficiency and reduce the inhibition of high-concentration substrate upon reduction. The optimum flux was found to be 20 ml h^-1. S-licarbazepine yield was 3.7678 mmol L^-1 d ^-1 in continuous reduction over four days. The enantiometric excess of S-licarbazepine was 100% for both batch and continuous reduction processes.

Full text of DOI:10.1016/j.molcatb.2011.07.004

Journal of Molecular Catalysis B: Enzymatic 72 (2011) 294–297
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Synthesis of S-licarbazepine by asymmetric reduction of oxcarbazepine with
Saccharomyces cerevisiae CGMCC No. 2266
Zhi-Min Oua,∗, Han-Bing Shib, Xing-Yuan Sunb, Wen-He Shenc
a Pharmaceuticals College, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, China
b The Third Affiliated Hospital of Qiqihar Medical College, Qiqihar, Heilongjiang 161000, China
c Zhejiang Jiuzhou Pharma Science & technology Co., Ltd., Hangzhou, Zhejiang 310051, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 May 2011
Received in revised form 18 June 2011
Accepted 4 July 2011
Available online 13 July 2011
S-licarbazepine was synthesized by asymmetric reduction of oxcarbazepine with CGMCC No. 2266. The
optimum batch reduction conditions were found to consist of a reaction time of 36 h, temperature of
30 ◦C, and initial pH value of 7.0. The optimum concentration of the glucose co-substrate was found to be
0.3 mol L−1. The addition of glucose contributed to in situ regeneration of NADPH in cells and improved
conversion. Conversion increased with the addition of more biomass and with a decrease in the initial
concentration of substrate. Within the membrane reactor, a continuous reduction process was used to
improve production efficiency and reduce the inhibition of high-concentration substrate upon reduction.
The optimum flux was found to be 20 ml h−1. S-licarbazepine yield was 3.7678 mmol L−1 d−1 in continu-
ous reduction over four days. The enantiometric excess of S-licarbazepine was 100% for both batch and
continuous reduction processes.
Keywords:
S-licarbazepine
Asymmetric reduction
Oxcarbazepine
Biotransformation
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
by asymmetric reduction of prochiral ketones through biotrans-
formation [19–22]. Asymmetric reduction with microorganism
S-licarbazepine (S-LC) can be used in synthesis of eslicar-
bazepine acetate by esterification with acetic acid. Eslicarbazepine
acetate [S-(−)-10-acetoxy-10,11-dihydro-5H-dibenz/-b,f/azepine-
5-car-Boxamide, ESL], formerly known as BIA 2-093, is a novel
central nervous system (CNS)-active compound with anticonvul-
sant activity [1–5]. It behaves as a voltage-gated sodium channel
blocker and is currently under clinical development for the treat-
ment of epilepsy and bipolar disorder [6–8]. In humans, ESL is more
rapidly reduced by liver esterases to the major active metabolite S-
licarbazepine than oxcarbazepine (OXC), a new antiepileptic drug
approved for the treatment of partial onset seizures and general-
ized tonic-clonic seizures [9–18].
catalysts has many advantages for the production of chiral com-
pounds. These include safety and reliability, low cost, scalability, a
lack of environmentally dangerous by-products, the fact that cell
culture facilities already exist in many laboratories, and the purity
of the final chiral product. In addition, the regeneration of coen-
zyme (NADH and NADPH) in situ can improve conversion because
of many kinds of enzyme exist in cells [23,24].
2. Experimental procedure
2.1. Reagents and instruments
This paper discusses in detail the synthesis of S-licarbazepine by
asymmetric reduction of oxcarbazepine with CGMCC No. 2266 as
a catalyst. This is the first report on biosynthesis of S-licarbazepine
using microorganism catalysts. Industry has adopted chemical syn-
thesis as a preferred method of production of S-licarbazepine. No
references exist on the enzymatic asymmetric reduction of OXC to
S-LC. S-licarbazepine can be obtained by asymmetric reduction of
oxcarbazepine within microorganisms (Fig. 1). Chiral alcohol, an
intermediate of many different pharmaceuticals, can be obtained
Oxcarbazepine and S-licarbazepine were purchased from Zhe-
jiang Jiuzhou Pharma Science & Technology Co., Ltd. (China). Agilent
HPLC 1200 equipped with Daicel OD-H chiral column from Daicel
Company was used in detection of substrate and product con-
tent and enantiometric excess of S-licarbazepine. The membrane
reactor and ultrafiltration membrane (MWCO 30 kDa) were from
Shanghai Mosu Company (China).
2.2. Microorganism cultivation
Saccharomyces cerevisiae CGMCC No. 2266 was obtained from
the soil in the vicinity of Hang Zhou West Lake Brewery and
preserved in China General Microbiology Culture Collection Cen-
∗
Corresponding author. Tel.: +86 0571 88320320; fax: +86 0571 88320320.
E-mail address: oozzmm@163.com (Z.-M. Ou).
1381-1177/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2011.07.004

Z.-M. Ou et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 294–297
295
O
HO
Reduction
N
N
NADPH
NADP
NH2
Oxcarbazepine
O
NH2
S-LC
O
Glucose dehydro-genase
Gluconic acid
Glucose
Esterification
O
CH3
Fig. 2. Continuous reduction process. (1) Substrate alcohol solution storage tank; (2)
membrane reactor for continuous reduction; (3) ultrafiltration membrane (MWCO
30 kDa); and (4) peristaltic pump.
O
N
acid before use. Percolating fluid from reactor 2 was pumped into
vessel 1 for cyclic utilization. Reaction volume, reaction tempera-
ture (28 ◦C), and stirring speed (160 r min−1) remained unchanged.
The flux pumped into vessel 1 was equal to that pumped out of ves-
sel 1. At the end of reduction, the liquid in vessel 1 was extracted
by ethyl acetate and the ethyl acetate layer was detected for deter-
mination of the substrate and product content by Agilent HPLC
1200.
NH2
O
ESL
Fig. 1. Synthesis of ESL by asymmetric reduction of OXC.
ter on November 26, 2007. Slant medium for strain storage was
composed of 10 g L−1 malt juice, −31g L−1 yeast extract, 5 g L−1 pep-
tone, 10 g L−1 glucose and 20 g L agar. The liquid yeast culture
med−iu1m was composed of 30 g L−1 glucose, 3 g L−1 yeast extract,
2.6. Production of S-licarbazepine by continuous reduction
5 g L
(NH4)2SO4, 0.25 g L−1 MgSO4, 1 g L−1 K2HPO4·3H2O, and
1 g L−1 KH2PO4. The strain picked from slant medium was inoc-
ulated into 100 ml liquid culture medium and cultivated in 30 ◦C
shaker (200 r min−1) for 24 h. Then 10 ml of the cell suspension was
inoculated into 100 ml liquid medium. After 24 h of cultivation, the
cells were harvested and used in reduction.
After 7.928 mmol L−1 of OXC dissolved in 100 ml alcohol solu-
tion was completely converted to S-LC, another 100 ml alcohol
solution containing 7.928 mmol L−1 of OXC was pumped into ves-
sel 2 for continuous reduction. The flux was 20 ml h−1. The process
was repeated until the cells in vessel 2 lost reduction activity.
2.3. Preparation of oxcarbazepine alcohol solution
3. Results and discussion
Oxcarbazepine alcohol solution was prepared by the addition of
oxcarbazepine to hot alcohol, which was then cooled to room tem-
perature. The concentration of the oxcarbazepine alcohol solutions
3.1. Influence of initial substrate concentration and reaction time
on batch reduction
were 0.793, 1.982, 3.964, 5.987, and 7.928 mmol L−1
.
Fig. 3 shows that conversion increased gradually with reaction
time and the rate of increase was very minor after 36 h. The opti-
mum reaction time is therefore 36 h. Conversion decreased with as
the concentration of initial substrate increased. This was probably
because the reduction activity of cell was inhibited by high con-
2.4. Batch reduction
The cultivated cells were separated from the culture medium
by centrifugation. The cells from the sediment were washed twice
with sterile water and separated by centrifugation before being
in reduction. The sediment obtained above and 10 ml oxcar-
bazepine alcohol solution were added to a flask containing 100 ml
of 0.2 mol L−1 phosphate buffered solution for reduction. The reac-
tion was stopped by centrifugation at the end of the reduction
period. The supernatant obtained above was extracted by ethyl
acetate. The ethyl acetate layer was detected for determination of
the substrate and product content by Agilent HPLC 1200.
100
90
80
70
60
2.5. Continuous reduction
50
-1
Initial substrate concentration 0.793 mmol·L
Initial substrate concentration 1.982 mmol·L
Initial substrate concentration 3.964 mmol·L
Initial substrate concentration 5.987 mmol·L
Fig. 2 shows the continuous reduction of oxcarbazepine in
the membrane reactor. One hundred milliliters of oxcarbazepine
alcohol solution was pumped from vessel 1 into a continuously
operated stirred membrane reactor (reactor 2). The flow rate of
oxcarbazepine alcohol solution was 10–50 ml h−1. The 200 ml of
reduction mixture in reactor 2 was composed of cells, 0.3 mol L−1
glucose, and phosphate buffer (pH 8.0). The cell dry weight in the
mixture was 28 g. The cells were trapped by an ultrafiltration mem-
brane (MWCO 30 kDa). The membrane was precoated with 1 mg
of bovine serum albumin per ml reactor volume to prevent cell
adsorption. The whole reactor was sterilized with 0.01% peracetic
-1
40
-1
-1
-1
30
Initial substrate concentration 7.928 mmol·L
20
10
20
30
40
50
60
70
Reduction time (h)
Fig. 3. Influence of initial substrate concentration and reaction time on conver-
sion 30 ◦C, initial pH value of reaction mixture 7.0, 160 r min−1, cell concentration
140 g L−1
.

296
Z.-M. Ou et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 294–297
Table 1
110
Effect of addition of glucose as co-substrate on conversion.
Glucose concentration (mol L−1
)
100
90
80
70
60
50
40
0
0.1
0.2
0.3
95.7
0.4
0.5
Conversion (%)
76.5
85.9
90.2
88.2
78.5
Initial substrate concentration 7.928 mmol L−1, 30 ◦C, initial pH value of reaction
mixture 7.0, 160 r min−1, 36 h, cell concentration 140 g L−1
.
Table 2
Effect of initial pH value of reaction mixture on reduction.
Initial pH value of reaction mixture
4
5
6
7
8
9
10
Conversion (%)
75.2
83.8
90.2
95.7
92.5
87.2
72.5
Initial substrate concentration 7.928 mmol L−1, 30 ◦C, 160 r min−1, 36 h, glucose con-
60
80
100
120
140
160
centration 0.3 mol L−1, cell concentration 140 g L−1
.
Cell concentration (%)
Fig. 4. Effect of cell concentration on reduction. Initial substrate concentration
Table 3
7.928 mmol L−1, 30 ◦C, initial pH value of reaction mixture 7.0, 160 r min−1, 36 h,
Effect of temperature on reduction.
glucose concentration 0.3 mol L−1
.
Temperature (◦C)
10
15
20
25
30
35
40
45
50
3.5. Effect of flux on continuous reduction
Conversion (%) 73.1 75.3 81.2 90.2 95.7 91.2 85.2 80.1 75.2
Initial substrate concentration 7.928 mmol L−1, initial pH value of reaction mixture
7.928 mmol L−1 of OXC dissolved in 100 ml alcohol solution was
completely converted to S-LC over 26 h of continuous reduction
at flux of 20 ml h−1. 7.928 mmol L−1 OXC in 10 ml alcohol solution
could not be completely converted to S-LC even after 36 h of batch
reduction. This showed continuous reduction to be more efficient
than batch reduction. The reduction activity of cells was inhibited
by high concentrations of substrate and product in batch reduction.
Continuous reduction did not encounter this problem because the
substrate and product were continuously pumped out of vessel 2.
Fig. 5 shows that the optimum flux in vessel 2 was 20 ml h−1. Com-
plete transformation of substrate was found to take longer when
flux was much higher or lower than 20 ml h−1. The enantiometric
excess of S-LC held steady at 100% and did not vary by flux.
7.0, 160 r min−1, 36 h, glucose concentration0.3 mol L−1, cell concentration 140 g L−1
.
centrations of substrate. The enantiometric excess of S-LC achieved
100% and was unaffected by the increase of initial substrate con-
centration.
3.2. Effects of addition of glucose as co-substrate on reduction
Glucose was selected as co-substrate to improve the conversion.
Table 1 shows that glucose helped to improve the conversion and
that the optimum glucose concentration was 0.3 mol L−1. Glucose
dehydrogenase in cells can convert glucose and NADP to gluconic
acid and NADPH, as shown in Fig. 1. NADPH can serve as a hydro-
gen donor and deliver H− to OXC to form S-LC. The addition of
glucose contributed to the regeneration of NADPH and improved
conversion. The enantiometric excess of S-LC was unaffected by the
addition of glucose and reached 100%.
3.6. Production of S-LC by continuous reduction for reuse of cells
Fig. 6 shows that cells used in continuous reduction processes
can be reused three times. Conversion reached 100% in the first,
second and third reductions and 80.2% in the fourth. This illus-
120
3.3. Effect of temperature and initial pH value on reduction
Tables 2 and 3 show the optimum temperature and initial pH
value of mixture to be conducive to high conversion rates. The initial
pH value of the mixture was adjusted by the addition of 0.2 mol L−1
of HCl and 0.2 mol L−1 of NaOH. The optimum temperature was
found to be 30 ◦C. The optimum initial pH value of mixture was 7.0.
The enantiometric excess of S-LC held steady at 100% under these
conditions.
Flux 50 ml·h -1
100
Flux 40 ml·h -1
Flux 30 ml·h -1
80
Flux 20 ml·h -1
Flux 10 ml·h -1
60
40
20
0
3.4. Effect of cell concentration on reduction
Fig. 4 shows that conversion increased with the concentration
of cells used in reduction. Higher conversion rates can be obtained
with more biomass because of the large amounts of reductase and
co-enzyme NADPH in these cells. The optimum cell concentration
was found to be 140 g L−1 (cell dry weight/mixture volume). The
mixture was too thick to be stirred well when the cell concentration
was higher than 140 g L−1. The enantiometric excess of S-LC held
steady at 100% and did not vary with cell concentration.
0
5
10
15
20
25
30
35
Reduction time (h)
Fig. 5. Effect of flux on continuous reduction. Initial substrate concentration
7.928 mmol L−1, 30 ◦C, initial pH value of reaction mixture 7.0, 160 r min−1, glucose
concentration 0.3 mol L−1
.

Z.-M. Ou et al. / Journal of Molecular Catalysis B: Enzymatic 72 (2011) 294–297
297
120
100
80
60
40
20
0
References
[1] L. Almeida, P. Soares-da-Silva, Neurotherapeutics 4 (2007) 88–96.
[2] A. Parada, P. Soares-da-Silva, Neurochem. Int. 40 (2002) 435–440.
[3] G. Alves, I. Figueiredo, M. Castel-Branco, A. Loureiro, A. Falcão, M. Caramona,
Anal. Chim. Acta 596 (2007) 132–140.
[4] M. Vaz-da-Silva, L. Almeida, A. Falcão, E. Soares, J. Maia, T. Nunes, P. Soares-da-
Silva, Clin. Ther. 32 (2010) 179–192.
[5] G. Sierra-Paredes, A. Nún˜ez-Rodriguez, A. Vázquez-López, T. Oreiro-García, G.
Sierra-Marcun˜o, Epilepsy Res. 72 (2006) 140–146.
[6] E. Perucca, J. French, M. Bialer, Lancet Neurol. 6 (2007) 793–804.
[7] D. Milovan, L. Almeida, M.K. Romach, T. Nunes, J.F. Rocha, M. Sokowloska, E.M.
Sellers, P. Soares-da-Silva, Epilepsy Behav. 18 (2010) 366–373.
[8] G. Sierra-Paredes, M.T. Oreiro-García, M.D. Vázquez-Illanes, G. Sierra-Marcun˜o,
Epilepsy Res. 77 (2007) 36–43.
[9] C. Di Resta, P. Ambrosi, G. Curia, A. Becchetti, Eur. J. Pharmacol. 643 (2010)
13–20.
[10] F. Donati, G. Gobbi, J. Campistol, G. Rapatz, M. Daehler, Y. Sturm, A.P.
Aldenkamp, Seizure 16 (2007) 670–679.
First reduction
Second reduction
Third reduction
Fourth reduction
0
20
40
60
80
100
120
140
160
180
200
Reduction time (h)
[11] J.S.E. Hellewell, J. Affect. Disorders 72 (2002) S23–S34.
[12] A.K. Kaddurah, G.L. Holmes, Epilepsy Behav. 8 (2006) 289–293.
[13] K. Linnet, A. Steentoft, K.W. Simonsen, A. Sabers, S.H. Hansen, Forensic Sci. Int.
177 (2008) 248–251.
Fig. 6. Production of ESL by continuous reduction with cell reuse.
[14] R.M. Stepanovic´-Petrovic´, M.A. Tomic´, S.M. Vucˇkovic´, G. Poznanovic´, N.D.
trates that the cells can be reused reasonably reliably for about
152 h. S-licarbazepine yield was 3.7678 mmol L−1 d−1 in continu-
ous reduction over four days. The enantiometric excess of ESL was
100% in continuous reduction process.
ˇ
Ugresˇic´, M.S. Prostran, B. Bosˇkovic´, Pharmacol. Biochem. Behav. 97 (2011)
611–618.
ˇ
[15] M.A. Tomic´, S.M. Vucˇkovic´, R.M. Stepanovic´-Petrovic´, N. Ugresˇic´, M.S. Prostran,
B. Bosˇkovic´, Pain 111 (2004) 253–260.
[16] J. Sawynok, A.R. Reid, B.B. Fredholm, Neurosci. Lett. 473 (2010) 178–181.
[17] W. Martinez, A. Ingenito, M. Blakeslee, G.L. Barkley, K. McCague, J. D’Souza,
Epilepsy Behav. 9 (2006) 448–456.
[18] M. Mazza, G. Della Marca, M. Di Nicola, G. Martinotti, G. Pozzi, L. Janiri, P. Bria,
S. Mazza, Epilepsy Behav. 10 (2007) 397–401.
4. Conclusions
[19] E.B. Kurbanoglu, K. Zilbeyaz, M. Ozdal, M. Taskin, N.I. Kurbanoglu, Bioresource
Technol. 101 (2010) 3825–3829.
[20] S. Bräutigam, D. Dennewald, M. Schürmann, J. Lutje-Spelberg, W.-R. Pitner, D.
Weuster-Botz, Enzyme Microb. Technol. 45 (2009) 310–316.
[21] T. Ema, H. Yagasaki, N. Okita, M. Takeda, T. Sakai, Tetrahedron 62 (2006)
6143–6149.
[22] B.B. Zhang, W.Y. Lou, M.H. Zong, H. Wu, J. Mol. Catal. B: Enzym. 54 (2008)
122–129.
[23] T. Matsuda, Y. Yamagishi, S. Koguchi, N. Iwai, T. Kitazume, Tetrahedron Lett. 47
(2006) 4619–4622.
S-licarbazepine was synthesized by batch and continuous
reduction of oxcarbazepine with CGMCC No. 2266. The optimum
reaction time and temperature were found to be 36 h and 30 ◦C. The
optimum initial reaction mixture pH was found to be 7.0. Conver-
sion increased with increased reaction time and the addition more
biomass. Conversion decreased with the addition more substrate.
In order to neutralize the inhibitive effect of high concentrations of
substrate, a continuous reduction process was used to produce S-LC
continuously. The S-licarbazepine yield was 3.7678 mmol L−1 d−1
after four days of continuous reaction. The enantiometric excess of
S-LC was 100% in both batch and continuous reduction process.
[24] Z. Ou, J. Wu, L. Yang, P. Cen, Korean J. Chem. Eng. 25 (2008) 124–128.
Acknowledgements
We thank the Natural Science Foundation of Zhejiang province
(Y4110130) for its financial support.