Stereoselective introduction of two chiral centers by a single diketoreductase: An efficient biocatalytic route for the synthesis of statin side chains

Article abstract of DOI:10.1007/s00726-009-0390-0

Statins, including atorvastatin (Lipitor), are the top-selling drugs in the world. The biocatalytic production of chiral side chains of statin drugs is of great interest to academia and industry. Stereoselective double reduction of a β,δ-diketo ester catalyzed by a diketoreductase offers a simple and efficient route for the preparation of statin side chains. Comparison of different cofactor regeneration systems resulted in an easy and cost-effective process for this enzymatic reduction.

Full text of DOI:10.1007/s00726-009-0390-0

Amino Acids (2010) 39:305–308
DOI 10.1007/s00726-009-0390-0
SHORT COMMUNICATION
Stereoselective introduction of two chiral centers by a single
diketoreductase: an efficient biocatalytic route for the synthesis
of statin side chains
•
•
•
Xuri Wu Lili Wang Shuzhen Wang
Yijun Chen
Received: 29 July 2009 / Accepted: 4 November 2009 / Published online: 16 November 2009
Ó Springer-Verlag 2009
Abstract Statins, including atorvastatin (LipitorÒ), are
the top-selling drugs in the world. The biocatalytic pro-
duction of chiral side chains of statin drugs is of great
interest to academia and industry. Stereoselective double
reduction of a b,d-diketo ester catalyzed by a diketore-
ductase offers a simple and efficient route for the prepa-
ration of statin side chains. Comparison of different
cofactor regeneration systems resulted in an easy and cost-
effective process for this enzymatic reduction.
product with ee value of 98.1% (Wolberg et al. 2000;
Wolberg et al. 2001). 2-Deoxy-D-ribose-5-phosphate
aldolase (DERA), on the other hand, showed the ability to
introduce two hydroxyl groups by two sequential aldol
reactions with high stereoselectivity (ee [ 99.9%,
de = 96.6%), and the biocatalytic process is under devel-
opment (Barbas et al. 1990; Gijsen and Wong 1994;
Greenberg et al. 2004). Meanwhile, an Acinetobacter
species SC 13874 was found to reduce both b- and
d-carbonyl groups of ethyl-6-(benzyloxy)-3,5-dioxohex-
anoate (1) to its corresponding ethyl 3R,5S-6-(benzyloxy)-
3,5-dihydroxy hexanoate (2) with 99% ee and 63% de (Guo
et al. 2006). Later, a ketoreductase responsible for the
conversion was cloned from Acinetobacter sp. SC 13874
and expressed in Escherichia coli to show excellent ste-
reoselectivity, and the study on this enzyme is being
undertaken (Goldberg et al. 2008). In addition, nitrilase has
been considered as another option for the synthesis of statin
side chains, but there are several challenges in this method,
such as the difficulty of synthesizing 3-hydroxyglutaronit-
rile and the problems associated with the isolation of the
product (Bergeron et al. 2006). Therefore, development of
a practical biocatalytic route for statin side chains is of
great interest to both academia and industry.
Keywords Biocatalysis Á Chiral side chain Á
Diketoreductase Á Oxidoreductase Á Statin
Introduction
Statins, such as atorvastatin (LipitorÒ), the top-selling
cholesterol-lowering drugs in the world (Mu¨ller 2005;
Thayer 2006), contain two chiral centers in the
b,d-dihydroxyheptanoate side chains. To pursue a practical
biocatalytic process for asymmetric synthesis of statin side
chains, a number of approaches have been explored. For
example, a strain of Lactobacillus brevis was reported to
convert d-keto group of b,d-diketo ester to d-hydroxy
We recently reported the cloning, expression and char-
acterization of a novel diketoreductase (GenBank acces-
sion no. EU273886) from Acinetobacter baylyi ATCC
33305. This recombinant diketoreductase (rDKR) is able to
stereoselectively reduce 1–2 with both de and ee greater
than 99.5% (Wu et al. 2008, 2009). Compound 2 is a
valuable intermediate in the synthesis of statin side chains
(Scheme 1). Herein, we present an efficient and practical
route for the preparation of statin side chains using whole
cells containing over-expressed rDKR with optimized
cofactor regeneration system. After purification of 2, ethyl
Electronic supplementary material The online version of this
article (doi:10.1007/s00726-009-0390-0) contains supplementary
material, which is available to authorized users.
X. Wu Á L. Wang Á S. Wang Á Y. Chen (&)
Laboratory of Chemical Biology,
School of Life Science and Technology,
China Pharmaceutical University, 24 Tongjia Street,
210009 Nanjing, Jiangsu, People’s Republic of China
e-mail: yjchen@cpu.edu.cn; yjchen_cpu@yahoo.com.cn
123

306
X. Wu et al.
3R,5S-6-hydroxy-3,5-O-isopropylidene-3,5-dihydroxy-
hexanoate (6) was obtained with high yield through
hydrogenolysis of ethyl 3R,5S-6-(benzyloxy)-3,5-O-isopro-
pyli-dene-3,5-dihydroxyhexanoate (5) that was synthesized
from 2 (Scheme 1).
1 ranging from 1 to 15 g/L, 0.5 mM NADP?, 5% 2-pro-
panol, 308 U rDKR in recombinant cells and 0.1 M potas-
sium phosphate buffer (pH 6.0) in a final volume of 100 ml.
All reactions were incubated at 25°C and 200 rpm for 18 h.
Samples were taken and analyzed by HPLC (Wu et al. 2008,
2009; Liu et al. 2008).
Materials and methods
Reduction of 1 by rDKR with glucose as co-substrate
Instruments and materials
Thirty grams of glucose and 1 g of 1 dissolved in 10 ml
ethanol were added to rDKR fermentation broth. The
reaction was incubated at 25°C and 200 rpm. Each 0.5 ml
of sample was taken and analyzed by HPLC. The reaction
mixture was extracted by ethyl acetate for three times.
Chromatography on silica gel (ethyl acetate/petroleum
ether 10:90 v/v) gave 2 as a white powder.
HPLC was performed on Shimadzu 2010A HT with UV
detector. NMR spectra were recorded at 300 K on a Bruker
AV-500 spectrometer (Bruker Optics, Germany). High
resolution mass spectra (HRMS) were recorded on Agilent
HPLC/ESI-Q-TOFMS spectrometer. Optical rotation was
obtained on a Pekin-Elmer 431 Polarimeter. NADH,
NADP? and IPTG (Isopropyl-b-D-thiogalactopyranoside)
were purchased from Sigma (USA). The expression of
rDKR in E. coli and activity assay were performed as
described previously (Wu et al. 2008, 2009).
Preparation of 5
The preparation of 5 was performed according to the
literature (Bode et al. 2002). 2, 2-dimethoxypropane
(20.8 g, 200 mmol) and a catalytic amount of camphorsul-
fonic acid at room temperature were added to a solution of 2
(5.64 g, 20 mmol) in acetone (50 ml). After stirring for 4.5 h,
the volatiles were evaporated. The residue was dissolved
in ethyl acetate and washed with saturated NaHCO3 and NaCl
solution, and dried over Na2SO4. Chromatography on sil-
ica gel (ethylacetate/petroleum ether 1:10v/v)gave5 (6.02 g,
93.5% yield) as a colorless oil.
Reduction of 1 by rDKR with 2-propanol
as co-substrate
The reaction catalyzed using crude rDKR contained 1 g/L
of 1, 0.5 mM NADP?, 2-propanol ranging from 1 to 5%,
0.7 U crude rDKR and 0.1 M potassium phosphate buffer
(pH 6.0) in a final volume of 0.5 ml. When whole cells were
used, reaction mixtures contained various concentrations of
Scheme 1 Biocatalytic
O
OH O
synthesis of statin side chain.
Compound 2 was obtained by
rDKR with both de and ee
values greater than 99.5%.
O
OEt
3
a 2, 2-dimethoxypropane, cat.
camphorsulfonic acid, acetone,
25°C, 7 h, 93.5%; b 20 atm H2,
cat. 10% Pd/C, ethanol, 25°C,
18 h, 100%
O
O
O
OH OH
O
O
O
OEt
OEt
2
1
ee > 99.5%, de > 99.5%
OH
O
O
O
O
OEt
4
O
O
O
O
O
(a)
(b)
HO
O
OEt
OEt
6
5
HMG-CoA reductase
inhibitors (vastatins)
123

Stereoselective introduction of two chiral centers
307
Scheme 2 Different cofactor
regeneration systems for the
biocatalytic process by rDKR.
a 2-Propanol was used to
regenerate cofactors by rDKR;
b glucose was used for cofactor
regeneration
(a)
OH OH
O
O
O
O
rDKR
O
O
OEt
OEt
O
2
1
NAD(P)+
OH
ee > 99.5%, de > 99.5%
NAD(P)H
rDKR
rDKR
(b)
OH OH
O
O
O
O
O
O
OEt
OEt
2
1
NAD(P)+
NAD(P)H
ee > 99.5%, de > 99.5%
Gluconic acid
D-Glucose
Escherichia coli cells
(Scheme 2a) in the presence of 2-propanol and NADP?,
resulting in high conversion within 3 h by resting cell
suspensions (Supplementary Fig. 1). In order to maximize
the catalytic capacity of this biocatalyst, different substrate
concentrations were evaluated for this cofactor regenera-
tion system, and we found that higher conversion rates
showed at relatively lower substrate concentrations, and
more accumulation of monohydroxy intermediates was
observed at early phase of the reaction. Nevertheless, at
10 g/L substrate concentration, the conversion was finally
reached to 93.5% (Supplementary Table 1).
Preparation of 6
The preparation of 6 was conducted according to reported
method (Sun et al. 2007). Pd/C (10%, 0.35 g) was added to
a solution of 5 (3.64 g, 11.3 mmol) in ethanol (100 ml)
contained in a 250 ml autoclave. The autoclave was filled
with H2 (20 atm). After stirring for 18 h at room temper-
ature, the catalyst was filtered and the solvent was evapo-
rated. The residue was a colorless oil 6 (2.62 g, 100%
yield).
To further facilitate the cofactor regeneration system
and to simplify the biocatalytic process, an in situ reduction
of 1 by whole cells containing over-expressed rDKR was
investigated for the preparation of 2 since such system does
not require exogenous addition of cofactors. Recombinant
E. coli cells were grown in a flask, and subsequently
induced by IPTG for the expression of rDKR. After 14 h,
cell mass was reached to approximately 10%. Thereafter,
glucose and substrate 1 dissolved in ethanol were directly
added to the fermentation broth for the biocatalytic con-
version. Given the fact that E. coli cells can simultaneously
regenerate both NADH and NADPH by glycolysis and
endogenous membrane-bound glucose dehydrogenase and
glucose-6-phosphate dehydrogenase (Scheme 2b), the cells
had enough capacity to regenerate the cofactors by serving
glucose as co-substrate. Thus, the biocatalytic process was
optimized to a conversion yield of 95% in 4 h without
additional cofactors (Supplementary Fig. 2).
Results and discussion
Previously, formate dehydrogenase (FDH) regeneration
system was used to regenerate NADH over the course of
conversion of 1, 93% of transformation yield was observed
within 3 h with the appearance of intermediates R-ethyl
6-(benzyloxy)-3-hydroxy-5-oxohexanoate (3) and S-ethyl
6-(benzyloxy)-5-hydroxy-3-oxohexanoate (4) (Wu et al.
2008, 2009). However, the requirement of FDH for NADH
regeneration in the biocatalytic reaction makes the process
costly and complicated. To overcome such shortcomings,
we took advantage of the dual cofactor specificity exhibited
by rDKR (Wu et al. 2008, 2009) to explore the possibility
of utilizing 2-propanol as a co-solvent for the substrate and
as a co-substrate for the regeneration of NAD(P)H (Inoue
et al. 2005; Findrik et al. 2005). At the beginning, only
\5% conversion of 1 was obtained when crude rDKR was
utilized as a catalyst to reduce 1 with 2-propanol (1–5% in
volume) mediated NADPH regeneration system (data not
shown). This low conversion was likely due to the insta-
bility of rDKR against organic solvents. Then, a different
approach with recombinant E. coli whole cells containing
over-expressed rDKR was taken to conduct the conversion
Since 6 is a more advanced intermediate in the synthesis
of statins, we conducted two more chemical steps to
demonstrate the feasibility of our approach for making the
side chains. After extraction and purification of 2 from the
biocatalytic reaction, acid-catalyzed protection reaction
was carried out to protect the hydroxyl groups to form 5.
123

308
X. Wu et al.
Gijsen HJM, Wong CH (1994) Unprecedented asymmetric aldol
reactions with three aldehyde substrates catalyzed by 2-deoxyr-
ibose-5-phosphate aldolase. J Am Chem Soc 116:8422–8423
Goldberg S, Guo ZW, Chen S, Goswami A, Patel RN (2008)
Synthesis of ethyl-(3R,5S)-dihydroxy-6-benzyloxyhexanoates
via diastereo- and enantioselective microbial reduction: cloning
and expression of ketoreductase III from Acinetobacter sp. SC
13874. Enzyme Microb Technol 43:544–549
Subsequently, hydrogenolysis of 5 yielded desired 6
(Bode et al. 2002; Sun et al. 2007). Compared to tert-butyl
3R,5S-6-hydroxy-3,5-O-isopropylidene-3,5-dihydroxyhex-
anoate (Sun et al. 2007), compound 6 showed similar value
of optical rotation.
Greenberg WA, Varvak A, Hanson SR, Wong K, Huang H, Chen P,
Burk M (2004) Development of an efficient, scalable, aldolase-
catalyzed process for enantioselective synthesis of statin inter-
mediates. J Proc Natl Acad Sci USA 101:5788–5793
Guo ZW, Chen YJ, Goswami A, Hanson RL, Patel RN (2006)
Synthesis of ethyl and t-butyl (3R, 5S)-dihydroxy-6-benzyloxy
hexanoates via diastereo- and enantioselective microbial reduc-
tion. Tetrahedron Asymmetry 17:1589–1602
Inoue K, Makino Y, Itoh N (2005) Purification and characterization of
a novel alcohol dehydrogenase from Leifsonia sp. strain S749: a
promising biocatalyst for an asymmetric hydrogen transfer
bioreduction. Appl Environ Microbiol 71:3633–3641
Conclusion
We have investigated two different cofactor regeneration
systems to optimize the biocatalytic process for rDKR
reduction. The results have demonstrated that rDKR can be
utilized as an effective biocatalyst for the preparation of
statin side chains. At present, we are expending the utility
of this unique biocatalyst for other chiral intermediates.
Acknowledgments This work was supported by the ‘‘111 Project’’
from the Ministry of Education of China and the State Administration
of Foreign Expert Affairs of China (No. 111-2-07) and the Key
Project of R&D from the Ministry of Education of China (No.
108070).
Liu JH, Yu BY, Chen YJ (2008) Determination of enantiomeric
excess of ethyl 3,5-dihydroxy-6-benzyloxy hexanoate by chiral
reverse phase high performance liquid chromatography. Chiral-
ity 25:51–53
Mu¨ller M (2005) Chemoenzymatic synthesis of building blocks for
statin side chain. Angew Chem Int Ed 44:362–365
Sun FL, Xu G, Wu JP, Yang LR (2007) A new and facile preparation
of tert-butyl (3R, 5S)-6-hydroxy-3, 5-O-isopropylidene-3,5-
dihydroxyhexanoate. Tetrahedron Asymmetry 18:2454–2461
Thayer AM (2006) Competitors want to get a piece of Lipitor. Chem
Eng News 84:26–27
References
Barbas CF, Wang YF, Wong CH (1990) Deoxyribose-5-phosphate
aldolase as a synthetic catalyst. J Am Chem Soc 112:2013–2014
Bergeron S, Chaplin DA, Edwards JH, Ellis BSW, Hill CL, Holt-
Tiffin K, Knight JR, Mahoney T, Osborne AP, Ruecrof G (2006)
Nitrilase-catalysed desymmetrisation of 3-hydroxyglutaronitrile:
preparation of a statin side-chain intermediate. Org Process Res
Dev 10:661–665
Wolberg M, Hummel W, Wandrey C, Mu¨ller M (2000) Highly regio-
and enantioselective reduction of 3,5-dioxocarboxylates. Angew
Chem Int Ed 39:4306–4308
Wolberg M, Hummel W, Mu¨ller M (2001) Biocatalytic reduction of
beta, delta-diketo esters: a highly stereoselective approach to all
four stereoisomers of a chlorinated beta, delta-dihydroxy hex-
anoate. Chem Eur J 7:4562–4571
Bode SE, Mu¨ller M, Wolberg M (2002) Diastereomer-differentiating
hydrolysis of 1,3-diol-acetonides: a simplified procedure for the
separation of syn- and anti-1,3-diols. Org Lett 4:619–621
Findrik Z, Vasic’-Racki D, Lu¨tz S, Daussmann T, Wandrey C (2005)
Kinetic modeling of acetophenone reduction catalyzed by
alcohol dehydrogenase from Thermoanaerobacter sp. Biotech-
nol Lett 27:1087–1095
Wu XR, Liu N, He YM, Chen YJ (2008) A bacterial enzyme
catalyzing double reduction of a b,d-diketo ester with unprec-
edented stereoselectivity. Nature Proc \http://hdl.handle.
net/10101/npre.2008.1697.1[
Wu XR, Liu N, He YM, Chen YJ (2009) Cloning, expression and
characterization of a novel diketoreductase from Acinetobacte
baylyi. Acta Biochim Biophys Sin 41:163–170
123