Stereospecific reduction of methyl o-chlorobenzoylformate at 300 g·L-1 without additional cofactor using a carbonyl reductase mined from Candida glabrata

Article abstract of DOI:10.1002/adsc.201100366

In order to search for oxidoreductases suitable for the preparation of methyl (R)-o-chloromandelate [(R)-CMM], the key intermediate for clopidogrel, the homologous proteins of Gre2p were expressed in Escherichia coli, among which CgKR1 showed the most satisfactory activity and stereoselectivity towards methyl o-chlorobenzoylformate (CBFM). Using the crude enzyme of CgKR1 and glucose dehydrogenase (GDH), as much as 300 g·L^-1 of CBFM was almost stoichiometrically converted to (R)-CMM with excellent enantiomeric excess (98.7% ee). More importantly, the reaction could be performed without external addition of an expensive cofactor. The substrate profile indicates that keto esters serve as the most suitable substrate, which was confirmed by gram-scale preparations. Homology modeling and docking analysis revealed the molecular basis for the high stereoselectivity of CgKR1. These demonstrate not only the feasibility of in silico mining of novel enzymes based on sequence homology but also the applicability of this new reductase for the practical production of optically active (R)-CMM. Copyright

Full text of DOI:10.1002/adsc.201100366

FULL PAPERS
DOI: 10.1002/adsc.201100366
Stereospecific Reduction of Methyl o-Chlorobenzoylformate
at 300 g·LÀ1 without Additional Cofactor using a Carbonyl
Reductase Mined from Candida glabrata
Hongmin Ma,a Linlin Yang,b Yan Ni,a Jie Zhang,a Chun-Xiu Li,a Gao-Wei Zheng,a
Huaiyu Yang,b,* and Jian-He Xua,*
a
Laboratory of Biocatalysis and Synthetic Biotechnology, State Key Laboratory of Bioreactor Engineering, East China
University of Science and Technology, 130 Meilong Road, Shanghai 200237, Peopleꢀs Republic of China
Fax : (+86)-21-6425-0840; e-mail: jianhexu@ecust.edu.cn
Drug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica,
b
Chinese Academy of Sciences, 555 Zuchongzhi Road, Shanghai 201203, Peopleꢀs Republic of China
Fax : (+86)-21-5080-7088; e-mail: hy_yang@mail.shcnc.ac.cn
Received: May 10, 2011; Revised: February 12, 2012; Published online: June 11, 2012
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201100366.
Abstract: In order to search for oxidoreductases suit- factor. The substrate profile indicates that keto
able for the preparation of methyl (R)-o-chloroman- esters serve as the most suitable substrate, which was
delate [(R)-CMM], the key intermediate for clopi- confirmed by gram-scale preparations. Homology
dogrel, the homologous proteins of Gre2p were ex- modeling and docking analysis revealed the molecu-
pressed in Escherichia coli, among which CgKR1 lar basis for the high stereoselectivity of CgKR1.
showed the most satisfactory activity and stereoselec- These demonstrate not only the feasibility of in silico
tivity towards methyl o-chlorobenzoylformate mining of novel enzymes based on sequence homolo-
(CBFM). Using the crude enzyme of CgKR1 and gy but also the applicability of this new reductase for
glucose dehydrogenase (GDH), as much as 300 g·LÀ1 the practical production of optically active (R)-
of CBFM was almost stoichiometrically converted to CMM.
(R)-CMM with excellent enantiomeric excess (98.7%
ee). More importantly, the reaction could be per- Keywords: bioinformatics; biotransformations; cofac-
formed without external addition of an expensive co- tors; oxidoreductases
Introduction
wide sales of Plavixꢁ (clopidogrel bisulfate) amount-
ed to $10 billion per year, which ranks second. To
The asymmetric reduction of ketones plays an impor- date, a series of routes for the synthesis of the key
tant role in synthetic chemistry since chiral secondary enantiomer, methyl (R)-o-chloromandelate [(R)-
alcohols are versatile synthons, especially for the CMM], have been developed, such as the synthesis of
preparation of enantiomerically pure pharmaceuticals. the corresponding acid by chemical resolution,[5] re-
Among the available strategies, biocatalytic reduction duction of the corresponding a-keto acid,[6] hydrocya-
has attracted great attention because of its inherent nation of o-chlorobenzaldehyde using hydroxyl-nitrile
advantages such as high conversion, excellent enantio- lyase,[7] hydrolysis of o-chloromandelonitrile using
selectivity, superior atom economy, mild reaction con- nitrilase,[8] and asymmetric reduction of methyl
ditions, etc.[1] To fulfill the demands for efficient bio- o-chlorobenzoylformate (CBFM).[9] Ema et al. have
catalysts for this vital transformation, some new ways recently identified a versatile carbonyl reductase,
to discover novel oxidoreductases, for instance, ge- Gre2p from Saccharomyces cerevisiae, and then coex-
nomic identification[2] and metagenomic screening,[3] pressed it with a glucose dehydrogenase (GDH) from
have been adopted, and many carbonyl reductases Bacillus megaterium in E. coli.[10] Using the recombi-
have recently been developed.[4]
nant E. coli and 1.3 mM NADP+, (R)-CMM was ob-
Clopidogrel is a platelet aggregation inhibitor tained at 198 g·LÀ1.[11]
widely used in the therapy for heart attack or stroke
We report herein the discovery of a new carbonyl
caused by the formation of a clot in the blood. World- reductase, CgKR1, which accepts CBFM as substrate
Adv. Synth. Catal. 2012, 354, 1765 – 1772
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1765

Hongmin Ma et al.
FULL PAPERS
Scheme 1. Bioreduction of CBFM at 300 g·LÀ1.
with high activity and stereoselectivity, by in silico
data mining based on sequence homology. Further-
more, we demonstrate the efficient asymmetric syn-
thesis of (R)-CMM as well as several other hydroxy
esters at very high concentrations even without addi-
tional cofactor (Scheme 1). Homology modeling and
docking analysis revealed a similar binding pattern in
the similar binding pocket responsible for the high
stereoselectivity, which might be the foundation for
this in silico mining strategy.
Figure 1. Specific activity (U·mgÀ1 total protein) and product
ee of the homologous proteins towards CBFM.
Table 1. Asymmetric reduction of CBFM with crude en-
zymes of CgKR1 and BsGDH.[a]
Entry
CBFM
[g·LÀ1]
[M]
NADP+
[mM]
Conv.
[%][b]
Yield
[%][c]
ee
[%]
N
1
2
3
4
5
6
200
200
200
200
300
400
1.0
1.0
1.0
1.0
1.5
2.0
1.3
0.50
0.13
0
0
0
>99
>99
>99
>99
>99
92
89
88
91
85
87
–
98.7
98.7
98.7
98.7
98.7
98.7
Results and Discussion
Discovery of CgKR1
In order to construct a library of oxidoreductases for
CBFM reduction, a pBLAST search was carried out
using Gre2p as the template (see the Supporting In-
formation, Figure S1), and 8 proteins possibly having
reductive activities from various yeasts were cloned.
Further expression in E. coli resulted in 6 soluble or
partially soluble enzymes and 2 absolutely insoluble
proteins. The cell-free extract (CFE) of the recombi-
nant E. coli was then subjected to a spectrophotomet-
ric activity assay and a bioreduction test of CBFM at
10 mM. As a result, all the 6 reductases were
NADPH-dependent, just the same as Gre2p, with
[a]
Reaction conditions: CBFM (2.0–4.0 g, 10–20 mmol), cell
lysate of E. coli BL21 (DE3) harboring pET28a-CgKR1
(1.0 g wet cells) in 9.4 mL phosphate buffer (100 mM,
pH 6.0), crude enzyme of BsGDH (0.30 g), glucose
(2 equiv. with respect to CBFM), NADP+ (0–1.3 mM),
EtOH (0.60 mL), 258C, 24 h.
Conversion determined by GC analysis (see Supporting
Information, Table S3) when Na2CO3 was no longer
added (6 h for entries 1, 2, 3 and 4; 14 h for entries 5 and
6).
[b]
[c]
Isolated yield, hereinafter the same.
greater or lesser reductive activities toward CBFM. enzyme was then renamed as CgKR1 (Candida glab-
Unexpectedly, most of the soluble proteins exhibited rata keto ester reductase No. 1). CgKR1 exhibited the
good to excellent stereoselectivities, among which highest activity around pH 5.8 (Supporting Informa-
CAG58832 protein, being soluble only when induced tion, Figure S2), but pH 6.0 was chosen as the com-
at 168C, showed the second highest activity and the monly compatible pH for both CgKR1 and the glu-
most satisfactory stereoselectivity (Figure 1 and Sup- cose dehydrogenase from Bacillus subtilis (BsGDH)
porting Information, Table S1). The amino acid se- employed for NADPH regeneration.
quence of CAG58832 reveals 58% identity to Gre2p,
which is the highest among the homologous proteins
(Supporting Information, Table S1). Further determi- Preparative Synthesis of (R)-CMM
nation of the specific activity and enantioselectivity of
the enzyme in the reduction of various ketones indi- Preparative bioreduction was firstly carried out with
cated that only some a- and b-keto esters turned out CBFM at 1M in a 10-mL aqueous buffer system con-
to be suitable substrates (Table 2), therefore the taining 0.3 g GDH (900 U), 1.0 g recombinant E. coli
1766
asc.wiley-vch.de
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 1765 – 1772

Stereospecific Reduction of Methyl o-Chlorobenzoylformate at 300 g·LÀ1
Table 2. Substrate profile of CgKR1.
Table 2. (Continued)
Entry Product (1b–18b)
Specific activity
ee
[U·mgÀ1 protein][a]
[%][b]
99
(S)
18
3.22
Entry Product (1b–18b)
Specific activity
ee
[U·mgÀ1 protein][a]
[%][b]
[a]
The enzyme activity was measured using the standard
assay protocol with purified enzyme.
The ee values were determined with chiral GC or HPLC
analysis (see Supporting Information, Table S3).
99
(R)
[b]
1
2
3
4
5
6
16.1
98
(R)
29.2
7.63
16.1
23.7
1.31
(subjected to ultrasonic cell disruption) and 1.3 mM
NADP+, during which the mixture was auto-titrated
with 2M Na2CO3 solution to neutralize the gluconic
acid formed. It was found that the reaction was com-
pleted in 6 h with >99% conversion (Table 1,
entry 1), and the synthetic intermediate for clopidog-
rel, (R)-CMM, was obtained with 98.7% ee. Such an
optical purity can be easily further improved by re-
crystallization of clopidogrel hydrogen sulphate. Since
the high amount of the cofactor added (even at
1.3 mM) makes the reaction less practically feasible
from the viewpoint of economic aspects, the subse-
quent optimization was carried out to reduce the
amount of NADP+. Similar conversions could be
achieved when the externally supplemented NADP+
was reduced from 1.3 mM to 0.5 or 0.13 mM (Table 1,
entries 2 and 3). Even without any addition of exter-
nal NADP+, the bioreduction could proceed smooth-
ly, giving full conversion within 5 h (Table 1, entry 4
and Figure 2). Therefore, extremely high total turn-
over numbers and space-time yields were achieved.
98
(R)
98
(R)
98
(R)
78
(R)
99
(S)
7
22.4
1.65
114
99
(S)
8
96
(R)
9
95
(R)
10
11
50.8
1.31
87
(R)
99
(S)
12
13
0.123
0.195
99
(S)
99
(S)
14
15
0.182
1.14
99
(S)
14
(R)
16
17
0.199
0.081
63
(S)
Figure 2. Bioreduction of CBFM at the concentration of
~
&
1.0M ( ) or 1.5M ( ), without any external cofactor using
crude enzymes of CgKR1 and BsGDH.
Adv. Synth. Catal. 2012, 354, 1765 – 1772
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1767

Hongmin Ma et al.
FULL PAPERS
Table 3. Asymmetric reduction of several keto esters at high substrate concentration.[a]
Entry
Substrate concentration
[g·LÀ1]
Conversion
[%]
Yield
[%]
ee
[%]
Substrate
G
[M]
1
2
3
4
CBFM
300
412
330
260
1.5
2.0
2.0
2.0
>99
>99
>99
>99
87
87
89
92
98.7 (R)
98.1 (R)
96.5 (R)
99.9 (S)
OPBE[b]
COBE[c]
EAA[d]
[a]
Reaction conditions: substrate (2.60–4.12 g, 15–20 mmol), cell lysate of E. coli BL21 (DE3) harboring pET28a-CgKR1
(1.0 g wet cells) in 9.4 mL phosphate buffer (100 mM, pH 6.0), crude enzyme of BsGDH (0.30 g), glucose (2 equiv, with
respect to each substrate), EtOH (0.60 mL), 258C, 24 h.
Ethyl 2-oxo-4-phenylbutyrate (5a).
Ethyl 4-chloro-3-oxo-butanoate (9a).
[b]
[c]
[d]
Ethyl acetoacetate (7a).
The prevalently low tolerance of biocatalysts in the ed in higher activities than that on the a-carbon posi-
presence of high substrate concentrations as well as tion (Table 2, entries 12–17). From the profile of b-
the cofactor dependency of oxidoreductases represent keto esters and aromatic esters, it could be concluded
major impediments on the way towards practical ap- that the reductase might have a small pocket to ac-
plicability on a large scale.[12] Therefore, it is essential commodate the small part of ketones, but not suffi-
to establish a bioreduction process at high substrate cient to accommodate that with larger substitution.
concentrations and, preferably, in the absence of addi-
Besides getting some information about the binding
tional cofactor, for the competitive application of bio- pocket, another important goal for mapping the sub-
catalytic technology on an industrial scale. As the strate profile is to determine which product of puta-
addtional cofactor has been completely abolished, fur- tive commercial application could be prepared by the
ther optimization was aimed at higher substrate con- enzyme. Following the procedures above, several a-
centrations to further improve the substrate/catalyst and b-keto esters were prepared at high substrate
(S/C) ratio. The full conversion was obtained after 6 h concentrations (Table 3). Among them, ethyl (R)-2-
at a CBFM concentration of 1.5M (Table 1, entry 5 hydroxy-4-phenylbutyrate [(R)-HPBE, 5b] is an im-
and Figure 2) although the further increase of CBFM portant precursor for angiotensin-converting enzyme
concentration to 2M resulted in 92% conversion (ACE) inhibitors, such as Enalapril, Lisinopril, Cilap-
(Table 1, entry 6). It should be noticed that the com- ril and Benazepril, which are widely used as antihy-
plete abolishment of external NADP+ and the high pertensive drugs and in the therapy for congestive
productivity of about 700 g·LÀ1·dÀ1 (Table 1, entry 5, heart failure.[13] Using CgKR1, (R)-HPBE was afford-
2.56 g product, 15 mL final volume, 6 h) could signifi- ed with 98% ee at about 412 g·LÀ1 (Table 3, entry 2).
cantly reduce the production costs of (R)-CMM, Interestingly, it was reported that the opposite enan-
making the process more economically feasible. The tiomer, (S)-HPBE was obtained using the template
outstanding performance of CgKR1 indicates its great reductase, Gre2p.[2a] In addition, ethyl (R)-4-chloro-3-
potential for industrial use.
hydroxybutyrate [(R)-CHBE, 9b] is an important syn-
thon for the synthesis of biologically and pharmaco-
logically important molecules, such as l-carnitine,
(R)-4-amino-3-hydroxybutyric acid, and (R)-4-hy-
droxy-2-pyrrolidone;[14] ethyl (S)-3-hydrobutyrate
Substrate Profile
The substrate profiles, including activity and stereose- [(S)-HBE, 7b] is a versatile building block for the syn-
lectivity towards a- and b-keto esters and aromatic thesis of pheromones,[15] antibiotics[16] and other com-
ketones, were explored. As shown in Table 2, CgKR1 pounds.[17] Both the compounds, (R)-CHBE and (S)-
exhibited broad substrate spectra, especially with high HBE, were produced at a 2M substrate concentration
activities and stereoselectivities towards most tested (Table 3, entries 3 and 4). Although the ee value of
keto esters. It is rather interesting that CgKR1 (R)-CHBE was less satisfactory, it is still higher than
showed higher activities towards ethyl esters than that reported before.[14a]
methyl esters (Table 2, entries 1–4). Varied activities
were observed as the substitution on the small part of
the ketones changed (Table 2, entries 7–11), and the Molecular Simulation
extremely high activity was obtained towards ethyl 4-
chloro-3-oxobutanoate (COBE, Table 2, entry 9). Homology modeling and docking analysis were per-
Among the aromatic ketones, activities were rather formed to gain insights into the high selectivity of the
poor, and the substitution on the aromatic ring result- enzymes. The 3D structural models of CgKR1 and
1768
asc.wiley-vch.de
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 1765 – 1772

Stereospecific Reduction of Methyl o-Chlorobenzoylformate at 300 g·LÀ1
Figure 3. Homology models of CgKR1 and Gre2p constructed based on the crystal structure of SSCR (PBD: 1Y1P) and the
binding models of CBFM. A: Complex of CgKR1 and CBFM. B: Complex of Gre2p and CBFM. C: Detailed view of the in-
teractions between CBFM and CgKR1. D: Detailed view of the interactions between CBFM and Gre2p. CBFM,
Gre2p were constructed based on the crystal structure bonyl oxygen atom of a ketone substrate forms hydro-
of Sporobolomyces salmonicolor carbonyl reductase gen bonds with both Tyr and Ser residues, and it is
(SSCR; Protein Data Bank accession: 1Y1P),[18] indi- protonated from the Tyr residue, followed by the at-
cating that their structures and binding pockets are tacking of a hydrogen atom from the C-4 atom of
very similar to each other (Figure 3, A, B and the NADPH to the carbonyl carbon atom of the sub-
Supporting Information, Figure S3).
strates. In the CgKR1/CBFM complex (Figure 3, C),
To reveal the mechanism of the high stereoselectiv- Ser-134 and Tyr-175 stabilize the substrate with hydro-
ity, the compound was docked into the pocket of the gen bonds. The phenyl ring is embedded in a hydro-
homology modeled structure followed by optimiza- phobic cavity next to the catalytic center, while the
tion. According to the proposed catalytic mechanism ester group is located in a small cavity mainly com-
of the short-chain alcohol dehydrogenase,[18] the car- posed of N224, F92, F94, Y175 as well as NADPH.
Adv. Synth. Catal. 2012, 354, 1765 – 1772
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1769

Hongmin Ma et al.
FULL PAPERS
was then cloned into pET28a(+) to obtain pET28a-CgKR1,
followed by transformation into E. coli BL21 (DE3) cells.
The recombinant E. coli was cultivated at 378C in LB
medium supplemented with kanamycin (50 mg·mLÀ1). When
the OD600 reached approximately 0.8, the enzyme expression
was induced by IPTG (final concentration=0.1 mM) and
the cells were allowed for cultivation at 168C for 18–24 h.
Cells were harvested by centrifugation (8,800 rpm, 5 min,
48C), washed with 0.85% NaCl solution, and resuspended in
a proper volume of sodium phosphate buffer (100 mM,
pH 6.0). The cell suspension was subjected to ultrasonic cell
disruption to obtain the cell lysate. The CgKR1 or GDH ac-
tivity was assayed spectrophotometrically at 308C by moni-
toring the decrease or increase in the absorbance of
NADPH at 340 nm. The standard assay mixture was com-
posed of 100 mM phosphate buffer (pH 6.0), 2.0 mM sub-
strate, 0.1 mM NADPH or NADP+ and an appropriate
amount of enzyme. One unit of enzyme activity was defined
as the amount of enzyme catalyzing the oxidation or forma-
tion of 1 mmol NADPH per minute.
Therefore, the NADPH donates its H at the C-4 posi-
tion of the nicotinamine ring to the carbonyl C from
the Re face and (R)-CMM is obtained so that CgKR1
follows the Prelog rule, as proven by the substrate
profile described above. Although Gre2p gives a very
similar situation (Figure 3, D), the slight difference of
the backbones of the enzymes leads to the much
smaller binding cavity than that of CgKR1, which
might contribute to the better stereoselectivity of
Gre2p. Such a binding pattern was closely similar to
the model of Gre2p docked with 3-chloro-1-phenyl-1-
propanone.[19] The similar binding pocket may be the
foundation for this successful in silico mining strategy.
Conclusions
In summary, we have demonstrated the power of ge-
nomic data mining as an emerging tool in the rapid
searching for novel and synthetically useful biocata-
lysts such as CgKR1. In addition, the development of
an efficient enzymatic process for (R)-CMM produc-
tion with a very high productivity of 700 g·LÀ1·dÀ1 in
the absence of any additional NADP+ has laid a solid
basis for promising applications in the future. Sub-
strate profiling guided the preparation of other hy-
droxy esters at very high substrate concentrations.
Preparation of CBFM
H2SO4 (98%, 15 mL) was added dropwise to CH3OH
(100 mL) at 0–48C. After adding o-chloromandelate (40 g),
the solution was refluxed with agitation for 4 h and then
washed with saturated Na2CO3 to a neutral pH. After re-
moval of CH3OH by rotary evaporation, the product was ex-
tracted with CH2Cl2. The organic layer was dried over
MgSO4 and concentrated to afford methyl o-chloromande-
Molecular simulation revealed the similar binding late as pale yellow oil; yield: 36.2 g (85%).
To stirred H2SO4 (23 mL), was gently added ground CrO3
pockets which are essential for the high stereoselec-
tivity. Further studies, including coexpression of
CgKR1 and GDH, high-expression fermentation of
engineered enzymes, are still in progress.
powder (26.7 g), and the mixture was stirred for another 2 h.
Then the solution was diluted with water to 100 mL. The re-
sulting Jones reagent was added dropwise to the solution of
methyl o-chloromandelate (30.3 g) in CH2Cl2 (100 mL) at 0–
108C. After stirring at room temperature for about 8 h, the
organic layer was washed with saturated NaHCO3 solution
and tap water, dried over MgSO4, and evaporated to give
methyl o-chlorobenzoylformate (CBFM) as a faint yellow
oil; yield: 26.7 g (89%, >99% purity as determined by GC
analysis).
Experimental Section
General Remarks
Compounds 2a and 4a were prepared previously,[20] similar
to CBFM preparation. All other compounds and biochemi-
cals were from standard commercial sources. The substrate
profile was mapped as previously described,[20] and condi-
tions for conversion and stereoselectivity determination are
listed in the Supporting Information, Table S3. PCR amplifi-
cation was carried out using Taq DNA polymerase with ge-
nomic DNA as a template and specific primers listed in the
Supporting Information, Table S2, and then verified by se-
quencing. Amino acid sequence alignments were performed
by the CLUSTALW method from Biology Workbench 3.2
software (http://workbench.sdsc.edu).
Asymmetric Bioreduction of CBFM and Other Keto
Esters
To the cell lysate of E. coli BL21ACTHUNTGRNEUNG(DE3) harboring pET28a-
CgKR1 (1.0 g wet cell, 600 U) in 9.4 mL phosphate buffer
(100 mM, pH 6.0), was added 5.40 g glucose (30.0 mmol),
0.30 g crude GDH (glucose dehydrogenase from Bacillus
subtilis, 900 U), 3.00 g CBFM (15.0 mmol) and 0.6 mL
EtOH (cosolvent). The mixture was stirred in a water bath
at 258C for 12 h, during which 2M Na2CO3 was auto-titrated
to neutralize the gluconic acid formed during the reaction.
Solid NaCl (5.5 g) was added, and the product was extracted
twice with EtOAc (20 mL each), whereby centrifugation
(8,800 rpm, 10 min) was carried out to facilitate the phase
separation. The combined organic layers were dried over
MgSO4, filtered, and concentrated under reduced pressure,
giving (R)-CMM as a faint yellow oil; yield: 2.56 g (85%,
>99% purity as determined by GC analysis). [a]2D5: À177.4
(c 1.0, CHCl3), 98.7% ee (R) {lit.[11b] [a]D25: À178.3 (c 1.3,
CHCl3), >99% ee (R)}; 1H NMR (CDCl3, 400 MHz): d=
Gene Cloning and Functional Expression of CgKR1
as well as Enzyme Activity Assay
The CgKR1-encoding gene was amplified by polymerase
chain reaction (PCR) using the genomic DNA of Candida
glabrata CGMCC 2.234 as the template and two synthetic
primers (primer 1, CATATGGCTTCTGATAACAGCAAC and
primer 2, GGATCCTTAATTAGAGTTCTTCTCGGC). The gene
1770
asc.wiley-vch.de
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 1765 – 1772

Stereospecific Reduction of Methyl o-Chlorobenzoylformate at 300 g·LÀ1
3.77 (s, 4H, ArCHOH, COOCH3), 5.36 (d, 1H, ArCHOH),
7.28–7.29, 7.39–7.40 (m, 4H, Ar-H).
Technology (No. 11431921600) and Innovation Program of
Shanghai Municipal Education Commission (No. 11CXY24).
Similarly, other keto esters were obtained as follows:
(R)-HPBE: yield: 87%; 98% ee; [a]2D5: À21.4 (c 1.0,
CHCl3); lit.[21] {[a]D25: À18.1 (c 1.0, CHCl3)}; 1H NMR
(CDCl3, 400 MHz): d=1.26–1.31 (t, 3H, OCH2CH3), 1.90–
2.00 (m, 1H, ArCH2CHH), 2.08–2.17 (m, 1H, ArCH2CHH),
2.70–2.82 (m, 2H, ArCH2), 4.15–4.26 (m, 3H, OCH2CH3,
CHOH), 7.16–7.32 (m, 5H, Ar-H).
References
[1] a) K. Nakamura, R. Yamanaka, T. Matsuda, T. Harada,
Tetrahedron: Asymmetry 2003, 14, 2659–2681; b) K. Na-
kamura, T. Matsuda, Curr. Org. Chem. 2006, 10, 1217–
1246; c) S. Panke, M. Wubbolts, Curr. Opin. Chem.
Biol. 2005, 9, 188–194; d) W. Kroutil, H. Mang, K.
Edegger, K. Faber, Curr. Opin. Chem. Biol. 2004, 8,
120–126; e) J.-H. Tao, J.-H. Xu, Curr. Opin. Chem.
Biol. 2009, 13, 43–50; f) R. N. Patel, Adv. Synth. Catal.
2001, 343, 527–546; g) Y. Ni, J.-H. Xu, Biotechnol. Adv.
2011, doi:10.1016/j.biotechadv.2011.10.007.
(R)-CHBE: yield: 89%; 96% ee; [a]2D5: +22.0 (c 1.0,
CHCl3); lit.[20e] {[a]D25: +22.3 (c 1.0,CHCl3)}; 1H NMR
(CDCl3, 400 MHz): d=1.26–1.29 (t, 3H, OCH2CH3), 2.57–
2.70 (m, 2H, CH2COOEt), 3.07 (s, 1H, CHOH), 3.58–3.65
(m, 2H, ClCH2CHOH), 4.16–4.20 (q, 2H, OCH2CH3), 4.21–
4.28 (m, 1H, ClCH2CHOH).
(S)-HBE: yield: 92%; 99.9% ee; [a]2D5: +45.2 (c 1.0,
CHCl3); lit.[22] {[a]D25: +46.0 (c 1.0, CHCl3)}; 1H NMR
(CDCl3, 400 MHz): d=1.21–1.23 (d, 3H, CH3CHOH), 1.25–
[2] a) I. Kaluzna, T. Matsuda, A. Sewell, J. Stewart, J. Am.
Chem. Soc. 2004, 126, 12827–12832; b) Y. Ni, C.-X. Li,
H.-M. Ma, J. Zhang, J.-H. Xu, Appl. Microbiol. Bio-
technol. 2011, 89, 1111–1118; c) D.-M. Zhu, H. Malik,
L. Hua, Tetrahedron: Asymmetry 2006, 17, 3010–3014.
[3] A. Knietsch, T. Waschkowitz, S. Bowien, A. Henneee,
R. J. Daniel, J. Mol. Microbiol. Biotechnol. 2003, 5, 46–
56.
[4] T. Matsuda, R. Yamanaka, K. Nakamura, Tetrahedron:
Asymmetry 2009, 20, 513–557.
[5] A. Bousquet, A. Musolino, PCT Int. Appl.
WO9918110A1, 1999.
[6] K. Mitsuhashi, H. Yamamoto, Eur. Pat. Appl.
EP1316613A2, 2003.
[7] a) A. Glieder, R. Weis, W. Skranc, P. Poechlauer, I.
Dreveny, S. Majer, M. Wubbolts, H. Schwab, K.
Gruber, Angew. Chem. 2003, 115, 4963; Angew. Chem.
Int. Ed. 2003, 42, 4815- 4818; b) L. M. van Langen,
R. P. Selassa, F. van Rantwijk, R. A. Sheldon, Org. Lett.
2005, 7, 327–329.
[8] a) Z.-J. Zhang, J. Pan, J.-F. Liu, J.-H. Xu, J. Biotechnol.
2011, 152, 24–29; b) C.-S. Zhang, Z.-J. Zhang, C.-X. Li,
H.-L. Yu, G. W. Zheng, J.-H. Xu, Appl. Microbiol. Bio-
technol. DOI: 10.1007/s00253-012-3993-4.
[9] a) J.-P. Genet, S. Juge, J.-A. Laffitte, C. Pinel, S. Mal-
lart, PCT Int. Appl. WO9401390A1, 1994; b) Y. Sun,
X. Wan, J. Wang, Q. Meng, H. Zhang, L. Jiang, Z.
Zhang, Org. Lett. 2005, 7, 5425–5427.
[10] a) T. Ema, Y. Sugiyama, M. Fukumoto, H. Moriya, J.-
N. Cui, T. Sakai, M. Utaka, J. Org. Chem. 1998, 63,
4996–5000; b) T. Ema, H. Moriya, T. Kofukuda, T.
Ishida, K. Maehara, M. Utaka, T. Sakai, J. Org. Chem.
2001, 66, 8682–8684; c) T. Ema, H. Yagasaki, N. Okita,
K. Nishikawa, T. Korenaga, T. Sakai, Tetrahedron:
Asymmetry 2005, 16, 1075–1078; d) T. Ema, H. Yagasa-
ki, N. Okita, M. Takeda, T. Sakai, Tetrahedron 2006, 62,
6143–6149.
1.28
(t,
3H,
OCH2CH3),
2.37–2.50
(m,
2H,
CHOHCH2COOEt), 2.90 (s, 1H, CH3CHOH), 4.17–4.22
(m, 1H, CH3CHOH), 4.14–4.16 (q, 2H, OCH2CH3).
Homology Modeling
The initial 3D structural models of CgKR1 and Gre2p were
constructed based on the crystal structure of Sporobolomy-
ces salmonicolor carbonyl reductase (SSCR; Protein Data-
Bank code: 1Y1P)[18] using INSIGHT II (Accelrys Software
Inc). The CgKR1 active site was examined by comparison
with the corresponding X-ray template of NADPH-bound
SSCR. After that, the overall structure of CgKR1/NADPH
was obtained by merging geometrically optimized NADPH
into CgKR1 and Gre2p.
Docking
Docking was done with the Glide module in Maestro,[23]
using CBFM as the ligand and the CgKR1/NADPH com-
plex or Gre2p/NADPH complex as the receptor molecule.
The ligand and the complex were prepared respectively
using the preparation facility in the LigPrep and the Protein
Preparation Wizard. The result was used as valid inputs for
the following Grid Generation and Ligand Docking. The
grid site was generated on the basis of the reported sub-
strate-binding domain.[24] Then, docking was carried out
with extra precision, outputing the top 10 conformations.
The binding energies of the conformations were calculated
by using the Prime MM-GBSA module. To obtain a more
relaxed model, the conformation with the lowest binding
energy were further put into energy minimization by using
the GROMACS with the GROMOS 53a6 force field.[25]
Acknowledgements
[11] a) T. Ema, N. Okita, S. Ide, T. Sakai, Org. Biomol.
Chem. 2007, 5, 1175–1176; b) T. Ema, S. Ide, N. Okita,
T. Sakai, Adv. Synth. Catal. 2008, 350, 2039–2044.
[12] a) F. Hollmann, I. W. C. E. Arends, K. Buehler, Chem-
CatChem 2010, 2, 762–782; b) N. St. Clair, Y.-F. Wang,
A. L. Margolin, Angew. Chem. 2000, 112, 388–391;
Angew. Chem. Int. Ed. 2000, 39, 380–383; c) W. Stamp-
fer, B. Kosjek, C. Moitzi, W. Kroutil, K. Faber, Angew.
This work was financially supported by the National Natural
Science Foundation of China (Nos. 20902023 & 31071604),
Ministry of Science and Technology, P. R. China (Nos.
2009CB724706 & 2011CB710800), China National Special
Fund for State Key Laboratory of Bioreactor Engineering
(No. 2060204) and Shanghai Leading Academic Discipline
Project (No. B505), Shanghai Committee of Science and
Adv. Synth. Catal. 2012, 354, 1765 – 1772
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1771

Hongmin Ma et al.
FULL PAPERS
Chem. 2002, 114, 1056–1059; Angew. Chem. Int. Ed.
2002, 41, 1014–1017.
[13] G. Iwasaki, R. Kimura, N. Numao, K. Kondo, Chem.
[18] S. Kamitori, A. Iguchi, A. Ohtaki, M. Yamada, K. Kita,
J. Mol. Biol. 2005, 352, 551–558.
[19] Y.-H. Choi, H.-J. Choi, D. Kim, K.-N. Uhm, H.-K. Kim,
Appl. Microbiol. Biotechnol. 2010, 87, 185–193.
[20] a) Y. Ni, J.-H. Xu, J. Mol. Catal. B: Enzym. 2002, 18,
233–241; b) W. Yang, J.-H. Xu, Y. Xie, Y. Xu, G. Zhao,
G.-Q. Lin, Tetrahedron: Asymmetry 2006, 17, 1769–
1774; c) Y. Xie, J.-H. Xu, W.-Y. Lu, G.-Q. Lin, Biore-
sour. Technol. 2009, 100, 2463–2468; d) L.-J. Wang, C.-
X. Li, Y. Ni, J. Zhang, X. Liu, J.-H. Xu, Bioresour.
Technol. 2011, 102, 7023–7028; e) Y. Ni, C.-X. Li, L.-J.
Wang, J. Zhang, J.-H. Xu, Org. Biomol. Chem. 2011, 9,
5463–5468; f) N.-D. Shen, Y. Ni, H.-M. Ma, L.-j. Wang,
C.-X. Li, G.-W. Zheng, J. Zhang, J.-H. Xu, Org. Lett.
2012, 14, 1982-1985.
Pharm. Bull. 1989, 37, 280–283.
[14] a) M. Kataoka, K. Yamamoto, H. Kawabata, M. Wada,
K. Kita, H. Yanase, S. Shimizu, Appl. Microbiol. Bio-
technol. 1999, 51, 486–490; b) J. Grunwald, B. Wirz,
M. P. Scollar, A. M. Klibanov, J. Am. Chem. Soc. 1986,
108, 248–258; c) W. R. Shieh, A. S. Gopalan, C. J. Sih, J.
Am. Chem. Soc. 1985, 107, 2993–2994; d) C.-H. Wong,
D. G. Drueckhammer, H. M. Sweers, J. Am. Chem. Soc.
1985, 107, 4028–4031; e) B. Zhou, A. S. Gopalan, F. Van
Middlesworth, W. R. Shieh, C. J. Sih, J. Am. Chem. Soc.
1983, 105, 5925–5926.
[15] a) T. Sakai, K. Mori, Agric. Biol. Chem. 1986, 50, 177–
183; b) T. Katsuki, M. Yamaguchi, Tetrahedron Lett.
1987, 28, 651–654; c) J. Ferreira, F. Simonelli, Tetrahe-
dron 1990, 46, 6311–6318; d) A. Tai, N. Morimoto, M.
Yoshikawa, K. Uehara, T. Sugimura, T. Kikukawa,
Agric. Biol. Chem. 1990, 54, 1753–1762.
[16] a) G. Cainelli, M. Contento, D. Giacomini, M. Panun-
zio, Tetrahedron Lett. 1985, 26, 937–940; b) D. J. Hart,
D.-C. Hua, Tetrahedron Lett. 1985, 26, 5493–5496; c) D.
Seebach, B. Seuring, H.-O. Kalinowski, W. Lubosch, B.
Renger, Angew. Chem. 1977, 89, 270–271; d) G. E.
Keck, D. F. Kachensky, E. J. Enholm, J. Org. Chem.
1984, 49, 1462–1464.
[21] N. W. Fadnavis, K. R. Radhika, Tetrahedron: Asymme-
try 2004, 15, 3443–3447.
[22] A. Wolfson, C. Dlugy, D. Tavor, J. Blumenfeld, Y. Shot-
land, Tetrahedron: Asymmetry 2006, 17, 2043–2045.
[23] a) R. A. Friesner, J. L. Banks, R. B. Murphy, T. A.
Halgren, J. J. Klicic, D. T. Mainz, M. P. Repasky, E. H.
Knoll, M. Shelley, J. K. Perry, D. E. Shaw, P. Francis,
P. S. Shenkin, J. Med. Chem. 2004, 47, 1739–1749;
b) T. A. Halgren, R. B. Murphy, R. A. Friesner, H. S.
Beard, L. L. Frye, W. T. Pollard, J. L. Banks, J. Med.
Chem. 2004, 47, 1750–1759.
[24] U. C. Oppermann, C. Filling, H. Jornvall, Chem.-Biol.
[17] a) K. Mori, Z.-H. Qian, Bull. Soc. Chim. Fr. 1993, 130,
382–387; b) K. Mori, M. Ikunaka, Tetrahedron 1987, 43,
45–58.
Interact. 2001, 130–132, 699–705.
[25] L. Schuler, X. Daura, W. Van Gunsteren, J. Comput.
Chem. 2001, 22, 1205–1218.
1772
asc.wiley-vch.de
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2012, 354, 1765 – 1772