Enantioselective microbial reduction of substituted acetophenones

Article abstract of DOI:10.1016/j.tetasy.2004.02.024

The chiral intermediate (S)-1-(2′-bromo-4′-fluoro phenyl)ethanol 2 was prepared by the enantioselective microbial reduction of 2-bromo-4-fluoro acetophenone 1. Organisms from genus Candida, Hansenula, Pichia, Rhodotorula, Saccharomyces, Sphingomonas and B

Full text of DOI:10.1016/j.tetasy.2004.02.024

(S) H d
Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 1247–1258
Enantioselective microbial reduction of substituted
acetophenones
Ramesh N. Patel,* Animesh Goswami, Linda Chu, Mary Jo Donovan, Venkata Nanduri,
Steven Goldberg, Robert Johnston, Prasad J. Siva, Brent Nielsen, Junying Fan,
WeiXuan He, Zhongping Shi, Kwok Y. Wang, Ronald Eiring,
Dana Cazzulino, Ambarish Singh and Richard Mueller
Process Research & Development, Bristol-Myers Squibb Pharmaceutical Research Institute,
PO Box 191, New Brunswick, NJ 08903, USA
Received 13 November 2003; revised 1 December 2003; accepted 6 February 2004
Abstract—The chiral intermediate (S)-1-(20-bromo-40-fluoro phenyl)ethanol 2 was prepared by the enantioselective microbial
reduction of 2-bromo-4-fluoro acetophenone 1. Organisms from genus Candida, Hansenula, Pichia, Rhodotorula, Saccharomyces,
Sphingomonas and Baker’s yeast reduced 1 to 2 in >90% yield and 99% enantiomeric excess (ee). In an alternative approach, the
enantioselective microbial reductions of methyl, ethyl, and tert-butyl 4-(20-acetyl-50-fluorophenyl) butanoates 3, 5, and 7, respec-
tively, were demonstrated using strains of Candida and Pichia. Reaction yields of 40–53% and ee’s of 90–99% were obtained for the
corresponding (S)-hydroxy esters 4, 6, and 8. The reductase, which catalyzed the enantioselective reduction of ketoesters was
purified to homogeneity from cell extracts of Pichia methanolica SC 13825. It was cloned and expressed in Escherichia coli with
recombinant cultures used for the enantioselective reduction of keto methyl ester 3 to the corresponding (S)-hydroxy methyl ester 4.
On a preparative scale, a reaction yield of 98% and an ee of 99% was obtained.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
review articles3–10 have been published on the use of
enzymes in organic synthesis.
Recently, much attention has been focused on the
interaction of small molecules with biological macro-
molecules. Selective enzyme inhibitors and receptor
agonists/antagonists are keys for target-oriented
research in the pharmaceutical industry.1 Chirality is a
key factor in the efficiency of many drug products, and
as a result the production of single enantiomers of
molecules has become increasingly important in the
pharmaceutical industry.2 Single enantiomers can be
produced by chemical or chemo-enzymatic synthesis.
The advantages of biocatalysis over chemical synthesis
are that enzyme-catalyzed reactions are often highly
enantioselective and regioselective. They can be carried
out at ambient temperature and atmospheric pressure,
thus avoiding the use of more extreme conditions, which
could cause problems due to isomerization, racemiza-
tion, epimerization, or rearrangement. A number of
Herein we report the enzymatic synthesis of (S)-1-(20-
bromo-40-fluorophenyl)-ethanol
2
and (S)-hydroxy
esters 4, 6, and 8 (Fig. 1A and B), as potentially useful
intermediates in the synthesis of pharmaceutical prod-
ucts.11–16
F
Br
F
Br
Microorganisms
Microorganisms
O
OH
Ketone 1
(S)-Alcohol 2
F
F
CO2R
CO2R
OH
O
(S)-Hydroxyester
R= Methyl 4, Ethyl 6, t-Butyl 8
Ketoester
R= Methyl 3, Ethyl 5, t-Butyl 7
* Corresponding author. Tel.: +1-732-227-6213; fax: +1-732-227-3994;
e-mail: patelr@bms.com
Figure 1.
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.02.024

1248
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
Table 2. Enantioselective microbial reduction of keto methyl ester 3
2. Results and discussion
Microorganism
% Conversion
to (S)-alcohol 4
Ee of (S)-alcohol
4 (%)
A number of microorganisms were screened for the
enantioselective reduction of ketone 1 to alcohol 2.
Results (yields, and ee’s of product 2) obtained with the
best cultures are as shown in Table 1. Many cultures
from Candida, Hansenula, and Pichia gave high reaction
yields (99–100 M %) and ee’s (>99%) of desired (S)-
alcohol 2. Furthermore, commercially available active
dry yeast (Red Star) gave a reaction yield of 90% and
an ee of 99.9% for desired alcohol 2.
Pichia methanolica SC
13825
40
41
33
11
6
99.8
Pichia methanolica SC
13860
99
Pichia methanolica SC
16116
96
Candida boidinii SC
13821
99.4
99
Geotrichum candidum SC
16010
The dry yeast was investigated further with the reduc-
tion process carried out at a 1-L and 100-L scale as
described in the Experimental section. A reaction yield
of 90% and an ee of 99.5% were obtained for (S)-alcohol
2 in each experiment. At the end of the reaction, product
2 was adsorbed onto XAD-16 resin and, after filtration,
recovered in 84% yield from the resin by acetonitrile
extraction and silica gel chromatography.
Microbial cultures were suspended in 10 mL of 100 mM phosphate
buffer (pH 7.0) at 20% (w/v, wet cells) cell concentration supplemented
with 1.0 mg/mL substrate 3 and 25 mg/mL of glucose. Reductions were
carried out at 28 ꢁC and 250 rpm on a rotary shaker for 24 h.
Table 3. Enantioselective microbial reduction of keto ethylester 5
Microorganism
% Conversion
to (S)-alcohol 6
Ee of (S)-alcohol
6 (%)
Various microbial cultures were screened for the enantio-
selective reduction of keto methyl ester 3. Results
(yields, and ee’s of product) obtained from the five best
cultures are as shown in Table 2. Three strains of Pichia
gave reaction yields of 33–41 M % and ee’s of >96% of
the desired (S)-alcohol 4. The lower reaction yield rel-
ative to 1 was due to the hydrolysis of keto methyl ester
3 to the corresponding acid, which is not a substrate for
reduction.
Pichia methanolica SC
13825
18
51
33
93
Pichia methanolica SC
13860
>99.9
98.5
Candida boidinii SC
13821
Microbial cultures were suspended in 10 mL of 100 mM phosphate
buffer (pH 7.0) at 20% (w/v, wet cells) cell concentration supplemented
with 1.0 mg/mL substrate 5 and 25 mg/mL of glucose. Reductions were
carried out at 28 ꢁC and 250 rpm on a rotary shaker for 24 h.
We also evaluated various cultures for the enantio-
selective reduction of ethyl ester 5 and tert-butyl ester 7.
Results with the three best cultures are as shown in
Tables 3 and 4. As observed with the methyl ester,
hydrolysis of the ethyl ester to the corresponding acid
occurred with lower reaction yields being obtained. The
tert-butyl ester proved more stable, with formation of
the acid being less then 7%; however, a lower yield was
obtained due to poor substrate specificity.
In order to circumvent the problem of ester hydrolysis,
we decided to purify the reductase from Pichia met-
Table 1. Enantioselective microbial reduction of 2-bromo-4-fluoroacetophenone 1
Microorganism
% Conversion to (S)-alcohol 2
Ee of (S)-alcohol 2 (%)
Candida sonorensis SC 16117
Candida Guilliermondi SC 13861
Candida boidinii SC 13821
100
100
100
99
99.2
99
97.4
99.6
97.6
99.9
99
Candida utilis SC 13983
Candida parapsilosis SC 16346
Rhodotorula glutinis SC 16267
Hansenula fabianii SC 13894
Hansenula polymorpha SC 13824
Hansenula saturnus SC 13829
Nocardia salmonicolor SC 6310
Pichia anomala SC 16142
98
100
94
100
100
100
100
100
100
99
99.8
99
99.3
99
Pichia methanolica SC 13860
Pichia pinus SC 13864
Pichia stiptis SC 13863
99
99
99
99
Pichia capsulata SC 16306
Pichia silvicola SC 16159
100
99
99
99
Sphingomonas paucimobilis SC 16113
Saccharomyces cerevisiae SC 13902
Active Dry Yeast (Red Star Co.)
100
93
99.9
99.9
90
Microbial cultures were suspended in 10 mL of 100 mM potassium phosphate buffer (pH 7.0) at 20% (w/v, wet cells) cell concentration supplemented
with 1.5 mg/mL substrate 1 and 25 mg/mL of glucose. Reductions were carried out at 28 ꢁC and 250 rpm on a rotary shaker for 24 h.

R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
Table 4. Enantioselective microbial reduction of keto tert-butyl ester 7
1249
Microorganism
% Conversion
to (S)-alcohol 8
Ee of (S)-alcohol
8 (%)
Mucor rouxii SC
13920
53
43
12
>99
Mucor hiemalis SC
13974
93
Pichia methanolica SC
16116
92.6
Microbial cultures were suspended in 10 mL of 100 mM phosphate
buffer (pH 7.0) at 20% (w/v, wet cells) cell concentration supplemented
with 1.0 mg/mL substrate 7 and 25 mg/mL of glucose. Reductions were
carried out at 28 ꢁC and 250 rpm on a rotary shaker for 24 h.
Figure 2.
hanolica SC 13825 and clone and overexpress this
enzyme in a suitable host. The enzyme was purified 247-
fold from the cell extract using Hi-Trap Blue affinity
column chromatography (Table 5). The purified protein
gave a single band on SDS-PAGE with a molecular
weight of 40,000 Da. The molecular weight of the puri-
fied protein, as determined by size-exclusion column
chromatography, was 38,000–40,000, indicating that the
reductase was a monomeric protein.
120
100
80
Substrate 2 g/L
Substrate 4 g/L
Substrate 6 g/L
60
40
The purified protein required NADPH or NADH as a
cofactor for the reduction of keto methyl ester 3.
Reaction yields of 96% and 89% were obtained for
product 4 with 0.1 mM NADP or NAD at 1 g/L sub-
strate input, respectively, using glucose dehydrogenase
as the cofactor recycling enzyme. The N-terminal and
internal peptide sequences (generated by Lys-peptidase
treatment) of the purified reductase were determined to
allow the synthesis of oligonucleotide probes for clon-
ing. The reductase was expressed as Escherichia coli
strain SC 16445. The complete sequence of the keto-
reductase is covered in Ref. 17.
20
0
0
10
20
30
40
50
Bioconversion time (Hrs)
Figure 3.
phosphate buffer pH 7.0 at 10% (w/v) cell concentration.
Cell suspensions were supplemented with nicotinamide
adenine dinucleotide phosphate (NADP), glucose, glu-
cose dehydrogenase, and 4.5 g/L substrate. Biotrans-
formations were carried out at 500 rpm and 28 ꢁC as
described in the Experimental section. The reaction went
to completion in 20 h with 95% reaction yield and 99.9%
ee of product 4. Significant hydrolysis of substrate 3 was
not observed, with <2% of keto acid 5 being formed.
This process was scaled-up to 500 L to prepare 2.0 kg of
product 4.
The production of cloned ketoreductase in a 250-L fer-
mentor was conducted as described in the Experimental
section. Growth was completed in 48 h and about 50–
52 g/L cells were obtained from the fermentation broth
(Fig. 2). Cells harvested from the fermentor were used to
conduct the bioreduction of keto methyl ester 3.
The effect of substrate concentration was evaluated in
flasks using 10% (w/v, wet cells) cell concentration of
E. coli strain SC 16445. Their reaction went to com-
pletion in 24–30 h at 2–4 g/L substrate input. At 6 g/L
substrate input, 48 h were required to complete the
reaction (Fig. 3).
The dehydrogenases from yeast,18–21 horse liver,22 and
T. brockii23 transfer the pro-R hydride to the re-face of
the carbonyl to give (S)-alcohols, a process described by
Prelog’s rule.24 In contrast, dehydrogenases from Lac-
tobacillus kefir25 and two Pseudomonas sp.26 exhibit anti-
Prelog specificity, transferring the pro-R hydride to form
(R)-alcohols. Previously we have used various dehy-
drogenases to catalyze the enantioselective reduction
and reductive amination of ketones and a-ketoacids to
The biotransformation process was scaled up to 1-L and
500-L fermentors. Cells were suspended in 1 L of 50 mM
Table 5. Purification of ketoreductase from Pichia methanolica SC 13825
Step
Volume (mL)
Total protein
(mg)
Total units
Sp. activity
(units/mg protein)
Purification (fold)
Cell extracts
HiTrap affinity column (pH gradient)
HiTrap affinity column (NADP gradient)
265
51
2
2170
19.5
0.06
9.4
1.2087
0.0638
0.0043
0.0619
1.063
1
14.4
247
The reaction mixture for the enzyme assay contained 1 mg substrate 3, 4 mM NADP, 4 units glucose dehydrogenase in 1 mL. The reaction was carried
out at 30 ꢁC for 5 h, and the concentration of product 4 determined by HPLC. One unit is defined as 1 lmol of 4 formed/h.

1250
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
prepare alcohols and chiral amines required for the
synthesis of antihypertensive, anticholesterol, antican-
cer, and antiviral drugs.27–30
and solvents were used as received. Thin-layer chroma-
tography was performed on silica gel F254, 2.5 · 7.5 cm
plates, obtained from EM Science. The HPLC condi-
tions used for 10, 11, 12, 13, and 3 were as follows:
Column: YMC ODS-AQ, 4.6 · 150 mm; UV detection at
225 nm; solvent A: water–CH3CN (60:40 v/v); solvent B:
water–CH3CN (20:80 v/v), gradient program starting
with 100% A for 10 min, changing to 100% B over 5 min,
holding at 100% B for 15 min, changing to 100% A over
5 min and final equilibration with 100% A for 5 min;
flow rate: 0.8 mL/min to follow the formation of 10 (rt
8.6 min) and 1.2 mL/min to follow the formation of 11
(rt 4.6 min), 12 (rt 17.3 min), 13 (rt 19.7 min), and 3 (rt
6.5 min). The HPLC conditions used for 15 and 19 were:
Column: YMC S3 ODS-A, 6.0 · 150 mm; UV detection
at 220 nm; solvent A: 0.2% H3PO4; solvent B: water–
CH3CN (10:90 v/v); gradient program starting at 30% of
B and changing over to 100% B over 15 min and holding
at 100% B for 5 min, changing over to 30% B over 5 min
and final equilibration with 30% B for 5 min; flow rate:
1.5 mL/min; rt for 15 was 12.2 min and 14.5 min for 19.
Nuclear magnetic resonance (NMR) spectra were run
on a Bruker AC-300 spectrometer at 300 MHz for pro-
ton and 75 MHz for carbon (Scheme 1).
3. Conclusion
Herein, we have described the enantioselective reduction
of 2-bromo-4-fluoro-acetophenone 1 to the corre-
sponding (S)-2-bromo-4-fluoro-phenyl ethanol 2 by
Baker’s yeast. We have also demonstrated the enantio-
selective microbial reduction of ketoesters 3, 5, and 7 to
the corresponding (S)-hydroxy esters 4,6, and 8. The
keto reductase from P. methanolica SC 13825 was
cloned and expressed in E. coli. Recombinant E. coli
expressing the keto reductase was used to catalyze the
enantioselective reduction of keto methyl ester 3 to the
corresponding (S)-hydroxy methyl ester 4 on pre-
parative scale.
4. Experimental
4.1. Materials
4.2.1. Preparation of 4-(5-fluoro-2-hydroxy-phenyl)-4-
oxo-butyric acid methyl ester 10. A 20-L reactor was
charged with succinic anhydride (618 g, 6.17 mol) and
aluminum chloride (2210 g, 16.57 mol). After inerting
the vessel with nitrogen, 4-fluoroanisole 9 (650 g,
5.15 mol) dissolved in methylene chloride (2800 mL) was
added under slow agitation while maintaining the reac-
tion temperature below 35 ꢁC. The transfer line was
rinsed with additional methylene chloride (400 mL). The
slurry was heated to 37–42 ꢁC (reflux) and the reflux
temperature maintained until the reaction was judged to
have gone to completion by HPLC (about 9 h) to obtain
the intermediate keto acid. The reaction mixture was
cooled to room temperature and diluted with methylene
chloride (3250 mL).
The 2-bromo-4-fluoroacetophenone was obtained from
Lancaster Chemicals. Starting substrates 3, 5, 7 and
reference compounds 2, 4, 6, 8 were synthesized by
colleagues in the Process Research and Development
Department, Bristol-Myers Squibb Pharmaceutical
Research Institute. The physico-chemical properties,
including spectral characteristics (1H NMR, 13C NMR,
mass spectra), were in full accord for all compounds.
4.2. Preparation of substrates 3, 5, and 7
All reactions were conducted under nitrogen unless
mentioned otherwise. Commercially available reagents
O
F
F
OCH3
F
OCH3
a
b
O
O
OMe
OH
OH
11
9
10
F
OCH3
F
F
OCH3
c
d
O
O
OSO2CF3
On-Bu
12
13
OCH3
e
O
O
3
Scheme 1. Reagents and conditions: (a) CH2Cl2, AlCl3, succinic anhydride; MeOH and HCl (g); (b) H2, Pd(OH)2, MeOH; (c) CH2Cl2, Tf2O,
DIPEA; (d) Pd(OAc)2, DPPP, DIPEA, n-butyl vinyl ether; (e) 2 M HCl.

R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
1251
A separate 25-L reactor was charged with methanol
(13,000 mL) and cooled to <10 ꢁC at which point agi-
tation commenced. The methylene chloride solution of
the reaction mixture obtained above was charged slowly
while maintaining the batch at <35 ꢁC. Anhydrous HCl
gas (650 g, 18 mol) was bubbled into the reaction mix-
ture at ꢀ10–25 g/min while maintaining the batch at
<25 ꢁC. The resulting slurry was then stirred at ambient
temperature until the reaction to obtain the keto ester
was determined to be complete by HPLC (18–24 h). The
batch was concentrated to ꢀ6500 mL, cooled to <10 ꢁC,
and then treated slowly with 1 M HCl (13,000 mL) and
stirred for an additional 4 h. The solid obtained was
filtered, washed with water (ꢀ6500 mL), and then dried
in a vacuum oven at 645 ꢁC for 24 h to obtain 1064 g of
the keto ester 10 as a pale yellow solid in 91.3% yield; 1H
NMR d 7.5–6.8 (m, 3H), 3.6 (s, 3H), 3.3 (t, 2H,
J ¼ 6:5 Hz), 2.7 (t, 2H, J ¼ 6:5 Hz) ppm; 13C NMR
d 203.6, 173.2, 158.7, 156.7, 153.6, 124.4, 120.2, 116.9,
52.3, 33.5, and 27.8 ppm; FT-IR (KBr) 1746, 1649, 1490,
consisted of the preparation of the triflate ester 12. To a
reactor containing the phenol 11 (650 g), diisopropyl-
ethyl amine (554.2 g, 747 mL), and dichloromethane
(6500 mL) were added and the mixture cooled to )30 ꢁC.
Triflic anhydride (1039.4 g, 620 mL) was added slowly
(ꢀ45 min) while maintaining the batch at <)20 ꢁC. The
reaction mixture was warmed and stirred at 20–22 ꢁC for
15 min. The reaction was followed by TLC. After ꢀ2 h,
DMF (1950 mL) was charged slowly to the reactor and
dichloromethane removed under vacuum while keeping
the batch temperature at 50–60 ꢁC until no condensation
of dichloromethane was observed.
4.2.4. Palladium coupling to prepare the vinyl ether 13.
Another reactor was charged with Pd(OAc)2 (37.25 g),
DPPP (70.86 g), diisopropylethyl amine (554.8 g,
747.7 mL), DMF (1.86 kg, 1.94 L) and agitated under
nitrogen for 15 min. The triflate 12 prepared in the first
step was added to the catalyst mixture over 5 min fol-
lowed by n-butyl vinyl ether (462.58 g, 597.60 mL). The
reaction mixture was heated to 83–85 ꢁC over 90 min
and the temperature held until the reaction was judged
to have gone to completion by HPLC (ꢀ2 h). Water
(1100 mL) was charged to the reaction mixture and the
mixture stirred well. The product was extracted into
heptane (1 · 3400 mL, 1 · 2000 mL, and 3 · 1500 mL).
The combined heptane extract was washed with water
(833 mL) and then concentrated under reduced pressure
to remove solvent while maintaining the batch temper-
ature at 50–60 ꢁC to obtain 889 g of crude vinyl ether 13.
MTBE (4300 mL) was charged to the reactor containing
13 and agitated well to obtain a solution. HCl (2 M,
3430 mL) was added and the biphasic mixture was agi-
tated vigorously at ambient temperature for about 3 h,
at which time HPLC indicated that the hydrolysis was
complete. The organic phase was separated, washed
with water (3 · 1200 mL, final pH of the aqueous phase
3.1), and then concentrated under reduced pressure to
remove solvent. The methyl ketone 3, 769 g, was
obtained as a light amber oil containing ꢀ6 wt % of
n-BuOH as a contaminant (NMR determination). The
oil was further purified using an EVAPOR thin-film
evaporator to remove n-BuOH. After two passes, the
n-BuOH content was reduced to 2.2 wt % and the methyl
ketone 3, 671.8 g, was obtained as a light yellow liquid in
1449, 1378, 1285, 1173, 1009, 845, 794, and 676 cmꢁ1
;
Anal. Calcd for C11H11O4FÆ0.04H2O: C, 58.21; H, 4.92.
Found: C, 57.90; H, 4.68.
4.2.2. Preparation of 4-(5-fluoro-2-hydroxy-phenyl)-buty-
ric acid methyl ester 11. A Buchi hydrogenator flask was
charged with keto ester 10 (1049 g, 4.63 mol) and
methanol (4140 mL) and stirred well. Pearlman’s cata-
lyst (105 g, 10 wt % loading, Degussa type, 50 wt % wet
catalyst) was added to the resulting solution. Hydroge-
nation was conducted at ꢀ60 psi for ꢀ16 h. HPLC
indicated that the reaction was complete. The reaction
mixture was filtered on a bed of 40 g of Celite to remove
the catalyst. The Celite bed was washed with methanol
(6 · 200 mL). The combined washings and the initial
product-rich filtrate were concentrated under reduced
pressure, keeping the bath temperature at 645 ꢁC to
obtain crude 11 as a solid (938 g). Toluene (1000 mL)
was added and the slurry heated to ꢀ50 ꢁC to obtain a
solution. After a polish filtration to remove some fine
suspended particles, heptane (1000 mL) was added over
a period of 60 min to the filtrate maintained at ꢀ50 ꢁC.
The resulting hazy solution was stirred at the same
temperature for 60 min and then treated with another
portion of heptane (1000 mL). The cloudy solution was
stirred for an additional hour at the same temperature
and then for two more hours at ambient temperature.
The solid obtained was filtered, washed with a toluene–
heptane mixture (150 mL: 850 mL) and dried in a vac-
uum oven at ambient temperature for 24 h to obtain
759 g of 3 as a gray solid in 77.3% yield; 1H NMR d 6.7
(m, 3H), 3.6 (s, 3H), 2.6 (t, 2H, J ¼ 6:5 Hz), 2.3 (t, 2H,
J ¼ 6:5 Hz), 1.9 (m, 2H) ppm; 13C NMR d 175.9, 158.7,
155.5, 150.6, 129.2, 114.9, 52.4, 33.2, 29.7, and 25.0 ppm;
FT-IR (KBr) 1716, 1511, 1439, 1367, 1337, 1204, 1147,
1091, and 850 cmꢁ1; Anal. Calcd for C11H13O3FÆ
0.11H2OÆ0.1MeOH: C, 61.32; H, 6.32. Found: C, 61.21;
H, 5.88.
1
90% yield (corrected for n-BuOH content). H NMR d
7.8 (m, 1H), 7.0 (m, 2H), 3.6 (s, 3H), 2.9 (dd, 2H), 2.5 (s,
3H), 2.4 (dd, 2H), and 1.9 (m, 2H) ppm; 13C NMR d
200.3, 174.1, 166.1, 162.8, 146.2, 133.9, 132.6, 115.8,
51.8, 33.9, 33.7, 29.9, and 26.7 ppm; FT-IR (film) 1741,
1690, 1608, 1588, 1501, 1439, 1362, 1245, 1163, 1111,
1070, 983, and 825 cmꢁ1
;
Anal. Calcd for
C13H15FO3Æ0.17n-BuOH: C, 65.52; H, 6.50. Found: C,
64.86; H, 6.40.
4.2.5. Preparation of 4-(2-acetyl-5-fluoro-phenyl)-butyric
acid ethyl ester 5. A dry 1 L, round-bottomed flask was
charged with PdCl2(PPh3)2 (1.62 g). After flushing with
nitrogen, DMA (120 mL) was charged to the flask fol-
lowed by the addition of the bromo ketone 14 (10 g,
46.07 mmol). To the resulting suspension, 4-ethoxy-4-
4.2.3. Preparation of 4-(2-acetyl-5-fluoro-phenyl)-butyric
acid methyl ester 3. Preparation of 12: The first step

1252
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
oxobutylzinc bromide (0.5 M solution in THF, 120 mL,
60 mmol) was added over a period of 20 min. The
reaction mixture was stirred at ambient temperature for
ꢀ60 h. After cooling in an ice bath, the reaction mixture
was quenched with 1 M HCl (100 mL) and then stirred
for an hour at ambient temperature. THF was removed
under reduced pressure and the resulting mixture
extracted with MTBE (2 · 200 mL). The MTBE extract
was washed with water (2 · 200 mL) followed by brine
(200 mL) and dried over Na2SO4, and then concentrated
under reduced pressure to remove solvent. The crude
product was purified by flash chromatography on silica
gel to obtain 9.2 g of ethyl ester 5 as a yellow liquid in
79.2 M % yield. 1H NMR d 7.8 (m, 1H), 6.9 (m, 2H), 4.1
(q, 2H, J ¼ 7 Hz), 2.9 (dd, 2H), 2.6 (s, 3H), 2.4 (t, 2H),
1.9 (m, 2H), and 1.1 (t, 3H, J ¼ 6:8 Hz) ppm; 13C NMR
d 200.3, 173.7, 166.1, 162.8, 146.3, 134.0, 132.5, 115.8,
60.6, 34.2, 33.7, 30.0, 26.7, and 14.59 ppm; FT-IR (film):
1741, 1685, 1613, 1582, 1495, 1429, 1357, 1234, 1188,
1111, 800, and 973 cmꢁ1; Anal. Calcd for C14H17FO3: C,
66.65; H, 6.79; F, 7.53; Found: C, 66.57; H, 6.59; F, 7.32
(Scheme 2).
4.2.7. Preparation of 4-(2-acetyl-5-fluoro-phenyl)-butyric
acid tert-butyl ester 7. A 250-mL, three-necked, round-
bottomed flask was charged with zinc powder (1.7 g,
26.9 mmol) and lithium iodide (0.6 g, 4.5 mmol), then
heated under vacuum at 100 ꢁC for 1 h. After cooling to
ambient temperature, dimethylacetamide (30 mL) was
added to the flask followed by dibromoethane (0.25 g,
1.3 mmol). The reaction mixture was heated and held at
70 ꢁC for 10 min. After cooling it back to ambient tem-
perature, trimethylsilyl chloride (0.15 g, 1.3 mmol) was
added and the reaction mixture was stirred for an
additional 30 min. Bromo ester 17 (1.0 g, 4.48 mmol) was
added and the mixture heated to 75 ꢁC. The rest of the
bromo ester (4.0 g, 17.92 mmol) was added over a 30 min
period and heating continued for an additional 5 h. The
reaction mixture was then cooled to ambient tempera-
ture and transferred over 15 min to a flask containing
the bromo ketone 14 (3.0 g, 13.82 mmol) and
PdCl2(PPh3)2 (0.48 g, 0.69 mmol) in anhydrous THF
(15 mL) and stirred for ꢀ18 h. HCl (1M, 50 mL)
was added and stirring continued. The reaction mixture
was extracted with MTBE (100 mL). The organic phase
was separated and the aqueous phase back-extracted
with MTBE (50 mL). The combined organic phase was
washed with water (2 · 100 mL) and brine (2 · 50 mL)
and then stirred with deactivated carbon (2 g) for ꢀ2 h.
After drying over Na2SO4, the organic phase was con-
centrated under vacuum to a brown oil. The crude
product was further purified by chromatography on
silica gel to obtain 2.4 g of 7 as a yellow oil in 62% yield.
1H NMR d 7.8 (m, 1H), 6.8 (m, 2H), 2.8 (dd, 2H), 2.5 (s,
3H), 2.3 (t, 2H), 1.9 (m, 2H), 1.4 (s, 9H) ppm; 13C NMR
d 200.3, 173.0, 166.1, 162.7, 146.4, 134.0, 132.4, 115.8,
80.4, 35.4, 33.6, 30.0, 28.4, and 26.8 ppm; FT-IR (film):
F
O
F
Br
O
a
O
O
14
5
Scheme 2. Reagents and conditions: (a) DMA, PdCl2(PPh3)2,
BrZn(CH2)3COOEt.
1726, 1690, 1603, 1593, 1367, 1239, 1157, and 968 cmꢁ1
;
4.2.6. Preparation of 4-bromo-butyric acid tert-butyl ester
17. Trifluoroacetic anhydride (6 mL) was added slowly
to a solution of 4-bromo butyric acid 16 (3.4 g,
20.48 mmol) in dry THF (30 mL), pre-cooled to )40 ꢁC.
The resulting solution was stirred at the same tempera-
ture for ꢀ30 min. tert-Butyl alcohol (25 mL) was added
and the solution allowed to warm to ambient tempera-
ture and stirred for an additional 16 h. The reaction
mixture was poured slowly into a mixture of crushed ice
and saturated sodium bicarbonate solution (50 mL). The
product was extracted into EtOAc (100 mL) and the
organic phase was washed with water (100 mL) and
brine (100 mL). After drying over Na2SO4, the solution
was concentrated under reduced pressure to obtain an
oil. The oil was dissolved in MTBE (100 mL) and filtered
through a short pad of silica-gel. The silica pad was
washed with MTBE (100 mL). The combined MTBE
filtrate and washing were concentrated under reduced
pressure to obtain 4.1 g of 17 as a liquid in 90 M % yield
(Scheme 3).
Anal. Calcd for C16H21FO3: C, 68.55; H, 7.55; F, 6.77;
Found: C, 68.42; H, 7.46; F, 6.68 (Scheme 4).
F
O
F
Br
O
a
O
O
7
14
Scheme 4. Reagents and conditions: (a) PdCl2(PPh3)2, THF, and then
BrZn(CH2)3COOtBu.
4.3. Microorganisms
Microorganisms (Tables 1–4) were obtained from the
culture collection of the Bristol-Myers Squibb Pharma-
ceutical Research Institute. Microbial cultures were
stored at )90 ꢁC in vials.
O
O
a
b
OH
BrZn
Br
Br
O
O
O
18
16
17
Scheme 3. Reagents and conditions: (a) (CF3CO)2O, t-BuOH, THF; (b) DMA, Zn, Lil, Br(CH2)Br, TMSCI.

R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
1253
4.4. Growth of microorganisms
similar conditions, this process was scaled-up to 200-L
scale to reduce 1 kg of ketone 1.
For screening purposes, one vial (1 mL) of each culture
was used to inoculate 100 mL of medium A (1% malt
extract, 1% yeast extract, 2% glucose, and 0.3% pep-
tone). The medium was adjusted to pH 6.8 before steri-
lization. Cultures were grown at 28 ꢁC and 200 rpm for
48 h. Cultures were harvested by centrifugation at
18,000g for 15 min, washed with 100 mM potassium
phosphate buffer (pH 7.0), and suspended in 10 mL
of the same buffer, and then used for reduction
studies.
4.7. Recovery of product 2
At the end of the biotransformation (20 h), 50 g of
XAD-16 resin (previously washed with 500 mL of water
containing 50% methanol and then with 2 · 500 mL of
water) was added to 1 L of the reaction broth. The
mixture was stirred at room temperature at 300 rpm for
3 h and then filtered through 40 mesh stainless steel
sieve. The collected resin was washed with 50 mL of
water containing 20% methanol and filtered through a
stainless steel sieve. The product-rich resin was treated
with 50 mL of methyl tert-butyl ether (MTBE) to exude
the desired alcohol 2. The solution was dried over
anhydrous sodium sulfate, filtered, and the solvent
removed in vacuo to provide 4.25 g of an oil in 85%
overall yield. HPLC analysis showed 98.9% HI for
product 2 with an ee of 99.8%. The same procedure was
used to recover alcohol 2 from a 200-L biotransform-
ation batch. After adsorption of the product on XAD-
16 and subsequent desorption of the product from the
rich resin by methyl tert-butyl ether extraction and
evaporation, about 1.03 kg of oil was recovered. The
crude product was purified by silica gel chromatography
to provide 707.4 g of product 2 in 99.6% purity with an
ee of 99.9%.
4.5. Reduction of 2-bromo-4-fluoro acetophenone 1
Cells of various microorganisms were suspended sepa-
rately in 100 mM potassium phosphate buffer (pH 7.0) at
20% (w/v, wet cells) cell concentration and supple-
mented with 1.5 mg/mL of acetophenone 1 and 25 mg/
mL of glucose. Reduction was conducted at 28 ꢁC and
150 rpm. Periodically, samples (1 mL) were taken and
extracted with 4 mL of acetonitrile. The sample was fil-
tered through a 0.2 lm LID/X filter and analyzed using
a Hewlett Packard 1070 high pressure liquid chro-
matograph (HPLC) to determine the substrate 1 and
product 2 concentration. The remaining solution was
evaporated to dryness under a stream of nitrogen and
the residue taken up in 1 mL of ethanol, filtered and
analyzed by HPLC to determine the enantiomeric excess
of product 2. A phenylhexyl (150 · 4.6 mm, 5 lm Phe-
nomenex) column was used. The mobile phase consisted
of 1:1 acetonitrile and water, and was used at a flow rate
of 1.0 mL/min. The detection wavelength was 210 nm
and the column temperature 50 ꢁC. The retention times
for substrate 1 and product 2 were 6.3 and 5.4 min,
respectively. Separation of the two enantiomers of the
racemic product 2 was achieved on a Chiralpak AD
(4.6 mm · 250 mm, 10 lm, Diacel Chemical Industry
Ltd.) column. The mobile phase consisted of 0.249%
absolute ethanol in hexane. The flow rate was 1 mL/min
and the detection wavelength 210 nm. The retention
times for the desired (R)-enantiomer was 48 min and
that for the undesired (S)-enantiomer 54 min.
4.8. Reduction of methyl-, ethyl, and tert-butyl 4-(20-
acetyl-50-fluorophenyl) butanoate
Various microbial cultures (1 mL) were inoculated into
100 mL of medium A or medium B (1% malt extract, 1%
yeast extract, 0.3% peptone, and 2% glycerol adjusted to
pH 7) in a 500-mL flask and incubated at 28 ꢁC and
200 rpm on a shaker for 48 h. Cells were harvested by
centrifugation and suspended in 10 mL of 100 mM
potassium phosphate buffer (pH 7.0) at 20% (w/v, wet
cells) concentration. Glucose (25 mg/mL) and keto
methyl ester 3 (1 mg/mL) were added to the cell sus-
pensions. The reduction was carried out at 28 ꢁC and
200 rpm on a shaker for 24–48 h. At predetermined
times, the reaction mixtures were quenched with four
volumes of acetone, mixed on a vortex mixer, filtered
through a 0.2 lm filter, and collected, with a 1 mL
sample analyzed by HPLC to determine the keto methyl
ester 3 and hydroxy methyl ester 4 concentrations. The
remaining solution was evaporated to dryness under a
stream of nitrogen and the residue taken up in 1 mL of
ethanol, filtered, and analyzed by chiral HPLC to
determine the ee of product 4. A YMC ODS-A
(150 · 4.6 mm, 5 lm) column was used. The mobile
phase consisted of a 40% acetonitrile and 60% H3PO4
(0.3%) mixture used at a flow rate of 1.0 mL/min. The
detection wavelength was 210 nm and the column tem-
perature 50 ꢁC. The retention times for keto methyl ester
3 and hydroxy methyl ester 4 were 11.1 and 6.7 min,
respectively. Separation of the two enantiomers of the
racemic hydroxy methyl ester was achieved on a Chir-
alpak AD (4.6 mm · 250 mm; 10 lm) column (Diacel
4.6. Reduction of 2-bromo-4-fluoro acetophenone 1 using
Baker’s yeast
A 3-L bioreactor (Braun Biostat B) was equipped with a
pH electrode and the impeller speed and temperature set
at 500 rpm and 28 ꢁC, respectively. Phosphate buffer
(800 mL of 10 mM, pH 6.0) was added to the bioreactor
and 150 g of Baker’s yeast (from Red Star) added slowly
over a 30 min period. The contents of the bioreactor
were stirred at 500 rpm, with the temperature main-
tained at 28 ꢁC, and the pH maintained at 6.0. Aceto-
phenone 1 (5 g) and 192 mL of glucose solution (25%)
were added to the reactor at the rate of 8 mL/h over a
period of 24 h. Foaming was controlled by addition of
0.5 mL of SAG antifoam as required. Samples (1 mL)
were taken at intervals and analyzed for substrate 1 and
product 2 concentrations and ee of product 2. Under

1254
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
Chemical Industry Ltd). The mobile phase consisted of
95% solvent A (hexane/absolute ethanol, 98.5%:1.5%)
and 5% solvent B (hexane/isopropanol, 95:5) mixture.
The flow rate was 1 mL/min and the detection wave-
length was 210 nm. The retention time for the desired
(R)-enantiomer was 23.6 min and that for the undesired
(S)-enantiomer was 31.1 min. Using similar conditions
microbial reductions of keto ethyl ester 5 and keto tert-
butyl ester 7 were carried out. The analysis of keto ethyl
ester 5 and the hydroxy ethyl ester was carried out using
a phenylhexyl (150 · 4.6 mm, 5 lm). The mobile phase
consisted of a 50% acetonitrile and 50% water mixture
used at a flow rate of 1.0 mL/min. The detection wave-
length was 210 nm. The retention times for keto ethyl
ester 5 and hydroxy ethyl ester 6 were 12.4 and 8.9 min,
respectively. Separation of the two enantiomers of
the racemic hydroxy ethyl ester was achieved on a
Chiralpak AD (4.6 mm · 250 mm; 10 lm) column
(Diacel Chemical Industry Ltd). The mobile phase
consisted of a hexane/absolute ethanol, and isopropanol
(96.9%:2.85%:0.25%) mixture. The flow rate was 1 mL/
min and the detection wavelength 210 nm. The retention
time for the desired (R)-enantiomer was 15.0 min and
that for the undesired (S)-enantiomer was 19.0 min. The
analysis of keto tert-butyl ester 7 and hydroxy tert-butyl
ester was carried out using a phenylhexyl (150 · 4.6 mm,
5 lm). The mobile phase consisted of a 50% acetonitrile
and 50% water mixture used at a flow rate of 1.0 mL/
min. The detection wavelength was 210 nm. The reten-
tion times for keto tert-butyl ester 7 and hydroxy tert-
butyl ester 8 were 28.3 and 19.2 min, respectively.
Separation of the two enantiomers of the racemic
hydroxy tert-butyl ester was achieved on a Chiralpak
AD (4.6 mm · 250 mm; 10 lm) column (Diacel Chemical
Industry Ltd). The mobile phase consisted of a hexane/
isopropanol (90%:10%) mixture. The flow rate was
1 mL/min and the detection wavelength 210 nm. The
retention time for the desired (R)-enantiomer was
23.6 min and that for the undesired (S)-enantiomer
32.8 min.
paste was collected on a Sharples centrifuge and stored
at )60 ꢁC for future use.
4.10. Purification of ketoreductase from P. methanolica
Preparation of cell extracts was carried out at 4–7 ꢁC.
P. methanolica SC 13825 cells were washed with 10 mM
potassium phosphate buffer pH 6.5 (buffer A). Washed
cells (55 g) were suspended in 275 mL of buffer A con-
taining 10% glycerol, 1 mM DTT, 0.5 mM phenylmeth-
ylsulfonyl fluoride (PMSF) and than homogenized. A
20% cell suspension was disintegrated with a Microflu-
idizer (Microfluidics, Inc.) at 12,000 psi (three passages)
and disintegrated cells were centrifuged at 25,000g for
30 min to obtain the cell extract. Protein in the cell
extracts was estimated by Bio-Rad protein reagent using
bovine serum albumin as standard. The assay mixture
contained 1–10 lL of enzyme fraction, 0.8 mL water,
and 0.2 mL Bio-Rad reagent. After mixing, the absor-
bance of the solution was measured at 595 nm. The
activity of the enzyme was measured by reduction of
keto methyl ester 3 to the corresponding hydroxy methyl
ester 4 using the HPLC system described previously. The
reaction mixture contained 5 mL of enzyme solution
supplemented with 2 mM reduced nicotinamide adenine
dinucleotide phosphate (NADPH), 20 units glucose
dehydrogenase, 25 mg glucose, and 2.5 mg keto methyl
ester 3. The reaction mixture was incubated at 28 ꢁC and
100 rpm. After 6 h incubation, the reaction mixtures
were quenched with four volumes of acetone, mixed on
a vortex mixer, and filtered through a 0.2 lm filter, with
a 1 mL sample was analyzed by HPLC. The cell extract
was loaded onto a Hi-Trap Blue-Sepharose affinity
column, which was equilibrated with buffer A contain-
ing 10% glycerol and 2 mM DTT. The protein was
eluted from the column by buffer A using a pH gradient
running from pH 6.5–8.5. Active fractions eluted from
the column were pooled and loaded again on to a Hi-
Trap Blue-Sepharose affinity column and eluted with
buffer A containing a NADP gradient running from 0.1
to 0.5 mM. Active fractions with similar specific activity
of enzyme were pooled and analyzed by sodium dodecyl
sulfate polyacrylamide (SDS/PAGE) gel electrophoresis
to check the purity of the protein.
4.9. Growth of Pichia methonica in fermentor
P. methanolica SC 13825 was grown in 380-L fermentors
containing 250 L of medium A containing 0.025% SAG
and 0.025% Dow Corning antifoam. Growth consisted
of two inoculum development stages and one fermen-
tation stage. Inoculum development consisted of F1 and
F2 stages. In the F1 stage, 1 mL of a culture of P. met-
hanolica SC 13825 was inoculated into 100 mL of med-
ium A contained in 500-mL flasks. Growth was carried
out at 28 ꢁC and 200 rpm for 24 h on a rotary shaker. In
the F2 stage, 100 mL of F1 stage culture of the organism
was inoculated into 1 L of medium A and incubated at
28 ꢁC and 200 rpm for 48 h. A fermentor containing
250 L of medium A was inoculated with 2 L of F2 stage
inoculum and grown at 28 ꢁC and 200 rpm agitation
with 250 LPM (liter per minute) aeration. During fer-
mentation, cells were harvested periodically by centri-
fugation from 200 mL of culture broth and assayed for
conversion of keto methyl ester 3 to hydroxy methyl
ester 4. At the end of the fermentation, 12 kg of wet cell
4.11. Sodium dodecyl sulfate polyacrylamide gel-electro-
phoresis
The active fractions from the second Hi-Trap Blue-
Sepharose affinity column were evaluated by SDS-
PAGE as described in the PhastSystemꢂ procedure by
Pharmacia31 using the homogeneous 12.5% Phastgel.
The enzyme samples were added to a buffer containing
10 mM Tris–HCl, 1 mM EDTA, pH 8, 2.5% SDS, and
5% b-mercaptoethanol. The mixture was heated at
100 ꢁC for 5 min, and bromophenol blue was added to
0.01%. Gels were stained with silver stain and destained
in 10% acetic acid solution. Markers with standard
molecular weights were phosphorylase b (94,000),
bovine serum albumin (67,000), ovalbumin (43,000),

R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
1255
carbonic anhydrase (30,000), soybean trypsin inhibitor
(20,100), and a-lactalbumin (14,400).
units was diluted into 1 mL LB medium and spread
evenly on the filter. The plate was incubated at 37 ꢁC for
24 h. Colonies were replicated onto two fresh filters,
which were placed onto LB containing kanamycin agar
medium and incubated at 37 ꢁC for 6 h. Lysis of cells and
neutralization of released DNA was performed accord-
ing to directions that were provided with the filters. The
DNA was crosslinked to the filters using a UV Strata-
linker 2400 unit (Stratagene, Inc., La Jolla, CA) in the
‘auto crosslink’ mode. Cell debris was removed by
placing the filters in a container with a solution of 3·
SSC (20· SSC contains, per liter, 173.5 g NaCl, 88.2 g
sodium citrate, pH adjusted to 7.0 with 10 N NaOH),
0.1% SDS and rubbing the lysed colonies with a wet
kimwipe. The filters were then incubated with the same
wash solution for at least 3 h at 65 ꢁC.
4.12. Determination of molecular weight
The molecular weight of the purified keto reductase was
determined by size-exclusion chromatography using a
Pharmacia Superose columnꢂ (15 cm · 1 cm). The col-
umn was equilibrated with buffer A. The reductase was
applied to the column and eluted with buffer A at a flow
rate of 0.4 mL/min. Fractions of 1 mL were collected. A
standard protein mixture containing thyroglobulin
(669,000 MW), ferritin (440,000 MW), human IgG
(150,000 MW), human transferrin (81,000 MW), oval-
bumin (43,000 MW), and myoglobin (17,600 MW) was
also applied to the column and eluted with buffer A.
4.13.2. Selection of clones containing the ketoreductase
gene. Mixed oligonucleotide primers based on partial
amino acid sequences of the purified P. methanolica
ketoreductase were prepared. All possible combinations
of sense and antisense primers were utilized in poly-
merase chain reactions (PCR). The reaction consisted of
0.05 M Tris–HCl (pH 8.3), 250 lg/mL bovine serum
albumin, 2% (w/v) sucrose, 0.1 mM cresol red, 0.2 mM
each dATP, dCTP, dGTP, dTTP, 4 mM MgCl2,
0.0005 mM each primer, 0.25 lL (0.625 U) Takara Z-
Taq DNA polymerase (PanVera, Madison, WI), and
0.1 lg P. methanolica chromosomal DNA in a total
volume of 0.05 mL. Amplification was carried out in a
Perkin–Elmer Model 480 Thermal Cycler under the
following conditions: Denaturation at 94 ꢁC for
4 min, followed by 30 cycles of 94 ꢁC, 1 min; 50 ꢁC,
1 min; 72 ꢁC, 1.5 min, and a final extension at 72 ꢁC for
5 min. Strong amplification of a 650- and 850-bp frag-
ment, respectively, was observed using oligonucleotide
pairs 183 and 186 and 185 and 188 after electrophoresis
of a sample of each reaction on a 1.0% TAE agarose gel.
4.13. Cloning and expression of ketoreductase from
P. methanolica
4.13.1. Construction of partial Sau3A1 library of P.
methanolica SC 13825. P. methanolica SC 13825 chro-
mosomal DNA was prepared from cultures grown on a
YPD medium (1% yeast extract, 2% peptone, and 2%
dextrose) in a 250-mL flask. The flask was incubated at
30 ꢁC with shaking at 250 rpm for 20 h. The procedure
for rapid isolation of Saccharomyces cerevisiae chro-
mosomal DNA was used to prepare P. methanolica
DNA. The precipitated DNA was washed with 70%
ethanol, air-dried, and resuspended to a final concen-
tration of 1.0 mg/mL in TE buffer (0.01 M Tris–HCl,
0.001 M EDTA, pH 8.0). DNA was partially cleaved
with restriction endonuclease Sau3A1. The DNA was
electrophoresed through a 0.8% agarose gel and the
region containing DNA fragments of ca. 5–10 kb iden-
tified by comparison to a 1 kb DNA ladder (Life Tech-
nologies, Gaithersburg, MD) and excised from the gel.
DNA was extracted using the QIAquick Gel Purifica-
tion Kit (Qiagen Inc., Valencia, CA) following the rec-
ommended protocol. An aliquot was electrophoresed on
a 0.8% TAE agarose gel for 20 h at 15 V to confirm that
the desired fragment size range had been obtained and
to determine the concentration of the fragment by
comparison to a DNA mass ladder (Life Technologies).
Fragments were isolated from the agaorse gel and
purified using the QIAquick Gel Extraction Kit. The
DNA was ligated to vector pCR2.1 (Invitrogen, Carls-
bad, CA) according to the manufacturer’s protocol and
transformed into E. coli DH10B by electroporation.
Cells were spread onto LB agar medium containing
50 lg/mL kanamycin and Bluo-gal (Life Technologies;
75 lL of a 2% [w/v] solution in dimethylformide) and
incubated at 37 ꢁC for 20 h. Five white colonies chosen
at random from each ligation/transformation were
inoculated into LB containing kanamycin liquid med-
ium and grown at 37 ꢁC, 250 rpm, for 20 h. Plasmid
DNA was prepared from each sample using the QIA-
prep Spin miniplasmid Kit (Qiagen). The presence of the
expected insert was confirmed by PCR using conditions
described above. Partial amino acid sequences obtained
from the purified enzyme but not used to synthesize
oligonucleotides were also found encoded within the
PCR fragments. Based on these results, digoxigenin-
labeled probes were prepared using two sets of primers
and the PCR DIG Probe Synthesis kit (Roche Bio-
chemicals, Indianapolis, IN) according to the manu-
facturer’s directions. Approximately 10 ng of the
The enriched P. methanolica DNA fragments (5–10 kb)
were ligated to BamHI-cleaved pZero2 in a 0.02 mL
reaction consisting of 0.1 lg chromosomal DNA,
0.03 lg plasmid DNA, 0.03 M Tris–HCl (pH 7.8),
0.01 M MgCl2, 0.01 M dithiothreitol, and 0.0005 M
adenosine-50-triphosphate and 3 Weiss units of T4 DNA
ligase (Promega). The reaction was carried out at 16 ꢁC
for 18 h. The ligated DNA was transformed into elec-
trocompetent DH10B cells (Life Technologies) accord-
ing to the vendor’s recommendations. Following
transformation, 0.96 mL of LB medium was added and
the cells grown at 37 ꢁC for 1 h. A 137 mm Hybond-N+
circle (Amersham-Pharmacia, Piscataway, NJ) was
placed on top of a 150 mm Petri dish containing 75 mL
LB agar containing kanamycin. An aliquot of the partial
Sau3A1 library sufficient to give 5000 colony forming

1256
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
isolated PCR fragment described above was included as
template DNA. Amplification conditions were: Dena-
turation at 94 ꢁC for 4 min, followed by 30 cycles of
94 ꢁC, 1 min; 50 ꢁC, 1 min, 72 ꢁC, 1 min. Incorporation of
the digoxigenin-dUTP nucleotide could be verified by a
significant increase in the molecular weight of the
labeled fragment. Superior incorporation was obtained
using oligonucleotides 183 and 186.Duplicate filters
containing lysed and denatured DNA from the P. met-
hanolica Sau3A1 library were incubated with 10 mL of
DIG EasyHyb solution (Roche) and 5 lL of the dena-
tured, labeled PCR fragment in a roller bottle.
Hybridization proceeded at 42 ꢁC for 18 h. Filters were
washed twice with 2· SSC (prepared from a 20· solu-
tion; 20· SSC contains, per liter, 173.5 g NaCl, 88.2 g
sodium citrate, pH adjusted to 7.0 with 10 N NaOH),
0.1% sodium dodecyl sulfate (SDS) at room temperature
for 5 min, then twice at 68 ꢁC with 0.5% SSC, 0.1% SDS
for 15 min. Anti-digoxigenin antibody binding, washing,
and detection were performed using the DIG Labeling
and Detection Kit reagents and protocols (Roche). The
membranes were placed on Whatman 3 mm paper to
remove excess liquid, covered with Saran Wrap, and
exposed to autoradiography film (Kodak X-OMAT
LS). A single hybridizing colony was picked from the
master filter and streaked onto LB agar medium con-
taining kanamycin and incubated at 37 ꢁC for 24 h. The
colony was grown LB liquid medium containing kana-
mycin for plasmid isolation using the QIAfilter Plasmid
Midi Kit (Qiagen). Restriction mapping indicated an
insert of ca. 5.0 kb was present in the recombinant
plasmid. Isolated DNA was tested for its ability to
support amplification using oligos 183 and 186 and 185
and 188, which was confirmed. Oligonucleotide primers
based on the DNA sequence of the isolated PCR frag-
ments were used for the analysis of the insert in pKR
5.0. An open reading frame of 1059 bp that encodes a
protein of 353 amino acids with a molecular weight of
39,800 Da was found. This is in near agreement with the
size of the isolated ketoreductase (40,000 Da by gel fil-
tration).
identical to those described previously, except pKR 5.0
DNA was used as a template.
The BspHI/BamHI-digested PCR fragment was ligated
to BspHI/BamHI-cleaved pBMS2000. The ligated DNA
was transformed into electrocompetent DH10B cells
(Life Technologies, Inc.) according to the vendor’s rec-
ommendations. Following transformation, 0.96 mL of
LB medium were added and the cells grown at 37 ꢁC for
1 h. An aliquot of the cells was spread onto the LB agar
medium containing 10 lg/mL neomycin sulfate (Sigma)
and the plate incubated at 37 ꢁC for 20 h. The presence
of the correct insert was established by PCR using
conditions described earlier except a portion of 16 ran-
domly chosen colonies was the source of template DNA.
Fourteen out of the 16 colonies amplified a fragment of
the correct size. The plasmid from one of these isolates
was named pBMS2000-PMKR (for ketoreductase).
pBMS2000-PMKR was transformed into competent
E. coli strain BL21-CodonPlus(DE3)-RIL cells (Strata-
gene) according the manufacturer’s protocol. Cells were
spread onto LB agar medium containing 30 lg/mL
chloramphenicol and 10 lg/mL neomycin. Four colonies
were randomly chosen and used as a source of template
DNA for PCR using the conditions described earlier.
All four reactions amplified a DNA fragment of the
correct size, demonstrating that the BL21-Codon-
Plus(DE3)-RIL had been successfully transformed. One
of these isolates was selected as the expression strain and
used for all further experiments. Vial lots of the
expression strain were prepared.
Expression of the ketoreductase gene was controlled by
D galactoside) induction of the
IPTG (isopropylthio-b-
ptac promoter that originated on plasmid pBMS2000.
The expression strain was grown in 25 mL of MT3/neo/
chlor in a 250-mL flask at 15 ꢁC, 225 rpm until it had
reached OD600 nm ꢀ0.7. At this point, IPTG (Life
Technologies) was added to a final concentration of
0.5 mM. The cultures were allowed to grow overnight
(ꢀ16 h) to allow complete induction of the ketoreduc-
tase gene and production of the ketoreductase protein.
Following overnight induction with IPTG, samples were
analyzed on protein gels. The results indicated the
strain’s ability to overexpress the heterologous protein.
4.13.3. Subcloning of the ketoreductase gene and expres-
sion in E. coli. The polymerase chain reaction was uti-
lized to amplify the complete ketoreductase gene
containing restriction sites suitable for cloning into
expression vector pBMS2000. The primers used are gi-
ven below:
4.14. Reduction of keto methyl ester by recombinant
E. coli
Cells of E. coli SC 16445 (expressing ketoreductase)
were grown in a 250-L fermentor. Growth consisted of
inoculum development stage, germinator stage and fer-
mentation stage. In inoculum development stage, 1 mL
culture of E. coli SC 16445 was inoculated into 1 L of
medium B (1.0% NZ Amine A, 2.0% Yeastamin, 2.0%
glycerol, 0.6% Na2HPO4, 0.3% KH2PO4, 0.125%
(NH4)2SO4, 0.0246% MgSO47H2O) containing a filtered
sterilized solution (10 mg in ethanol) of neomycin sul-
fate, and a filtered sterilized solution (33 mg in water) of
chloramphenicol. Growth was carried out in 4-L flasks
at 15 ꢁC and 150 rpm for 72 h. In a germinator stage, a
20-L fermentor containing 15 L medium B containing
BspHI
50 TGCTCATGAATTGGGAAAAAGTTCCACAAG 30
Nucleotides in bold indicate the recognition sequence
for the restriction enzyme BspHI;
BamHI
50 CTCGGATCCTTATAAAATTACAGAATATAAG 30
Nucleotides in bold indicate the recognition sequence
for the restriction enzyme BamHI; PCR conditions were

R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
1257
neomycin sulfate and chloramphenicol (150 mg neomy-
cin sulfate in 100 mL water and 495 mg chloramphenicol
in 50 mL ethanol were separately filter sterilized using a
0.2 lm cellulose nitrate filter and added to the germi-
nator after sterilizing the medium) was inoculated with
1 L inoculum. Cells were grown at 15 ꢁC, 20 LPM (liter
per min) aeration and 1000 rpm agitation for 24 h. A
380-L fermentor containing 250 L of medium B con-
taining neomycin sulfate and chloramphenicol chlor-
amphenicol (2.5 g neomycin sulfate in 600 mL water and
8.25 g chloramphenicol in 600 mL ethanol were sepa-
rately filter sterilized using a 0.2 lm cellulose nitrate
filter and added to the fermentor after sterilizing the
medium) was inoculated with 15 L germinator grown
inoculum. The growth was carried out at 15 ꢁC,
250 LPM aeration, and 200 rpm agitation. Induction of
the ketoreductase was carried out when the optical cell
density (OD at 600 nm) of the culture reached close to
3.0 by the addition of a filter-sterilized solution (14.9 g in
1 L of water) of isopropyl-b-thiogalactoside (IPTG).
Eight hours after addition of IPTG, the addition of
nutrient feed (20 L of feed containing 1% NZ Amine A,
1% yeastamin, 1% glycerol, and 0.02% antifoam
UCON) was initiated at 1 L/h over 20 h. During fer-
mentation, cells were periodically harvested by centri-
fugation from 200 mL of culture broth and assayed for
conversion of keto methyl ester 3 to hydroxy methyl
ester 4. Cells were harvested approximately 30 h after
induction when the highest ketoreductase activity was
observed. At the end of fermentation, the fermentor was
cooled to 6–8 ꢁC and cells collected in a Sharples cen-
trifuge. About 10–11 kg of cell paste was collected and
stored at )70 ꢁC until further used in the biotransfor-
mation process.
and filtered through a stainless steel sieve. The product-
rich resin was treated with 50 mL of methyl tert-butyl
ether (MTBE) to desorb the alcohol 4. The solution was
dried over anhydrous sodium sulfate and filtered and the
solvent was removed on the rotary evaporator to pro-
vide 3.9 g of an oil in 86% overall yield. HPLC analysis
showed 98.9 area percent for product 4 with an ee of
99.9%. The same procedure was used to recover alcohol
4 from a 500-L biotransformation batch.
Acknowledgements
We would like to acknowledge Dr. David Kronenthal
for reviewing this manuscript and providing valuable
suggestions.
References and notes
1. Patel, R. N. Enzyme Microb. Technol. 2002, 31, 804–826.
2. Food & Drug Administration. Chirality 1992, 4, 338–340.
3. Buckland, B. C.; Robinson, D. K.; Chartrain, M. Metab.
Eng. 2000, 2, 42–48.
4. Pesti, J. A.; Dicosimo, R. Curr. Opin. Discov. Dev. 2000, 3,
764–782.
5. Gotor, V. Biocat. Biotrans. 2000, 18, 87–103.
6. Stewart, D. Curr. Opin. Chem. Biol. 2001, 5, 120–129.
7. Zak, A. Curr. Opin. Chem. Biol. 2001, 5, 130–136.
8. Patel, R. N. In Stereoselective Biocatalysis; Patel, R. N.,
Ed.; Marcel Dekker: New York, 2000; pp 87–130.
9. Schnell, B.; Faber, K.; Kroutil, W. Adv. Synth. Catal.
2003, 345, 653–666.
10. Patel, R. N. Adv. Synth. Catal. 2001, 343, 527–546.
11. Churcher, I.; Nadin, A. J.; Owens, A. P. PCT Int. Appl.
2002; 56 pp. CAN 136:325575, AN 2002:293634.
12. Schenk, D.; Games, D.; Seubert, P. J. Mol. Neurosci. 2001,
17, 259–265.
13. Yang, M. G.; Liu, H. PCT Int. Appl. 2001; 170 pp. CAN
135:288707, AN 2001:747759.
14. Olson, R. E. PCT Int. Appl. 2001; 254 pp. CAN
136:20092, AN 2001:886081.
15. Felsenstein, K.; Smith, D. W.; Poss, M. A.; Chaturvedula,
P. C.; Sloan, C. P.; Charles, P. Eur. Pat. Appl. 1997; 30 pp.
CAN 127:109192, AN 1997:470004.
16. Rotella, D. P. Chemtracts 2000, 13, 626–629.
17. Patel, R. R. N.; Goswami, A.; Chu, L.; Nanduri, V.;
Goldberg, S.; Johnston, R.; Donovan, M.-J ; Mirfakhrae,
K. D. Patent Appl. WO 2002-US6783 20020306. CAN
137:277933. AN 2002:754608.
Cells were used to catalyze the bioconversion of keto
methyl ester 3. Cells were suspended in 1 L of 50 mM
pH 7.0 phosphate buffer at 10% (w/v, wet cells) cell
concentration. The suspensions were supplemented with
100 lM nicotinamide adenine dinucleotide phosphate
(NADP), 1 mM phenylmethanesulfonyl fluoride
(PMSF), 20 g glucose, 34,000 units glucose dehydroge-
nase, and 4.5 g substrate 3. Biotransformation was car-
ried out at 500 rpm and 28 ꢁC. The pH was maintained
at 6.8–7.0. Glucose feed (solution containing 20% glu-
cose, 1.92% K2HPO4, and 1.2% KH2PO4) was started
after 1 h at rate of 6 mL/h/L of reaction mixture. The
reaction was completed in 20 h with a 95% reaction yield
of 4. An ee of 99.9% was obtained for 4. The biotrans-
formation process was scaled up to 500 L to reduce 3 kg
of substrate 3.
18. Ward, O. P.; Young, C. S. Enzyme Microb. Technol. 1990,
12, 482–493.
19. Csuk, R.; Glanzer, B. I. Chem. Rev. 1991, 91, 49–97.
20. Sih, C. J.; Chen, C. S. Angew. Chem. 1984, 96, 556–566.
21. Kula, M.-R.; Kragl, U. In Stereoselective Biocatalysis;
Patel, R. N., Ed.; Marcel Dekker: New York, 2000; pp
839–866.
22. Jones, J. B.; Back, J. F. In Applications of Biochemical
Systems in Organic Synthesis; Jones, J. B., Jones, C. J.,
Perlman, S. D., Eds.; John Wiley & Sons: NY, 1976; pp
248–376.
4.15. Recovery of hydroxy methyl ester 4
At the end of the biotransformation, 40 g of XAD-16
resin (previously washed with 500 mL of a 1:1 (v/v)
water and methanol mixture, and then washed with
2 · 500 mL of water) was added to the 1 L of reaction
broth. The mixture was stirred at room temperature at
300 rpm for 8 h and finally filtered through a 40-mesh
stainless steel sieve. The collected resin was washed with
50 mL of a water and methanol mixture (80:20 by vol)
23. Keinan, E.; Hafeli, E. K.; Seth, K.; Lamed, R. J. Am.
Chem. Soc. 1986, 108, 162–168.
24. Prelog, V. Pure Appl. Chem. 1964, 9, 119–123.

1258
R. N. Patel et al. / Tetrahedron: Asymmetry 15 (2004) 1247–1258
25. Bradshaw, C. W.; Hummel, W.; Wong, C.-H. J. Org.
Chem. 1992, 57, 1532–1536.
Thottathil, J. K.; Szarka, L. J. Tetrahedron: Asymmetry
1993, 4, 2069–2084.
26. Bradshaw, C. W.; Fu, H.; Shen, G.-J.; Wong, C.-H.
J. Org. Chem. 1992, 57, 1526–1532.
27. Patel, R. N.; Banerjee, A.; McNamee, C. G.; Szarka, L.
J. Appl. Microbiol. Biotechnol. 1993, 40, 241–245.
28. Patel, R. N.; Banerjee, A.; Howell, J. M.; McNamee, C.
G.; Brzozowski, D.; Mirfakhrae, K. D.; Nanduri, V.;
29. Patel, R. N.; Banerjee, A.; Liu, M.; Hanson, R. L.; Ko, R.
Y.; Howell, J. M.; Szarka, L. J. Biotechnol. Appl. Biochem.
1993, 17, 139–153.
30. Patel, R. N.; Banerjee, A.; McNamee, C.; Brzozowski, D.;
Szarka, L. J. Tetrahedron: Asymmetry 1997, 8, 2547–2552.
31. Heukeshoven, J.; Dernick, R. Electrophoresis 1985, 6, 103–112.