Stereospecific microbial reduction of ethyl 1-benzyl-3-oxo-piperidine-4-carboxylate

Article abstract of DOI:10.1016/j.tetasy.2006.06.025

Microbial reduction of ethyl 1-benzyl-3-oxo-piperidine-4-carboxylate by the majority of evaluated microorganisms gave the ethyl cis-(3R,4R)-1-benzyl-3R-hydroxy-piperidine-4R-carboxylate as the major product in high diastereo- and enantioselectivities. The

Full text of DOI:10.1016/j.tetasy.2006.06.025

Tetrahedron: Asymmetry 17 (2006) 2015–2020
Stereospecific microbial reduction of ethyl
1-benzyl-3-oxo-piperidine-4-carboxylate
Zhiwei Guo, Bharat P. Patel, Richard M. Corbett, Animesh Goswami and Ramesh N. Patel*
Process Research and Development, Bristol-Myers Squibb Pharmaceutical Research Institute, New Brunswick, NJ 08903, USA
Received 7 June 2006; accepted 16 June 2006
Available online 18 July 2006
Abstract—Microbial reduction of ethyl 1-benzyl-3-oxo-piperidine-4-carboxylate by the majority of evaluated microorganisms gave the
ethyl cis-(3R,4R)-1-benzyl-3R-hydroxy-piperidine-4R-carboxylate as the major product in high diastereo- and enantioselectivities. The
3R,4R-hydroxy ester was produced in 97.4% diastereomeric excess (de) and 99.8% enantiomeric excess (ee) by Candida parapsilosis
SC16347, while 99.5% de and 98.2% ee were obtained from reduction by Pichia methanolica SC16415. A few of the evaluated micro-
organisms gave 10–40% of the ethyl trans-(3R,4S)-1-benzyl-3R-hydroxy-piperidine-4S-carboxylate as the minor product.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
NMR spectroscopy. An achiral HPLC method was devel-
oped for determining the extent of reduction and diastereo-
meric excess (de) of the hydroxy esters and chiral HPLC
method was established for determining the enantiomeric
excess of the hydroxy esters. In the chiral HPLC method,
racemic cis-hydroxy ester 2 showed two peaks of equal area
at retention times of 10.3 and 12.7 min, whose absolute
configurations were established to be cis-(3R,4R)-2a, and
cis-(3S,4S)-2b, respectively (see below). In the same chiral
HPLC method, racemic trans-hydroxy ester 3 showed
two peaks of equal area at 8.5 and 11.4 min, whose abso-
lute configurations were established to be trans-(3R,4S)-
3a, and trans-(3S,4R)-3b, respectively (see below).
Enantiomerically pure a-substituted-b-hydroxycarboxyl-
ates containing two chiral centers are useful building blocks
in organic synthesis. Microbial reduction has attracted
increasing attention in asymmetric synthesis in recent
years.1–6 It was reported that Baker’s yeast reduced the
title compound 1 diastereo- and enantioselectively to cis-
(3R,4R)-2a with de 73% and ee >95% under non-fermenting
conditions.7 The de (73%) of the product was low and sepa-
ration would be necessary to increase the de and ee to the
level required for the synthesis of most pharmaceutical
intermediates. The present work describes the microbial
reduction of ethyl 1-benzyl-3-oxopiperidine-4-carboxylate
1 to make the ethyl cis-(3R,4R)-1-benzyl-3R-hydroxypiperi-
dine-4R-carboxylate 2a in high diastereomeric and enantio-
meric excess (Scheme 1).
Three types of Baker’s yeast were evaluated for reduction
of 1 under non-fermenting conditions as described in the
literature.7 Sigma type I gave a negligible amount of prod-
uct. Both Sigma type II and Baker’s yeast from Red Star
gave one hydroxy ester isomer as the major product with
a retention time of 10.3 min in our chiral HPLC method.
Since the literature7 reported the predominant formation
of cis-(3R,4R) 2a with Baker’s yeast under the same condi-
tions, the 10.3 min peak in our chiral HPLC was assigned
the configuration cis-(3R,4R)-2a. The 12.7 min peak in
the chiral HPLC was thus assigned to the other cis-enantio-
mer cis-(3S,4S)-2b.
2. Results and discussion
The racemic cis 2a and 2b and trans 3a and 3b ethyl-1-benz-
yl-3-hydroxypiperidine-4-carboxylate were prepared by
reduction of ethyl 1-benzyl-3-oxopiperidine-4-carboxylate
1 with sodium borohydride followed by separation of the
two diastereomers by preparative HPLC. The assignment
of cis and trans stereochemistries was based on proton
Ninety one microorganisms in four multi-well plates were
screened at 2 g/L substrate input for the reduction of 1.
Most cultures gave cis-(3R,4R)-2a as the major product
and many of them showed high ee and high de. Results
*
Corresponding author. Fax: +1 732 227 3994; e-mail: ramesh.patel@
bms.com
0957-4166/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2006.06.025

2016
Z. Guo et al. / Tetrahedron: Asymmetry 17 (2006) 2015–2020
O
O
O
O
4
O
Microorganism
OH
3
N
N
Bn
Bn
1
O
O
O
O
O
O
O
O
OH
OH
OH
OH
N
N
N
N
Bn
Bn
Bn
Bn
trans-(3R,
4S)-3a, trans-(3S,4R)-3b
cis-(3R,4R)-2a, cis-(3S,4S)-2b,
trans-3
cis-2
Scheme 1. Microbial reduction of compound 1.
for the most promising microorganisms and Baker’s yeast
are shown in Table 1.
Microbial reductions by growing cultures of C. parapsilosis
SC16347 with no pH adjustment using 10 g/L substrate
input were complete (>99% substrate utilized) in 48 h to
give cis-(3R,4R)-2a as the major product. The yield was
determined to be 68% by HPLC, and the de and ee were
99% and 99.6%, respectively (Table 2). The pH of the
growing culture was approximately 5 and was not adjusted.
When the pH was adjusted to 7 just after addition of sub-
strate, the substrate utilized was 94% after 48 h to give 44%
product yield with 97% de (Table 2). Similar experiments
were carried out at 20 g/L substrate input. When the bio-
reduction was carried out at pH 5.0, 97% substrate was
utilized in 48 h and the cis-(3R,4R)-2a was obtained as
the major product in 59% estimated yield, 98% de and
99.6% ee. With adjustment of the pH to about 7, 80% sub-
strate was utilized in 48 h with 26% estimated yield and
94% de (Table 2). Thus, the native pH of approximately 5
is preferred for the reduction by C. parapsilosis SC16347,
and both the yield and de were lower at the higher pH of 7.0.
Table 1. Selectivity for cis-(3R,4R)-2a in the initial screening
Microorganism
SC number Conversion ee
de
(%)
(%) (%)
Pichia methanolica
16415
13865
13895
16347
100
100
100
100
68
98.2 99.5
98.7 99.0
99.8 98.6
99.8 97.4
79.9 97.0
93.9 35.3
Hansenula polymorpha
Hansenula polymorpha
Candida parapsilosis
Baker’s yeast Sigma type II
Baker’s yeast Red Star
42
The four cultures shown in Table 1 were also grown in
flasks, and microbial reductions were carried out at a higher
substrate input of 5 g/L. Pichia methanolica SC16415 and
Candida parapsilosis SC16347 gave cis-(3R,4R)-2a in high
yield with high de and high ee as seen in the initial screening
performed at a lower substrate concentration. Hansennula
polymorpha SC13865 gave cis-(3R,4R)-2a as the major
product but with lower de (83%), and H. polymorpha
SC13895 showed lower product yield (17%) to cis-
(3R,4R)-2a at the higher substrate input (Table 2).
In order to prepare cis-(3R,4R)-2a on gram scale, C. parap-
silosis SC16347 was grown in 1 L of medium. The growing
culture was used for reduction of 5 g of the hydrochloride
of 1. Conversion to cis-(3R,4R)-2a reached 84% after 56 h.
Table 2. Microbial reduction to cis-(3R,4R)-2a in 125-mL flasks
Microorganism
SC number
Substrate input (g/L)
pH
Substrate utilized (%)
Product formed (%)a
ee (%)
de (%)
Pichia methanolica
16415
13865
13895
16347
5
5
5
10
20
10
20
NA
NA
NA
ꢀ5
ꢀ5
7
99
85
18
99
97
94
80
100
78
17
68
59
44
26
97.8
ND
ND
99.6
99.6
ND
ND
99
83
99
99
98
97
94
Hansenula polymorpha
Hansenula polymorpha
Candida parapsilosis
7
Growing culture (20 mL), 28 ꢁC, 200 rpm, 48 h.
NA: pH not adjusted. ND: not determined.
a Yield estimated by HPLC.

Z. Guo et al. / Tetrahedron: Asymmetry 17 (2006) 2015–2020
2017
Both the ee and de were greater than 99%. The crude prod-
uct was purified by flash chromatography to give 2.2 g of
cis-(3R,4R)-2a (50% isolated yield) with AP 93, ee
99.3%, de 99.5%, and [a]D = +48.4 (c 2.35, CHCl3). The
specific rotation of a sample of cis-(3R,4R)-2a with de
73% and ee > 95% was reported to be [a]D = +27.35 (c
1.81, CHCl3) in the literature.7 Our sample of cis-
(3R,4R)-2a had higher de and ee, and thus had a higher
specific rotation (+48.4), as expected. The specific rotation
of the isolated product further confirmed our assignment of
the cis-(3R,4R)-2a.
The whole mixture was neutralized, and the hydroxy acids
generated were re-esterified back to the hydroxy ester. As
shown in Table 4, cis-(3R,4R)-2a was partially epimerized
to 18.6% trans-3a, with 81.4% remaining cis-(3R,4R)-2a.
Since the absolute configuration of cis-(3R,4R)-2a had been
established before, the absolute configuration of trans-3a
isolated from the microbial reduction was, therefore, estab-
lished as trans-(3R,4S)-3a.
Table 4. Epimerization of cis-(3R,4R)-2a
Reaction time (h)
Relative % and retention time in chiral HPLC
3.8 min 8.5 min 10.3 min
Acids mixture trans-(3R,4S)-3a cis-(3R,4R)-2a
In the initial screening, a few cultures gave trans-(3R,4S)-3a
as the major product with ee greater than 99% and moder-
ate de. However, the diastereoselectivity changed when the
biotransformation was conducted on a gram scale. The
reductions gave cis-(3R,4R)-2a as the major product and
trans-(3R,4S)-3a as the minor product with no detectable
trans-(3S,4R)-3b. Results are shown in Table 3. By re-
peated flash chromatography of the crude product from
several batches of gram scale reactions, 0.9 g of pure
trans-(3R,4S)-3a was obtained with AP 98, de 96%, ee
99.5%, and [a]D = +23.6 (c 1.68, CHCl3).
0
2
4
20
68
92
92a
Negligible
Negligible
Negligible
18
33.9
40.1
0.2
0.8
1.6
4.2
9.3
99.8
99.2
98.4
77.8
56.8
49.8
81.4
10.1
18.6
Negligible
a After conversion of hydroxyacids back to ethyl hydroxyesters.
Similarly, trans-(3R,4S)-3a was epimerized partially (Table
5). The relative proportion of cis-(3R,4R)-2a was increased
from 1.8 to 7.7. The cis-(3S,4S)-2b and trans-(3S,4R)-3b
compounds were not evident in the HPLC. Although the
Table 3. Change in diastereoselectivity when preparing trans-(3R,4S)-3a
Microorganism
SC number
Relative % of
trans-(3R,4S)-3a
Initial
Gram scale
screening
reaction
Table 5. Epimerization of cis-(3R,4R)-2a
Hansenula anomala
Pichia anomala
Pichia anomala
13830
16139
16143
70
73
88
25
42
13
Reaction time (h)
Relative % and retention time in chiral HPLC
3.8 min 8.5 min 10.3 min
Acids mixture trans-(3R,4S)-3a cis-(3R,4R)-2a
0
2
4
20
68
92
92a
Negligible
11.1
20.0
30.5
56.6
98.2
87.2
78.3
67.5
41.3
30.3
92.3
1.8
1.7
1.7
2.0
2.1
2.4
7.7
In order to establish the absolute configuration of the
trans-(3R,4S)-3a, two epimerization experiments were
carried out (Scheme 2) using cis-(3R,4R)-2a and trans-
(3R,4S)-3a isolated from microbial reduction experiments.
67.3
Negligible
The hydroxy esters were epimerized at C-4 and also par-
tially hydrolyzed to hydroxy acids by treatment with base.
a After conversion of hydroxyacids back to ethyl hydroxyesters.
O
O
4
R
OH
3
R
a.
a.
N
O
O
O
O
O
O
O
OH
Bn
OH
OH
OH
OH
b.
cis-(3
R
,4
R
)-2a
+
+
N
N
N
N
O
O
Bn
Bn
Bn
Bn
4
S
OH
3
R
trans-(3
R,4
S)-3a
cis-(3
R,4
R)-2a
N
Bn
trans-(3
Scheme 2. Epimerization of the stereogenic center at C-4. Reagents and conditions: (a) K2CO3, EtOH, rt, 92 h; (b) EtOH, DCC, DMAP, rt, 3 h.
R,4S)-3a

2018
Z. Guo et al. / Tetrahedron: Asymmetry 17 (2006) 2015–2020
epimerization did not reach equilibrium, these experiments
conclusively established the absolute configuration of
trans-(3R,4S)-3a.
NaBH4 (0.54 g, 14.27 mmol) was added portionwise over
30 min. After 30 min, the reaction was concentrated in
vacuo to a solid residue. The residue was partitioned
between ethyl acetate (80 mL) and water (40 mL). The
organic phase was collected, dried over Na2SO4, and fil-
tered. The filtrate was concentrated in vacuo to recover
4.12 g of crude product as a yellow oil. The oil was sub-
jected to flash chromatography using an ethyl acetate/hex-
ane elution to give 2.2 g of a mixture of racemic cis-2 and
trans-3 as a colorless oil. The cis-2 and trans-3 isomers were
separated using preparative HPLC. The assignment of cis-2
3. Experimental
3.1. Chemicals and general methods
Ethyl 1-benzyl-3-oxopiperidine-4-carboxylate hydrochlo-
ride was purchased from Aldrich. Other chemicals were
purchased from VWR and/or Sigma–Aldrich.
1
and trans-3 was based on H NMR shifts.
1
NMR spectra were recorded on a Bruker-300 or 400 NMR
spectrophotometer. LC–MS data were recorded on a Shi-
madzu LC–MS system. Rotation data were recorded on
a Perkin–Elmer 241 Polarimeter.
The racemic cis-2 was isolated as a white solid. H NMR
(CDCl3) d 7.29 (m, 5H), 4.2 (m, 1H), 4.18 (q, 2H), 3.53
(s, 2H), 3.10 (broad, OH), 2.98 (m, 1H), 2.88 (m, 1H),
2.38 (m, 1H), 2.22 (d, 1H), 2.02 (m, 2H), 1.77 (m, 1H),
1.26 (t, 3H) ppm; 13C NMR (CDCl3) d 173.3, 137.8,
129.0, 128.4, 127.3, 66.5, 62.5, 60.6, 59.0, 52.1, 45.6, 22.4,
14.2 ppm; HRMS (ES): exact mass calculated for
C15H22HO3 [M+H]+, 264.1600. Found: 264.1607. Elem.
Anal. Calcd for C15H21NO3: C, 68.41; H, 8.03; N, 5.31.
Found: C, 68.37; H, 7.90; N, 5.24.
3.2. HPLC methods
Analytical HPLC methods were performed with a gradient
of solvent A (0.01 M NH4OAc in 80:20 of water–methanol)
and solvent B (0.01 M NH4OAc in 5:20:75 of water–meth-
anol–acetonitrile) at ambient temperature and UV detec-
tion at 220 nm. Achiral HPLC was performed on a
Waters XTerra RP-18 column (3.5 lm, 150 · 4.6 mm) with
a flow rate of 1 mL/min and a gradient from 20% to 100%
solvent B over 5 min and 100% B for an additional 5 min.
The retention times for the substrate 1, cis-2, and trans-3
were 7.8, 4.2, and 5.5 min, respectively. Chiral HPLC was
1
The racemic trans-3 was isolated as a light yellow oil. H
NMR (CDCl3) d 7.29 (m, 5H), 4.17 (q, 2H), 3.96 (m,
1H), 3.52 (d, 2H), 3.00 (dd, 1H), 2.78 (dd, 1H), 2.28 (m,
1H), 2.0 (m, 3H), 1.75 (m, 1H), 1.26 (t, 3H) ppm; 13C
NMR (CDCl3) d 174.3, 137.9, 129.0, 128.3, 127.2, 68.1,
62.7, 60.8, 58.7, 52.3, 48.8, 26.4, 14.2 ppm; HRMS (ES):
exact mass calculated for C15H22NO3 [M+H]+, 264.1600.
Found: 264.1587.
performed on
a Chiralpak AD-RH column (5 lm,
150 · 4.6 mm) with a flow rate of 0.5 mL/min and a gradi-
ent from 60% to 70% solvent B over 15 min. The retention
times were 8.5 min for trans-(3R,4S)-3a, 10.3 min for cis-
(3R,4R)-2a, 11.4 min for trans-(3S,4R)-3b, and 12.7 min
for cis-(3S,4S)-2b.
3.5. Microbial reduction by Baker’s yeasts
Three types of Baker’s yeasts from Saccharomyces cerevi-
siae were used (Sigma YSC-1, type I; Sigma YSC-2, type
II; Red Star, active dry yeast). To 1 g of the Baker’s yeast
in a 50-mL flask was added 5 mL of tap water. After
40 min on a shaker at 200 rpm and 28 ꢁC, a solution of
the hydrochloride of 1 (10 mg) in DMSO (100 lL) was
added. The flasks were kept on the same shaker for 20 h.
The reaction mixture (0.5 mL) was withdrawn, mixed with
1 mL MeOH, filtered through a 0.2 lm filter, and subjected
to HPLC analysis.
3.3. Microorganisms and medium
Microorganisms were obtained from the Bristol-Myers
Squibb culture collection. The SC number denotes the ID
number in the culture collection. The multi-well plates were
prepared with the cultures and stored at À70 ꢁC before use.
One liter of F7 medium contained 10 g malt extract, 10 g
yeast extract, 1 g peptone, and 20 g dextrose, and was
adjusted to pH 7 and autoclaved for 20 min.
3.6. Screening of microorganisms for reduction of 1
3.4. Ethyl 1-benzyl-3-hydroxypiperidine-4-carboxylate
racemic cis-2 and trans-3
Each multi-well plate (Whatman) consists of a total of 24
deep wells (4 · 6). Four multi-well plates containing a total
of 91 yeast cultures were used. The plate was taken out of
the À70 ꢁC freezer and thawed to room temperature. One
milliliter of sterile F7 medium was added to each well.
A 250-mL flask equipped with a pH probe was charged
with ethyl 1-benzyl-3-oxo-piperidine-4-carboxylate hydro-
chloride (5.0 g, 16.79 mmol), water (75 mL), and ethyl ace-
tate (40 mL). The pH of the aqueous phase was adjusted to
pH 9 with 2 M K3PO4 solution (10.5 mL). The contents of
the flask were transferred to a separatory funnel to separate
the phases. The organic phase was collected. The aqueous
phase was extracted with additional ethyl acetate
(40 mL). The combined organic phases were dried over
Na2SO4 and filtered. The filtrate was concentrated in vacuo
to recover 4.46 g of a dark brown oil. The oil was dissolved
in ethanol (50 mL). The solution was cooled to 0 ꢁC.
The microorganisms were grown by shaking in
a
Thermomixer R at 600 rpm at 28 ꢁC for 24 h. A solution
of the hydrochloride of 1 (2 mg) in DMSO (20 lL) was
added to each well. Biotransformation was carried out by
shaking on the same shaker. After 24 h, methanol (1 mL)
was added to each well. The mixtures were filtered through
a 0.2 lm filter, and the filtrates were analyzed by achiral
HPLC to determine the conversion and de by relative area
counts and by chiral HPLC to determine the ee.

Z. Guo et al. / Tetrahedron: Asymmetry 17 (2006) 2015–2020
2019
3.7. Microbial reductions at 5, 10, and 20 g/L substrate
input in flasks
2.03 (m, 2H, 5-CH2-A and 6-CH2-B), 1.73 (m, 1H,
5-CH2-B), 1.24 (t, J = 7.2 Hz, 3H, CH3 in Et) ppm. 13C d
172.72, 137.02, 128.68 (2C), 127.92 (2C), 126.89, 65.99,
61.98, 60.08, 58.40, 51.56, 45.10, 21.85, 13.81 ppm.
The four cultures in Table 1 were grown in flasks for micro-
bial reduction at 5, 10, and 20 g/L substrate input. One
thawed frozen vial of microbial culture was inoculated in
100 mL F7 medium (500-mL flask). The flask was placed
on a shaker at 28 ꢁC and 200 rpm. After 64 h, the growing
culture was transferred (15% transfer) into fresh sterile F7
medium and shaken under the same conditions for 24 h.
The growing culture (20 mL) was transferred to a sterile
flask (125 mL). A solution or suspension of the substrate
(hydrochloride of 1) in water was added to make the sub-
strate input 5, 10, or 20 g/L. The pH was approximately 5
and was not adjusted. The flasks were placed on a shaker
at 28 ꢁC and 200 rpm for 48 h. For analytical samples,
0.5 mL of the reaction mixture was taken out and mixed
with 3.5 mL of acetonitrile, filtered through a 0.2 lm filter
and subjected to HPLC. The relative conversion was based
upon the HPLC area ratio of the product and the remaining
substrate. A standard sample of cis-(3R,4R)-2a was also
analyzed by the same HPLC method. The estimated yield
was the HPLC area of the product versus the standard.
3.9. Microbial reductions on gram scale to prepare
trans-(3R,4S)-3a
One thawed frozen vial of microbial culture of Pichia
anomala SC16143 was inoculated in 100 mL F7 medium
(500-L flask). The flask was placed on a shaker at 28 ꢁC
and 200 rpm. After 64 h, the growing culture was trans-
ferred (15% transfer) into fresh sterile F7 medium and
shaken under the same conditions for 24 h. The growing
culture was again transferred (15% transfer) into 1 L F7
medium (4-L flask) and shaken for 24 h at 28 ꢁC and
200 rpm. To the growing culture, a solution of the hydro-
chloride of 1 (5 g) in DMSO(50 mL) was added. After
24 h, work-up (as described in the previous section) gave
5.4 g of crude product, which contained cis-(3R,4R)-2a as
the major product with only 13% of trans-(3R,4S)-3a and
no detectable trans-(3S,4R)-3b. A similar result of only
28% trans-(3R,4S)-3a and predominantly the cis-(3R,4R)-
2a was obtained in another experiment at 2 g/L substrate
input.
Microbial reductions by growing cultures of C. parapsilosis
SC16347 were carried out under similar conditions using 10
and 20 g/L substrate input. For each substrate input, an
additional experiment was also carried out by adjusting
the pH to 7 just after substrate addition.
Microbial reductions were carried out at 2 g/L substrate
input under similar conditions with the following two
microorganisms. Pichia anomala SC16139 gave 2.8 g of
crude product containing cis-(3R,4R)-2a as the major prod-
uct with 42% of trans-(3R,4S)-3a and no detectable trans-
(3S,4R)-3b. Hansenula anomala SC13830 provided 2.7 g
of crude product with cis-(3R,4R)-2a as the major product
and 25% of trans-(3R,4S)-3a with no detectable trans-
(3S,4R)-3b.
3.8. Microbial reduction on gram scale to prepare
cis-(3R,4R)-2a
One thawed frozen vial of C. parapsilosis SC16347 was
inoculated in 100 mL F7 medium (500-mL flask). The flask
was placed on a shaker at 28 ꢁC and 200 rpm. After 64 h,
the growing culture was transferred (15% transfer) into
fresh sterile F7 medium and shaken under the same condi-
tions for 24 h. The growing culture was again transferred
(15% transfer) into 1 L F7 medium (4-L flask) and shaken
under the same condition for 24 h. A solution of the hydro-
chloride of 1 (5 g) in DMSO (50 mL) was added. The flask
was shaken under the same conditions for 56 h. The pH
was adjusted to 7.5 with 1 M NaOH. The mixture was ex-
tracted with EtOAc (3 · 0.8 L). The aqueous phase was
further adjusted to pH 9 and extracted with 1 L of EtOAc.
The combined organic extract was concentrated to dryness
to give 6.8 g of crude oily product. It was subjected to flash
chromatography (160 g silica gel) and eluted with dichloro-
methane–acetone–NH4OH (94:5:1) to give 4.3 g of prod-
uct. This was again subjected to flash chromatography
(140 g silica gel) and eluted with dichloromethane–MeOH
(97:3). The product fractions were combined, stirred with
charcoal overnight at room temperature, filtered through
a pad of Celite, and concentrated to dryness to give 2.2 g
of cis-(3R,4R)-2a (50% isolated yield) with AP 93, ee
99.3%, de 99.5%, and [a]D = +48.4 (c 2.35, CHCl3). LC–
Four batches of crude products from the above experi-
ments were combined and subjected to repeated flash
chromatography and eluted with dichloromethane–
MeOH–NH4OH (96.5:3:0.5) and heptane–EtOAc (1:1) to
give 0.9 g of trans-(3R,4S)-3a with AP 98, de 96%, ee
99.5%, and [a]D = +23.6 (c 1.68, CHCl3). LC–MS 264
1
1
(M+1). H, H–1H COSY and 13C NMR spectra were re-
1
corded in CDCl3. H d 7.2–7.4 (m, 5H, Ph), 4.16 (q, 2H,
J = 7.2 Hz, CH2 in Et), 3.96 (m, 1H, 3-CH), 3.56 (d,
J = 13.2, 1H, Ph–CH2-A), 3.52 (d, J = 13.2, 1H, Ph–
CH2-B), 3.01 (m, 1H, 2-CH2-A), 2.80 (m, 1H, 6-CH2-A),
2.27 (m, 1H, 4-CH), 1.9–2.1 (m, 3H, 2-CH2-B, 5-CH2-A
and 6-CH2-B), 1.74 (m, 1H, 5-CH2-B), 1.25 (t, 3H,
J = 7.2 Hz, CH3 in Et) ppm. 13C d 174.18, 137.30, 129.12
(2C), 128.18 (2C), 127.16, 67.73, 62.39, 60.66, 58.38,
51.96, 48.62, 26.15, 14.07 ppm.
3.10. Epimerization of cis-(3R,4R)-2a
The cis-(3R,4R)-2a (10 mg) was dissolved in 2 mL of etha-
nol. K2CO3 (20 mg) was added. The mixture was stirred at
room temperature. Analytical samples were taken periodi-
cally and subjected to chiral HPLC.
1
1
MS 264 (M+1). H, H–1H COSY and 13C NMR spectra
1
were recorded in CDCl3. H d 7.2–7.4 (m, 5H, Ph), 4.19
(m, 1H, 3-CH), 4.15 (q, J = 7.2 Hz, 2H, CH2 in Et), 3.52
(s, 2H, Ph–CH2), 2.96 (m, 1H, 2-CH2-A), 2.86 (m, 1H,
6-CH2-A), 2.35 (m, 1H, 4-CH), 2.20 (m, 1H, 2-CH2-B),
After 92 h, the reaction mixture was filtered. The filtrate
was neutralized with 1 M HCl. The whole mixture was

2020
Z. Guo et al. / Tetrahedron: Asymmetry 17 (2006) 2015–2020
dried under reduced pressure. The residue was dissolved in
5 mL of anhydrous dichloromethane, and 16 mg of DCC,
1 mg of DMAP, and 0.5 mL of EtOH were added. The
mixture was stirred at room temperature for 3 h. The mix-
ture was dried.
and subjected to chiral HPLC. The results are listed in
Table 4. The trans-(3S,4R)-3b (11.4 min peak) and
cis-(3S,4S)-2b (12.7 min peak) were negligible in all cases.
Acknowledgement
An analytical sample was taken and subjected to chiral
HPLC. The results are listed in Table 3. The trans-
(3S,4R)-3b (the 11.4 min peak in chiral HPLC) and cis-
(3S,4S)-2b (12.7 min peak) were negligible in all cases.
We acknowledge the evaluation of this manuscript by
Drs. Richard Mueller and Robert Waltermire and many
valuable suggestions provided by them.
3.11. Epimerization of trans-(3R,4S)-3a
References
The trans-(3R,4S)-3a (5 mg) was dissolved in 1 mL of eth-
anol. K2CO3 (10 mg) was added. The mixture was stirred
at room temperature. Analytical samples were taken out
periodically and subjected to chiral HPLC.
1. Hanson, R. L.; Goldberg, S.; Goswami, A.; Tully, T. P.; Patel,
R. N. Adv. Synth. Catal. 2005, 347, 1073–1080.
2. Ema, T.; Yagasaki, H.; Okita, N.; Nishikawa, K.; Korenaga,
O. Tetraahedron: Asymmetry 2005, 16, 1075–1078.
3. Ernst, M.; Kaup, B.; Mueller, M.; Bringer-Meyer, S.; Sahm,
H. Appl. Microbiol. Biotechnol. 2005, 66, 629–634.
4. Patel, R. N. Food Technol. Biotechnol. 2004, 42, 305–325.
5. Stewart, D. Curr. Opin. Chem. Biol. 2001, 5, 120–129.
6. Zak, A. Curr. Opin. Chem. Biol. 2001, 5, 130–136.
7. Seebach, D.; Roggo, S.; Maetzke, T.; Braunschweiger, H.;
Cercus, J.; Krieger, M. Helv. Chim. Acta 1987, 70, 1605–
1615.
After 92 h, the reaction mixture was filtered. The filtrate
was neutralized with 1 M HCl. The whole mixture was
dried under reduced pressure. The residue was dissolved
in 3 mL of anhydrous dichloromethane, and 8 mg of
DCC, 0.5 mg of DMAP, and 0.3 mL of EtOH were added.
The mixture was stirred at room temperature for 3 h. The
mixture was dried. An analytical sample was taken out