Ex Situ Enantioconvergent Approaches for the Effective Use of Undesired Isomers: Stereochemical Convergence of a Substrate with Multiple Chiral Centers and Recycling of a Decarboxylated Byproduct

Article abstract of DOI:10.1055/s-0035-1561620

Enzyme-mediated kinetic resolution of racemic starting materials is a valuable and convenient tool for the preparation of enantioenriched compounds. To overcome the 50% yield limitation in conventional kinetic resolution, diverse enantioconvergent approaches have been developed. After a brief introduction of the recently developed 'in situ deracemization' and 'ex situ enantioconvergent approach', we present unique ex situ enantioconvergent approaches to solve two difficult cases: 1) In the synthesis of ethyl (3R,4S,5R)-shikimate, a diastereomeric (3R?,4S?,5S?)-substrate containing multiple chiral centers was applied in an enzyme-catalyzed acetylation, and both the enzyme-catalyzed product and unreacted substrate converted into ethyl (3R,4S,5R)-shikimate via partial stereochemical inversions. 2) The enzyme-catalyzed kinetic resolution of a ranirestat precursor and the regeneration of the racemic substrate from a decarboxylated byproduct are described in detail. Since in the latter study, the products spontaneously decarboxylated after hydrolysis of the ester groups, the in situ regeneration of the racemic substrates was of significant difficulty. We successfully installed an ethoxycarbonyl group on the byproduct by ex situ sequential derivatization to overcome the 50% yield limitation. 1 Short Review of Enantioconvergent Approaches 2 Resolution of a Substrate with Multiple Chiral Centers 3 Resolution Based on Enzyme-Mediated Hydrolysis Accompanied by Nonenzymatic C-C Bond Cleavage 4 Conclusions.

Full text of DOI:10.1055/s-0035-1561620

SYNTHESIS
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X
© Georg Thieme Verlag Stuttgart · New York
2016, 48, A–J
feature
A
Y. Yamashita et al.
Feature
Synthesis
Ex Situ Enantioconvergent Approaches for the Effective Use of
Undesired Isomers: Stereochemical Convergence of a Substrate
with Multiple Chiral Centers and Recycling of a Decarboxylated
Byproduct
Yasunobu Yamashita
kinetic
resolution
O
1) BOMCl
DIPEA
2) (EtOCO)2O
KHMDS
chemical
Tohru Kurihara
Takanobu Horiguchi
Atsushi Miki
A. melleus
protease
buffer
N
regeneration
of racemic
substrate
NH
EtO2C
(pH 7.0)
O
N
(R)
O
3) Pd(OH)2
H2, Et3N
25% over 3 steps
Mitsuru Shoji
conv. 50% racemate
E > 200
NH
Takeshi Sugai
EtO2C
O
Kengo Hanaya*
O
O
50%, 98.6% ee
N
H
N
(S)
Faculty of Pharmacy, Keio University, 1-5-30 Shibakoen,
Minato-ku, Tokyo 105-8512, Japan
hanaya-kn@pha.keio.ac.jp
NH
NH
HO2C
spontaneous
O
O
50%
(R)-ranirestat
decarboxylation
CO2
Received: 05.01.2016
Accepted after revision: 16.03.2016
Published online: 11.05.2016
1 Short Review of Enantioconvergent Ap-
proaches
DOI: 10.1055/s-0035-1561620; Art ID: ss-2015-z0745-fa
Enzyme-mediated kinetic resolution is a valuable and
convenient tool for the preparation of enantiomerically en-
riched compounds on both laboratory and industrial pre-
parative scales if the racemic precursors are readily avail-
able. The principle of conventional kinetic resolution is
briefly illustrated in Scheme 1 (a). In the conventional ki-
netic resolution of a racemate, the yield of the required en-
Abstract Enzyme-mediated kinetic resolution of racemic starting ma-
terials is a valuable and convenient tool for the preparation of enantio-
enriched compounds. To overcome the 50% yield limitation in conven-
tional kinetic resolution, diverse enantioconvergent approaches have
been developed. After a brief introduction of the recently developed ‘in
situ deracemization’ and ‘ex situ enantioconvergent approach’, we pres-
ent unique ex situ enantioconvergent approaches to solve two difficult
cases: 1) In the synthesis of ethyl (3R,4S,5R)-shikimate, a diastereomer-
ic (3R*,4S*,5S*)-substrate containing multiple chiral centers was ap-
plied in an enzyme-catalyzed acetylation, and both the enzyme-cata-
lyzed product and unreacted substrate converted into ethyl (3R,4S,5R)-
shikimate via partial stereochemical inversions. 2) The enzyme-cata-
lyzed kinetic resolution of a ranirestat precursor and the regeneration
of the racemic substrate from a decarboxylated byproduct are de-
scribed in detail. Since in the latter study, the products spontaneously
decarboxylated after hydrolysis of the ester groups, the in situ regenera-
tion of the racemic substrates was of significant difficulty. We success-
fully installed an ethoxycarbonyl group on the byproduct by ex situ se-
quential derivatization to overcome the 50% yield limitation.
(a)
(b)
slow
slow
A
P
A
P
enzyme-
catalyzed
reaction
+
I
Q
B
fast
B
Q
A and Q: up to 50% yield
fast
Q: up to 100% yield
I: intermediate
A, B: substrate enantiomers
P, Q: product enantiomers
1
2
3
Short Review of Enantioconvergent Approaches
Resolution of a Substrate with Multiple Chiral Centers
Resolution Based on Enzyme-Mediated Hydrolysis Accompanied
by Nonenzymatic C–C Bond Cleavage
(c)
(d)
A
A
slow
P
4
Conclusions
fast
I
slow
B
Key words enzyme catalysis, kinetic resolution, convergence, unde-
sired stereoisomers
fast
B
P: up to 100% yield
A: up to 100% yield
Scheme 1 Principles of enzyme-catalyzed kinetic resolution and dera-
cemization approaches
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

B
Y. Yamashita et al.
Feature
Synthesis
antiomer is limited to a maximum of 50%. To overcome this
inevitable constraint, in situ deracemization approaches in
which the product is converted into a single enantiomer
have been intensively studied.1–19 These approaches are
classified into three types: dynamic kinetic resolution
(Scheme 1, b), in situ stereoinversion (Scheme 1, c), and en-
antioconvergent (Scheme 1, d).20 Interestingly, it has re-
cently been reported that sulfatases and epoxide hydrolases
can be applied in an in situ enantioconvergent approach, re-
sulting in the convergence of racemic sulfate esters and ep-
oxides into the corresponding enantioenriched alcohols
and diols, respectively.20
limitation. In lipase-catalyzed kinetic resolutions starting
from both esters and alcohols, the convergence of products
into a single enantiomer of the esters or alcohols can be ac-
complished via the introduction of a leaving group and suc-
cessive stereoinversion (Scheme 2).
Herein, we present resolution, ex situ enantioconver-
gent approaches to overcome the 50% yield limitation in
two difficult cases where the above-mentioned deracem-
ization approach cannot be successfully applied. The first
case is the convergence of a racemate containing multiple
chiral centers into a single product via several partial ste-
reoinversions (Section 2). The second case involves nonen-
zymatic C–C bond cleavage in the product, via decarboxyl-
ation, immediately after the enzyme-catalyzed hydrolysis
(Section 3).
The ex situ inversion of the stereochemistry in an unde-
sired stereoisomer after the completion of a kinetic resolu-
tion is another possible method to overcome the 50% yield
Biographical Sketches
Yasunobu Yamashita (A) graduated
from the Faculty of Pharmacy, Keio Uni-
versity in 2012. He is currently a PhD
student in the Graduate School of Phar-
maceutical Sciences, Keio University and
is expected to complete his PhD in
March 2017.
From 1999 to 2001, he worked in the re-
search group of Prof. K. C. Nicolaou as a
Postdoctoral Fellow at The Scripps Re-
search Institute, USA. In 2001, he be-
came an Assistant Professor at Tokyo
University of Science, while working with
Prof. Yujiro Hayashi. He was promoted to
a Lecturer position in the laboratory of
Prof. Minoru Ueda at Tohoku University
in 2006. In 2009, he moved to Keio Uni-
versity as an Associate Professor. His re-
search interests include the total
synthesis of natural products and the in-
novation of synthetic methodologies.
michi Ohta at the Faculty of Science and
Technology, Keio University, and was
promoted to Associate Professor in the
laboratory of Prof. Shigeru Nishiyama
from 1988 to 2008. From 1991 to 1992,
he worked in the research group of Prof.
C.-H. Wong as a Research Associate at
The Scripps Research Institute, USA. In
2008, he moved to the Faculty of Phar-
macy, Keio University as a Full Professor.
His research interests include prepara-
tive bioorganic chemistry.
Tohru Kurihara (B) graduated from the
Graduate School of Pharmaceutical Sci-
ences, Keio University in 2012.
Takanobu Horiguchi (C) graduated
from the Faculty of Pharmacy, Keio Uni-
versity in 2014.
Kengo Hanaya (G) received his PhD in
2012 from the Graduate School of Phar-
maceutical Sciences, Tokyo University of
Science under the supervision of Prof.
Shin Aoki. In 2012, he became an Assis-
tant Professor at Keio University. His re-
Takeshi Sugai (F) graduated from the
Faculty of Agriculture, University of To-
kyo in 1981. He quit the PhD course of
the Graduate School of the University of
Tokyo in 1984, where he became an As-
sistant in 1984 and received his PhD de-
gree in 1987 under the supervision of
Prof. Kenji Mori. He then became an In-
structor in the laboratory of Prof. Hiro-
Atsushi Miki (D) graduated from the
Faculty of Pharmacy, Keio University in
2015 and is currently a student in the
Graduate School of Pharmaceutical Sci-
ences, Keio University.
search
interests
include
the
development of tools for bioorganic
chemistry, enzyme-mediated organic
synthesis, and the preparation of novel
enzymes.
Mitsuru Shoji (E) received his PhD in
1999 from Tohoku University under the
supervision of Prof. Masahiro Hirama.
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

C
Y. Yamashita et al.
Feature
Synthesis
(3R,4S,5R)-2a and (3R,4R,5R)-2c, respectively, thus estab-
lishing a new ex situ enantioconvergent route, as shown in
Scheme 3.21
(a) convergence into fast enantiomer
1st step
lipase-
OH
OH
X
catalyzed
acetylation
2nd
step
L
S
L
L
S
L
L
S
(slow)
NH2
OH
OH
OAc
OAc
5
NHAc
O
4
OH
OH
L
S
S
S
EtO2C
EtO2C
3
(fast)
X = leaving group
OMs, OP Ar3, etc.
X
(3R,4S,5R)-1a
oseltamivir
ethyl shikimate
3rd step
path A
3rd step
path B
L
L
S
OH
OAc
OAc
L
S
L
S
OH
O
AcO as
nucleophile
OH
hydrolysis
1) Tf2O, DMAP
pyridine
S
O
O
2) KNO2
(b) convergence into slow enantiomer
O
EtO2C
EtO2C
OAc
1st step
OAc
OAc
(3R,4S,5S)-2a
(slow enantiomer)
(3R,4S,5R)-2a
lipase-
catalyzed
hydrolysis
2nd
step
L
S
(slow)
L
S
L
L
S
Dowex 50W-X8
(H+ form), EtOH
OAc
OH
X
OH
L
S
L
S
S
OH
OH
B. cepacia
lipase
(fast)
OAc
(
)-2a
EtO2C
vinyl acetate
E > 200
(3R,4S,5R)-1a
ethyl shikimate
3rd step
path B
3rd step
path A
L
L
S
OH
OAc
X
L
S
L
S
hydrolysis
AcO as
nucleophile
1) Tf2O, DMAP
pyridine
2) KNO2
S
3) H2SO4
EtOH
Scheme 2 Convergence of racemic secondary alcohols (a) and their
esters (b) into a single enantiomer in three steps involving lipase-cata-
lyzed kinetic resolution (1st step), introduction of a leaving group (2nd
step), and stereoinversion/hydrolysis (3rd step); L: large substituents, S:
small substituents.
OAc
O
OAc
1) FeCl3·6H2O
2) SOCl2, Et3N
OH
3) CsOAc
O
EtO2C
EtO2C
OAc
(3S,4S,5R)-2b
(from fast enantiomer)
(3R,4R,5R)-2c
2 Resolution of a Substrate with Multiple
Chiral Centers
Scheme 3 Convergence of (3R,4S,5S)-2a (slow enantiomer) and
(3S,4S,5R)-2b (from fast enantiomer) into ethyl shikimate [(3R,4S,5R)-
1a]
Ethyl shikimate [(3R,4S,5R)-1a], an important interme-
diate for the synthesis of oseltamivir phosphate, has three
chiral centers. Therefore, the above-mentioned deracem-
ization approaches including a simple stereoinversion at a
specific secondary alcohol afford only the diastereomer of
the desired stereoisomer, not the single enantiomer. We de-
signed a secondary alcohol (3R*,4S*,5S*)-(±)-2a, having a
diastereomeric relationship with ethyl shikimate, as a sub-
strate for lipase-mediated kinetic resolution.21 Alcohol
(3R*,4S*,5S*)-(±)-2a was prepared on a large scale by the
Diels–Alder reaction of furan with acryloyl chloride and the
subsequent introduction of oxygen functional groups.
Among the three commercially available lipases [Burk-
holderia cepacia lipase (Amano, PS-IM), Candida antarctica
lipase B (Novozym 435), and Candida rugosa lipase (Meito
OF)], the use of B. cepacia lipase resulted in an excellent E
value (>200) and robust catalytic activity under the reac-
tion conditions. Both products obtained from the B. cepacia
lipase-catalyzed resolution, alcohol (3R,4S,5S)-2a and ace-
tate (3S,4S,5R)-2b, were merged into (3R,4S,5R)-1a via
3 Resolution Based on Enzyme-Mediated
Hydrolysis Accompanied by Nonenzymatic
C–C Bond Cleavage
In the enzyme-mediated hydrolysis of esters bearing
adjacent electron-withdrawing groups, such as in β-oxo es-
ters, subsequent decarboxylation can occur during the
progress of the enzyme-mediated hydrolysis. The sequen-
tial reaction can be advantageous for the desymmetrization
approach from 3a to 422 (Scheme 4).
On the other hand, in kinetic resolution, in situ decar-
boxylation is disadvantageous for maximum yield; the de-
carboxylated products are difficult to reconvert in situ into
the racemic substrates by the above-mentioned deracem-
ization approaches. In the typical examples shown in
Scheme 5,α-nitro ester 5a23 and oxo ester 6a24 were hydro-
lyzed and subsequently decarboxylated, producing nitroal-
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

D
Y. Yamashita et al.
Feature
Synthesis
For the effective use of a waste decarboxylated product,
the regeneration of the racemic substrate by introducing
the lost functional groups after the kinetic resolution is a
powerful approach. In the preparation of the synthetic pre-
cursor of ranirestat (AS-3201, Scheme 6), enzymes were
screened, and the substrate structure was optimized in the
kinetic resolution. Finally, the undesired products in the ki-
netic resolution were recycled. These results are now de-
scribed in detail.
pig liver
esterase
(PLE)
N
N
Bn
Bn
n-BuO2C
CO2n-Bu
HO2C
CO2n-Bu
O
O
3a
3b
N
Bn
CO2n-Bu
CO2
O
4
Ranirestat has been developed as a drug for the treat-
ment of diabetic neuropathy,25 a complication of diabetes,
by Sumitomo Dainippon Pharma Co., Ltd. It is a highly po-
tent aldose reductase inhibitor that suppresses the accu-
mulation of sorbitol in neural tissues.25a,b Ranirestat has a
unique stereocenter on its succinimide ring; the nitrogen
atom of a pyrrole ring is directly attached to the tetrasubsti-
tuted stereogenic center. The (R)-isomer of ranirestat shows
10 times more potent inhibitory activity against aldose re-
ductase in vitro and 500 times more potent in vivo inhibito-
ry activity than the (S)-isomer.25a Thus, enantiomerically
enriched 2,2-disubstituted succinimide (R)-9a (Scheme 6)
is crucial for the synthesis of (R)-ranirestat.25a
Scheme 4 Enzyme-catalyzed hydrolysis of β-oxo ester 3a with sponta-
neous decarboxylation22a
kane (±)-7 and ketone (±)-8, respectively. Because only slow
enantiomers are available for the subsequent synthesis, the
fast enantiomer-derived products were generally discard-
ed.
(a)
Ar
Ar
O
O
MeO2C
N
O
MeO2C
N
O
(S)-5a
(S)-5a
α-chymotrypsin
(slow enantiomer)
O
NH
Ar
MeO2C
N
Ar
O
O
F
O
O
N
O
CbzHN
EtO2C
O
HO2C
N
N
N
O
NH
O
O
(R)-5a
NH
(R)-5b
EtO2C
(fast enantiomer)
O
O
(R)-10a
(R)-9a
CO2
Br
(R)-ranirestat
Ar =
Ar
N
H
O
H
N
Scheme 6 (R)-Ranirestat and its synthetic precursors
O
)-7
(
(b)
To access (R)-9a with a high enantiomeric excess, Ne-
goro and co-workers prepared enantioenriched (R)-10a, the
precursor of (R)-9a, by recrystallization of the cinchonidine
salt.25a Owing to the high therapeutic potency of ranirestat,
more practical synthetic methods were needed, and alter-
native asymmetric syntheses of (R)-9a have been devel-
oped. Shibasaki and co-workers reported the asymmetric
amination of a succinimide derivative with a lanthanum-
based ternary catalyst.26 Seki and Kawase reported the
asymmetric alkylation of an α-cyano ester using a phase-
transfer catalyst.27 On the other hand, Kudo and Yamada ac-
complished the kinetic resolution of 10a by pig liver ester-
ase (PLE)-mediated hydrolysis (Scheme 7), in which (S)-10a
(the fast enantiomer) was hydrolyzed more rapidly than
(R)-10a (the slow enantiomer), and a moderate E value of
16.6 was obtained.28 Although (R)-10a was obtained in
98.8% ee, no data exist for (S)-10b. This compound may
have spontaneously decarboxylated to afford (±)-11 under
the reaction conditions.
NC(H2C)2
EtO2C
NC(H2C)2
EtO2C
O
O
(R)-6a
(R)-6a
PLE
(slow enantiomer)
NC(H2C)2
NC(H2C)2
HO2C
EtO2C
O
O
(S)-6b
(S)-6a
(fast enantiomer)
CO2
NC(H2C)2
H
O
(
)-8
Scheme 5 Kinetic resolution by enzyme-catalyzed hydrolysis of α-nitro
ester 5a23 (a) and oxo ester 6a24 (b) accompanied by the decarboxyl-
ation of the products derived from the fast enantiomers
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

E
Y. Yamashita et al.
Feature
Synthesis
providing nitrile 13c. Palladium-catalyzed hydration of the
cyano group with acetamide29 afforded amide 13d. Using
our procedure, amide 13d can be prepared in two steps, one
step shorter than the method used by Negoro and co-work-
ers.25a Finally, the subsequent base-mediated intramolecu-
lar cyclization of 13d afforded succinimide (±)-9a. Com-
pared to previous reports on the syntheses of kinetic reso-
lution substrates, succinimide (±)-9a can be prepared from
commercially available diethyl aminomalonate (13a) in
fewer steps; the total yield was >60% over four steps. More-
over, all the reactions are easy to manipulate, indicating
suitability for large-scale synthesis.
O
O
NH
CbzHN
EtO2C
CbzHN
NH
EtO2C
O
O
(R)-10a
(R)-10a
PLE
98.8% ee
(slow)
64% conv.
E = 16.6
O
O
NH
CbzHN
EtO2C
CbzHN
HO2C
NH
O
O
(S)-10a
(S)-10b
(fast)
CO2
O
O
MeO
OMe
NH3
CbzHN
N
Cl
NaOAc
AcOH
77%
NH
EtO2C CO2Et
EtO2C CO2Et
H
O
13a
13b
(
)-11
PdCl2
AcNH2
BrCH2CN
NaH
Scheme 7 Kudo and Yamada’s kinetic resolution of 10a
N
CN
THF, H2O
88%
DMF
92%
EtO2C CO2Et
Because the pyrrole ring in 9a is smaller than the Cbz
group in 10a,28 reducing the steric repulsion between 9a
and enzymes, an improvement in the reactivity and enanti-
oselectivity of the kinetic resolution of 9a was expected.
First, the synthesis of racemic substrate (±)-9a and the ki-
netic resolution of 9a with several enzymes were investi-
gated. Second, the effects of the size of the ester moiety on
the enantioselectivity and reaction rate were studied. Final-
ly, the chemical regeneration of (±)-9a from the undesired
(S)-isomer by ex situ convergence via the ethoxycarbonyla-
tion of (±)-12a was investigated (Scheme 8).
13c
O
K2CO3
N
N
EtO2C
CONH2
acetone
96%
NH
EtO2C CO2Et
13d
O
(
)-9a
Scheme 9 Preparation of racemic substrate 9a
The enzyme-mediated kinetic resolutions of (±)-9a
were conducted under hydrolytic conditions (Table 1). In all
cases, (S)-9a was hydrolyzed preferentially, as in the case of
(S)-10a, affording 12a via the spontaneous decarboxylation
of (S)-9b. The stereochemistry of the slow enantiomer, (R)-
9a, was confirmed by comparing optical rotation values
[(R)-9a: [α]D25 –56.4 (Lit.25a [α]D26 –59.5)]. When (±)-9a was
treated with PLE, the E value (11.3) was slightly lower than
that of 10a, contrary to our expectations. The hydrolysis of
9a proceeded faster than that of 10a, indicating that the re-
activity of the (R)-isomer increased by more than that of
the (S)-isomer upon replacing the NHCbz group by the pyr-
role ring. Thus, three other enzymes, C. antarctica lipase B
(Novozym 435), B. cepacia lipase (Amano, PS-IM), and As-
pergillus melleus protease (Nagase, XP-488), were applied.
The two lipases are solid-supported enzymes that are high-
ly stable in aqueous and organic solvents and can be easily
removed from reaction mixtures by simple filtration. A.
melleus protease has been developed for the modification of
food properties by Nagase ChemteX Corporation (Japan)
and applied in the kinetic resolution of ethyl tetrahydrofur-
ancarboxylate and dihydropyridine derivatives,30 which are
structurally similar to 9a. In contrast to the sluggish hydro-
O
chemical
N
regeneration
of racemic
substrate
enzyme-mediated
kinetic resolution
NH
EtO2C
O
O
(
)-9a
N
NH
EtO2C
O
O
N
H
(R)-9a
NH
O
N
O
NH
(
)-12a
HO2C
(R)-ranirestat
O
spontaneous
decarboxylation
(S)-9b
CO2
Scheme 8 Our kinetic resolution and reuse of the undesired isomer
(ex situ convergence)
The racemic substrate 9a used for the kinetic resolution
was prepared as shown in Scheme 9. Malonate 13b, pre-
pared by the Clauson-Kaas reaction of diethyl amino-
malonate (13a),25a was alkylated with bromoacetonitrile,
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

F
Y. Yamashita et al.
Feature
Synthesis
lysis of 9a with the two lipases, A. melleus protease showed
fast reaction rate and excellent enantioselectivity
Table 2 Effect of the Structure of the Ester Moiety on Kinetic Resolu-
tion
a
(E > 200).
O
O
N
N
Table 1 Kinetic Resolution of (±)-9a by Enzyme-Mediated Hydrolysis
NH
NH
RO2C
RO2C
O
O
O
O
N
N
A. melleus
protease
(R)
(R)
NH
NH
EtO2C
EtO2C
O
O
O
O
N
H
N
(R)-9a
(R)-9a
enzymes
(slow)
NH
NH
RO2C
O
O
O
O
N
N
(
)-12a
(S)
NH
NH
EtO2C
H
(
Substrate
R
Time (h)
Conv. (%)
E
O
O
(S)-9a
)-12a
9a
9c
9d
Et
1.5
2.0
2.0
50.0
47.8
29.8
>200
>200
5.4
(fast)
i-Bu
i-Pr
Enzyme
Time (h)
1.0
Conv. (%)
62.6
E
Pig liver esterase
(PLE, Sigma)
11.3
NDa
50% yield limitation, the regeneration of racemic substrate
(±)-9a from (±)-12a was investigated. Our regeneration
strategy for (±)-9a was as follows: protection of the imide
nitrogen in (±)-12a; subsequent ethoxycarbonylation at the
C-3 position and deprotection to afford (±)-9a.
C. antarctica lipase B
(Novozym 435)
24
24
1.5
0
B. cepacia lipase
(Amano, PS-IM)
0
NDa
A. melleus protease
(Nagase, XP-488)
50.0
>200
Because the acidic proton of the imide nitrogen would
interfere with deprotonation of the C-3 position in the
ethoxycarbonylation step, a Boc group was introduced on
the imide nitrogen, affording 12b (Scheme 10, a). The imide
ring in 12b was expected to be unstable owing to the elec-
tron-withdrawing properties of the Boc group; thus, 12b
was used in the subsequent ethoxycarbonylation step with-
out further purification. The reaction conditions were ad-
opted from a procedure developed by Crider and co-work-
ers,32 in which a similar compound bearing a cyclic imide
group and an aromatic ring was used. Unfortunately, this
resulted in decomposition of the imide ring by the nucleo-
philic attack of ethoxide anion. This result indicated that
electron-withdrawing groups such as carbamate-type pro-
tecting groups are not suitable as protecting groups for the
imide nitrogen, and that bulky bases should be used for the
ethoxycarbonylation.
Next, various protecting groups, such as a tert-butyl, p-
methoxyphenyl, p-methoxybenzyl, or allyl group, were in-
troduced on the nitrogen atom of the imide ring. Disap-
pointingly, the removal of these groups was difficult owing
to the high sensitivity of the pyrrole ring under various re-
action conditions, including acidic, oxidative, and transi-
tion-metal-mediated conditions. These findings prompted
us to use the benzyloxymethyl (BOM) group, which can be
removed by hydrogenolysis; BOM imide 12c was prepared
from 12a in a high yield (Scheme 10, b). When weakly nuc-
leophilic bases such as NaH and KH were used with 12c,
a ND: not determined.
To further establish the relationship between the reac-
tivity, selectivity, and substrate structures in A. melleus pro-
tease-mediated kinetic resolution, other substrates with a
more hindered ester group, such as isobutyl and isopropyl,
were used (Table 2). It was expected that the reaction rate
of the slow enantiomers would decrease owing to the steric
hindrance of their alkoxy moieties, thus improving the en-
antioselectivity. Substrates 9c and 9d were prepared by the
Ti(Oi-Pr)4-catalyzed transesterification of 9a with the cor-
responding alcohols.31 In the case of isobutyl ester 9c, the
enantioselectivity seemed to be slightly lower than that of
ethyl ester 9a based on a simple calculation of the data. The
bulkier isopropyl ester group in 9d significantly decreased
both the enantioselectivity (E = 5.4) and reactivity. The fast
enantiomer was more affected by the steric hindrance of
the ester moiety than the slow enantiomer. These results
proved that (±)-9a is a better substrate for A. melleus prote-
ase-mediated kinetic resolution in terms of both synthetic
accessibility and enantioselectivity.
As noted above, the kinetic resolution of (±)-9a (E > 200)
was successfully carried out using A. melleus protease, af-
fording (R)-9a in 50% yield and with 98.6% ee. While (R)-9a
was used for the synthesis of (R)-ranirestat, undesired (±)-
12a would normally be discarded. Therefore, for the effec-
tive use of undesired product (±)-12a and to overcome the
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

G
Y. Yamashita et al.
Feature
Synthesis
Table 3 Hydrogenolysis of the BOM Group in 14b
(a)
(Boc)2O
DIPEA
O
O
N
H
N
H
O
O
O
N
N
NH
N
N
conditions
THF
+
Boc
N
NH
N
O
O
EtO2C
EtO2C
EtO2C
BOM
BOM
(
)-12a
12b
O
O
O
14b
(
)-9a
16
(3:1 mixture with 15)
ClCO2Et
NaOEt
O
N
Entry Catalyst (wt%)
Et3N (mol%) Ratio 9a/16a Yield (%) of 9a
N
toluene
EtO2C
Boc
O
1
2
3
4
5
6
7
10% Pd/C (50)
–
1:2
1:5
1:2
1:2
0:1
NDb
NDb
NDb
NDb
NDb
NDb
NDb
56
(
)-14a
20% Pd(OH)2/C (50)
ASCA-2®c (200)
–
–
(b)
10% Pd/HP20d (80)
8% Pd/CR11e (80)
10% Pd/C (116)
–
BOMCl
DIPEA
O
O
–
N
H
N
80
60
NH
N
CH2Cl2
96%
H
BOM
20% Pd(OH)2/C (213)
63
O
O
a Determined from the 1H NMR spectrum of the crude mixture.
b ND: not determined.
(
)-12a
12c
c N.E. CHEMCAT Ltd.
EtO2C
N
(EtOCO)2O
KHMDS
d Pd on Diaion® HP20 polymer (Mitsubishi Chemical Co. Ltd., on polysty-
rene–polyvinylbenzene-based synthetic adsorbent).
e Pd on Diaion® CR11 polymer (Mitsubishi Chemical Co. Ltd., on iminodiace-
tate-based chelating resin).
O
O
N
+
N
N
toluene
59%
EtO2C
BOM
BOM
O
O
3:1
(
)-14b
( )-15
4 Conclusions
Scheme 10 (a) Attempted ethoxycarbonylation of N-Boc-succinimide
12a, and (b) ethoxycarbonylation of imide 12c bearing a BOM group
In this study, we have attempted to solve difficult cases
of kinetic resolutions where in situ deracemization ap-
proaches cannot be applied to overcome the 50% yield lim-
itation. These examples prove that not only in situ deracem-
ization, but also ex situ synthetic derivatization after kinetic
resolution, namely a ex situ enantioconvergent approach,
can play an important role in the efficient chemoenzymatic
transformation of organic compounds. In Section 2 we de-
scribed the enzyme-catalyzed kinetic resolution of a ‘dia-
stereomeric substrate’ bearing multiple chiral centers and
chemical convergence to the desired stereoisomer. The de-
sign of a racemic substrate having a diastereomeric rela-
tionship with ethyl shikimate was the key for the success of
the enantioconvergent approach. In Section 3 we described
the regeneration of the racemic substrate from a decarbox-
ylated product in an enzyme-catalyzed hydrolytic kinetic
resolution. Although the total yield of this regeneration se-
quence was as low as 25%, for the first time the lost func-
tional group was reintroduced into the undesired product
derived from the fast enantiomer, which would normally be
discarded.
ethoxycarbonylated compound was obtained in negligible
amounts. Finally, the reaction was accomplished using po-
tassium hexamethyldisilazide and diethyl dicarbonate, af-
fording 14b as a 3:1 inseparable mixture with regioisomer
15 in a moderate yield.
For the removal of the BOM group, initially, hydrogeno-
lysis conditions were applied by using a palladium catalyst
adsorbed on carbon (Table 3, entries 1 and 2) or synthetic
polymers33 (entries 3 and 4). These attempts, however, re-
sulted in reduction of the pyrrole ring, producing the unde-
sired pyrrolidine 16 as the major product. To suppress the
reduction of the pyrrole ring, palladium on Diaion® CR11
polymer34 bearing an iminodiacetate-chelating group was
used; however, imide 9a was not obtained at all (Table 3,
entry 5). The addition of Et3N35 suppressed the reduction of
the pyrrole ring, affording (±)-9a (Table 3, entry 6). After
further investigation of the reaction conditions, a combina-
tion of Pd(OH)2 and Et3N was found to be a superior catalyst
for the selective hydrogenolytic cleavage of the BOM group
(Table 3, entry 7). The inseparable impurity (±)-15 was also
reduced under these conditions, and the reduced com-
pound could be separated in this step. Using this method,
the ex situ convergence of (±)-12a into (±)-9a was accom-
plished. The overall yield was 25% over three steps, and the
cycle for the effective use of the undesired isomer was com-
pleted.
We hope that our findings will inspire chemists and
more unique chemoenzymatic transformations are devel-
oped, thus not waiting for the development of better en-
zymes by enzymologists.
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

H
Y. Yamashita et al.
Feature
Synthesis
Melting points were recorded on a Yanaco MP-J3 micro melting point
apparatus and are uncorrected. IR spectra were measured as ATR on a
Jeol FT-IR SPX60 spectrometer. 1H NMR spectra were measured at 400
MHz on a Varian 400-MR or at 500 MHz on an Agilent Inova-500
spectrometer, and 13C NMR spectra were measured at 100 MHz on a
Varian 400-MR or at 125 MHz on an Agilent Inova-500 spectrometer.
High-resolution mass spectra were recorded on a Jeol JMS-T100LP Ac-
cuTOF spectrometer. HPLC data were recorded using Jasco MD-2010
multichannel detectors and a Shimadzu SPD-20A diode array detec-
tor. Optical rotation values were recorded on a Jasco P-1010 polarime-
ter. Merck silica gel 60 F254 thin-layer plates (1.05744, 0.5-mm thick-
ness) were used for preparative TLC. Silica gel 60 (spherical and neu-
tral; 100–210 μm, 37560-79) from Kanto Chemical Co. and SNAP
silica gel cartridges and an Isolera One flash purification system from
Biotage (Sweden) were used for column chromatography. Aspergillus
melleus protease (XP-488), Candida antarctica lipase B (Novozym
435), and Burkholderia cepacia lipase (PS-IM) were gifts from Nagase
& Co., Novozymes Japan, and Amano Enzyme Inc., respectively.
layer was dried over Na2SO4 and concentrated in vacuo. The residue
was purified by silica gel column chromatography (2.5 g; hexane/EtOAc,
5:1) to afford 9a (400 mg, 96%) as a colorless oil.
(±)-9a: HPLC [Daicel Chiralpak AY-H, 0.46 cm × 25 cm; hexane/i-PrOH
(5:1), 0.5 mL/min; detected at 206 nm]: tR (min) = 21.1 [(S)-9a], 25.0
[(R)-9a].
1H NMR (400 MHz, CDCl3): δ = 1.26 (t, J = 7.2 Hz, 3 H), 3.35 (d, J = 17.9
Hz, 1 H), 3.61 (d, J = 17.9 Hz, 1 H), 4.28 (q, J = 7.2 Hz, 2 H), 6.27 (dd, J =
2.1, 2.2 Hz, 2 H), 6.93 (dd, J = 2.1, 2.2 Hz, 2 H), 8.10 (br s, 1 H); the 1H
NMR spectrum was identical with that reported previously.25a
Isobutyl 2,5-Dioxo-3-(1H-pyrrol-1-yl)pyrrolidine-3-carboxylate
(9c)
To a stirred solution of ethyl ester 9a (200 mg, 0.85 mmol) in i-BuOH
(2.0 mL) was added Ti(Oi-Pr)4 (110 μL, 0.37 mmol), and the mixture
was refluxed for 2.5 h. After cooling, the mixture was filtered through
a short column of Celite® and the column was washed with EtOAc.
The combined filtrate and washings was concentrated in vacuo. The
residue was dissolved in EtOAc, washed with brine, dried over Na2SO4,
and concentrated in vacuo. The residue was purified by silica gel col-
umn chromatography (1.5 g; hexane/EtOAc, 10:1) to afford 9c (164
mg, 73%) as a colorless solid; mp 69.0–70.0 °C.
Diethyl 2-(Cyanomethyl)-2-(1H-pyrrol-1-yl)malonate (13c)
To a stirred suspension of NaH (252 mg, 10.5 mmol) in anhydrous
DMF (7.0 mL) was added ester 13b25a (2.00 g, 8.88 mmol) at 0 °C, and
the mixture was stirred for 1 h at r.t. Then, to the mixture was added
dropwise bromoacetonitrile (700 μL, 10.0 mmol) at 0 °C, and the mix-
ture was further stirred at r.t. for 4 h. The reaction was quenched with
saturated aq NH4Cl solution, and the organic materials were extracted
with EtOAc (3 ×). The combined organic layer was washed with H2O
and brine, dried over Na2SO4, and concentrated in vacuo. The residue
was purified by silica gel column chromatography (100 g; hex-
ane/EtOAc, 5:1) to afford 13c (2.15 g, 92%) as a colorless oil.
(±)-9c: HPLC [under the same conditions as for (±)-9a]: tR (min) = 19.5
[(S)-9c], 25.7 [(R)-9c].
IR (ATR): 3209, 3082, 2962, 2873, 2779, 1716 cm–1
.
1H NMR (500 MHz, CDCl3): δ = 0.85 (d, J = 6.7 Hz, 6 H), 1.90 (ddqq, J =
6.4, 6.8, 6.7, 6.7 Hz, 1 H), 3.37 (d, J = 18.0 Hz, 1 H), 3.62 (d, J = 18.0 Hz,
1 H), 3.97 (dd, J = 6.8, 10.8 Hz, 1 H), 4.00 (dd, J = 6.4, 10.8 Hz, 1 H), 6.28
(dd, J = 2.0, 2.5 Hz, 2 H), 6.95 (dd, J = 2.0, 2.5 Hz, 2 H), 8.49 (br s, 1 H).
13C NMR (100 MHz, CDCl3): δ = 18.6, 27.5, 41.9, 68.6, 73.4, 110.0,
120.1, 167.0, 170.6, 172.8.
IR (ATR): 2985, 1741, 1250, 721 cm–1
.
1H NMR (400 MHz, CDCl3): δ = 1.31 (t, J = 7.1 Hz, 6 H), 3.35 (s, 2 H),
4.35 (q, J = 7.1 Hz, 2 H), 4.36 (q, J = 7.1 Hz, 2 H), 6.25 (dd, J = 2.1, 2.3 Hz,
2 H), 6.81 (dd, J = 2.1, 2.3 Hz, 2 H).
Anal. Calcd for C13H16N2O4: C, 59.08; H, 6.10; N, 10.60. Found: C,
59.18; H, 6.14; N, 10.48.
13C NMR (125 MHz, CDCl3): δ = 13.8, 26.2, 63.8, 68.9, 110.0, 115.0,
120.0, 165.2.
Isopropyl 2,5-Dioxo-3-(1H-pyrrol-1-yl)pyrrolidine-3-carboxylate
Anal. Calcd for C13H16N2O4: C, 59.08; H, 6.10; N, 10.60. Found: C,
59.08; H, 6.11; N, 10.59.
(9d)
In the same manner as described for 9c, 9a (300 mg, 1.28 mmol) was
treated with Ti(Oi-Pr)4 (140 μL, 0.47 mmol) in i-PrOH (3.0 mL) at re-
flux for 2.5 h. After a similar workup, the residue was purified by sili-
ca gel column chromatography (2 g; hexane/EtOAc, 10:1) to afford 9d
(256 mg, 80%) as a colorless solid; mp 94.0–94.5 °C.
Diethyl 2-(Carbamoylmethyl)-2-(1H-pyrrol-1-yl)malonate (13d)
A solution of 13c (500 mg, 1.89 mmol), PdCl2 (42 mg, 0.24 mmol), and
acetamide (1.12 g, 19.0 mmol) in a mixture of THF and H2O (1:1, 15
mL) was stirred at r.t. for 24 h. The mixture was filtered through a
short column of Celite® and the column was washed with EtOAc. The
combined filtrate and washings was washed with H2O and brine,
dried over Na2SO4, and concentrated in vacuo. The residue was puri-
fied by silica gel column chromatography (4 g; hexane/EtOAc, 5:1) to
afford 13d (468 mg, 88%) as a colorless solid; mp 94.0–95.0 °C.
1H NMR (400 MHz, CDCl3): δ = 1.27 (t, J = 7.2 Hz, 6 H), 3.33 (s, 2 H),
4.29 (q, J = 7.2 Hz, 2 H), 4.30 (q, J = 7.2 Hz, 2 H), 5.34 (br s, 1 H), 5.56
(br s, 1 H), 6.17 (dd, J = 2.2, 2.4 Hz, 2 H), 6.84 (dd, J = 2.2, 2.4 Hz, 2 H);
the 1H NMR spectrum was identical with that reported previously.25a
(±)-9d: HPLC [under the same conditions as for (±)-9a]: tR (min) =
15.5 [(S)-9d], 18.6 [(R)-9d].
IR (ATR): 3240, 3101, 2987, 1708 cm–1
1H NMR (400 MHz, CDCl3): δ = 1.22 (d, J = 6.2 Hz, 3 H), 1.25 (d, J = 6.2
Hz, 3 H), 3.34 (d, J = 18.0 Hz, 1 H), 3.55 (d, J = 18.0 Hz, 1 H), 5.08 (qq,
J = 6.2, 6.2 Hz, 1 H), 6.25 (dd, J = 2.0, 2.2 Hz, 2 H), 6.92 (dd, J = 2.0, 2.2
Hz, 2 H), 8.73 (br s, 1 H).
.
13C NMR (100 MHz, CDCl3): δ = 21.3, 21.4, 41.9, 68.7, 72.3, 110.1,
120.0, 166.2, 170.4, 172.5.
Anal. Calcd for C12H14N2O4: C, 57.59; H, 5.64; N, 11.19. Found: C,
57.70; H, 5.66; N, 11.04.
Ethyl 2,5-Dioxo-3-(1H-pyrrol-1-yl)pyrrolidine-3-carboxylate (9a)
To a stirred solution of 13d (500 mg, 1.77 mmol) in acetone (7.5 mL)
was added K2CO3 (25 mg, 0.18 mmol), and the mixture was refluxed
for 24 h. After cooling, the mixture was filtered and the filtrate was
concentrated in vacuo. The residue was dissolved in EtOAc and
washed with saturated aq NH4Cl solution, H2O, and brine. The organic
PLE-Catalyzed Hydrolysis of (±)-9a
To a solution of (±)-9a (20.9 mg, 88.5 μmol) in a mixture of phosphate
buffer (0.1 M, pH 7.0; 200 μL) and acetone (50 μL), PLE (Sigma, E2884;
50 μL) was added, and the mixture was stirred for 1 h at 30 °C. The
progress of the reaction was monitored by TLC analysis [silica gel,
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

I
Y. Yamashita et al.
Feature
Synthesis
hexane/EtOAc (1:1), by the detection of the deethoxycarbonylated
product 12a (Rf = 0.38)]. The reaction was quenched with 2 M HCl to
pH 2, and the organic materials were extracted with EtOAc (3 ×). The
combined extract was dried over Na2SO4 and concentrated in vacuo.
The residue (16.7 mg) was purified by preparative TLC (hexane/EtOAc,
1:1) to afford (R)-9a (7.8 mg, 33 μmol, 37.3%) and 12a (9.1 mg, 55
μmol, 62.6%).
A. melleus Protease-Catalyzed Hydrolysis of (±)-9d
In a similar manner, by the treatment of (±)-9d (19.8 mg, 79.1 μmol)
with A. melleus protease for 2 h, (R)-9d (13.9 mg, 58.8 μmol, 55.5%)
was obtained.
(R)-9d: [α]D26 –16.2 (c 0.53, MeOH).
HPLC: tR (min) = 16.3 [(S)-9d, 37.0%], 19.6 [(R)-9d, 63.0%]; 26.0% ee.
(R)-9a: [α]D25 –56.4 (c 0.32, MeOH) [Lit.25a [α]D26 –59.5 (c 1.0, MeOH),
for (R)-9a].
1-(tert-Butoxycarbonyl)-3-(1H-pyrrol-1-yl)pyrrolidine-2,5-dione
(12b)
HPLC: tR (min) = 22.5 [(S)-9a, 3.3%], 26.1 [(R)-9a, 96.7%]; 93.4% ee.
The 1H NMR spectrum of (R)-9a was identical with that of (±)-9a.
A solution of 12a (12 mg, 0.074 mmol), Boc2O (40 μL, 0.154 mmol),
and DIPEA (40 μL, 0.23 mmol) in THF (100 μL) was stirred for 20 h at
40 °C. The reaction was quenched with H2O, and the organic materials
were extracted with EtOAc (3 ×). The combined extract was washed
with H2O and brine, dried over Na2SO4, and concentrated in vacuo to
give 12b.
1H NMR (400 MHz, CDCl3): δ = 1.56 (s, 9 H), 3.03 (dd, J = 6.4, 18.5 Hz, 1
H), 3.35 (dd, J = 9.5, 18.5 Hz, 1 H), 5.06 (dd, J = 6.4, 9.5 Hz, 1 H), 6.24
(dd, J = 2.0, 2.2 Hz, 2 H), 6.70 (dd, J = 2.0, 2.2 Hz, 2 H).
Identification of 12a
An authentic sample of 12a was prepared in the following manner:
Ester 9a (289 mg, 1.23 mmol) was dissolved in THF (2 mL). After cool-
ing to 0 °C, aq NaOH solution (2.0 M; 700 μL, 1.40 mmol) was added,
and the mixture was stirred for 27 h at 0 °C. The reaction was
quenched with saturated aq NH4Cl solution, and the organic materials
were extracted with EtOAc (3 ×). The combined extract was washed
with H2O and brine, dried over Na2SO4, and concentrated in vacuo.
The residue was purified by silica gel column chromatography (15 g;
hexane/EtOAc, 3:1) to afford 12a (129 mg, 64%).
13C NMR (100 MHz, CDCl3): δ = 27.6, 36.8, 56.7, 87.1, 110.3, 119.6,
145.7, 146.7, 169.1, 169.6.
1-(Benzyloxymethyl)-3-(1H-pyrrol-1-yl)pyrrolidine-2,5-dione
(12c)
IR (ATR): 3240, 3095, 1785, 1702, 1172 cm–1
.
1H NMR (500 MHz, CDCl3): δ = 3.00 (dd, J = 5.7, 18.6 Hz, 1 H), 3.34 (dd,
J = 9.6, 18.6 Hz, 1 H), 5.08 (dd, J = 5.7, 9.6 Hz, 1 H), 6.25 (dd, J = 2.1, 2.2
Hz, 2 H), 6.92 (dd, J = 2.1, 2.2 Hz, 2 H), 8.19 (br s, 1 H); the 1H NMR
spectrum of the sample derived from the PLE-catalyzed hydrolysis
was identical with that of this authentic sample.
To a solution of 12a (602 mg, 3.67 mmol) and DIPEA (959 μL, 5.51
mmol) in CH2Cl2 (15 mL) was added dropwise benzyl chloromethyl
ether (554 μL, 4.03 mmol) over 15 min at 0 °C, and the mixture was
stirred for 30 min at 0 °C. Then, the mixture was washed with aq
NH4Cl solution, H2O, and brine (20 mL each). The organic layer was
dried over Na2SO4 and concentrated in vacuo. The residue was puri-
fied by silica gel column chromatography (SNAP Ultra, 25 g; hexane/
EtOAc, 9:1 to 0:1) to afford 12c (1000 mg, 96%) as a colorless solid.
13C NMR (125 MHz, CDCl3): δ = 37.8, 57.7, 110.3, 119.5, 173.1, 173.6.
Anal. Calcd for C8H8N2O2: C, 58.53; H, 4.91; N, 17.06. Found: C, 58.65;
H, 4.99; N, 16.85.
IR (ATR): 3128, 3008, 2945, 2926, 2879, 1794, 1757, 1716, 1387, 1365,
1323, 1194, 1072 cm–1
.
Comparison with Other Hydrolytic Enzymes
1H NMR (500 MHz, CDCl3): δ = 2.86 (dd, J = 18.5, 5.8 Hz, 1 H), 3.17 (dd,
J = 18.5, 9.5 Hz, 1 H), 4.64 (d, J = 12.8 Hz, 1 H), 4.68 (d, J = 12.8 Hz, 1 H),
4.81 (dd, J = 9.5, 5.8 Hz, 1 H), 5.11 (s, 2 H), 6.23 (dd, J = 2.1, 2.2 Hz, 2
H), 6.60 (dd, J = 2.1, 2.2 Hz, 2 H), 7.29–7.38 (m, 5 H).
13C NMR (125 MHz, CDCl3): δ = 36.6, 56.5, 68.2, 72.5, 110.2, 119.5,
127.6, 128.0, 128.5, 137.4, 173.1, 173.4.
A 2-mL sample tube was charged with (±)-9a (20 mg), phosphate buf-
fer (0.1 M, pH 7.0; 0.20 mL), and acetone (50 μL). To this, a 20-mg por-
tion of one of the hydrolytic enzymes, A. melleus protease (Nagase,
XP-488), C. antarctica lipase B (Novozymes, Novozym 435), or B. cepa-
cia lipase PS (Amano, PS-IM), was added, and the mixture was stirred
for 24 h at 30 °C. The reaction was monitored by TLC analysis, as de-
scribed above. Of the three enzymes tested, only A. melleus protease
showed a certain progress of the hydrolysis.
HRMS (ESI+): m/z [M + Na]+ calcd for C16H16N2NaO3: 307.1059; found:
307.1074.
A. melleus Protease-Catalyzed Hydrolysis of (±)-9a
Ethyl 1-(Benzyloxymethyl)-2,5-dioxo-3-(1H-pyrrol-1-yl)pyrroli-
dine-3-carboxylate (14b) and Ethyl 1-(Benzyloxymethyl)-2,5-di-
oxo-4-(1H-pyrrol-1-yl)pyrrolidine-3-carboxylate (15)
In a similar manner as described for the hydrolysis with PLE, to a
solution of (±)-9a (21.4 mg, 90.6 μmol) in a mixture of phosphate buf-
fer (0.1 M, pH 7.0; 200 μL) and acetone (50 μL), A. melleus protease (20
mg) was added. The mixture was stirred for 1.5 h at 30 °C. By the
same workup and purification, ester (R)-9a (10.7 mg, 45.3 μmol, 50%)
and 12a (8.1 mg, 49.3 μmol, quant.) were obtained.
To a solution of 12c (352 mg, 1.24 mmol) in anhydrous toluene (10
mL) was added KHMDS solution in toluene (0.5 M; 2.73 mL, 1.36
mmol), and the mixture was stirred at r.t. for 5 min. To the reaction
mixture was added diethyl dicarbonate (206 μL, 1.42 mmol), and the
mixture was stirred at r.t. for 20 min. The mixture was then diluted
with EtOAc (30 mL), and the solution was washed with aq NH4Cl solu-
tion, H2O, and brine. The organic layer was dried over Na2SO4 and
concentrated in vacuo. The residue was purified by silica gel column
chromatography (15 g; hexane/EtOAc, 10:1 to 5:1) to afford an insep-
arable mixture of 14b and 15 (3:1) as a yellow oil (262 mg, 59%).
(R)-9a: [α]D25 –61.5 (c 0.38, MeOH).
HPLC: tR (min) = 22.3 [(S)-9a, 0.7%], 26.1 [(R)-9a, 99.3%]; 98.6% ee.
A. melleus Protease-Catalyzed Hydrolysis of (±)-9c
In a similar manner, by the treatment of (±)-9c (20.3 mg, 76.8 μmol)
with A. melleus protease for 2.0 h, (R)-9c (10.6 mg, 40.1 μmol, 52.2%)
was obtained.
(R)-9c: [α]D26 –42.9 (c 0.48, MeOH).
IR (ATR): 3027, 2981, 2939, 2908, 2871, 1797, 1759, 1716, 1344, 1325,
1227, 1086, 1076 cm–1
.
HPLC: tR (min) = 18.6 [(S)-9c, 4.2%], 24.3 [(R)-9c, 95.8%]; 91.6% ee.
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J

J
Y. Yamashita et al.
Feature
Synthesis
1H NMR (400 MHz, CDCl3): δ (signals attributed to 14b) = 1.23 (t, J =
7.2 Hz, 3 H), 3.23 (d, J = 7.8 Hz, 1 H), 3.56 (d, J = 7.8 Hz, 1 H), 4.26 (q, J =
7.2 Hz, 2 H), 4.61 (s, 2 H), 5.10 (d, J = 10.6 Hz, 1 H), 5.43 (d, J = 6.6 Hz, 1
H), 6.28 (dd, J = 2.2, 2.1 Hz, 2 H), 6.95 (dd, J = 2.2, 2.1 Hz, 2 H), 7.27–
7.38 (m, 5 H).
(6) Burton, S. G.; Dorrington, R. A. Tetrahedron: Asymmetry 2004,
15, 2737.
(7) Turner, N. J. Curr. Opin. Chem. Biol. 2004, 8, 114.
(8) Gruber, C. C.; Lavandera, I.; Faber, K.; Kroutil, W. Adv. Synth.
Catal. 2006, 348, 1789.
(9) Ahn, Y.; Ko, S.-B.; Kim, M.-J.; Park, J. Coord. Chem. Rev. 2008, 252,
647.
(10) Kamal, A.; Azhar, M. M.; Krishnaji, T.; Malik, M. S.; Azeeza, S.
Coord. Chem. Rev. 2008, 252, 569.
1H NMR (400 MHz, CDCl3): δ (signals attributed to 15) = 1.34 (t, J = 7.2
Hz, 3 H), 3.82 (d, J = 6.6 Hz, 1 H), 4.28 (q, J = 7.2 Hz, 2 H), 4.65 (d, J =
13.2 Hz, 1 H), 4.68 (d, J = 13.2 Hz, 1 H), 5.09 (d, J = 10.6 Hz, 1 H), 5.14
(d, J = 10.6 Hz, 1 H), 5.43 (d, J = 6.6 Hz, 1 H), 6.28 (dd, J = 2.3, 2.2 Hz, 2
H), 6.95 (dd, J = 2.3, 2.2 Hz, 2 H), 7.27–7.38 (m, 5 H).
(11) Pellissier, H. Tetrahedron 2008, 64, 1563.
(12) Holt, J.; Hanefeld, U. Curr. Org. Synth. 2009, 6, 15.
(13) Lee, J. H.; Han, K.; Kim, M.-J.; Park, J. Eur. J. Org. Chem. 2010, 999.
(14) Ahmed, M.; Kelly, T.; Ghanem, A. Tetrahedron 2012, 68, 6781.
(15) Marcos, R.; Martín-Matute, B. Isr. J. Chem. 2012, 52, 639.
(16) Rachwalski, M.; Vermue, N.; Rutjes, F. P. J. T. Chem. Soc. Rev.
2013, 42, 9268.
(17) Akai, S. Chem. Lett. 2014, 43, 746.
(18) de Miranda, A. S.; Miranda, L. S. M.; de Souza, R. O. M. A. Biotech-
nol. Adv. 2015, 33, 372.
(19) Verho, O.; Bäckvall, J.-E. J. Am. Chem. Soc. 2015, 137, 3996.
(20) Schober, M.; Faber, K. Trends Biotechnol. 2013, 31, 468.
(21) Yamashita, Y.; Hanaya, K.; Sugai, T.; Mizushima, T.; Shoji, M. Tet-
rahedron 2013, 69, 6527.
(22) (a) Node, M.; Nakamura, S.; Nakamura, D.; Katoh, T.; Nishide, K.
Tetrahedron Lett. 1999, 40, 5357. (b) Katoh, T.; Kakiya, K.; Nakai,
T.; Nakamura, S.; Nishide, K.; Node, M. Tetrahedron: Asymmetry
2002, 13, 2351.
Regeneration of (±)-9a
To a solution of 14b and 15 (75 mg) in EtOH (1 mL) was added 20%
Pd(OH)2/C (50% in H2O, 61 mg) and Et3N (6 μL, 0.042 mmol). The mix-
ture was stirred for 3.5 h at r.t. under hydrogen atmosphere, and then
was filtered. The filtrate and washings were concentrated in vacuo,
and the residue was dissolved in EtOH (1 mL). To the solution was
again added 20% Pd(OH)2/C (50% in H2O, 114 mg) and Et3N (12 μL,
0.084 mmol), and the resulting mixture was stirred for a further 7.5 h
under hydrogen atmosphere. The mixture was filtered, and the com-
bined filtrate and washings was concentrated. The residue was puri-
fied by silica gel column chromatography (4 g; hexane/EtOAc, 5:1) to
afford (±)-9a (23 mg, 63%) as a colorless oil.
The 1H NMR spectrum was identical with that of authentic 9a.
(23) Laronde, J. J.; Bergbrieter, D. E.; Wong, C.-H. J. Org. Chem. 1988,
53, 2323.
Acknowledgment
(24) Westerman, B.; Scharmann, H. G.; Kortman, I. Tetrahedron:
Asymmetry 1993, 4, 2119.
This study was supported by the Platform Project for Supporting Drug
Discovery and Life Science Research (Platform for Drug Discovery, In-
formatics, and Structural Life Science) from the Ministry of Educa-
tion, Culture, Sports, Science (MEXT) and Japan Agency for Medical
Research and Development (AMED) and Keio Gijuku Academic Devel-
opment Funds and The Science Research Promotion Fund from Pro-
motion and Mutual Aid Corporation for Private Schools of Japan. We
thank Drs. Masahiro Takeda and Naoki Shirasaka of Nagase & Co. Ltd.
for the gift of A. melleus protease XP-488, Dr. Yoichi Suzuki of Novo-
zymes Japan for C. antarctica lipase B Novozym 435, Dr. Yoshihiko
Hirose of Amano Enzyme Inc. for B. cepacia lipase PS-IM, Mr. Hiroyuki
Uchiyama of Meito Sangyo Co. Ltd. for C. rugosa lipase Meito OF, and
Profs. Hironao Sajiki and Yasunari Monguchi of Gifu Pharmaceutical
University for hydrogenation catalysts and helpful suggestions. We
also thank Prof. Masaya Ikunaka of Yasuda Women’s University for
discussion and encouragement, and Mr. Yuuki Tatsumi for his efforts
in the early phase of this study.
(25) (a) Negoro, T.; Murata, M.; Ueda, S.; Fujitani, B.; Ono, Y.;
Kuromiya, A.; Komiya, M.; Suzuki, K.; Matsumoto, J. J. Med.
Chem. 1998, 41, 4118. (b) Bril, V.; Buchanan, R. Diabetes Care
2004, 27, 2369. (c) Bril, V.; Buchanan, R. Diabetes Care 2006, 29,
68. (d) Matsumoto, T.; Ono, Y.; Kuromiya, A.; Toyosawa, K.;
Ueda, Y.; Bril, V. J. Pharmacol. Sci. 2008, 107, 340.
(e) Matsumoto, T.; Ono, Y.; Kurono, M.; Kuromiya, A.;
Nakamura, K.; Bril, V. J. Pharmacol. Sci. 2008, 107, 231. (f) Ota,
A.; Kakehashi, A.; Toyoda, F.; Kinoshita, N.; Shinmura, M.;
Takano, H.; Obata, H.; Matsumoto, T.; Tsuji, J.; Dobashi, Y.;
Fujimoto, W. Y.; Kawakami, M.; Kanazawa, Y. J. Diabetes Res.
2013, 2013, 1.
(26) Mashiko, T.; Kumagai, N.; Shibasaki, M. Org. Lett. 2008, 10, 2725.
(27) Seki, M.; Kawase, Y. WO 2009/051216 A1, 2009.
(28) Kudo, Y.; Yamada, O. WO 2008/035735 A1, 2008.
(29) Maffioli, S. I.; Marzorati, E.; Marazzi, A. Org. Lett. 2005, 7, 5237.
(30) Chikusa, Y.; Hirayama, Y.; Ikunaka, M.; Inoue, T.; Kamiyama, S.;
Moriwaki, M.; Nishimoto, Y.; Nomoto, F.; Ogawa, K.; Ohno, T.;
Otsuka, K.; Sakota, K.; Shirasaka, N.; Uzura, A.; Uzura, K. Org.
Process Res. Dev. 2003, 7, 289.
(31) Ballatore, C.; Brunden, K. R.; Piscitelli, F.; Piscitelli, F.; James, M.
J.; Crowe, A.; Yao, Y.; Hyde, E.; Trojanowski, J. Q.; Lee, V. M.-Y.;
Smith, A. B. III J. Med. Chem. 2010, 53, 3739.
(32) Crider, A. M.; Sylvestri, S. C.; Tschappat, K. D.; Dick, R. M.;
Leader, W. G. J. Heterocycl. Chem. 1988, 25, 1407.
(33) Monguchi, Y.; Fujita, Y.; Endo, K.; Takao, S.; Yoshimura, M.;
Takagi, Y.; Maegawa, T.; Sajiki, H. Chem. Eur. J. 2009, 15, 834.
(34) Monguchi, Y.; Ichikawa, T.; Nozaki, K.; Kihara, K.; Yamada, Y.;
Miyake, Y.; Sawama, Y.; Sajiki, H. Tetrahedron 2015, 71, 6499.
(35) Monguchi, Y.; Kume, A.; Hattori, K.; Maegawa, T.; Sajiki, H. Tet-
rahedron 2006, 62, 7926.
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0035-1561620.
S
u
p
p
ortiInfogrmoaitn
S
u
p
p
ortioInfgrmoaitn
References
(1) Stecher, H.; Faber, K. Synthesis 1997, 1.
(2) Azerad, R.; Buisson, D. Curr. Opin. Biotechnol. 2000, 11, 565.
(3) Faber, K. Chem. Eur. J. 2001, 7, 5004.
(4) May, O.; Verseck, S.; Bommarius, A.; Drauz, K. Org. Process Res.
Dev. 2002, 6, 452.
(5) Pálmies, O.; Bäckvall, J.-E. Chem. Rev. 2003, 103, 3247.
© Georg Thieme Verlag Stuttgart · New York — Synthesis 2016, 48, A–J