Resolution of non-proteinogenic amino acids via microbial lipase-catalyzed enantioselective transesterification

Article abstract of DOI:10.1139/V07-154

A number of non-proteinogenic amino acids bearing aliphatic side chains were resolved with moderate to good enantioselectivities (E = 15-42) through the Burkholderia cepacia lipase-catalyzed enantioselective transesterification of the 2,2,2-trifluoroethyl esters of their N-benzyloxycarbonyl derivatives with methanol as a nucleophile in diisopropyl ether.

Full text of DOI:10.1139/V07-154

219
Resolution of non-proteinogenic amino acids via
microbial lipase-catalyzed enantioselective
transesterification
Toshifumi Miyazawa, Motoe Mio, Yuko Watanabe, and Takashi Yamada
Abstract: A number of non-proteinogenic amino acids bearing aliphatic side chains were resolved with moderate to
good enantioselectivities (E = 15–42) through the Burkholderia cepacia lipase-catalyzed enantioselective
transesterification of the 2,2,2-trifluoroethyl esters of their N-benzyloxycarbonyl derivatives with methanol as a
nucleophile in diisopropyl ether.
Key words: non-proteinogenic amino acids, microbial lipases, Burkholderia cepacia lipase, 2,2,2-trifluoroethyl ester,
enantioselective transesterification, enantioselectivity.
Résumé : On a résolu un certain nombre d’acides aminés non-protéinogéniques portant des chaînes latérales aliphati-
ques, avec des énantiosélectivités allant de modérées à bonnes (E = 15–42), en faisant appel à une transestérification
énantiosélective catalysée par la lipase Burkholderia cepacia des esters 2,2,2-trifluoroéthyles de leurs dérivés N-benzy-
loxycarbonyles et en utilisant le méthanol comme nucléophile dans l’éther diisopropylique comme solvant.
Mots-clés : acides aminés non-protéinogéniques, lipases microbiennes, lipase Burkholderia cepacia, ester 2,2,2-
trifluoroéthylique, transestérification énantiosélective, énantiosélectivité.
[Traduit par la Rédaction] Miyazawa et al. 224
Introduction
paya lipase-catalyzed enantioselective transesterification of
their N-benzyloxycarbonyl (Z) esters (4). In the present pa-
per, we describe the details of our study on the resolution of
Lipases (triacylglycerol hydrolases, EC 3.1.1.3) have been
employed as a very attractive group of enzymes for synthetic
purposes (1) because they are stable and easy to handle and
they have a rather broad substrate specificity. Moreover, as
they are easily available from a variety of sources (especially
bacteria and fungi), there must be a fair chance of finding a
suitable enzyme for each transformation of interest in terms
of catalytic activity and (or) selectivity. Since the first dis-
covery of enzymatic activity in anhydrous organic media (2),
1ipase-catalyzed reactions, such as esterification and trans-
esterification, have attracted an increasing attention in the
field of organic synthesis. Thus, a large number of reports
have dealt with the optical resolution of alcohols through
lipase-catalyzed reactions in organic solvents. On the other
hand, fewer examples have been accumulated for the resolu-
tion of carboxylic acids by means of the lipase-catalyzed
esterification or transesterification procedure in organic milieu.
We have previously reported on the resolution of 2-
substituted carboxylic acids via the microbial lipase-
catalyzed irreversible transesterification of their vinyl esters
(3). We also have recently reported on the resolution of non-
proteinogenic aliphatic amino acids through the Carica pa-
non-proteinogenic
amino
acids
via
the
similar
enantioselective transesterification procedure mediated by
microbial lipases (5). Homochiral non-proteinogenic amino
acids are useful as building blocks for the synthesis of
analogs of biologically active peptides (6) and as chiral aux-
iliaries for other synthetic purposes (7). We have already re-
ported on the resolution of non-proteinogenic amino acids
via the enantioselective hydrolysis of their N-Z-protected es-
ters mediated by microbial and mammalian lipases (8) and
via the enantioselective hydrolysis of their N-unprotected es-
ters catalyzed by microbial lipases (9).
Results and discussion
Initially, the transesterification between N-Z-DL-norvaline
(2-aminopentanoic acid) 2,2,2-trifluoroethyl ester (DL-1c)
and methanol in diisopropyl ether was chosen as a model re-
action (Scheme 1; R = n-C3H7; R′ = CH3), and a dozen li-
pases of microbial and mammalian origins were screened:
Burkholderia (Pseudomonas) cepacia (Amano P), Rhizopus
javanicus (Amano F), Mucor javanicus (Amano M),
Rhizopus japonicus (Rhilipase A5), Chromobacterium
viscosum (Asahikasei LP), Aspergillus niger (Amano A),
Candida rugosa (Sigma Type VII, Meito MY, Amano AY),
and porcine pancreas (Sigma Type II). Part of the results are
summarized in Table 1. We already reported that in the
Carica papaya lipase-catalyzed enantioselective transesteri-
fication the 2,2,2-trifluoroethyl ester was an ester of choice
(4). Also in the present case, the transesterification reactions
Received 22 October 2007. Accepted 17 December 2007.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 13 February 2008.
T. Miyazawa,1 M. Mio, Y. Watanabe, and T. Yamada.
Department of Chemistry, Faculty of Science and
Engineering, Konan University, Kobe 658–8501, Japan.
1Corresponding author (e-mail: miyazawa@konan-u.ac.jp).
Can. J. Chem. 86: 219–224 (2008)
doi:10.1139/V07-154
© 2008 NRC Canada

220
Can. J. Chem. Vol. 86, 2008
Scheme 1. Microbial lipase-catalyzed enantioselective transesterification between an N-Z-DL-amino acid 2,2,2-trifluoroethyl ester (DL-1)
and an alcohol. See Table 4 for R.
H
R
R
CO2CH2CF3
CO2R'
NHZ
L-2
CO2CH2CF3
NHZ
D-1
lipase
R
H
R'OH
+
+
org. solvent
NHZ
DL-1
Table 1. Enantioselective transesterification between N-Z-DL-
norvaline 2,2,2-trifluoroethyl ester (DL-1c) and methanol with dif-
ferent lipases in diisopropyl ether.
Table 2. B. cepacia lipase-catalyzed transesterification of N-Z-
DL-norvaline 2,2,2-trifluoroethyl ester (DL-1c) with different alco-
hols as nucleophiles in diisopropyl ether.
Lipase source
Conversion (%) Time
ee (%)a
E
Alcohol
Conversion (%)
Time (h)
ee (%)a
E
Burkholderia cepaciab 38
10 h
12 h
7 h
106 h
38.5 h 68
91
88
67
77
37
27
8.0
12
8.4
Methanol
Ethanol
Butanol
Heptanol
Tetradecanol
38
40
37
40
43
10
11
13
18
24
91
90
90
88
86
37
35
32
28
26
Rhizopus javanicusc
Mucor javanicusd
Rhizopus japonicuse
Chromobacterium
viscosumf
Aspergillus nigerg
Porcine pancreash
Candida rugosai
38
41
37
43
Note: Conditions: 0.4 mmol of DL-1c, 1.2 mmol of an alcohol, and 120
mg of B. cepacia lipase in 0.8 mL of diisopropyl ether at 25 °C.
aEnantiomeric excess of the newly formed ester.
21
31
34
21 d
88 d
92 d
61
51
7.6
4.9
3.8
1.2
Note: Conditions: 0.4 mmol of DL-1c, 1.2 mmol of methanol, and
catalyzed transesterification of vinyl 2-phenoxypropanoate
(3), though the differences among the alcohols were rather
small in the present case. A reverse tendency, i.e., an in-
crease in enantioselectivity with increasing chain length of
the alcohol, has been reported in the C. rugosa lipase-
catalyzed esterification of 2-(4-chlorophenoxy)propanoic
acid (12). The discrepancy may be attributable to the nature
of the lipases used and (or) a difference in the type of reac-
tions (transesterification vs. esterification).
120 mg of a lipase preparation in 0.8 mL of diisopropyl ether at 25 °C.
aEnantiomeric excess of the newly formed methyl ester (2c).
bAmano P.
cAmano F.
dAmano M.
eRhilipase A5 from Nagase ChemteX Corporation.
fAsahi Chemical Industry LP.
gAmano A.
hSigma Type II.
iSigma Type VII.
Among the external factors influencing the catalytic activ-
ity and enantioselectivity of enzymatic reactions, much at-
tention has recently been focused on the solvent effect (2).
However, it is still difficult to predict the solvent effect, es-
pecially on the enantioselectivity of a certain reaction, and
the best way is still to select an appropriate solvent after the
screening experiment. Among many solvent parameters pro-
posed so far is the hydrophobicity parameter, log P (13), and
a correlation has sometimes been observed between the E
values of an enzymatic resolution carried out in organic sol-
vents and their log P values. We also have reported such a
correlation, though rough, in the A. niger lipase-catalyzed
transesterification of vinyl 2-phenoxypropanoate with meth-
anol: the larger the log P value, the higher the E value (3). A
number of solvents having different log P values were
screened next in the B. cepacia lipase-catalyzed trans-
esterification between DL-1c and methanol (Table 3). In the
table, the solvents are arranged in decreasing order of log P.
The stereochemical preference remained to be L with all the
solvents examined. Good-to-excellent E values were ob-
tained in most of the solvents except methanol and N,N-
dimethylformamide. Thus, it is difficult to find a correlation
between the E value and the log P value. The solvent effect
on enantioselectivity observed here resembles that reported
previously in the transesterification of 1-nitro-2-propanol
with vinyl acetate cata1yzed by a lipase from Pseudomonas
sp. (14) in that tetrahydrofuran, 1,4-dioxane, and benzene
give excellent enantioselectivities. In the A. niger lipase-
catalyzed transesterification of vinyl 2-phenoxypropanoate
of other esters, such as the 2-chloroethyl and 2,2,2-
trichloroethyl esters, were extremely slow, and moreover,
such an active ester as N-hydroxysuccinimide ester pro-
ceeded with almost no enantioselectivity. As can be seen
form Table 1, the rate and enantioselectivity observed varied
largely with the enzyme employed, while the stereochemical
preference remained unchanged, the preferred absolute con-
figuration of the newly formed methyl ester (2c) being L.
Among the enzymes tested, 1ipases from B. cepacia and
R. javanicus showed a rather high reaction rate together with
a good enantioselectivity, as judged from the value of
enantiomeric ratio, E (10). With M. javanicus lipase, the
fastest reaction was accompanied by a poor enantio-
selectivity. With other lipases, the reactions were slow, and
enantioselectivities were practically unacceptable. This in-
cludes C. rugosa lipase, which is known to show a high
enantioselectivity toward carboxylic acids (11).
Next, the effect of alcohols as a nucleophile was exam-
ined on the B. cepacia lipase-catalyzed transesterification of
DL-1c in diisopropyl ether. The results are shown in Table 2.
The preferred configuration of the newly formed esters was
L, irrespective of the alcohol employed. There was almost no
difference between methanol and ethanol as the nucleophile
in terms of the reaction rate and enantioselectivity. There
was a tendency for the enantioselectivity, together with the
reaction rate, to decrease with the chain length of the alco-
hol. This is similar to that observed in the A. niger lipase-
© 2008 NRC Canada

Miyazawa et al.
221
Table 3. B. cepacia lipase-catalyzed transesterification between N-Z-DL-norvaline 2,2,2-
trifluoroethyl ester (DL-1c) and methanol in different solvents.
Solvent
Log Pa
Conversion (%)
Time (h)
% eeb
E
Cyclohexane
Toluene
Benzene
3.2
2.5
2.0
2.0
1.9
0.85
0.71
0.49
–0.33
–0.76
–1.0
–1.1
41
39
40
37
38
36
30
40
36
37
43
40
6.3
75
71
14 d
10
33
10 d
129
15.5
34
23
95
90
90
92
91
91
91
77
92
89
6
36
34
45
36
37
35
11
45
28
Chloroform
Diisopropyl ether
Diethyl ether
Pyridine
Tetrahydrofuran
Acetonitrile
Methanol
1.2
1.0
39
N,N-Dimethylformamide
1,4-Dioxane
0
91
Note: Conditions: 0.4 mmol of DL-1c, 1.2 mmol of methanol, and 120 mg of B. cepacia lipase in 0.8
mL of an anhydrous organic solvent at 25 °C.
aHydrophobicity parameters are cited from ref. 13.
bEnantiomeric excess of the newly formed methyl ester (2c).
with methanol (3) and the C. papaya lipase-catalyzed trans-
esterification of N-Z-amino acid esters (4), aliphatic hydro-
carbons, especially cyclohexane, gave high enantioselectivities.
Thus, the solvent effect on the enantioselectivity of lipase-
catalyzed transesterifications of carboxylic acid esters is
likely to be dependent largely on the lipase source. On the
other hand, the solvent had a larger effect on the reaction
rate of the present lipase-catalyzed transesterification. The
reaction was fast in an aliphatic hydrocarbon, cyclohexane,
while it was much slower in the aromatic hydrocarbons. It
was extremely retarded in a halogenated methane, chloro-
form. In the acyclic ethers, the reaction proceeded rather
smoothly. Although cyclohexane seemed to be the solvent of
choice in terms of the reaction rate as well as enantio-
selectivity, it had a poor solubilizing power towards some of
the substrates examined here. Thus, diisopropyl ether was
chosen as the most practical solvent, taking into account not
only the reaction rate and enantioselectivity but also the sol-
ubility of the substrates.
Based on these results, the transesterification between the
2,2,2-trifluoroethyl esters (DL-1) of the N-Z-derivatives of a
number of non-proteinogenic amino acids and methanol in
diisopropyl ether was examined using the two microbial li-
pases, B. cepacia lipase and R. javanicus lipase. The results
are summarized in Tables 4 and 5, respectively, which
include also those with proteinogenic amino acids for the
purpose of comparison. The B. cepacia lipase-catalyzed
transesterifications proceeded rather smoothly. The esters
(DL-1) of amino acids carrying aliphatic side chains reacted
generally with high enantioselectivities. One exception is the
result with the alanine derivative (DL-1a). The enantio-
selectivity did not depend directly on the length of the side
chain of amino acids. As the E values were larger than 15 in
all these cases except 1a, the recovered, unreacted substrate
can be obtained with a practically high enantiomeric excess
(ee) value when the reaction is allowed to proceed near or
over ~60% conversion (10). In practice, this was confirmed
experimentally in a gram-scale resolution of N-Z-norvaline
mediated by B. cepacia lipase as described later. The
enantiodiscrimination by R. javanicus lipase was poorer than
that by B. cepacia lipase for each substrate. In comparison
with the results with aliphatic amino acids, however, the
enantioselectivities observed with aromatic amino acid de-
rivatives were lower in the presence of both the microbial li-
pases. It is interesting to note that R. javanicus lipase
showed better enantiodiscrimination than B. cepacia lipase
for each aromatic substrate. In all the cases mentioned ear-
lier, the preferential reaction of the L-enantiomers was con-
firmed by comparison with authentic samples prepared from
the optically active amino acids, if available, on HPLC or
suggested from the regularity of elution order of the enantio-
mers on HPLC (15). This stereochemical preference is the
same as that observed in the lipase-catalyzed hydrolysis of
the 2-chloroethyl or 2,2,2-trifluoroethyl esters of the N-Z-
derivatives of amino acids (8) and that observed in the
C. papaya lipase-catalyzed transesterification of the N-Z-
amino acid 2,2,2-trifluoroethyl esters (4).
Furthermore, a gram-scale resolution of N-Z-DL-norvaline
was carried out via the B. cepacia lipase-catalyzed trans-
esterification of DL-1c with methanol in diisopropyl ether.
After incubation of 9 h (33% conversion), the product
methyl ester was isolated by column chromatography on sil-
ica gel, which after saponification yielded N-Z-L-norvaline
with 92% ee. In another run, N-Z-D-norvaline with 97% ee
was obtained from the remaining trifluoroethyl ester sub-
strate separated after incubation of 35 h (55% conversion).
In the C. papaya lipase-catalyzed transesterification of N-
Z-amino acid 2,2,2-trifluoroethyl esters, excellent enantio-
selectivities were obtained in cyclohexane. However, the
enantioselectivity deteriorated by switching the solvent from
cyclohexane to diisopropyl ether in which the starting esters
were more soluble (4). On the other hand, although the
enantioselectivities observed in the B. cepacia lipase-
catalyzed transesterification of the same substrates were
moderate to good, the fact that diisopropyl ether can be used
as a solvent without lowering the enantioselectivity as compared
with that in cyclohexane should make the present methodol-
ogy applicable to more non-proteinogenic amino acids.
© 2008 NRC Canada

222
Can. J. Chem. Vol. 86, 2008
Table 4. B. cepacia lipase-catalyzed transesterification between N-Z-DL-amino acid 2,2,2-
trifluoroethyl esters (DL-1) and methanol in diisopropyl ether.
Compound
R
Conversion (%)
Time (h)
ee (%)a
E
1a
1b
1c
1d
1e
1f
1g
1h
1i
CH3
C2H5
n-C3H7
n-C4H9
n-C5H11
n-C6H13
CH2=CHCH2
C6H5CH2
3-FC6H4CH2
4-FC6H4CH2
4-Thiazolylmethyl
41
38
38
38
35
36
42
42
37
43
36
2.1
2.5
10
32
37
18
2.5
14
22
18
69
92
91
89
87
90
79
73
68
55
73
8.7
42
37
30
23
31
15
11
7.7
5.1
9.5
1j
1k
11
Note: Conditions: 0.4 mmol of DL-1, 1.2 mmol of methanol, and 120 mg of B. cepacia lipase in 0.8
mL of diisopropyl ether at 25 °C.
aEnantiomeric excess of the newly formed methyl ester (2).
Experimental
Table 5. R. javanicus lipase-catalyzed transesterification between
N-Z-DL-amino acid 2,2,2-trifluoroethyl esters (DL-1) and methanol
in diisopropyl ether.
General
Compound
Conversion (%)
Time (h)
ee (%)a
E
Thin-layer chromatography was run on precoated silica
gel plates (Merck). Melting points were determined on a
Yamato MP-21 apparatus and are uncorrected. Optical rota-
1a
1b
1c
1d
1e
1f
1g
1h
1i
37
37
38
32
44
37
40
38
38
41
36
31
32
8
56
85
88
90
83
78
79
84
73
70
73
4.8
20
27
29
21
13
14
19
9.9
9.1
9.5
tions were measured with
a JASCO DIP-4 digital
1
polarimeter. H NMR spectra were obtained at 300 MHz on
a Varian Unity 300 spectrometer using chloroform-d as a
solvent with TMS as an internal standard. The liquid
chromatograph employed was a Shimadzu LC-5A instru-
ment, equipped with a Rheodyne 7125 sample injector and a
Shimadzu SPD-2A variable wavelength UV monitor. A
Shimadzu C-R1A data processor was used for data acquisi-
tion and processing.
12
18
20
17
14
21
14
11
1j
1k
Note: Conditions: 0.4 mmol of DL-1, 1.2 mmol of methanol, and 120
mg of R. javanicus lipase in 0.8 mL of diisopropyl ether at 25 °C.
aEnantiomeric excess of the newly formed methyl ester (2).
Chemicals
The racemic amino acids were benzyloxycarbonylated using
benzyloxycarbonyl chloride (Z-Cl) under the usual Schotten–
Baumann conditions (16) to give N-Z-DL-amino acids, which
were purified by recrystallization from appropriate solvents (e.g.,
ethyl acetate–petroleum ether) (8). They were converted to the
2,2,2-trifluoroethyl, 2-chloroethyl or 2,2,2-trichloroethyl esters
by the EDC [l-ethy1-3-(3-dimethylaminopropyl)carbodiimide]-
DMAP (4-dimethylaminopyridine) method (17) using the corre-
sponding alcohols. All alcohols and organic solvents employed
in this study were obtained commercially, and they were distilled
following standard protocols and dried over molecular sieves
prior to use.
Conclusion
In this paper, we have demonstrated that a number of non-
proteinogenic amino acids carrying aliphatic side chains can
be resolved with moderate-to-good enantioselectivities
through the microbial lipase-catalyzed enantioselective
transesterification of the 2,2,2-trifluoroethyl esters of their
N-Z derivatives. Although C. rugosa lipase is known to show
high enantioselectivity toward carboxylic acids, B. cepacia
lipase is a preferable enzyme in the present case, with meth-
anol being the nucleophile in diisopropyl ether. This, to-
gether with the results reported earlier on the resolution of
2-substituted carboxylic acids via the lipase-catalyzed irre-
versible transesterification where lipases from A. niger and
B. cepacia were utilized, demonstrates that the microbial
lipase-catalyzed transesterification procedure in organic mi-
lieu must be useful for the resolution of various racemic
carboxylic acids, taking into account that these lipase prepa-
rations are available easily and inexpensively.
Enzymes
Lipases from porcine pancreas (Type II) and Candida
rugosa (Type VII) were purchased from Sigma-Aldrich
Chemical Co. (USA). Lipases from Candida rugosa (MY),
Chromobacterium viscosum (LP) and Rhizopus japonicus
(Rhilipase A5) were supplied by Meito Sangyo Co. (Japan),
Asahi Chemical Industry (Japan), and Nagase ChemteX
Corporation (Japan), respectively. All other enzymes used
were supplied by Amano Pharmaceutical Co. (Japan).
© 2008 NRC Canada

Miyazawa et al.
223
Table 6. TLC and HPLC separations of the 2,2,2-trifluoroethyl esters (1) and methyl esters
(2, R′ = CH3) of N-Z-amino acids.
2,2,2-Trifluoroethyl ester
Methyl ester
Ra
TLC, Rfb
HPLC, tR (min)c
TLC, Rfb
HPLC, tR (min)c
CH3
C2H5
0.33
0.38
0.47
0.50
0.54
0.52
0.45
2.39
2.69
3.21
4.42
5.39
5.59
2.92
0.22
0.29
0.35
0.38
0.40
0.41
0.31
2.99
3.50
4.32
5.78
7.86
11.37
3.86
n-C3H7
n-C4H9
n-C5H11
n-C6H13
CH2=CHCH2
aThe side chain of an amino acid [–NHCH(R)CO–].
bSolvent: hexane–diethyl ether (2:1, v/v).
cHPLC conditions: column, Cosmosil 5C18 (4.6 mm i.d. × 150 mm); eluent, 75% aq. MeOH; flow
rate, 1.0 mL min–1; column temperature, 30 °C; detection, UV at 254 nm.
The ee value for each compound compiled in Table 4 is that
obtained by the HPLC procedure utilizing the diastereomeric
separation.
The TLC and HPLC separations of the 2,2,2-trifluoroethyl
esters (1) and methyl esters (2, R′ = CH3) of N-Z-amino ac-
ids carrying aliphatic side chains are shown in Table 6.
Reactions
General procedure for the lipase-catalyzed
transesterification
In a typical experiment, a solution of the 2,2,2-
trifluoroethyl ester of a racemic N-Z-DL-amino acid (DL-1;
0.4 mmol) and an alcohol (1.2 mmol) in an anhydrous or-
ganic solvent (0.8 mL) was stirred with a lipase preparation
(120 mg) at 25 °C. The progress of the reaction was moni-
tored by reversed-phase HPLC on a Cosmosil 5C18 column
(4.6 mm i.d. × 150 mm; Nacalai Tesque, Japan) using aq.
methanol as an eluent. When the desired degree of conver-
sion (~40 % based on the racemic starting ester) had been
achieved, the enzyme was filtered off and washed with di-
ethyl ether. The error of conversion determined by the HPLC
method was estimated as ~ 1%. The filtrate and washing
were combined and concentrated to a small volume, and the
newly formed ester (2) was separated from the unchanged
starting ester (1) by preparative TLC using hexane–diethyl
ether (mainly 2:1, v/v) as a developing solvent. After the ab-
sence of the contaminant starting ester (1) had been con-
firmed by reversed-phase HPLC, the ester formed (2) was
saponified by 0.1 mol/L NaOH in methanol as usual. The ee
value of the resolved N-Z-amino acids (Z-AA) was deter-
mined by the following HPLC procedure (18): the obtained
Z-AA was coupled with Gly-L-Phe-OMe to afford the
diastereomeric tripeptide Z-AA-Gly-L-Phe-OMe, which was
separated by reversed-phase HPLC. The peak area of each
diastereomer separated was measured to calculate the ee
value of the resolved amino acid. When the amino acid AA
bears an aromatic ring, Z-AA was converted to Z-AA-Sar-L-
Phe-OMe (Sar = sarcosine or N-methylglycine) for a better
diastereomeric separation. The kinetic resolution during the
couplings was found to be negligible, probably because
achiral Gly or Sar occupies the coupling site in the amino
component. When the product of the transesterification was
the methyl ester, another procedure was also employed. The
enantiomers of the methyl ester product were separated di-
rectly by normal-phase HPLC on a chiral column, Chiralcel
OD (4.6 mm i.d. × 250 mm; Daicel Chemical Industry,
Japan), using hexane–2-propanol as an eluent (15). The ee
values obtained by these different procedures were consis-
tent with each other within the experimental error ( 2%).
Gram-scale resolution of N-Z-norvaline
The 2,2,2-trifluoroethyl ester of racemic N-Z-DL-norvaline
(DL-1c; 1.40 g, 4.2 mmol) was dissolved in diisopropyl ether
(8 mL), followed by the addition of methanol (520 µL,
12.8 mmol) and then B. cepacia lipase (1.2 g). The reaction
mixture was stirred at 25 °C. The reaction was stopped after
9 h (33% conversion) by removing the enzyme powder by
filtration. The enzyme was washed with diethyl ether. Evap-
oration of the solvent in vacuo from the combined filtrate
and the washing afforded a pale yellow oil, from which the
product (2c) was isolated by column chromatography on sil-
ica gel (Wakogel C-300) using hexane–diethyl ether (10:1,
v/v) as an eluent. Thus, the unreacted trifluoroethyl ester
(1c) eluted first, and further elution afforded the methyl ester
(2c) as a colorless oil. The latter was saponified by
0.1 mol/L NaOH, and the raw material was recrystallized
from EtOAc–petroleum ether to give N-Z-L-norvaline as
colorless crystals: 290 mg (27% yield); mp 87–88.5 °C; [α]D25
–4.2° (c 1.0, acetone). N-Z-L-norvaline thus obtained was re-
esterified with diazomethane, and the resulting methyl ester
was analyzed by HPLC on a Chiralcel OD column: 92% ee.
In another run, the reaction was stopped after 35 h (55%
conversion), and the unreacted trifluoroethyl ester (1c) was
separated from the newly formed methyl ester (2c) by col-
umn chromatography in the same manner as mentioned ear-
lier. The trifluoroethyl ester (1c) thus obtained was saponified
by 0.1 mol/L NaOH, and the raw material was recrystallized
from EtOAc–petroleum ether to give N-Z-D-norvaline as col-
orless crystals: 410 mg (39% yield); mp 89.5–90 °C; [α]D25
+4.6° (c 1.0, acetone); 97% ee by HPLC. 1H NMR
(300 MHz, CDCl3) δ: 0.94 (3H, t. J = 7.4 Hz, CH3), 1.30–
1.50 (2H, m, CH2CH3), 1.61–1.95 (2H, m, CHCH2), 4.41
(1H, q-like, CH), 5.12 (2H, apparent s, PhCH2), 5.27 (1H, d,
J = 8.1 Hz, NH), 7.26–7.41 (5H, m, Ph), 10.48 (1H, br, OH).
Anal. calcd. for C13H17NO4: C, 62.14; H, 6.82; N, 5.57%.
Found: C, 62.30; H, 6.89; N, 5.58%.
© 2008 NRC Canada

224
Can. J. Chem. Vol. 86, 2008
7. G.M. Coppola and H.F. Schuster. Asymmetric synthesis: con-
struction of chiral molecules using amino acids. Wiley, New
York. 1987.
Acknowledgments
This work was partially supported by the Science Re-
search Promotion Fund from Japan Private School Promo-
tion Foundation. The donation of lipases by Amano
Pharmaceutical Co., Meito Sangyo Co., Asahi Chemical In-
dustry, and Nagase ChemteX Corporation is gratefully ac-
knowledged.
8. T. Miyazawa, H. Iwanaga, S. Ueji, and T. Yamada. Biocatal.
Biotrans. 17, 445 (2000).
9. T. Miyazawa, K. Imagawa, H. Minowa, T. Miyamoto, and T.
Yamada. Tetrahedron, 61, 10254 (2005).
10. C.-S. Chen, Y. Fujimoto, G. Girdaukas, and C.J. Sih. J. Am.
Chem. Soc. 104, 7294 (1982).
11. (a) I.J. Colton, S.N. Ahmed, and R.J. Kazlauskas. J. Org.
Chem. 60, 212 (1995); (b) J.J. Lalonde, C. Govardhan, N.
Khalaf, A.G. Martinez, K. Visuri, and A.L. Margolin. J. Am.
Chem. Soc. 117, 6845 (1995).
12. S.-H. Pan, T. Kawamoto, T. Fukui, K. Sonomoto, and A.
Tanaka. Appl. Microbiol. Biotechnol. 34, 47 (1990).
13. C. Laane, S. Boeren, K. Vos, and C. Veeger. Biotechnol.
Bioeng. 30, 81 (1987).
References
1. (a) U.T. Bornscheuer and R.J. Kazlauskas. Hydrolases in or-
ganic synthesis: regio- and stereo-selective biotransformations.
2nd ed. Wiley-VHC, Weinheim, Germany. 2006. p. 84. (b) K.
Faber. Biotransformations in organic chemistry. 5th ed.
Springer, Berlin. 2004. pp. 94–123, 344–367; (c) H. J. Gais and
F. Theil. In Enzyme catalysis in organic synthesis. Edited by K.
Drauz and H. Waldmann. Wiley-VHC, Weinheim, Germany.
2002. pp. 335–578.
14. K. Nakamura, Y. Takebe, T. Kitayama, and A. Ohno. Tetrahe-
dron Lett. 32. 4941 (1991).
15. T. Miyazawa, Y. Shindo, and T. Yamada. Chromatographia,
60, 419 (2004).
16. M. Bodanszky and A. Bodanszky. The practice of peptide syn-
2. A.M.P. Koskinen and A.M. Klibanov (Editors). Enzymatic re-
actions in organic media. Blackie A & P, Glasgow, UK. 1996.
3. T. Miyazawa, S. Kurita, M. Shimaoka, S. Ueji, and T. Yamada.
Chirality, 11, 554 (1999).
thesis. Springer, Berlin. 1984. p. 12.
17. M.K. Dhaon, R.K. Olsen, and K. Ramasamy. J. Org. Chem.
47, 1962 (1982).
4. T. Miyazawa, K. Onishi, T. Murashima, T. Yamada, and S.-W.
Tsai. Tetrahedron Asymmetry, 16, 2569 (2005).
18. T. Miyazawa, H. Iwanaga, T. Yamada, and S. Kuwata.
5. T. Miyazawa, M. Mio, Y. Watanabe, T. Yamada, and S.
Kuwata. Biotechnol. Lett. 14, 789 (1992).
Chirality, 4, 427 (1992).
6. D.C. Roberts and F. Vellaccio. In The peptides. Vo1. 5. Edited
by E. Gross and J. Meienhofer. Academic Press, New York.
1983. pp. 341–449.
© 2008 NRC Canada