Reduction of β-keto esters with a reductase: Construction of plural stereocenters remote from the reaction center

Article abstract of DOI:10.1006/bioo.1998.1106

The reduction of sec-alkyl 2-methyl-3-oxobutyrate with a keto ester reductase from bakers' yeast (YKER-I) is accompanied by simultaneous dynamic and static resolution of chiral centers affording the corresponding (2R,3S,1'R)-hydroxy esters preferentially. Thus, the enzyme discriminates three chiral centers simultaneously in high stereoselectivity producing useful chiral building blocks. To study the effect of the alcohol moiety which is located at a remote position from the reaction center, upon the interaction between the enzyme and a substrate, steady-state kinetic parameters, K(m) and k(cat), of YKER-I for each (1'R)- and (1'S)-substrate have been determined. The results reveal that the stereochemistry at the alcohol moiety affects K(m) rather than k(cat).

Full text of DOI:10.1006/bioo.1998.1106

Bioorganic Chemistry 27, 3–19 (1999)
Article ID bioo.1998.1106, available online at http://www.idealibrary.com on
Reduction of ␤-Keto Esters with a Reductase: Construction of
Plural Stereocenters Remote from the Reaction Center
Yasushi Kawai,1 Kouichi Hida, and Atsuyoshi Ohno
Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
Received June 2, 1998
The reduction of sec-alkyl 2-methyl-3-oxobutyrate with a keto ester reductase from bakers’
yeast (YKER-I) is accompanied by simultaneous dynamic and static resolution of chiral centers
affording the corresponding (2R,3S,1ЈR)-hydroxy esters preferentially. Thus, the enzyme dis-
criminates three chiral centers simultaneously in high stereoselectivity producing useful chiral
building blocks. To study the effect of the alcohol moiety which is located at a remote position
from the reaction center, upon the interaction between the enzyme and a substrate, steady-state
kinetic parameters, Km and kcat, of YKER-I for each (1ЈR)- and (1ЈS)-substrate have been
determined. The results reveal that the stereochemistry at the alcohol moiety affects Km rather
than kcat
.
᭧ 1999 Academic Press
INTRODUCTION
A number of enzymes are now commercially available and are used for asymmetric
syntheses extensively (1). Most of the enzymes, however, discriminate the stereochem-
istry at the reaction center or its vicinity only. Generally, the further the chiral center
is kept away from the reaction center, the less the ability of enzyme to discriminate
the chirality. Although several examples of enzymatic discrimination of a chiral
center three or more bonds apart from the reaction center have been reported, most
examples are seen in hydrolytic reactions (2–6). Although a few alcohol dehydroge-
nases that discriminate such chiral centers have been reported, all the examples
available are those that operate on cyclic compounds with a rigid skeleton (7, 8). No
alcohol dehydrogenase has been reported on the discrimination of remote and multi-
chiral centers in acyclic compounds that have flexible skeletons.
Hudlicky et al. reported the reduction of sec-alkyl acetoacetate with bakers’ yeast
(9). It is noteworthy that bakers’ yeast discriminates stereochemistry not only at the
reaction center but also at the alcohol moiety which is four bonds apart from the
reaction center, although the stereoselectivity of this reaction is not satisfactory. The
fact suggests that bakers’ yeast contains at least one enzyme which discriminates the
chiral center in the alcohol moiety in the reduction of ␤-keto esters.
Keto ester reductases have been isolated from bakers’ yeast (10–12). YKER-I
1 To whom correspondence should be addressed. Fax: ϩ81-774-38-3209. E-mail: kawai@scl.
kyoto-u.ac.jp.
3
0045-2068/99 $30.00
Copyright ᭧ 1999 by Academic Press
All rights of reproduction in any form reserved.

4
KAWAI, HIDA, AND OHNO
(yeast keto ester reductase - I), an NADPH-dependent enzyme, is one of those enzymes
isolated and reported previously as -enzyme-1 (11). The enzyme has wide specificity
L
in substrates with extremely high enantioselectivity (13). Although the enzyme reduces
various keto esters, its reaction rate highly depends on the structure of alcohol moiety
in the keto esters. For example, in the reduction of alkyl 2-methyl-3-oxobutyrate,
the reaction rate of the neopentyl ester is about 60 times larger than that of the t-
butyl ester. Kinetic parameters of YKER-I for these esters have been studied and it
was found that Km for the t-butyl ester is 20 times larger than that for the neopentyl
ester, whereas kcat for the t-butyl ester is only one-third of that for the neopentyl ester
(14). Thus, the contribution of Km is larger than that of kcat in the sensitivity toward
the change in structure of the alcohol moiety: the variation in substrate structure does
not affect kcat appreciably. The finding suggests that anchoring of a substrate into
the pocket of YKER-I is largely affected by the structure in the alcohol moiety, or
YKER-I might have the ability to discriminate the stereochemistry in the alcohol
moiety, and the reduction of ␣-alkyl ␤-keto ester of a secondary alcohol will construct
plural asymmetric centers in the product simultaneously (Eq. [1]) (15).
[1]
RESULTS AND DISCUSSIONS
Asymmetric Reduction of ␣-Methyl ␤-Keto Esters with YKER-I
␣-Methyl ␤-keto esters, 1, were prepared by allowing the corresponding alcohols
to react with diketene and succeeding methylation at the ␤-position (16). The results

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
5
from the reductions of 1 with YKER-I are listed in Table 1, in which the stereoselectiv-
ity in the carboxylic acid moiety is evaluated on the basis of enantiomeric and
diastereomeric excesses and that in the alcohol moiety in the product is evaluated
from E values (17). ␤-Hydroxy esters, 2, obtained as the products have three chiral
centers in the molecules and, therefore, have eight stereoisomers. After the reduction,
the unreacted substrate and the products formed were separated by column chromatog-
raphy on silica gel, and then each compound was hydrolyzed. The free alcohols were
converted into the corresponding acetyl (18) or ␣-methoxy-␣-trifluoromethylphenyla-
cetyl (MTPA) (19) esters, and enantiomeric excesses in the esters were measured by
GC or HPLC analysis. The enantiomeric and diastereomeric excesses in the ␤-
hydroxybutyric acid were determined by GC analysis of the corresponding ethyl ester
(Scheme 1).
In the reduction of sec-alkyl 2-methyl-3-oxobutyrate, the chiral carbon at the 2-
position remains racemic throughout the reaction due to enolization (dynamic resolu-
tion) (20), and only (2R)-1 is reduced with YKER-I to the corresponding (2R)-2 in
nearly 100% chemical and enantiomeric yields. On the other hand, the chiral center
at the 1Ј position does not racemize throughout the reaction (static resolution). There-
fore, if YKER-I discriminates the chiral center at the 1Ј position, the reaction must
cease at 50% conversion of the substrate. To regulate the conversion of the substrate,
a coupling system with glucose-6-phosphate (G6P) Ϫ glucose-6-phosphate dehydro-
genase (G6PDH) was employed to regenerate the appropriate amount of coenzyme
TABLE 1
Asymmetric Reduction of ␣-Methyl ␤-Keto Esters with YKER-I
Recovered
(%)
eea
(%)
Yield
(%)
eeb
(%)
dec
(%)
eed
(%)
R1
R2
Ee
a
b
c
d
e
Hex (0.73)f
c-Hexyl (0.87)f
Ph (1.66)f
Ph
2-Cl-Ph
4-Cl-Ph
2-Me-Ph
4-Me-Ph
4-NO2-Ph
2-Py
Me
Me
64
63
52
61
59
64
46
58
56
58
56
62
57
32
40
68
36
39
36
72
58
55
59
71
50
68
33
35
48
40
35
29
49
40
36
38
43
36
42
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
Ͼ98
91
66
78
74
54
70
77
75
75
80
80
86
91
91
6.5
12.2
13.3
4.6
Me (0.52)f
Et (0.56)f
Me
Ͼ98
Ͼ98
Ͼ98
Ͼ98
93
Ͼ98
Ͼ98
98
8.2
f
Me
Me
Me
Me
Me
Me
Me
Me
10.7
14.6
12.2
15.9
15.9
29.0
35.8
41.4
g
h
i
j
k
m
n
3-Py
4-Py
1,3-Dithianyl
Ͼ98
87
98
a ee at the 1Ј position of 1 recovered.
b Excess of (2R,3S)-2 over (2S,3R)-2.
c Excess of (2R,3S) over other diastereomers of 2.
d ee at the 1Ј position of 2.
e E value was calculated from enantiomeric excesses at the 1Ј positions in the substrate recovered
unreacted and the product (17).
f Steric parameter, v, of a substituent is defined as its van der Waals radius (22).

6
KAWAI, HIDA, AND OHNO
SCHEME 1
(Scheme 2) (21), and the procedure was found to be effective to obtain the product,
(2R,3S,1ЈR)-2, in moderate yield.
The majority of the materials isolated from the reaction mixture were (2R,3S,1ЈR)-
2 and (1ЈS)-1 which remained unreacted. The fact reveals that YKER-I discriminates
SCHEME 2

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
7
the stereochemistry at the 1Ј-position as well as those at the reaction center and the
2-position. It is noteworthy that YKER-I is quite unique in the sense that it discrimi-
nates a chiral center four bonds apart from the reaction center.
Stereoselectivity in the acid moiety is very high (Ͼ98% ee and Ն98% de) in most
of the products except for 2b (91% de), 2g (93% de), and 2m (87% de). However,
the stereoselectivity in the alcohol moiety changes depending on the property of the
substituent. For instance, in the series of 1a–1c, steric bulk of R1 is larger in 1b and
1c than that in 1a on the basis of the steric parameter, v (22), and the stereoselectivity
at the 1Ј-position in the enzymatic reduction of 1b and 1c is higher than that of
1a, suggesting that steric bulk is one of the important factors for controlling the
stereoselectivity in the alcohol moiety. Particularly, the reduction of 1n affords (2R,
3S, 1ЈR)-2n in more than 94% purity out of a possible eight stereoisomers. Although
the steric bulk of R2 also affects stereoselectivity at the 1Ј position, its direction is
opposite to that of R1 (2c and 2d); when R2 is bulkier than the methyl group, a
decrease in stereoselectivity is observed. The effect of the substituent on the phenyl
group, however, is not remarkable (2c, 2e–2i); E value does not depend on the position
and electronic property of the substituent appreciably. Remarkable improvement of
the stereoselectivity at the 1Ј position is observed when R1 has an electronegative
heteroatom as in the 3-pyridyl, 4-pyridyl group (2k–2m). Usually, an electronegative
heteroatom acts as an electron donor and interacts with an electron acceptor such as
a proton. Therefore, the fact suggests that YKER-I has an electron acceptor at the
binding site and a heteroatom in R1 interacts with this electron acceptor to stabilize
the enzyme–substrate complex affording the product in high stereoselectivity (Fig 1).
Substituent Effect on Kinetic Parameters
The mechanism of discrimination of a chiral center by an enzyme has often been
discussed on the basis of active-site models such as diamond lattice (23), cubic section
(7,24), and so on (25–30). The models have been proposed on the basis of size and
shape of the substrate, which would be accommodated by the active site of the
FIG. 1. Schematic model for the active site of YKER-I.

8
KAWAI, HIDA, AND OHNO
enzyme. The relative location of catalytic residues of the enzyme from the reaction
center of the substrate is the crucial factor for controlling the reactivity. Although
examples are few, the mechanism of enantiomer discrimination has been explored by
means of kinetic parameters, Km and kcat (8,31–36). For instance, in the optical
resolution of secondary alcohols by a lipase, the discrimination was explained to stem
from the dominant contribution of kcat over Km (31–35), whereas, in the oxidation of
some alcohols by horse liver alcohol dehydrogenase, the discrimination is due to the
dominancy of Km (36). Thus, kinetic studies may provide important information for
the discussion on the mechanism of enantiomer discrimination.
Lee et al. demonstrated, using chiral inhibitors, that serine proteases have the
ability to discriminate remote chiral centers (37). (R)- and (S)-3-phenylbutanols, chiral
transition-state-analog inhibitors of ␣-chymotrypsin, have an 88-fold difference in Ki
value in favor of the (R)-enantiomer. Thus, kinetic study has elucidated a potential
of ␣-chymotrypsin quantitatively for the discrimination of a chiral center two bonds
apart from the reaction center.
As noted above, YKER-I discriminates the stereochemistry of sec-alkyl 2-methyl-
3-oxobutyrate at the 2 and 1Ј positions as well as the stereochemistry at the reaction
center yielding (2R,3S,1ЈR)-hydroxy esters preferentially. YKER-I is the first reduc-
tase, to the authors’ best knowledge, that has been found to discriminate three chiral
centers of acyclic compounds simultaneously.
When attention is focused on E values listed in Table 1, it appears that a 3- or 4-
pyridyl substituent in the alcohol moiety of the ester exerts much higher selectivity
than the ester with a 2-pyridyl substituent, despite the fact that their structures are
quite similar. To elucidate the origin of this difference, we investigated the kinetics
of the reduction with these substrates. Optically pure secondary alcohols with an
appropriate pyridyl substituent were prepared by kinetic resolution of the correspond-
ing racemic alcohols by means of a lipase and employed for the preparation of esters
with a chiral alcohol moiety, (1ЈR)- and (1ЈS)-1j, (1ЈR)- and (1ЈS)-1k, and (1ЈR)- and
(1ЈS)-1m. Steady-state kinetic parameters, Km and kcat, of YKER-I for each substrate
were determined from [S] Ϫ [S]/v plots and the results are listed in Table 2.
When ␦⌬GÞ is defined as the difference between the free energy of activation in
the reaction of the (1ЈR)-isomer (⌬GÞR ) and that of the (1ЈS)-isomer (⌬GÞS ), the quantity
is given by
TABLE 2
Kinetic Parameters of YKER-I for 1Ј-Optically Pure Esters
Substrate
(1ЈR)-1j
R1
R2
Km, mM
kcat, sϪ1
2-Py
Me
1.4 Ϯ 0.1
10.3 Ϯ 0.8
0.89 Ϯ 0.1
9.6 Ϯ 0.9
0.56 Ϯ 0.1
8.8 Ϯ 1.1
3.5 Ϯ 0.5
1.3 Ϯ 0.04
3.7 Ϯ 0.2
1.3 Ϯ 0.1
3.8 Ϯ 0.1
1.3 Ϯ 0.09
(1ЈS)-1j
Me
2-Py
Me
3-Py
Me
(1ЈR)-1k
(1ЈS)-1k
(1ЈR)-1m
(1ЈS)-1m
3-Py
Me
4-Py
Me
4-Py

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
9
TABLE 3
The Difference in Activation Energy between the Reactions of (R)- and (S)-Isomers
Þ a
ϪRT ln kcatR /kcatS
ϪRT ln KmS /KmR
Ϫ␦⌬G
Ϫ␦⌬GEÞb
kin
Substrate
(kcal/mol)
(kcal/mol)
(kcal/mol)
(kcal/mol)
1c
1j
1k
1m
—
—
—
1.56
1.67
2.03
2.15
0.60
0.62
0.64
1.20
1.43
1.66
1.80
2.05
2.30
a Calculated from kinetic parameters.
b Calculated from E value.
␦⌬GÞ ϭ ϪRT ln E
ϭ ϪRT ln(kcat /Km)R /(kcat /Km)S
ϭ ϪRT ln(kcat R /kcat S) Ϫ RT ln(KmS /KmR),
[2]
where R, T, and E are gas constant, reaction temperature in K, and E value, respectively.
Suffixes R and S represent (1ЈR)- and (1ЈS)-isomers.
The first and second terms in the right-hand side of the last equality in Eq. [2] are
summarized in Table 3. The differences in free energy of activation between the
reactions of (1ЈR)- and (1ЈS)-isomers were calculated from the E value which was
determined from enantiomeric excesses in the product and the reactant that remained
unreacted (␦⌬GÞE ), and those calculated from kinetic parameters (␦⌬GÞkin) agree with
each other quite well within a limit of experimental error. The fact guarantees that
all the experimentally observed quantities are reasonable.
Table 3 also reveals that the contribution of kcat for ␦⌬GÞ is smaller than the
contribution of Km. Thus, it is apparent that the contribution of Km dominates the
contribution of kcat in discrimination of the chiral center at the alcohol moiety. In
other words, the active site of YKER-I accommodates the (1ЈR)-isomer better than
the (1ЈS)-counterpart. However, after the substrate is accommodated in the enzyme
pocket, the chemical reaction between the enzyme and the substrate proceeds with
similar rates for the (1ЈR)- and (1ЈS)-isomers.
E value has a wide distribution depending on the structure of R1, and here a question
TABLE 4
The Difference in Activation Energy between the Reactions with 1j and 1k or 1m
ϪRT ln kcatB /kcatA
ϪRT ln KmA /KmB
(kcal/mol)
(kcal/mol)
Substrates
1ЈR
1ЈS
1ЈR
1ЈS
1k–1j
1m–1j
0.03
0.05
0.00
Ϫ0.00
0.27
0.55
Ϫ0.04
Ϫ0.09

10
KAWAI, HIDA, AND OHNO
arises: Which of the (1ЈR)- and (1ЈS)-isomers is responsible for exerting different
sensitivities for different compounds? To answer this question, we selected 1j as a
representative low selective compound (E ϭ 16), and 1k (E ϭ 29) and 1m (E ϭ 36)
as candidates for highly selective compounds. Because these compounds are similar
in structure, it is safe to take into account the position of the nitrogen atom only as
the difference in substrate structure.
The relationship shown in Eq. [2] also holds for two different substrates, A and
B, of the same configuration, giving
␦⌬GÞ ϭ ⌬GÞB Ϫ ⌬GÞA
ϭ ϪRT ln(kcatB/kcatA) Ϫ RT ln(KmA/KmB),
[3]
where suffixes A and B represent the substrates A and B. The results are listed in
Table 4.
The difference in energetic contribution of Km between the reactions of (1ЈR)-1j
and (1ЈR)-1k or (1ЈR)-1j and (1ЈR)-1m is 0.27 or 0.55 kcal/mol, respectively, whereas
the value is one order of magnitude smaller in the corresponding series of the (1ЈS)-
counterparts (Ϫ0.04 and Ϫ0.09, respectively). On the other hand, the difference in
kcat remains almost unchanged for the two isomers.
It is worth emphasizing that 1c exerts ␦⌬GÞE of 1.56 cal/mol, a value similar to
that of 1j (1.67 kcal/mol), which suggests that the nitrogen in the 2-pyridyl group
exerts no role in binding the substrate in the large pocket; the 2-pyridyl and phenyl
groups behave similarly as substituents. The observation may be interpreted by means
of the ability of a substrate to act as a hydrogen-bond acceptor. The active site of
YKER-I has two pockets, a large and a small which accommodate the alcohol moiety.
The large pocket may have enough volume to accommodate R1 (large) of both (R)-
and (S)-configurations of a substrate without causing appreciable steric stress for the
enzyme. The stress-free enzyme is expected from the constancy of kcat terms for the
(1ЈR)- and (1ЈS)-isomers. However, of course, the position of R1 in the (1ЈR)-isomer
in the pocket is different from the position of R1 in the (1ЈS)-isomer, provided the
volume of the small pocket is nearly the same as that of the small substituent, R2 (38).
If one assumes a hydrogen-bonding site at the end of the large pocket that can
form a hydrogen bond with the nitrogen in the 3- or 4-pyridyl group, but not with
the 2-pyridyl group, the large stability of the ES complex with 1k or 1m compared
with the stability of 1c or 1j is understood. The strength of hydrogen-bond in a
protein has been reported to be about 0.5–1.5 kcal/mol (39) and the difference in
binding energy observed may be accounted for on the basis of a hydrogen bond.
Thus, we propose that the active site of YKER-I has a large and a small pocket
and there exists a hydrogen-bonding residue at the end of the large pocket, although
the pockets are essentially hydrophobic.
EXPERIMENTAL
Instruments
1H NMR spectra were recorded on a Varian VXR-200 spectrometer in CDCl3.
Capillary gas chromatograms (GC) were recorded on a Shimadzu GC-9A or GC-14B

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
11
gas chromatograph with a Shimadzu C-R6A Chromatopac. Liquid chromatograms
(HPLC) were recorded on a Hitachi 655 liquid chromatograph with a Hitachi D-2000
Chromato-Integrator.
Materials
NADPH was purchased from Kohjin Co., Ltd. YKER-I was isolated from cells
of bakers’ yeast (Oriental Yeast Co., Ltd) and purified as described previously (11).
Preparation of sec-Alkyl 2-Methyl-3-oxobutyrate
Diketene (3.00 g, 36.0 mmol) and 30.0 mmol of a secondary alcohol were dissolved
into 20 ml of benzene, and then the solution was stirred for 12 h at room temperature
after addition of few drops of triethylamine. Then, 1,1,3,3-tetramethylguanidine (3.50
g, 30.0 mmol) was added to the reaction mixture and 3.40 g (24.0 mmol) of iodometh-
ane was dropped into the solution, successively, at room temperature. After being
stirred for 2 days at room temperature, the solution was washed with 30 ml of 2 M
aqueous hydrochloric acid and 0.1 M aqueous sodium thiosulfate, successively. The
organic layer was dried over anhydrous sodium sulfate and the solvent was removed
under reduced pressure. The residue was purified by column chromatography on
silica gel. The yields of isolated products and their spectral and elemental analyses
data are as following.
1
2-Octyl 2-methyl-3-oxobutyrate (1a): 34% yield. H NMR (␦ from TMS in CDCl3,
a mixture of diastereomers) 0.83–0.92 (6H, m), 1.18–1.38 (20H, m), 1.22 (6H, d,
J ϭ 6.3 Hz), 1.33 (6H, d, J ϭ 7.1 Hz), 2.24 (6H, s), 3.47 (2H, q, J ϭ 7.1 Hz), and
4.95 (2H, 6, J ϭ 6.3 Hz). IR (neat) 1740 and 1717 cmϪ1. Anal. Calcd for C13H24O3:
C, 68.38; H, 10.59%. Found: C, 68.56; H, 10.68%.
1
1-Cyclohexylethyl 2-methyl-3-oxobutyrate (1b): 41% yield. H NMR (␦ from TMS
in CDCl3, a mixture of diastereomers) 0.84–1.86 (22H, m), 1.18 (6H, d, J ϭ 6.4 Hz),
1.34 (3H, d, J ϭ 7.0 Hz), 1.34 (3H, d, J ϭ 7.2 Hz), 2.24 (6H, s), 3.48 (1H, q, J ϭ
7.2 Hz), 3.49 (1H, q, J ϭ 7.2 Hz), and 4.78 (2H, 5, J ϭ 6.4 Hz). IR (neat) 1742 and
1717 cmϪ1. Anal. Calcd for C13H22O3: C, 68.99; H, 9.80%. Found: C, 68.77; H, 9.78%.
1
1-Phenylethyl 2-methyl-3-oxobutyrate (1c) 24% yield. H NMR (␦ from TMS in
CDCl3, a mixture of diastereomers) 1.33 (3H, d, J ϭ 7.0 Hz), 1.34 (3H, d, J ϭ 7.0
Hz), 1.56 (3H, d, J ϭ 6.6 Hz), 1.56 (3H, d, J ϭ 6.6 Hz), 2.13 (3H, s), 2.21 (3H, s),
3.51 (2H, q, J ϭ 7.0 Hz), 5.93 (2H, q, J ϭ 6.6 Hz), 5.98 (2H, q, J ϭ 6.6 Hz), and
7.21–7.45 (5H, m). IR (neat) 1738 and 1717 cmϪ1. Anal. Calcd for C13H16O3: C,
70.89; H, 7.32%. Found: C, 70.80; H, 7.41%.
1
1-Phenylpropyl 2-methyl-3-oxobutyrate (1d): 44% yield. H NMR (␦ from TMS
in CDCl3, a mixture of diastereomers) 0.89 (6H, t, J ϭ 7.0 Hz), 1.33 (6H, d, J ϭ
7.0 Hz), 1.74–2.05 (4H, m), 2.12 (3H, s), 2.20 (3H, s), 3.52 (2H, q, J ϭ 7.0 Hz),
5.70 (2H, t, J ϭ 6.9 Hz), and 7.20–7.42 (10H, m). IR (neat) 1740 and 1717 cmϪ1.
Anal. Calcd for C14H18O3: C, 71.77; H, 7.74%. Found: C, 71.49; H, 7.85%.
1
1-(2-Chlorophenyl)ethyl 2-methyl-3-oxobutyrate (1e): 54% yield. H NMR (␦ from
TMS in CDCl3, a mixture of diastereomers) 1.36 (3H, d, J ϭ 7.0 Hz), 1.37 (3H, d,
J ϭ 7.2 Hz), 1.55 (6H, d, J ϭ 6.4 Hz), 2.21 (3H, s), 2.25 (3H, s), 3.55 (2H, q,
J ϭ 7.2 Hz), 6.27 (1H, q, J ϭ 6.6 Hz), 6.28 (1H, q, J ϭ 6.6 Hz), and 7.15–7.46

12
KAWAI, HIDA, AND OHNO
(8H, m). IR (neat) 1744 and 1717 cmϪ1. Anal. Calcd for C13H15ClO3: C, 61.30;
H, 5.94%. Found: C, 61.60; H, 6.20%.
1
1-(4-Chlorophenyl)ethyl 2-methyl-3-oxobutyrate (1f): 37% yield. H NMR (␦ from
TMS in CDCl3, a mixture of diastereomers) 1.33 (3H, d, J ϭ 7.2 Hz), 1.34 (3H, d,
J ϭ 7.0 Hz), 1.53 (3H, d, J ϭ 6.8 Hz), 1.54 (3H, d, J ϭ 6.6 Hz), 2.15 (3H, s), 2.21
(3H, s), 3.51 (1H, q, J ϭ 7.2 Hz), 3.51 (1H, q, J ϭ 7.2 Hz), 5.89 (2H, q, J ϭ 6.6
Hz), and 7.22–7.37 (8H, m); IR (neat) 1742 and 1717 cmϪ1. Anal. Calcd for
C13H15ClO3: C, 61.30; H, 5.94%. Found: C, 61.04; H, 6.03%.
1
1-(2-Methylphenyl)ethyl 2-methyl-3-oxobutyrate (1g): 39% yield. H NMR (␦ from
TMS in CDCl3, a mixture of diastereomers) 1.34 (3H, d, J ϭ 7.4 Hz), 1.35 (3H, d,
J ϭ 7.2 Hz), 1.52 (6H, d, J ϭ 6.4 Hz), 2.16 (3H, s), 2.23 (3H, s), 2.38 (6H, s), 3.52
(2H, q, J ϭ 7.2 Hz), 6.12 (1H, q, J ϭ 6.6 Hz), 6.13 (1H, q, J ϭ 6.6 Hz), 7.11–7.26
(6H, m), and 7.30–7.39 (2H, m). IR (neat) 1740 and 1717 cmϪ1. Anal. Calcd for
C14H18O3: C, 71.77; H, 7.74%. Found: C, 71.82; H, 7.90%.
1
1-(4-Methylphenyl)ethyl 2-methyl-3-oxobutyrate (1h): 51% yield. H NMR (␦ from
TMS in CDCl3, a mixture of diastereomers) 1.32 (3H, d, J ϭ 7.2 Hz), 1.33 (3H, d,
J ϭ 7.2 Hz), 1.54 (3H, d, J ϭ 6.6 Hz), 1.54 (3H, d, J ϭ 6.6 Hz), 2.13 (3H, s), 2.20
(3H, s), 2.34 (6H, s), 3.49 (2H, q, J ϭ 7.2 Hz), 5.90 (2H, q, J = 6.6 Hz), and
7.11–7.28 (8H, m). IR (neat) 1740 and 1717 cmϪ1. Anal. Calcd for C14H18O3: C,
71.77; H, 7.74%. Found: C, 71.74; H, 7.72%.
1
1-(4-Nitrophenyl)ethyl 2-methyl-3-oxobutyrate (1i): 29% yield. H NMR (␦ from
TMS in CDCl3, a mixture of diastereomers) 1.37 (3H, d, J ϭ 7.2 Hz), 1.38 (3H, d,
J ϭ 7.2 Hz), 1.58 (6H, d, J ϭ 6.6 Hz), 2.20 (3H, s), 2.24 (3H, s), 3.57 (1H, q, J ϭ
7.2 Hz), 3.58 (1H, q, J ϭ 7.2 Hz), 5.97 (1H, q, J = 6.6 Hz), 5.98 (1H, q, J ϭ 6.6
Hz), 7.52 (4H, m), and 8.24 (4H, m). IR (neat) 1744, 1717, 1524, and 1348 cmϪ1.
Anal. Calcd for C13H15NO5: C, 58.86; H, 5.70; N 5.28%. Found: C, 58.66; H, 5.60;
N 5.22%.
1
1-(2-Pyridyl)ethyl 2-methyl-3-oxobutyrate (1j): 38% yield. H NMR (␦ from TMS
in CDCl3, a mixture of diastereomers) 1.37 (3H, d, J ϭ 7.2 Hz), 1.38 (3H, d, J ϭ
7.2 Hz), 1.62 (6H, d, J ϭ 6.6 Hz), 2.23 (3H, s), 2.27 (3H, s), 3.61 (2H, q, J ϭ 7.2
Hz), 5.97 (2H, q, J ϭ 6.6 Hz), 7.18–7.39 (4H, m), 7.71 (2H, m), and 8.59 (2H, m).
IR (neat) 1742, 1717, 1591, and 1437 cmϪ1. Anal. Calcd for C12H15NO3: C, 65.14;
H, 6.83; N 6.33%. Found: C, 64.92; H, 6.85; N 6.44%.
1
1-(3-Pyridyl)ethyl 2-methyl-3-oxobutyrate (1k): 52% yield. H NMR (␦ from TMS
in CDCl3, a mixture of diastereomers) 1.35 (3H, d, J ϭ 7.2 Hz), 1.36 (3H, d, J ϭ
7.2 Hz), 1.60 (6H, d, J ϭ 6.6 Hz), 2.16 (3H, s), 2.23 (3H, s), 3.54 (2H, q, J ϭ 7.2
Hz), 5.96 (2H, q, J ϭ 6.6 Hz), 7.30 (2H, m), 7.67 (2H, m), 8.57 (2H, m), and 8.62
(2H, m). IR (neat) 1742 and 1717 cmϪ1. Anal. Calcd for C12H15NO3: C, 65.14;
H, 6.83; N 6.33%. Found: C, 64.98; H, 6.71; N 6.30%.
1
1-(4-Pyridyl)ethyl 2-methyl-3-oxobutyrate (1m): 39% yield. H NMR (␦ from TMS
in CDCl3, a mixture of diastereomers) 1.38 (3H, d, J ϭ 6.8 Hz), 1.39 (3H, d, J ϭ
7.4 Hz), 1.55 (6H, d, J ϭ 6.6 Hz), 2.21 (3H, s), 2.25 (3H, s), 3.58 (2H, q, J ϭ 7.1
Hz), 5.88 (2H, q, J ϭ 6.6 Hz), 7.23 (4H, m), and 8.60 (4H, m). IR (neat) 1742 and
1717 cmϪ1. Anal. Calcd for C12H15NO3: C, 65.14; H, 6.83; N 6.33%. Found: C,
65.04; H, 6.84; N 6.43%.

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
13
1
1-(1,3-Dithian-2-yl)ethyl 2-methyl-3-oxobutyrate (1n): 57% yield. H NMR
(␦ from TMS in CDCl3, a mixture of diastereomers) 1.36 (6H, d, J ϭ 6.6 Hz), 1.43
(3H, d, J ϭ 6.6 Hz), 1.44 (3H, d, J ϭ 6.4 Hz), 1.81–2.17 (4H, m), 2.28 (3H, s), 2.30
(3H, s), 2.18–3.02 (8H, m), 3.53 (1H, q, J ϭ 7.0 Hz), 3.55 (1H, q, J ϭ 7.2 Hz), 4.04
(1H, d, J ϭ 6.6 Hz), 4.06 (1H, d, J ϭ 6.6 Hz), 5.28 (1H, 5, J ϭ 6.4 Hz), and 5.29
(1H, 5, J ϭ 6.6 Hz). IR (neat) 1740 and 1715 cmϪ1. Anal. Calcd for C11H18S2O3:
C, 50.35; H, 6.91%.Found: C, 50.52; H, 6.91%.
Preparation of (S)- and (R)-1-Pyridylethanols
1-(2-Pyridyl)ethanols. A lipase (Amano PS, 3.85 g, 23100 unit; standard: triglycer-
ide) was added to a stirred 400-ml benzene solution containing 9.99 g of racemic 1-
(2-pyridyl)ethanol (81.1 mmol) and 34.91 g of vinyl acetate (400 mmol). The reaction
mixture was stirred for 2 days at 35ЊC. The mixture was filtrated and the filtrate
was concentrated under reduced pressure to give a mixture of the (S)-alcohol and
acetate of the (R)-alcohol. The mixture was subjected to silica gel column chromatog-
raphy with an eluent of ethyl acetate to give 4.77 g of the alcohol (38.7 mmol) and
6.93 g of the acetate (42.0 mmol) in 47.7 and 51.7% yields, respectively. Enantiomeric
excesses in the alcohol and the acetate were determined to be 95.0 and 91.9%,
respectively, by the procedure noted below (vide infra).
The (S)-1-(2-pyridyl)ethanol (4.77 g, 38.7 mmol) thus obtained was enzymatically
acetylated with the same procedure mentioned above once more and 3.15 g of the
(S)-alcohol (25.7 mmol, 62.3% overall yield) was obtained in Ͼ99.9% ee;
[␣]D24 Ϫ59.1 (c ϭ 1.00, EtOH): lit. (40) [␣]D Ϫ56.1 (c ϭ 0.5, EtOH).
(R)-1-(2-Pyridyl)ethyl acetate (6.93 g, 42.0 mmol) was hydrolyzed with a lipase
(Amano PS) in 100 ml of 0.1 M potassium phosphate buffer (pH 7.0, 25ЊC) for 12
h. The reaction mixture was filtrated and the filtrate was made alkaline by 1 M
aqueous sodium hydroxide. Then, the resulting solution was extracted with ethyl
acetate. The organic layers collected were dried over anhydrous sodium sulfate and
the solvent was removed under reduced pressure. The residue was subjected to silica
gel column chromatography with an eluent of ethyl acetate to give 3.70 g of the (R)-
alcohol (30.0 mmol, 71.6% overall yield) in 99.6% ee; [␣]D24 ϩ60.6 (c ϭ 1.00, EtOH).
(3-Pyridyl)ethanols. The procedure noted above gave 4.19 g of (S)-1-(3-pyridyl)-
ethanol (34.0 mmol, 74.0% overall yield) in Ͼ99.5% ee; [␣]D24 Ϫ55.8 (c ϭ 1.02,
CHCl3): lit. (41) [␣]D Ϫ53.5 (c ϭ 1.09, CHCl3) and 4.71 g of the (R)-alcohol (38.2
mmol, 83.1% overall yield) in 99.4% ee; [␣]D24 ϩ55.1 (c ϭ 1.00, CHCl3).
(4-Pyridyl)ethanols. The procedure noted above gave 3.55 g of (S)-1-(4-pyridyl)-
ethanol (28.8 mmol, 68.7% overall yield) in 99.9% ee; [␣]D24 ϭ Ϫ43.6 (c ϭ 1.51,
EtOH): lit. (40) [␣]D Ϫ43.4 (c ϭ 0.5, EtOH) and 3.40 g of the (R)-alcohol (27.6
mmol, 65.9% overall yield) in Ͼ99.9% ee; [␣]D24 ϩ43.4 (c ϭ 1.52, EtOH).
General Procedure for Enzymatic Reduction of sec-Alkyl 2-Alkyl-3-oxobutyrates.
A substrate (0.50 mmol), 10 mg of NADPH, 80 mg of G6P (ca. 0.2 mmol), 0.5
mg of G6PDH, and 17 ml of buffer solution (0.05 M, Mopso, pH 7.0) were placed
in a 50-ml flask, and 3.0 ml of aqueous solution containing YKER-I (0.7 unit) was
added to the solution. The reaction mixture was stirred overnight at 30ЊC in the dark,
and then the organic materials were extracted with ethyl acetate. The organic layers

14
KAWAI, HIDA, AND OHNO
collected were dried over anhydrous sodium sulfate and the solvent was removed
under reduced pressure. The residue was subjected to column chromatography on
silica gel with an eluent of a hexane:ethyl acetate (3:1) mixture to give the product
and the substrate remained unreacted.
1
2-Octyl 3-hydroxy-2-methylbutyrate (2a): 33% yield. H NMR (␦ from TMS in
CDCl3) 0.78–0.94 (3H, m), 1.05–1.40 (17H, m), 1.45–1.69 (2H, m), 2.47 (1H, q, d,
J ϭ 7.2 Hz, J ϭ 3.9 Hz), 2.68 (1H, br), 3.98–4.12 (1H, m) and 4.94 (1H, 6,
J ϭ 6.3 Hz).
1
1-Cyclohexylethyl 3-hydroxy-2-methylbutyrate (2b): 35% yield. H NMR (␦ from
TMS in CDCl3) 0.85–1.87 (11H, m), 1.18 (3H, d, J ϭ 6.4 Hz), 1.19 (6H, d, J ϭ 7.2
Hz), 2.48 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.8 Hz), 2.68 (1H, br), 4.07 (1H, m), and 4.77
(1H, 5, J ϭ 6.4 Hz).
1
1-Phenylethyl 3-hydroxy-2-methylbutyrate (2c): 48% yield. H NMR (␦ from TMS
in CDCl3) 1.12 (3H, d, J = 6.4 Hz), 1.17 (3H, d, J ϭ 7.3 Hz), 1.54 (3H, d, J ϭ 6.6
Hz), 2.51 (1H, q, d, J ϭ 7.3 Hz, J ϭ 4.0 Hz), 2.58 (1H, br), 3.97–4.11 (1H, m),
5.90 (1H, t, J ϭ 6.6 Hz), and 7.23–7.42 (5H, m).
1
1-Phenylpropyl 3-hydroxy-2-methylbutyrate (2d): 41% yield. H NMR (␦ from
TMS in CDCl3) 0.90 (3H, t, J ϭ 7.4 Hz), 1.12 (3H, d, J ϭ 6.4 Hz), 1.23 (3H, d,
J ϭ 7.2 Hz), 1.73–2.04 (2H, m), 2.53 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.8 Hz), 2.55 (1H,
br), 4.05 (1H, m), 5.69 (1H, t, J ϭ 6.9 Hz) and, 7.20–7.44 (5H, m).
1
1-(2-Chlorophenyl)ethyl 3-hydroxy-2-methylbutyrate (2e): 35% yield. H NMR
(␦ from TMS in CDCl3) 1.17 (3H, d, J ϭ 6.6 Hz), 1.24 (3H, d, J ϭ 7.2 Hz), 1.54
(3H, d, J ϭ 6.4 Hz), 2.43 (1H, br), 2.55 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.8 Hz), 4.08
(1H, q, d, J ϭ 6.4 Hz, J ϭ 4.0 Hz), 6.25 (1H, q, J ϭ 6.6 Hz), and 7.18–7.49 (4H, m).
1
1-(4-Chlorophenyl)ethyl 3-hydroxy-2-methylbutyrate (2f): 29% yield. H NMR
(␦ from TMS in CDCl3) 1.14 (3H, d, J ϭ 6.4 Hz), 1.21 (3H, d, J ϭ 7.2 Hz), 1.53
(3H, d, J ϭ 6.6 Hz), 2.41 (1H, br), 2.51 (1H, q, d, J ϭ 7.2 Hz, J ϭ 4.0 Hz), 4.04
(1H, q, d, J ϭ 6.4 Hz, J ϭ 4.0 Hz), 5.87 (1H, q, J ϭ 6.6 Hz), and 7.24–7.37 (4H, m).
1
1-(2-Methylphenyl)ethyl 3-hydroxy-2-methylbutyrate (2g): 49% yield. H NMR
(␦ from TMS in CDCl3)1.14 (3H, d, J ϭ 6.4 Hz), 1.22 (3H, d, J ϭ 7.2 Hz), 1.52
(3H, d, J ϭ 6.6 Hz), 2.38 (3H, s), 2.52 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.8 Hz), 2.53
(1H, br), 4.16 (1H, m), 6.16 (1H, q, J ϭ 6.6 Hz), and 7.11–7.42 (4H, m).
1
1-(4-Methylphenyl)ethyl 3-hydroxy-2-methylbutyrate (2h): 40% yield. H NMR
(␦ from TMS in CDCl3) 1.13 (3H, d, J ϭ 6.6 Hz), 1.21 (3H, d, J ϭ 7.2 Hz), 1.54
(3H, d, J ϭ 6.6 Hz), 2.34 (3H, s), 2.51 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.8 Hz), 2.52
(1H, br), 4.04 (1H, m), 5.88 (1H, q, J ϭ 6.6 Hz), and 7.17-7.29 (4H, m).
1
1-(4-Nitrophenyl)ethyl 3-hydroxy-2-methylbutyrate (2i): 36% yield. H NMR
(␦ from TMS in CDCl3) 1.17 (3H, d, J ϭ 6.4 Hz), 1.24 (3H, d, J ϭ 7.2 Hz), 1.58
(3H, d, J ϭ 6.6 Hz), 2.28 (1H, br), 2.56 (1H, q, d, J ϭ 7.2 Hz, J ϭ 4.0 Hz), 4.08
(1H, m), 5.96 (1H, q, J ϭ 6.6 Hz), 7.52 (2H, m), and 8.23 (2H, m).
1
1-(2-Pyridyl)ethyl 3-hydroxy-2-methylbutyrate (2j): 38% yield. H NMR (␦ from
TMS in CDCl3) 1.15 (3H, d, J ϭ 6.6 Hz), 1.17 (3H, d, J ϭ 7.2 Hz), 1.61 (3H, d,
J ϭ 6.8 Hz), 2.74 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.6 Hz), 4.30 (1H, m), 4.62 (1H, br),
5.97 (1H, q, J ϭ 6.6 Hz), 7.16–7.31 (2H, m), 7.69 (1H, m), and 8.50 (1H, m).
1
1-(3-Pyridyl)ethyl 3-hydroxy-2-methylbutyrate (2k): 43% yield. H NMR (␦ from

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
15
TMS in CDCl3) 1.15 (3H, d, J ϭ 6.4 Hz), 1.23 (3H, d, J ϭ 7.2 Hz), 1.58 (3H, d,
J ϭ 6.6 Hz), 2.53 (1H, q, d, J ϭ 7.2 Hz, J ϭ 4.2 Hz), 2.70 (1H, br), 4.06 (1H, q,
d, J ϭ 6.6 Hz, J ϭ 4.2 Hz), 5.94 (1H, q, J ϭ 6.6 Hz), 7.30 (1H, m), 7.69 (1H, m),
8.55 (1H, m), and 8.63 (1H, m).
1
1-(4-Pyridyl)ethyl 3-hydroxy-2-methylbutyrate (2m): 36% yield. H NMR (␦ from
TMS in CDCl3) 1.18 (3H, d, J ϭ 6.6 Hz), 1.25 (3H, d, J ϭ 7.0 Hz), 1.54 (3H, d,
J ϭ 6.6 Hz), 2.57 (1H, q, d, J ϭ 7.2 Hz, J ϭ 4.0 Hz), 4.10 (1H, q, d, J ϭ 6.6 Hz,
J ϭ 4.0 Hz), 4.74 (1H, br), 5.86 (1H, q, J ϭ 7.0 Hz), 7.25 (2H, m), and 8.59 (2H, m).
1
1-(1,3-Dithian-2-yl)ethyl 3-hydroxy-2-methylbutyrate (2n): 44% yield. H NMR
(␦ from TMS in CDCl3) 1.18 (3H, d, J ϭ 7.2 Hz), 1.20 (3H, d, J ϭ 6.6 Hz), 1.43
(3H, d, J ϭ 6.4 Hz), 1.83–2.19 (2H, m), 2.60 (1H, q, d, J ϭ 7.2 Hz, J ϭ 3.3 Hz),
2.70–3.01 (4H, m), 4.12 (1H, d, J ϭ 6.2 Hz), 4.21 (1H, br), and 5.29 (1H, 5, J ϭ
6.4 Hz).
Determination of Enantiomeric Excesses at the 1Ј Position of Substrate and
Product
A sec-alkyl 3-hydroxy-2-methylbutyrate was added to a solution containing each
1 ml of methanol and 1 M aqueous sodium hydroxide. The solution was stirred for
30 min at 60ЊC, and then the methanol was removed under reduced pressure. The
organic materials were extracted with ethyl acetate. The solvent was evaporated
under reduced pressure and the residue was subjected to a short column of anhydrous
sodium sulfate and silica gel with ethyl acetate as an eluent giving an alcohol from
the ester. The alcohol was modified to the corresponding acetate (18) or MTPA ester
(19). Unreacted substrate was treated with the same procedure as above and the
enantiomeric excess in the alcohol was determined similarly.
Absolute configurations of the esters, except for 1n and 2n, were determined by
a comparison of the retention times of their corresponding acetates or MTPA esters
(for 1i and 2i) on GC with those of the authentic samples from the (R)- and (S)-
alcohols. The (R)- and (S)-alcohols were prepared by optical resolution of the
corresponding racemic alcohols with appropriate lipases (31,41). Absolute configura-
tions of MTPA esters from 1n and 2n were determined by a measure of the optical
rotations of the corresponding alcohols. The optical rotations were compared with
those reported in the literature (42). Thus, absolute configurations of the alcohols
from 2 and 1 that remained unreacted were established as R and S, respectively.
The chromatographic conditions are as follows.
All acetates of the alcohols: GC with CP-cyclodextrin-B-236-M12, 0.25 mm ϫ 25
m; temperature depends on the material.
MTPA esters of the alcohols from 1i and 2i; GC with HR-1701, 0.25 mm ϫ 25
m, 210ЊC.
MTPA esters of the alcohols from 1n and 2n; HPLC with Chiralcel OD, 25 cm;
eluent: hexane-iPrOH = 200:1. After separation of the isomers, optical rotations were
observed: 1-(1-hydroxyethyl)-1,3-dithian from 2n, [␣]D24 ϩ5.2 (c ϭ 1.79, MeOH); 1-
(1-hydroxyethyl)-1,3-dithian from 1n, [␣]D24 Ϫ3.6 (c ϭ 2.19, MeOH). lit (42), [␣]D
ϩ5.75 (c ϭ 1, MeOH) for the (R)-isomer.

16
KAWAI, HIDA, AND OHNO
TABLE 5
Conditions for Kinetic Measurement and Kinetic Results
a
Substrate
[E]0 ϫ 105, mM
[S]0, mM
␯ ϫ 108 (M и sϪ1
)
(1ЈR)-1j
3.4
1.0
2.0
3.0
4.0
5.0
1.0
2.0
3.0
4.0
5.0
3.0
5.0
7.0
10
6.0 Ϯ 0.2
8.2 Ϯ 0.2
8.5 Ϯ 0.1
10.2 Ϯ 0.2
10.6 Ϯ 0.04
7.3 Ϯ 0.5
9.2 Ϯ 0.2
10.9 Ϯ 0.2
12.3 Ϯ 0.5
13.1 Ϯ 0.1
1.7 Ϯ 0.04
2.1 Ϯ 0.05
2.8 Ϯ 0.1
3.6 Ϯ 0.05
4.6 Ϯ 0.04
2.1 Ϯ 0.07
2.5 Ϯ 0.07
3.1 Ϯ 0.05
3.9 Ϯ 0.02
5.4 Ϯ 0.2
7.2 Ϯ 0.3
10.4 Ϯ 0.2
12.0 Ϯ 0.2
16.8 Ϯ 0.3
18.3 Ϯ 0.7
10.0 Ϯ 0.3
16.1 Ϯ 0.2
21.1 Ϯ 1.0
26.6 Ϯ 1.1
29.7 Ϯ 1.0
2.0 Ϯ 0.07
3.1 Ϯ 0.05
3.5 Ϯ 0.1
4.4 Ϯ 0.2
5.4 Ϯ 0.2
3.2 Ϯ 0.09
4.5 Ϯ 0.1
5.5 Ϯ 0.2
6.7 Ϯ 0.1
8.0 Ϯ 0.1
10.0 Ϯ 0.4
13.1 Ϯ 0.4
16.3 Ϯ 0.2
17.6 Ϯ 0.2
20.0 Ϯ 0.4
(1ЈR)-1j
(1ЈS)-1j
(1ЈS)-1j
(1ЈR)-1k
(1ЈR)-1k
(1ЈS)-1k
(1ЈS)-1k
(1ЈR)-1m
5.4
5.4
20
6.7
3.0
5.0
7.0
10
20
6.4
0.33
0.67
1.0
2.0
3.0
0.33
0.67
1.0
2.0
3.0
3.0
5.0
6.7
10
10.7
6.4
15
10.7
6.4
3.0
5.0
6.7
10
15
0.33
0.67
1.0
1.3
2.0

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
17
TABLE 5—Continued
a
Substrate
[E]0 ϫ 105, mM
[S]0, mM
␯ ϫ 108 (M и sϪ1
)
(1ЈR)-1m
10.7
0.33
0.67
1.0
1.3
2.0
3.0
5.0
7.0
10
14.4 Ϯ 0.8
21.2 Ϯ 0.4
25.6 Ϯ 0.3
28.4 Ϯ 0.3
31.7 Ϯ 0.4
1.8 Ϯ 0.08
2.6 Ϯ 0.08
3.1 Ϯ 0.09
3.7 Ϯ 0.1
4.7 Ϯ 0.08
2.2 Ϯ 0.01
3.2 Ϯ 0.02
3.8 Ϯ 0.1
4.5 Ϯ 0.1
6.2 Ϯ 0.2
(1ЈS)-1m
(1ЈS)-1m
5.4
6.7
20
3.0
5.0
7.0
10
20
a
␯ ϭ k[E]0.
Determination of Enantiomeric and Diastereomeric Excesses in sec-Alkyl 3-
Hydroxy-2-methylbutyrate
After the ester was partially hydrolyzed, the aqueous layer was separated form the
reaction mixture and acidified with 2 M aqueous hydrochloric acid. Then, the mixture
was extracted with ethyl acetate. The organic layers collected were dried over
anhydrous sodium sulfate and the solvent was removed under reduced pressure. The
residue and 1 mg of p-toluenesulfonic acid were dissolved into 10 ml of ethanol, and
then the solution was refluxed for overnight. The ethanol was removed under reduced
pressure, and the residue was subjected to a short column on anhydrous sodium sulfate
and silica gel with ethyl acetate as an eluent to give ethyl 3-hydroxy-2-methylbutyrate.
Enantiomeric and diastereomeric excesses in ethyl 3-hydroxy-2-methylbutyrate thus
obtained were determined by GC (Chiraldex G-TA, 0.25 mm ϫ 30 m, 70ЊC).
Absolute configuration of the isomer corresponding to each signal of the ester
obtained from sodium borohydride reduction was determined by a comparison of its
retention time on the GC with those of the authentic samples prepared by antiselective
methylation of ethyl (S)- and racemic-3-hydroxy-butyrate (43) to obtain the (2S,3S)-
isomer and a mixture of (2S,3S)- and (2R,3R)-isomers, respectively. The (2R,3S)-
isomer was obtained by the reduction of ethyl 3-oxo-2-methylbutyrate with bakers’
yeast (16). Absolute configuration of ethyl 3-hydroxy-2-methylbutyrate obtained
herein was established to be 2R,3S.
Kinetic Measurement
A solution containing an appropriate amount of YKER-I was added to 0.1 M
potassium phosphate buffer containing an appropriate amount of a substrate and
NADPH (0.092 mM) to give a 3.00 ml solution. The reaction was followed spectropho-
tometrically at 30ЊC by an observation of the decrease in absorbance of NADPH at

18
KAWAI, HIDA, AND OHNO
340 nm. Kinetic parameters, Km and kcat, were elucidated from [S] Ϫ [S]/v plots
with least-squares analyses. Kinetics were run at least three times for a particular
concentration of a substrate. Substrate concentrations employed for the measurement
as well as kinetic results are listed in Table 5. Km and kcat thus calculated are listed
in Table 2.
REFERENCES
1. Wong, C.-H., and Whitesides, G. M. (1994) Enzymes in Synthetic Organic Chemistry, Pergamon
Press, Oxford.
2. Hughes, D. L., Bergan, J. J., Amato, J. S., Bhupathy, M., Leazer, J. L., McNamara, J. M., Sidler, D.
R., Reider, P. J., and Grabowski, E. J. J. (1990) J. Org. Chem. 55, 6252.
3. Ladner, W. E., and Whitesides, G. M. (1984) J. Am. Chem. Soc. 106, 7250.
4. Mohr, P., Waespe-Sarcevic, N., and Tamm, C. (1983) Helv. Chim. Acta 66, 2501.
5. Jones, J. B., and Marr, P. W. (1973) Tetrahedron Lett. 3065.
6. Cohen, S. G., and Weinstein, S. Y. (1964) J. Am. Chem. Soc. 86, 725.
7. Jakovac, I. J., Goodbrand, H. B., Lok, K. P., and Jones, J. B. (1982) J. Am. Chem. Soc. 104, 4659.
8. Lee, K. M., Dahlhauser, K. F., and Plapp, B. V. (1988) Biochemistry 27, 3528.
9. Hudlicky, T., Tsunoda, T., and Gadamasetti, K. G. (1991) J. Org. Chem. 56, 3619.
10. Shieh, W., Gopalan, A. S., and Sih, C. J. (1985) J. Am. Chem. Soc. 107, 2993.
11. Nakamura, K., Kawai, Y., Nakajima, N., and Ohno, A. (1991) J. Org. Chem. 56, 4778.
12. Nakamura, K., Kondo, S., Kawai, Y., Nakajima, N., and Ohno, A. (1994) Biosci. Biotech. Biochem.
58, 2236.
13. Kawai, Y., Tsujimoto, M., Kondo, S., Takanobe, K., Nakamura, K., and Ohno, A. (1994) Bull. Chem.
Soc. Jpn. 67, 524.
14. Nakamura, K., Kawai, Y., Miyai, T., Honda, S., Nakajima, N., and Ohno, A. (1991) Bull. Chem.
Soc. Jpn. 64, 1467.
15. Kawai, Y., Hida, K., Nakamura, K., and Ohno, A. (1995) Tetrahedron Lett. 36, 591.
16. Nakamura, K., Miyai, T., Nagar, A., Oka, S., and Ohno, A. (1989) Bull. Chem. Soc. Jpn. 62, 1179.
17. Chen, C.-S., Fujimoto, Y., Girdaucas, G., and Sih, C. J. (1982) J. Am. Chem. Soc. 104, 7294.
18. Weber, H., and Khorana, H. G. (1972) J. Mol. Biol. 72, 219.
19. Dale, J. A., Dull, D. L., and Mosher, H. S. (1969) J. Org. Chem. 34, 2543.
20. Kitamura, M., Tokunaga, M., and Noyori, R. (1993) J. Am. Chem. Soc. 115, 144.
21. Wong, C.-H., and Whitesides, G. M. (1981) J. Am. Chem. Soc. 103, 4890.
22. Charton, M. (1975) J. Am. Chem. Soc. 97, 1552.
23. Prelog, V. (1964) Pure. Appl. Chem. 9, 119.
24. Jones, J. B. and Jakovac, I. J. (1982) Can. J. Chem. 60, 19.
25. Toone, E. J., Werth, M. J., and Jones, J. B. (1990) J. Am. Chem. Soc. 112, 4946
26. Itoh, T., Kuroda, K., Tomosada, M., and Takagi, Y. (1991) J. Org. Chem. 56, 797.
27. Kazlauskas, R. J., Weissfloh, A. N. E., Rappaport, A. T., and Cuccia, L. A. (1991) J. Org. Chem.
56, 2656.
28. Burgess, K., and Jennings, L. D. (1991) J. Am. Chem. Soc. 113, 6129.
29. Parida, S., and Dordick, J. S. (1993) J. Org. Chem. 58, 3238.
30. Cygler, M., Grochulski, P., Kazlauskas, R. J., Schrag, J. D., Bouthillier, F., Rubin, B., Serreqi, A.
N., and Gupta, A. K. (1994) J. Am. Chem. Soc. 116, 3180.
31. Nakamura, K., Kawasaki, M., and Ohno, A. (1996) Bull. Chem. Soc. Jpn. 69, 1079.
32. Ema, T., Kobayashi, J., Maeno, S., Sakai, T., and Utaka, M. (1998) Bull. Chem. Soc. Jpn. 71, 443.
33. Nishizawa, K., Ohgami, Y., Matsuo, N., Kishida, H., and Hirohara, H. (1997) J. Chem. Soc. Perkin
Trans. 2, 1293.
34. Secundo, F., Riva, S., and Carrea, G. (1992) Tetrahedron: Asymmetry 3, 267.
35. Okahata, Y., Fujimoto, Y., and Ijiro, K. (1995) J. Org. Chem. 60, 2244.
¨
36. Adolph, H. W., Maurer, P., Schneider-Bernlohr, H., Sartorius, C., and Zeppezauer, M. (1991) Eur.
J. Biochem. 201, 615.
37. Lee, T. and Jones, J. B. (1996) J. Am. Chem. Soc. 118, 502.

ENZYMATIC CONSTRUCTION OF MULTISTEREOCENTERS
19
38. Large E-value for 1n may suggest that the volume of the large pocket is such that which does not
allow free accommodation of (S)-1n.
39. Fersht, A. R., Shi, J.-P., Knill-Jones, J., Lowe, D. M., Wilkinson, A. J., Blow, D. M. , Brick, P.,
Carter, P., Waye, M. M. Y., and Winter, G. (1985) Nature 314, 235.
40. Gottarelli, G. and Samori, B. (1974) J. Chem. Soc. Perkin Trans. 2, 1462.
41. Seemayer, R. and Schneider, M. P. (1992) Ttrahedron. Asymm. 3, 827.
42. Bianchi, O., Cesti, P., and Golini, P. (1989) Tetrahedron 45, 869.
´
¨
¨
43. Frater, G., Muller, U., and Gunter, W. (1984) Tetrahedron 40, 1269.