Action of L-aminoacylase and L-amino acid oxidase on 2-amino-3- methylpent-4-enoic acid [Δ(4)-dehydroisoleucine and alloisoleucine] stereoisomers: An alternative route to a stereochemically pure compound and the application to the synthesis of (R)-2-methylbutan-1-ol

Article abstract of DOI:10.1055/s-1999-3565

The action of L-aminoacylase and L-amino acid oxidase on the stereoisomers of 2-amino-3-methylpent-4-enoic acid was examined, to secure stereochemically pure compounds related to L-alloisoleucine. A scalemic mixture [(2S,3R)- (90.0%), (2S,3S)(3.2%), (2R,3S)- (5.9%), (2R,3R)- (0.9%)] of the N-trifluoroacetyl derivatives of the title compound, which was prepared by an asymmetric Claisen rearrangement, was submitted to Aspergillus L-aminoacylase-catalyzed hydrolysis. Only the (2S,3R)- and (2S,3S)isomers were hydrolyzed to give the corresponding amino acids. The subsequent treatment of these isomers with Crotalus adamanteus L-amino acid oxidase afforded the pure (2S,3R)-isomer, related to L-alloisoleucine, in more than 99% stereochemically pure form. Moreover, the product was converted to L- alloisoleucine and (R)-2-methylbutan-1-ol via a chemoenzymatic procedure.

Full text of DOI:10.1055/s-1999-3565

PAPER
1671
Action of L-Aminoacylase and L-Amino Acid Oxidase on 2-Amino-3-
methylpent-4-enoic acid [D(4)-Dehydroisoleucine and alloisoleucine]
Stereoisomers: An Alternative Route to a Stereochemically Pure Compound
and the Application to the Synthesis of (R)-2-Methylbutan-1-ol
a
a
b
a
Mikio Bakke, Hiromichi Ohta, Uli Kazmaier, Takeshi Sugai *
aDepartment of Chemistry, Keio University, 3–14–1, Hiyoshi, Yokohama 223–8522, Japan
Fax +81(45)5631141; E-mail: sugai@chem.keio.ac.jp
bOrganisch-Chemisches Institut der Universität, Im Neuenheimer Feld 270, D-69120 Heidelberg, Germany
Fax +49(6221)544205; E-mail: ck1@ix.urz.uni-heidelberg.de
Received 12 April 1999; revised 13 May 1999
2
material for isostatine. Moreover, by taking advantage of
Abstract: The action of L-aminoacylase and L-amino acid oxidase
on the stereoisomers of 2-amino-3-methylpent-4-enoic acid was ex-
amined, to secure stereochemically pure compounds related to L-al-
loisoleucine. A scalemic mixture [(2S,3R)- (90.0%), (2S,3S)-
the terminal double bond, 1a is expected to be a potential
precursor of homologous dehydroamino acids by means
of a recently developed alkene metathesis reaction
(
3.2%), (2R,3S)- (5.9%), (2R,3R)- (0.9%)] of the N-trifluoroacetyl (Scheme 1).3
derivatives of the title compound, which was prepared by an asym-
metric Claisen rearrangement, was submitted to Aspergillus L-ami-
noacylase-catalyzed hydrolysis. Only the (2S,3R)- and (2S,3S)-
isomers were hydrolyzed to give the corresponding amino acids.
For a long time, enantiomers of amino acids whose stere-
ochemistry was related to alloisoleucine had been pre-
pared as follows: a combination of Strecker amino acid
The subsequent treatment of these isomers with Crotalus adaman- synthesis and the subsequent L-aminoacylase-catalyzed
teus L-amino acid oxidase afforded the pure (2S,3R)-isomer, related enantiomeric resolution of the corresponding N-acyl de-
to L-alloisoleucine, in more than 99% stereochemically pure form.
Moreover, the product was converted to L-alloisoleucine and (R)-2-
methylbutan-1-ol via a chemoenzymatic procedure.
4
rivatives. The first step of this process, however, yields
all four possible stereoisomers in nearly the same quanti-
ties, and a separation of the diastereomers requires a labo-
rious procedure.
Key words: L-aminoacylase, L-amino acids oxidase, chelation-
controlled enolate Claisen rearrangement, L-alloisoleucine, (R)-2-
methylbutan-1-ol
Among several reports on the efforts for stereoselective
5
synthesis, a decisive clue to a solution of this issue has re-
cently been presented by one of these authors (Kazmai-
6
er). A metal-chelated asymmetric Claisen rearrangement
Among the amino acids with branched side chains, isole-
ucine and alloisoleucine play important roles in biologi-
cally active oligopeptides and depsipeptides such as
7
of N-trifluoroacetylglycine (E)-crotyl ester 4 provides a
highly scalemic stereoisomeric mixture in which (2S,3R)-
b comprises ca. 90% of the compounds. Any desired
enantiomer of 1b is obtained at will, by changing the ab-
solute configuration of amine base as the additive. The
stereoisomers as impurities could be removed by crystal-
lization of the corresponding diastereomeric salt with α-
phenethylamine.
1
1
antibiotics. From the view of synthetic organic chemistry,
it has been well recognized that these amino acids bearing
the nature-made chirality which consists of methyl and
ethyl substituents [CH CH (CH )CH] serve as enantio-
3
2
3
merically enriched building blocks for natural product
syntheses. In this context, the potentiality of 2-amino-3-
8
methylpent-4-enoic acid (1a), a γ, δ-unsaturated amino Current studies of one of the authors (Sugai) prompted us
acid, closely related to alloisoleucine in terms of its struc- to study enzyme actions and substrate specificities on the
tural backbone and stereochemistry, has been gaining in- stereoisomers of 1, hopefully to establish an alternative,
terest. Indeed, it has recently been utilized as a starting facile enzyme-mediated way for the purification of the
one carbon
degradation
hydrogenation
CO2H
OH
NH2
L-alloisoleucine 2a
(R)-3a
CO2H
NH2
2S,3R)-1a
(
olefin metathesis
CO2H
NH2
R
Grubbs' catalyst
R
Scheme 1
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

1
672
M. Bakke et al.
PAPER
(
2S,3R)-isomer. At first, we focused on the beneficial while (2S,3S)-1b was completely hydrolyzed, and these
character of the substrate, the N-trifluoroacetamide group, results indicated that the hydrolysis of the (2S,3R)-isomer
which is essentially required as the chelation-controlled was substantially slower than that of the (2S,3S)-isomer.
Claisen rearrangement to give the (2S*,3R*)-isomer This conclusion was supported by an earlier observation
through the most thermodynamically stable transition by Chibata that the hydrolysis of N-acetyl-L-alloisoleu-
state. During the present study, we became aware that the cine was slower (ca. 1:15) than that of N-acetyl-L-isoleu-
N-trifluoroacetamide group indeed worked very well un- cine (Scheme 2).11a,b
der Aspergillus aminoacylase-catalyzed hydrolysis condi-
The second task was the removal of the obstinate (2S,3S)-
tions for the enantiomeric kinetic resolution of
isomer. As the conventional recrystallization proved inef-
9
phenylglycine derivatives. So far, in some kinetic stud-
fective, we attempted the kinetic resolution of the diaste-
ies, porcine kidney aminoacylase-catalyzed hydrolysis of
reomers based on the difference in the rates of
the N-trifluoroacetamide group has been reported to be
aminoacylase-catalyzed hydrolysis between (2S,3R)-1b
and (2S,3S)-1b. The hydrolysis was stopped at a lower
1
0
very fast, compared with those of the conventional N-
1
1
acetyl and/or N-chloroacetyl groups. The high rate of
(
(
16%) conversion, expecting the most highly reactive
2S,3S)-isomer would be almost hydrolyzed, as men-
hydrolysis especially observed in the case of amino acids
1
0a
with a branched chain suggested to us a more efficient
resolution of enantiomers due to an exhaustive hydrolysis
of the highly reactive enantiomer.
tioned above. In the case of the present experiment, the ra-
tio of the reactivity between (2S,3R)-1b and (2S,3S)-1b
was estimated to be 1:7. To our disappointment, the hy-
Based on the foregoing facts, a scalemic substrate mixture drolysis of (2S,3S)-isomer did not go to completion and a
[
(
(2S,3R)-1b (90.0%), (2S,3S)-1b (3.2%), (2R,3S)-1b small portion was still observed in the fraction of 1b;
5.9%), (2R,3R)-1b (0.9%)] was submitted to Aspergillus 90.6% of (2S,3R), 1.3% of (2S,3S) and 8.1% of the D-iso-
aminoacylase-catalyzed hydrolysis. The reaction pro- mers.
ceeded smoothly, and the conversion reached 83% within
At this stage, we turned our attention to another enzyme,
L-amino acid oxidase from snake venom. Through ex-
tensive studies on the substrate specificity, the relative
rate of oxidation on isoleucine and alloisoleucine stereoi-
somers has been estimated, and L-isoleucine reacts ten to
1
day. The desired (2S,3R)-1b and another L-isomer,
12
(
2S,3S)-1b were preferentially hydrolyzed while two D-
isomers, (2R,3R)-1b and (2R,3S)-1b remained intact. In
this case, the major (2S,3R)-1b isomer was recovered to
some extent in the fractions of N-trifluoroacetamides 1b,
CO
2
R
CO R
2
NHCOCF
2S,3R)-1b
(90.0)
3
NHCOCF3
(
(2S,3S)-1b
(3.2)
a)
O
F
3
COCHN
O
2
CO R
2
CO R
(E)-4
NHCOCF
2R,3S)-1b
5.9)
3
NHCOCF
(2R,3R)-1b
(0.9)
3
(
(
1
b: R = H
1c: R = Me
CO
2
H
CO
2
H
2
CO H
NHCOCF
(2S,3R)-1b
10.9)
3
NH
2
NH2
b)
(2S,3R)-1a
(2S,3S)-1a
+
(
(79.0)
(3.0)
CO
2
H
CO
2
H
NHCOCF
2R,3S)-1b
6.1)
3
NHCOCF
3
(
(2R,3R)-1b
(1.0)
(
a) LHMDS, quinidine, Al(OiPr) / THF, –80 °C → r.t., 1 day, 90% yield. b) Aspergillus L-aminoacylase, CoCl , pH 7.0, 30 °C, 1 day, 18%
3
2
conv., 16% yield. The values in parentheses show the distribution of stereoisomers (total = 100).
Scheme 2
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
An Alternative Route to a Stereochemically Pure Compound of (R)-2-Methyl-1-butanol
1673
one hundred times faster than L-alloisoleucine.12a The tained before was incubated with this enzyme for 2 days
mixture of (2S,3R)-1a and (2S,3S)-1a [96.4:3.6] was sub- at pH 7.2, expecting to avoid the epimerization of the
jected to this enzyme-catalyzed oxidation. The reaction chiral center adjacent to the carbonyl group present in the
worked well to give highly pure (2S,3R)-1a (82%) as the α-oxo acid intermediate. Acid (R)-6 was obtained in 82%
unchanged recovered product. Initially formed 3-methyl- yield and was immediately reduced with LiAlH41 to give
4d
2
-oxopent-4-enoic acid was isomerized in situ to the cor- the desired alcohol (R)-3a in 73% yield. The enantiomeric
1
responding conjugated 2-oxo acid 5, and this could easily excess (ee) was determined to be 96% by an H NMR
be separated from (2S,3R)-1a by anion exchange [Dowex measurement of the corresponding (R)-MTPA ester
1
–
14b,d
x8−50 (OH )] chromatography (Scheme 3).
3b.
Judged from the ee of the starting material (2a), ra-
cemization as low as 1% was observed even after pro-
longed incubation (Scheme 4).
CO2H
CO H
a)
2
NH2
2S,3R)-1a
96.4)
NH2
a)
b)
CO2H
(2S,3R)-1a
(
(2S,3S)-1a
(
(3.6)
NH2
2S,3R)-2a
(
CO2H
CO2H
2
CO H
+
+
MeO CF3
CO
c)
(
R)-MTPA:
NH2
2S,3R)-1a
99.5)
NH2
(2S,3S)-1a
(0.5)
O
OR
(R)-3a: R = H
CO2H
R)-6
(
5
(
d)
(
3
b: R = (R)-MTPA
a) Crotalus L-amino acid oxidase, catalase, tris-maleate buffer (pH
.8), KCl, 30 °C, 3 h, 8% conv., 82% yield. The values in parentheses
7
.
show the distribution of stereoisomers (total = 100).
b)
CO2H
CO2H
Scheme 3
(2S,3R)-1a
O
O
9
5
With the highly pure (2S,3R)-1a in hand, we further em-
barked upon the conversion to (R)-3a. The alcohol (R)-3a
and its precursor, (R)-2-methylbutyric acid (6) are impor-
+
CO2H
CO2H
7
8
1
3,14a–d
tant starting materials for natural products
and liq-
(75)
(25)
1
4e
uid crystals syntheses, and methods for preparation of
1
4
15
(
R)-3a or (R)-6 were extensively studied. Toward this
b)
(2S,3S)-2a
(S)-6
end, the double bond of 1a was saturated; the hydrogena-
tion smoothly proceeded to give L-alloisoleucine
a) PtO , H O, H , r.t., 15 h, 98% yield. b) Crotalus L-amino acid oxi-
2
2
2
[
(2S,3R)-2] without any epimerization of the chiral center
dase, tris-HCl buffer (pH 7.2), O , 33 °C, 1–2 days, 82% yield. for
(R)-6; 91% yield. for (S)-6. c) LAH, Et O, r.t., 4 h, 73% yield. d) MT-
2
PACl, pyridine, r.t., 20 h. The values in parentheses show the distri-
bution of stereoisomers (total = 100).
2
adjacent to the double bond. Further one-carbon degrada-
tion of L-alloisoleucine ultimately might lead to (R)-2-me-
thylbutan-1-ol (3a), which is a mirror image of naturally
occurring (S)-3a, isolated from fusel oil as the yeast-me-
diated metabolite of L-isoleucine.
Scheme 4
Although the oxidative one-carbon degradation of a close-
1
6
So far, Crotalus amino acid oxidase-catalyzed oxidation
of branched chain α-amino acids have been performed on
a small scale, only for the kinetic resolution of racemate.
It is noteworthy that some results on the related substrates
are described here. In the case of unsaturated substrate
ly related substrate mediated by AgO was reported, the
yield was recorded to be as low as 50%. The moderate
yield could be raised by an elaborated enzymatic reaction.
When an L-amino acid oxidase-catalyzed reaction was
performed, concomitantly formed hydrogen peroxide at-
tacked the α-oxo group, under the special conditions that
no catalase was added to the reaction. Subsequently, the
(
2S,3R)-1a, the desired one-carbon degraded product 717
was contaminated with an α, β-unsaturated acid 8 (ca.
5%). This was probably due to the migration of the dou-
2
C–C bond cleavage and the loss of carbon dioxide ulti-
mately led to the lower homologous carboxylic acid.12b In
ble bond to a stable conjugated position at the stage of in-
termediates (from 9 to 5), even under mild reaction
conditions. In contrast, the oxidation was proved to be an
excellent tool for the preparation of (S)-6. Oxidation of L-
isoleucine afforded (S)-6 (>99% e.e., determined at the
stage of (S)-3b] in 91% yield. In this case, the enzyme/
this case, Crotalus adamanteus (rattlesnake) venom
1
2a,c
oxidase
was the best choice among some commercial-
ly available candidates, such as those from Crotalus atrox
and Bothrops atrox, in terms of the the reaction rate of ox-
idation and the availability. L-Alloisoleucine (2a), as ob-
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

1
674
M. Bakke et al.
PAPER
substrate ratio could certainly be suppressed, as the oxida- gel, 1.5 g, hexane–EtOAc (1:0 to 30:1)] to afford 1c (24.7 mg, 87%)
2
3
as colorless oil, [α]
D
= +56.2 (c = 1.13, CHCl3).
tion of L-isoleucine was much faster than that of L-alloiso-
leucine, as mentioned earlier.
IR (film): ν = 3329, 3086, 2977, 1721, 1643, 1548, 1440, 1165,
96, 927, 861, 771, 726, 670, 519, 444, 423 cm .
–
1
9
1
In conclusion, the action and rate of L-aminoacylase and
L-amino acid oxidase to the stereoisomers of 2-amino-3-
methylpent-4-enoic acid derivatives were clarified in a
preparative scale, and a new alternative way to secure
highly isomerically pure 1a was established.
H NMR (270 MHz, CDCl ): δ = 1.10 (d, 3 H, J = 6.9 Hz, CHCH ),
.75 (m, 1 H, CHCH=CH ), 3.79 (s, 3 H, CO CH ), 4.65 (dd, 1 H,
J = 5.0, 8.6 Hz, CHNHCO), 5.10 (dd, 1 H, J = 1.3, 16.8 Hz,
CH=CH ), 5.14 (dd, 1 H, J = 1.3, 10.2 Hz, CH=CH ), 5.67 (ddd, 1
H, J = 7.6, 10.2, 16.8 Hz, CH=CH ), 6.80 (br s, 1 H, NHCO).
3
3
2
2 2 3
2
2
2
1
3
C NMR (CDCl ): δ = 15.3, 40.6, 52.6, 56.2, 115.6 (q, J = 288 Hz),
3
1
17.3, 137.2, 156.6 (q, J = 38.1 Hz), 170.2.
All mps were uncorrected. IR spectra were measured on a FT/IR-
1
13
4
10 spectrometer. H and C NMR spectra were measured at 270
HRMS for C H F NO calc. 239.0768, found 239.0756.
9 12 3 3
MHz on a JEOL JNM EX-270 or 400 MHz on a JEOL JNM α-400
spectrometer unless otherwise stated. Mass spectra were recorded
on a Hitachi M-80B spectrometer at 70 eV. Optical rotations were
recorded on a Jasco DIP 360 polarimeter. Silica gel 60 K070-WH
The ee and de of 1b was estimated by a GC analysis of 1c.19 GC
column, CHROMPACK CHIRASIL-L-VAL (0.25 mm x 25 m);
Oven, 80 °C; Inj, 250 °C; Det, 250 °C]: Rt = 7.3 min [(2R,3R)-1c,
.9%], 7.9 min [(2R,3S)-1c, 5.9%], 8.8 min [(2S,3S)-1c, 3.2%], 9.9
[
0
(
70–230 mesh) of Katayama Chemical Co. was used for column
min [(2S,3R)-1c, 90.0%].
chromatography. Amicon Stirred Ultra-filtration Cell 8200 was
used for ultra-filtration.
L-Amino Acylase-Catalyzed Hydrolysis of 1b. Preparation of
the (2S,3R)-(+)-2-Amino-3-methylpent-4-enoic Acid (1a)
N-TFA-Amino acid 1b (90% isomerically pure, 3.50 g, 15.5 mmol)
(
2S,3R)-(+)-3-Methyl-2-(trifluoroacetylamino)pent-4-enoic
Acid (1b)
6
a
was dissolved in H O by adding aq KOH (ca. 870 mg, 15.5 mmol
This was prepared according to the published procedure with a
slight modification. LHMDS solution was prepared by adding BuLi
2
in H O) until the pH of the mixture exactly became 7.0, and then to-
2
tal volume was adjusted to 145 mL by the addition of H O. To the
(
°
39.8 mL of a 1.71 M solution in hexane, 68.1 mmol, 5.1 eq.) at –20
C under Ar to hexamethyldisilazane (16.1 mL, 79.1 mmol, 5.9 eq.)
in THF (38 mL) and the mixture was stirred for 1 h. A homogeneous
2
−4
mixture was added CoCl ⋅6H O (1.5 mg, 6.3 µmol, 3.9 x 10 eq)
2
2
and L-aminoacylase (L-acylase ACU06514, Aspergillus, Amano
pharmaceutical Co., 1.08 g, 70 mg/mmol of the substrate), and the
mixture was stirred for 1 day at 30 °C. The progress of the hydrol-
ysis was confirmed by a TLC analysis [silica gel, developed with
7
,18
18c
solution of the N-TFA glycine crotyl ester (E)-4 (E/Z = 10/1,
3
.00 g, 13.3 mmol), Al(O-i-Pr) (3.27 g, 16.0 mmol, 1.2 eq.) and
3
quinidine (10.9 g, 33.6 mmol, 2.5 eq.) in THF (135 mL) was stirred
for 10 min under Ar at r.t., and cooled to –78 °C. To the mixture, the
freshly prepared LHMDS solution was added slowly, and the mix-
ture was stirred overnight under Ar at r.t. The progress of the hydro-
lysis was confirmed by a TLC analysis [silica gel, developed with
BuOH–CH CO H–H O (4:1:1) and detected with I2, R : product,
3
2
2
f
0.4; 1b, 0.9]. The enzyme was removed from the mixture (pH 6.7)
2
by using ultra-filtration (Amicon YM10 membrane, 4.5 kg/m ), and
1
the filtrate was concentrated in vacuo. The H NMR spectrum (400
hexane–EtOAc (3:1) and CHCl –MeOH–HOAc (60:6:4), detected
MHz, D O) of the crude product indicated that it was a mixture of
3
2
with I or KMnO solution]. The mixture was cooled at 0 °C again,
N-acylamino acid 1b (δ 4.32, d, 18%) and free amino acid 1a (δ
3.79, d, 82%) by judging from the signals of α-proton. The mixture
was charged on a cation-exchange resin column [Amberlite IR-
2
4
successively the reaction was quenched slowly by sat. aq potassium
hydrogensulfate solution, and the mixture was stirred for 10 min at
r.t. The organic layer was diluted with EtOAc, washed with sat. aq
potassium hydrogensulfate solution. The product was extracted
+
120B (H ), volume: 280 mL], and the column was washed with
H O. The washings were acidified with solid potassium hydrogen-
2
with sat. aq NaHCO The basic solution was subsequently acidified
3.
sulfate to pH 1, and extracted with EtOAc (3 × 20 mL). The com-
with solid potassium hydrogensulfate to pH 1 and extracted with
EtOAc (3 x 20 mL). The combined extracts were dried (Na SO )
bined extracts were dried (Na SO ) and concentrated in vacuo. The
2
4
2
4
residue was chromatographed [silica gel, 15 g, hexane–EtOAc
20:1 to 10:1)] to give the recovery of 1b (570 mg, 16%) as yellow
and concentrated in vacuo. The residue was chromatographed [sili-
(
ca gel, 40 g, hexane–EtOAc (20:1 to 10:1)] to give 1b (2.71 g, 90%)
solid. A small portion of this was converted to the methyl ester and
analyzed by GC as above: the recovery consisted of 60.5% of
(2S,3R)-1b, 34.1% of (2R,3S)-1b and 5.4% of (2R,3R)-1b.
as brownish oil; [α]24 = +70.3 (c = 0.98, CHCl3).
D
IR (film): ν = 3301, 3091, 2981, 1714, 1552, 1456, 1419, 1382,
–
1
1
171, 996, 927, 771, 729, 698, 607, 520, 444, 424 cm .
Then the ion-exchange resin was eluted with 1 M aq NH and the
3
1
H NMR (400 MHz, CDCl ): δ = 1.15 (d, 3 H, J = 7.3 Hz, CHCH ),
eluent was concentrated in vacuo to give a light brownish solution.
Decoloration with Norit, and concentration in vacuo afforded 1a in
a colorless aqueous solution.
3
3
2
.82 (m, 1 H, CHCH=CH ), 4.69 (dd, 1 H, J = 4.8, 8.8 Hz, CHNH-
2
CO), 5.15 (dd, 1 H, J = 1.1, 16.9 Hz, CH=CH ), 5.17 (dd, 1 H,
2
J = 1.1, 10.3 Hz, CH=CH ), 5.71 (ddd, 1 H, J = 7.0, 10.3, 16.9 Hz,
A small portion of this was N-trifluoroacetylated20 and converted to
the corresponding methyl ester 1c in the following manner. A mix-
ture of dried 1a (8.7 mg, 67.4 µmol), small portions of anhyd
2
CH=CH ), 6.89 (d, 1 H, J = 8.8 Hz, NHCO), 10.30 (br s, 1 H,
2
CO H).
2
1
H NMR (400 MHz, D O): δ = 1.05 (d, 3 H, J = 6.8 Hz, CHCH ),
MeOH, Et N, and ethyl trifluoroacetate was stirred under Ar at r.t.
3
2
3
2
5
1
.78 (m, 1 H, CHCH=CH ), 4.28 (d, 1 H, J = 5.9 Hz, CHNHCO),
until the amino acid had completely dissolved. The mixture was
concentrated in vacuo, diluted with EtOAc and washed with sat. aq
2
.07 (dd, 1 H, J = 1.5, 10.5 Hz, CH=CH ), 5.13 (dd, 1 H, J = 1.5,
2
6.3 Hz, CH=CH ), 5.82 (ddd, 1 H, J = 7.3, 10.5, 16.3 Hz,
KHSO . The organic layer was dried (Na SO ) and concentrated in
4
2
4
2
CH=CH2).
vacuo. The residue was chromatographed [silica gel, 1.5 g, hexane–
EtOAc (20:1 to 10:1)] to give 1b (12.6 mg, 78%). Additions of
MeOH and TMS-diazomethane solution, and concentration in vac-
uo afforded methyl ester 1c. Based on the GC analysis of this as
above, the ee and de of (2S,3R)-1a was estimated >99.8% and
92.8% respectively.
1
3
C NMR (CDCl ): δ = 15.0, 40.1, 55.9, 115.5 (q, J = 288 Hz),
3
1
17.7, 136.8, 157.0 (q, J = 38.0 Hz), 174.4.
A small portion of this was converted to the corresponding methyl
ester 1c in the following manner. To the mixture of 1b (27.0 mg,
1
20 µmol) and MeOH, TMS-diazomethane solution was added and
concentrated in vacuo. The residue was chromatographed [silica
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
An Alternative Route to a Stereochemically Pure Compound of (R)-2-Methyl-1-butanol
1675
Lyophilization of (2S,3R)-1a in aqueous solution afforded a color-
idue was washed with H O. The combined filtrate and washings
2
less solid (1.47 g, 73%).
were decolorated with Norit and lyophilized to give 2a (802 mg,
1
98%).
H NMR (400 MHz, D O): δ = 1.12 (d, 3 H, J = 7.3 Hz, CH ), 2.89
2
3
(
1
m, 1 H, CHCH=CH ), 3.77 (d, 1 H, J = 4.0 Hz, CHNH ), 5.26 (dd,
A small portion of this was N-trifluoroacetylated and converted to
the corresponding methyl ester 2b in the same manner as described
above.
2
2
H, J = 1.0, 16.2 Hz, CH=CH ), 5.27 (dd, 1 H, J = 1.0, 10.6 Hz,
2
CH=CH ), 5.86 (ddd, 1 H, J = 6.6, 10.6, 16.2 Hz, CH=CH ).
2
2
1
This was employed for the next step without further purification.
H NMR (270 MHz, CDCl ): δ = 0.88 (d, 3 H, J = 6.9 Hz, CHCH ),
3
3
0
.93 (t, 3 H, J = 7.4 Hz, CH CH ), 1.16 (m, 1 H, CH CH ), 1.40 (m,
2
3
2
3
L-Amino Acid Oxidase-Catalyzed Oxidation of 1a. Preparation
of the (2S,3R)-(+)-2-Amino-3-methylpent-4-enoic Acid (1a)
This experiment was carried out according to the published proce-
dure.1 Amino acid 1a (>99.8% e.e., 92.8% d.e., 1.45 g, 11.2
mmol) was dissolved in tris-maleate buffer (185 mL, 0.05 M, pH
1 H, CH
2
CH
3
), 2.01 (m, 1 H, CHCH
3
), 3.76 (s, 3 H, CO CH ), 4.69
2 3
(dd, 1 H, J = 4.0, 8.9 Hz, CHNHCO), 6.89 (br s, 1 H, NHCO). The
NMR spectrum was identical with an authentic sample derived from
commercially available D-alloisoleucine (Tokyo Chemical Industry
A0212).
2d
7
.8). To this mixture was added KCl (1.40 g, 18.8 mmol) followed
GC analysis of the 2b was compared with an authentic sample de-
rived from DL-alloisoleucine, D-alloisoleucine, DL-isoleucine, and
L-isoleucine,: Rt = 10.1 min [(2S,3R)-2b, 99.5%], 10.8 min
by L-amino acid oxidase (Crotalus adamanteus venom, Sigma
A9253; type I, 0.31 units/mg, 36 mg) and catalase (Tokyo Chemical
Industry C0052, 10 mg). The mixture was vigorously stirred for 3 h
at 30 °C. The progress of the oxidation was confirmed by a TLC
analysis [silica gel, developed with BuOH–CH CO H–H O (4:1:1)
and detected with phosphomolybdic acid–EtOH, R : 1a, 0.4; prod-
uct, 0.5]. Oxidase and catalase were removed from the mixture by
using ultra-filtration (Amicon YM10 membrane, 4.5 kg/m ), and
the filtrate was concentrated in vacuo. The H NMR spectrum of the
21
[
[
(2S,3S)-2b, 0.5%]. No peak on 8.2 min [(2R,3S)-2b] or 9.2 min
(2R,3R)-2b] were observed.
3
2
2
Recrystallization of (2S,3R)-2a from H O–EtOH gave an analytical
2
f
2
4
sample (217 mg, 26.4%) as needles, mp 219–221 °C, [α] +14.7
D
2
2
2
(c = 0.99, H
2
O); lit. : [α]
D
+15.7 (c = 2, H O), authentic sample of
–14.1 (c = 1.00, H O).
2
2
2
4
1
D-alloisoleucine: [α]
D
1
crude product indicated that it was a mixture of amino acid 1a
H NMR (270 MHz, CDCl ): δ = 0.97 (d, 3 H, J = 6.9 Hz, CHCH ),
3 3
(92%) and 2-oxo acid 5 (8%).
0.99 (t, 3 H, J = 7.6 Hz, CH CH ), 1.42 (m, 2 H, CH CH ), 2.11 (m,
2 3 2 3
1
1 H, CHCH ), 3.76 (d, 1 H, J = 3.3 Hz, CHNH ). This NMR spec-
3 2
trum was identical with authentic sample of D-alloisoleucine.
H NMR (400 MHz, D O): δ = 1.11 (d, 3 H, J = 7.0 Hz, 1a, CH ),
2
3
1
2
.77 (s, 3 H, 5, CH CCO), 1.95 (d, 3 H, J = 6.6 Hz, 5, CH CH=C),
3
3
.89 (m, 1 H, 1a, CHCH=CH ), 3.77 (d, 1 H, J = 4.0 Hz, 1a,
2
Oxidase-Catalyzed Oxidation of (2S,3R)-2a: Preparation of (R)-
CHNH ), 5.26 (d, 1 H, J = 17.0 Hz, 1a, CH=CH ), 5.27 (d, 1 H,
J = 11.7 Hz, 1a, CH=CH ), 5.85 (ddd, 1 H, J = 6.8, 11.7, 17.0 Hz,
1
2
2
(
–)-2-Methylbutyric Acid (6)
A solution of L-alloisoleucine 2a (>99.8% e.e., 99.0% d.e., 100 mg,
.76 mmol) in tris-hydrochloride buffer (0.1 M, pH 7.2, 30 mL) was
2
a, CH=CH ), 6.86 (q, 1 H, J = 6.6 Hz, 5, CH CH=C).
2
3
0
The mixture was charged on an anion-exchange resin column
[
washed with H O. The subsequent elution with 1 M HOAc and con-
centration in vacuo gave a mixture of amino acid with some impu-
rities. The mixture was charged on a cation-exchange resin column
[
washed with H O. The subsequent elution with 1 M aq NH and
concentration in vacuo gave an amino acid solution. The solution
was decolorised with Norit and concentrated in vacuo. Lyophiliza-
tion afforded L-(2S,3R)-1a (1.19 g, 82%) as a colorless solid.
filtered through a cellulose acetate filter (Advantec TOYO, DIS-
MIC-25CS020AS). To the mixture was added L-amino acid oxidase
–
Dowex 1X8−50 (OH ), volume:420 mL], and the column was
2
(
2
Crotalus adamanteus venom, as descibed above, 0.53 units/mg,
00 mg), sodium azide (3 mg, 46 µmol) and a small portion of an-
tifoam, and the mixture was vigorously stirred for 1 day at 30 °C. A
stream of oxygen through a cellulose acetate filter (Advantec
TOYO, DISMIC-25JP020) was continuously bubbled into the solu-
+
Amberlite IR-120B (H ), volume:280 mL], and the column was
2
3
3
tion (7 cm /min). To the mixture, L-amino acid oxidase (200 mg)
and sterilized H O (3 mL) were added and the mixture was further
2
stirred for 1 day at 30 °C with oxygen bubbling. The consumption
of the starting material was confirmed by a TLC analysis [silica gel,
developed with BuOH–CH CO H–H O (4:1:1) and detected with
A small portion of this was N-trifluoroacetylated and converted to
the corresponding methyl ester 1c in the same manner as described
above. Based on the GC analysis of this, the ee and de of (2S,3R)-
3
2
2
ninhydrin reagent]. Oxidase was removed from the mixture by us-
2
1
a was estimated >99.8% and 99.0% respectively.
ing ultra filtration (Amicon YM10 membrane, 4.5 kg/m ), and the
1
filtrate was concentrated in vacuo. The H NMR spectrum of the
Recrystallization of (2S,3R)-1a from H O–EtOH gave an analytical
2
2
4
crude product indicated that it was a mixture of amino acid (2S,3R)-
sample (181 mg, 13%) as needles, mp 214–216 °C, [α]D +16.0
c = 1.02, H O). Its de was not changed by recrystallization.
2
a (ca. 10%) and product 6 (ca. 90%). In the present experiment, 2-
(
2
oxo-3-methylpentanoic acid (9) was not detected.
IR (KBr disc): ν = 3422, 3082, 2983, 2944, 2621, 2117, 1611, 1585,
1
H NMR (400 MHz, D O): δ = 0.79 (t, 3 H, J = 7.5 Hz, 6, CH CH ),
2
3
2
1
1
5
511, 1418, 1392, 1353, 1322, 1265, 1244, 1187, 1131, 1075, 1057,
023, 1001, 936, 915, 881, 854, 807, 795, 766, 715, 669, 573, 546,
0
.88 [d, 3 H, J = 7.0 Hz, (2S,3R)-2a, CH CH], 0.91 [t, 3 H, J = 7.0
3
–
1
Hz, (2S,3R)-2a, CH CH ], 0.98 (d, 3 H, J = 7.0 Hz, 6, CH CH), 1.31
3 2 3
04, 465, 402 cm .
[
m, 1 H, (2S,3R)-2a, CH CH ], 1.32 (m, 1 H, 6, CH CH ), 1.37 [m,
3
2
3
2
1
3
C NMR (D O): δ = 13.8, 38.5, 59.1, 118.3, 138.3, 174.2.
2
1 H, (2S,3R)-2a, CH CH ], 1.41 (m, 1 H, 6, CH CH ), 2.02 [m, 1 H,
3 2 3 2
(
2S,3R)-2a, CH CH], 2.14 (m, 1 H, 6, CH CH), 3.69 [d, 1 H, J = 3.3
3 3
Anal. calcd. for C H NO : C, 55.80; H, 8.58; N, 10.85 (129.16).
Found: C, 55.79; H, 8.48; N, 10.92.
6
11
2
Hz, (2S,3R)-2a, CHNH ]. An authentic sample of 9 showed
δ = 2.90 (m, 1 H, CH CH).
2
3
(
2S,3R)-(+)-2-Amino-3-methylpentanoic Acid (2a) (L-Alloiso-
The crude mixture was cooled at 0 °C, successively the mixture was
leucine)
acidified to pH 1 with concentrated hydrochloric acid. Extraction
Unsaturated L-amino acid (2S,3R)-1a (807 mg, 6.25 mmol) was dis-
with Et O (7 x 10 mL) and the combined ethereal extracts were
2
solved in H O (10 mL) and catalytic amount of platinum oxide was
dried (Na SO ). Ethereal solution was concentrated under atmo-
2
2
4
added. After stirring for 15 h under H at r.t., the disappearance of
the starting material was confirmed by a TLC analysis [silica gel,
spheric pressure. The residue was distilled in vacuo to give 6 as col-
2
2
8
orless oil (63.8 mg, 82%), bp 120 °C / 12 mmHg, [α] –18.7
D
1
5h
detected with I ]. The mixture was filtered by a suction and the res-
(c = 0.76, EtOH) [lit. : [α] –20.7 (c = 0.77, EtOH)].
D
2
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

1
676
M. Bakke et al.
PAPER
1
H NMR (270 MHz, CDCl ): δ = 0.95 (t, 3 H, J = 7.4 Hz, CH CH ), References
3
3
2
1
1
.18 (d, 3 H, J = 7.3 Hz, CH CH), 1.50 (m, 1 H, CH CH ), 1.71 (m,
3 3 2
(
1) a) Bodanszky, M.; Perlman, D. Science 1969, 163, 352.
b) Yajima, T.; Grigg, M. A.; Katz, E. Arch. Biochem. Biophys.
972, 151, 565.
H, CH CH ), 2.40 (m, 1 H, CH CH). Its NMR spectrum was iden-
3
2
3
1
5c
tical with that reported previously.
1
Although Bothrops atrox (Sigma A4257; type II, 0.64 units/mg) and
Clotalus atrox (Sigma A5147; type VI, 0.2 units/mg) oxidases gave
the same product under the same condition, the reactions were very
slow compared with that of Crotalus adamanteus.
c) Shoji, J.; Hinoo, H. J. Antibiot. 1975, 28, 60.
d) Helynck, G.; Dubertert, C.; Frechet, D.; Leboul, J. J.
Antibiot. 1998, 51, 512.
(
(
2) a) Kazmaier, U.; Krebs, A. Tetrahedron Lett. 1999, 40, 479.
b) Castejón, P.; Moyano, A.; Pericàs, M. A.; Riera, A. Chem.
Eur. J. 1996, 2, 1001.
3) a) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413.
b) Gibson, S. E.; Gibson, V. C.; Keen, S. P. Chem. Commun.
1997, 1107.
(
R)-(+)-2-Methylbutan-1-ol (3a)
1
4d
Alcohol (3a) was prepared according to the published procedure
in 73% yield as colorless oil, bp 80 °C / 130 mm Hg, [α] +6.48
(
2
1
D
1
4a
22
c = 0.92, CHCl ); [lit. :82% ee, [α]D +5.89 (c = 1.08, CHCl )].
3
3
1
c) Rutjes, F. P. J. T.; Schoemaker, H. E. Tetrahedron Lett.
H NMR (270 MHz, CDCl ): δ = 0.91 (t, 3 H, J = 7.3 Hz, CH CH ),
3
3
2
1997, 38, 677.
0
(
.91 (d, 3 H, J = 6.6 Hz, CH CH), 1.13 (m, 1 H), 1.43 (m, 1 H), 1.56
3
d) Zumpe, F. L.; Kazmaier, U. Synlett 1998, 1199.
4) a) Mori, K.; Iwasawa, H. Tetrahedron 1980, 36, 2209.
b) Senda, S.; Mori, K. Agric. Biol. Chem. 1987, 51, 1379.
m, 1 H), 1.70 (brs, 1 H, OH), 3.43 (dd, 1 H, J = 6.4, 10.4 Hz,
(
CH OH), 3.52 (dd, 1 H, J = 5.8, 10.4 Hz, CH OH). Its NMR spec-
trum was almost identical with that reported previously with an ex-
ception of the chemical shift of OH signal.
2
2
1
4b
(5) a) Oppolzer, W.; Pedrosa, R.; Moretti, R. Tetrahedron Lett.
986, 27, 831.
1
A small portion of this was converted to the corresponding (R)-
MTPA ester 3b. The ee of (R)-3a was estimated by H NMR spec-
trum (270 MHz, CDCl ) of 3b.
δ = 4.08 (dd, 1H, J = 6.6, 10.6 Hz, CH O–MTPA) and 4.25 (dd, 1
H, J = 5.6, 10.6 Hz, CH O–MTPA) for (R)-3a (98%) and δ = 4.17
(
was estimated to be 96%.
b) Tsunoda, T.; Tatsuki, S.; Shiraishi, Y.; Akasaka, M.; Ito, S.
Tetrahedron Lett. 1993, 34, 3297.
c) Kelly, N. M.; Reid, R. G.; Willis, C. L.; Winton, P. L.
Tetrahedron Lett. 1996, 37, 1517.
1
1
4b, d
Judging from the protons at
3
2
2
(
6) a) Kazmaier, U.; Krebs, A. Angew. Chem. Int. Ed. Engl. 1995,
d, 2 H, J = 6.3 Hz, CH O–MTPA) for (S)-3a (2%), the ee of (R)-3a
2
34, 2012.
b) Review: Kazmaier, U. Liebigs Ann./Recueil 1997, 285.
c) Review: Kazmaier, U. Amino Acids 1996, 11, 283.
d) Review: Ito, H.; Taguchi, T. Chem. Soc. Rev. 1999, 38, 43.
7) Bartlett, P. A.; Barstow, J. F. J. Org. Chem. 1982, 47, 3933.
8) Sugai, T. Curr. Org. Chem. 1999, 3, 373.
Oxidase-Catalyzed Oxidation of (2S,3R)-1a
The reaction was carried out in the same manner as described for
(
(
(
(
(
1
2S,3R)-2a. A mixture of 7 and 8 was obtained as colorless oil
70%).
9) Bois-Choussy, M.; Zhu, J. J. Org. Chem. 1998, 63, 5662.
H NMR (270 MHz, CDCl ): δ = 1.30 (t, 3 H, J = 6.9 Hz, 7,
(10) a) Fones, W. S.; Lee, M. J. Biol. Chem. 1954, 210, 227.
b) Fones, W. S.; Lee, M. J. Biol. Chem. 1953, 201, 847.
(11) a) Chibata, I. Nutrition and Foods 1969, 22, 511; Chem. Abstr.
1970, 72, 117761b.
3
CH CH), 1.82 (d, 3 H, 8, J = 7.9 Hz, CH CCO H), 1.83 (s, 3 H, 8,
CH C=CCO H), 3.19 (m, 1 H, 7, CH CH), 5.15 (d, 1 H, J = 10.2
Hz, 7, CH =CH), 5.18 (d, 1 H, J = 17.2 Hz, 7, CH =CH), 5.94 (ddd,
3
3
2
3
2
3
2
2
1
H, J = 7.3, 10.2, 17.2 Hz, 7, CH =CH), 7.01 (m, 1 H, 8). The spec-
b) Chibata, I.; Yamada, S.; Ito, H. Pharmaceutical Studies
1960, 32, 247; Chem. Abstr. 1961, 55, 8522e.
c) Birnbaum, S. M.; Levintow, L.; Kingsley, R. B.;
Greenstein, J. P. J. Biol. Chem. 1952, 194, 455.
d) Greenstein, J. P.; Levintow, L.; Baker, C. G.; White, J. J.
Biol. Chem. 1951, 188, 647.
e) Greenstein, J. P.; Birnbaum, S. M.; Levintow, L. Biochem.
Prepn. 1953, 3, 84.
f) Chenault, H. K.; Dahmer, J.; Whitesides, G. M. J. Am.
Chem. Soc. 1989, 111, 6354.
2
1
7d
trum of 7 (ca. 75%) was identical with that reported previously,
and that of tiglic acid 8 (ca. 25%) was in agreement with the authen-
tic sample.
Oxidase-Catalyzed Oxidation of (2S,3S)-2a: Preparation of (S)-
6
In a similar manner as described for (2S,3R)-2a, the oxidation of
(
2S,3S)-2a (L-isoleucine) was carried out. The reaction completed
within 2 days with a use of less amount of enzyme (1 mg of enzyme
12.5 mg of substrate), and (S)-6 was isolated in 91% yield. This
(
12) a) Greenstein, J. P.; Birnbaum, S. M.; Otey, M. C. J. Biol.
/
Chem. 1953, 204, 307.
b) Parikh, J. R.; Greenstein, J. P.; Winitz, M.; Birnbaum, S. M.
J. Am. Chem. Soc. 1958, 80, 953.
was converted to (S)-3b, and no signals due to (R)-3b were detected
in H NMR analysis.
1
c) Meister, A. Nature 1951, 168, 1119.
Acknowledgement
d) Sun, G.; Slavica, M.; Uretsky, N. J.; Wallace, L. J.; Shams,
G.; Weinstein, D. M.; Miller, J. C.; Miller, D. D. J. Med.
Chem. 1998, 41, 1034.
The authors thank Amano pharmaceutical Co., for the generous gift
of L-aminoacylase. This work was supported by a Grant-in-Aid for
Scientific Research (No. 10125238) from the Ministry of Educati-
on, Science, Sports and Culture, Japan, and T. S. thanks Professor
Isao Kuwajima of Kitasato Institute, for his encouragement. This
work was also supported by the 'Research for the Future' Program
(
(
13) a) Ichihara, A.; Miki, S., Kawagishi, H., Sakamura, S.
Tetrahedron Lett. 1989, 30, 4551.
b) Trost, B. M.; Probst, G. D.; Schoop, A. J. Am. Chem. Soc.
1998, 120, 9228.
c) Martischonok, V.; Melikyan, G. G.; Mineif, A.;
Vostrowsky, O.; Bestmann, H. J. Synthesis 1991, 560.
14) a) Geresh, S.; Valiyaveettil, T. J.; Lavie, Y.; Shani, A.
Tetrahedron: Asymmetry 1998, 9, 89.
(
JSPS-RFTF 97I00302) from the Japan Society for the Promotion of
Science. M. B. thanks to JAPAN Scholarship Foundation, FURU-
KAWA Memorial Funds and Fujiwara Scholarship Fund for
Schlorships.
b) Mori, K.; Takikawa, H. Liebigs Ann. Chem. 1991, 497.
c) Hjalmarsson, M.; Hˆgberg, H.-E. Acta. Chem. Scand. B
1985, 39, 793.
d) Brown, H. C.; Naik, R. G.; Bakshi, R. K.; Pyun, C.;
Singaram, B. J. Org. Chem. 1985, 50, 5586.
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
An Alternative Route to a Stereochemically Pure Compound of (R)-2-Methyl-1-butanol
1677
e) Tius, M. A.; Gu, X.-q.; Truesdell, J. W.; Savariar, S.;
Crooker, P. P. Synthesis 1988, 36.
m) Meyers, A. I.; Knaus, G., Kamata, K., Ford, M. E. J. Am.
Chem. Soc. 1976, 98, 567.
(
15) a) Saburi, M.; Shao, L.; Sakurai, T.; Uchida, Y. Tetrahedron
Lett. 1992, 33, 7877.
(16) Clarke, T. G.; Hampson, N. A.; Lee, J. B.; Morley, J. R.;
Scanlon, B. J. Chem. Soc. (C) 1970, 815.
(17) a) Lane, J. F.; Roberts, J. D.; Young, W. G. J. Am. Chem. Soc.
1944, 66, 543.
b) Yoshikawa, K.; Murata, M.; Yamamoto, N.; Inoguchi, K.;
Achiwa, K. Chem. Pharm. Bull. 1992, 40, 1072.
c) Chu, K. S.; Negrete, G. R.; Konopelski, J. P.; Lakner, F. J.;
Woo, N.-T.; Olmstead, M. M. J. Am. Chem. Soc. 1992, 114,
b) Sato, F.; Iijima, S.; Sato, M. J. Chem. Soc., Chem. Commun.
1981, 180.
1
800.
c) Oppolzer, W.; Kündig, E. P.; Bishop, P. M.; Perret, C.
Tetrahedron Lett. 1982, 23, 3901.
d) Bloodworth, A. J.; Curtis, R. J.; Spencer, M. D.; Tallant, N.
A., Tetrahedron 1993, 49, 2729.
d) Brown, J. M.; Brunner, H.; Leitner, W.; Rose, M.
Tetrahedron: Asymmetry 1991, 2, 331.
e) Engel, K.-H. Tetrahedron: Asymmetry 1991, 2, 165.
f) Genet, J. P.; Mallart, S.; Pinel, C.; Juge, S.; Laffitte, J. A.
Tetrahedron: Asymmetry 1991, 2, 43.
g) Holmberg, E.; Holmquist, M.; Hedenström, E.; Berglund,
P. Norin, T.; Högberg, H.-E.; Hult, K. Appl. Microbiol.
Biotechnol. 1991, 35, 572.
(18) a) Weygand, F.; Leising, E. Chem. Ber. 1954, 87, 248.
b) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978,
17, 522.
c) Brouwer, H.; Stothers, J. B. Can. J. Chem. 1972, 50, 1361.
(19) Parr, W.; Yang, C.; Bayer, E.; Gil-Av, E. J. Chromatog. Sci.
1970, 8, 591.
h) Negrete, G. R.; Konopelski, J. P. Tetrahedron: Asymmetry
1
991, 2, 105.
(20) Curphey, T. J. J. Org. Chem. 1979, 44, 2805.
(21) Kazmaier, U. Angew. Chem., Int. Ed. Engl. 1994, 33, 998.
(22) Huffman, W. A. H.; Ingersoll, A. W. J. Am. Chem. Soc. 1951,
73, 3366.
i) Yamashita, M.; Naoi, M.; Imoto, H.; Oshikawa, T. Bull.
Chem. Soc. Jpn. 1989, 62, 942.
j) Ohta, T.; Takaya, H.; Kitamura, M.; Nagai, K.; Noyori, R.
J. Org. Chem. 1987, 52, 3174.
k) Quoqiang, L.; Hjalmarsson, M.; Högberg, H.-E.; Jernstedt,
K.; Norin, T. Acta. Chem. Scand. B 1984, 38, 795.
l) Evans, D. A.; Takacs, J. M. Tetrahedron Lett. 1980, 21,
Article Identifier:
1437-210X,E;1999,0,09,1671,1677,ftx,en;F12499SS.pdf
4233.
Synthesis 1999, No. 9, 1671–1677 ISSN 0039-7881 © Thieme Stuttgart · New York