Asymmetric synthesis of (+)-febrifugine and (+)-isofebrifugine using yeast reduction

Article abstract of DOI:10.1016/S0040-4020(00)01109-1

The antimalarial agents febrifugine (D-1) and isofebrifugine (D-2) were synthesized from chiral 3-piperidinol (D-4), which was asymmetrically prepared by the yeast reduction of 3-piperidone derivatives (DL-3), with dynamic optical resolution.

Full text of DOI:10.1016/S0040-4020(00)01109-1

TETRAHEDRON
Pergamon
Tetrahedron 57 (2001) 1213±1218
Asymmetric synthesis of (1)-febrifugine and (1)-isofebrifugine
using yeast reduction
Yasuo Takeuchi,p Kumiko Azuma, Kentaro Takakura, Hitoshi Abe, Hye-Sook Kim,
Yusuke Wataya and Takashi Harayama
Faculty of Pharmaceutical Sciences, Okayama University, Tsushima-naka 1-1-1, Okayama 700-8530, Japan
Received 16 October 2000; accepted 20 November 2000
AbstractÐThe antimalarial agents febrifugine (d-1) and isofebrifugine (d-2) were synthesized from chiral 3-piperidinol (d-4), which was
asymmetrically prepared by the yeast reduction of 3-piperidone derivatives (dl-3), with dynamic optical resolution. q 2001 Elsevier Science
Ltd. All rights reserved.
1. Introduction
was found.4a,b We previously synthesized the racemates of
d-1 and d-2 via the unusual Claisen rearrangement and
highly diastereoselective reduction.5 In this paper, we
describe the asymmetric synthesis of d-1 and d-2 based
on our synthetic method.
Malaria is the world's most important tropical parasitic
disease and there are an estimated 300±500 million cases
of malaria each year. The emergence of multi-drug resistant
strains of the parasite is exacerbating the situation. Febrifu-
gine (d-1), which was isolated from Dichroa febrifuga and
Hydrangea umbellata along with isofebrifugine (d-2),1a±c is
a well-known candidate antimalarial agent. The plane struc-
ture2a of d-1 and d-2 was ®rst proposed in 1950. Subse-
quently, their relative2b and absolute2c structures were
proposed, based on Baker's synthetic work.3a±c The relative
con®guration2d of d-1 was corrected in 1973 and then the
absolute structures2e of d-1 and d-2 were corrected in 1999,
as shown in Fig. 1. These repeated errors and corrections
have caused much confusion in the study of the relationship
between the structure and antimalarial activity of febrifu-
gine derivatives, and have resulted in sacri®ce of many test
animals. However, investigation of the antimalarial activity
of febrifugine derivatives is beginning anew, since it was
reported that d-1 had higher activity than clinically used
antimalarial drugs, and a derivative more potent than d-1
2. Results and discussion
Considering green chemistry, we selected the yeast reduc-
tion of 3-piperidone derivatives (dl-3)5 protected by a Cbz
group as a key asymmetric reaction. The reaction of dl-3
with baker's yeast and sucrose in EtOH and water afforded
cis-3-piperidinol (d-4) with high selectivity. We determined
that the absolute con®guration of l-3 is certainly 2R from
the optical rotation of a known compound (l-6)6 via reduc-
tion7 of the tosylhydrazone (l-5) with NaBH3CN and ZnCl2
from l-3. The reduction of dl-3 with NaBH4 is known to
give the corresponding cis alcohol (dl-4) with complete
diastereoselectivity in high yield.5 Reduction of l-3 with
1
NaBH4 gave l-4, which had the same H-NMR as d-4 and
the opposite optical rotation. These results showed that the
absolute con®guration of d-4 must be 2S,3S (Scheme 1).
Fortunately, a mixture of dl-3 and dl-4 produced four
completely separate peaks on HPLC with a chiral column.
Using HPLC, we examined the conditions that optimized
the yield and enantiomer excess of d-4 having the same
con®guration as d-2 in the yeast reduction of dl-3. Under
the conditions shown in Fig. 2, dl-3 afforded the best results
for d-4 on HPLC. This supported the prediction that the
yeast reduction of dl-3 would consume only d-3 according
to Prelog's rule8 to give d-4 and l-3 by optical resolution.
We think that the decrease in the enantiomer excess of l-3
observed with time might be due to the effect of yeast
empimerase, since the epimerization of l-3 was not
Figure 1. The ®nally corrected structure of d-1 and d-2.
Keywords: asymmetric synthesis; optical resolution; reduction; antimalarial
compounds.
p
Corresponding author. Tel.: 181-86-2517964; fax: 181-86-2517963;
e-mail: take@pharm.okayama-u.ac.jp
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)01109-1

1214
Y. Takeuchi et al. / Tetrahedron 57 (2001) 1213±1218
Scheme 1.
observed in the reaction of l-3, even after 168 h, unless
there was yeast present. Successful isolation of these
compounds was achieved by column chromatography on
SiO2 to give l-3 in 31% yield (93% ee) and d-4 in 41%
yield (97% ee) (method A). From a practical perspective, we
estimate that this reaction has an E value9 of 215.
room temperature for 4 h in MeOH gave dl-3 in 67% yield.
Stirring a mixture of dl-3, baker's yeast, sucrose, and
K2CO3 in ethanol and water at room temperature for 90 h
afforded d-4 in 62% yield (97% ee) along with l-3 in 14%
yield (14% ee) (method B). However, the yield of d-4 in this
reaction varied markedly with small changes in the activity
of the yeast, temperature, or K2CO3 concentration. We are
further investigating the reaction conditions that produce a
steady yield. Whereas yield of d-4 is lower in method A than
it is in Method B, repeating the reaction in method A using
dl-3 which was easily obtained by treatment of l-3 with the
base, certainly gave d-4 in high yield.
The dynamic optical resolution using a substrate that was
enolized easily demonstrated that the product yield in the
yeast reduction would exceed 50%, maintaining high
enantioselectivity.10 If we could cause the epimerization
of l-3 quickly enough to maintain the reductive activity of
the yeast, we thought that reductive dynamic optical resolu-
tion would occur, since epimerization of l-3 with K2CO3 at
Isofebrifugine (d-2) was prepared by improving our
Figure 2. Yeast reduction of dl-3. A stirring mixture of dl-3 (0.4 mmol), yeast (1.0 g), and sucrose (1.0 g) in EtOH (1 ml) and water (20 ml) at 258C was
measured by HPLC with a chiral column.

Y. Takeuchi et al. / Tetrahedron 57 (2001) 1213±1218
1215
Scheme 2. Reagents and conditions: (a) NBS, MeCN, rt, 0.5 h, 87%; (b) (i) t-BuOK, THF, 08C, 0.25 h; (ii) NBS, MeOH, rt, 1 h, 90%; (c) (i) H1, MeCN, rt, 1 h;
(ii) 4(3H)-quinazolinone, K2CO3, DMF, rt, 1 h, 75%; (d) H2, 20%-Pd(OH)2/C, MeOH, rt, 4 h, 62%; (e) (i) H2O, 808C, 15 min; (ii) H1, 73%.
Figure 3. Isomerization of d-1 and d-2. A mixture of d-1 or d-2 in EtOH at 808C was measured by HPLC.
previous method (Scheme 2). The intramolecular
bromoetheri®cation of d-4 using NBS afforded hexahydro-
furo[3,2-b]pyridine (7) in 87% yield. The HPLC data indi-
cated that this was a 3:1 mixture of the diastereomeric
isomers. To improve the yield and reproducibility of prepar-
ing Z-protected isofebrifugine (9) from 7, we designed a
2-methoxy derivative (8). The new intermediate (8) could
be prepared by a 2-step reaction in high yield (90%) as a 4:1
mixture of the diastereomeric isomers. The steps consisted
of dehydrobromination using potassium tert-butoxide and
bromoetheri®cation using NBS and methanol. Deacetaliza-
tion of 8 followed by a coupling reaction with 4(3H)-quina-
zolinone afforded 9 in 75% yield. The hydrogenolysis of 9
gave isofebrifugine (d-2) in 62% yield as a crystalline solid.
The melting point,1a 1H NMR data,4a and optical rotation1a
of d-2 agreed with reported values for the natural product.
produced the largest ratio (2:1) of d-1 to d-2. However,
even after 8 h, no isomerization of d-2 was observed in
10% HCl aq4b (Table 1). Based on these results, we isolated
pure d-1 as the hydrochloride salt, and its physicochemical
properties and spectral data were identical to those reported
for the natural product.1c Herein, we prepared d-1 asym-
metrically from dl-3 with a clearly determined absolute
con®guration. Our results strongly support the corrected
structures of d-1 and d-2.
Although both d-1 and d-2 exhibited potent antimalarial
activities in vitro as high as that of drugs in clinical use,
the hydrochloride of d-1 was half as potent as the free base
(d-1) and the activity of d-2 was 1/10 that of d-1 (Table 2).
While screening the antimalarial activity of febrifugine
derivatives, we have been troubled by the decrease in
activity with time. We are certain that this results from
In our previous method of synthesizing dl-febrifugine
(dl-1), the large differences in the melting point and solu-
bility of dl-1 and dl-2 facilitated the isolation of dl-1.5
Isomerization was naturally observed in the reaction of
d-1 at 808C in EtOH, and the reaction reached equilibrium
within 1 h to give a 1:1 mixture of d-1 and d-2 (Fig. 3).
However, we could not isolate pure d-1 in a satisfactory
yield. Although isomerization of the congeners of d-2 is
known,11,12 there are no reports on d-2 itself. We examined
the transformation of d-2 to d-1 in various solvents, includ-
ing toluene, AcOEt, EtOH, DMSO, water, and 10% HCl aq.
An equilibrium mixture of d-1 and d-2 was observed in
HPLC by heating d-2 at 808C in most solvents. The time
to reach equilibrium and the ratio of d-1 and d-2 changed
with the solvent. Of the solvents tested, heating in water
Table 1. Isomerization of d-1 and d-2 in heating solvent at 808C
Solvent
Time until
reaching
equilibrium
(h)
Product
ratio
(d-1:d-2)a
Toluene
AcOEt
EtOH
DMSO
H2O
10% HCl aq.
4
4
1
1:5
1:1
1:1
2:1
2:1
d-2 only
4
0.25
b
±
a Determined by HPLC.
b Even after 8 h, no isomerization was observed.

1216
Y. Takeuchi et al. / Tetrahedron 57 (2001) 1213±1218
Table 2. Antimalarial activity and toxic selectivity of d-1
Compound
FM3A EC50, mM
P. falciparum EC50, nM
Toxic selctivity
Febrifugine (d-1)
Febrifugine´2HCl (d-1´2HCl)
Isofebrifugine (d-2)
Quinine
Chloroquine
Pyrimethamine
Artemisinin
0.18
0.14
0.74
100
32
0.91
1.8
9.0
110
18
198
78
82
909
1778
120
1266
0.12
10
1.0
7.9
isomerization. We propose that the hydrochloric salt be
administered to screen the antimalarial activity of febrifu-
gine derivatives.
room temperature; eluent, hexane/isopropyl alcoholˆ37:3;
¯ow rateˆ1.5 ml/min; wavelength, 254 nm; tRˆ6.6 and
in the literature.5 1H NMR (60 MHz, CDCl3, rotomers) d
1.40±1.90 (4H, m), 2.20±2.80 (3H, m), 2.41 (1H, s), 3.64±
4.09 (2H, m), 4.34±4.68 (1H, m), 4.87±5.81 (3H, m), 5.10
(2H, s), 7.33 (5H, s). HRMS (FAB): m/z calcd for
C16H22NO3 (M111) 276.1599, found 276.1571.
1
8.2 min. The H NMR spectrum agreed with that reported
3. Experimental
3.1. General
3.1.2. Yeast reduction (method B). A mixture of dl-3
(1.00 g, 3.66 mmol), K2CO3 (3.00 g), yeast (10 g), and
sucrose (30 g) in EtOH (30 ml) and water (300 ml) was
stirred at 158C for 90 h. The reaction mixture was treated
in the same way as described above. The eluant (AcOEt/
hexaneˆ1:7) gave l-3 (0.14 g, 14%, 14% ee based on
HPLC with a chiral column) as a colorless oil. The second
eluant (AcOEt/ hexaneˆ1:3) gave d-4 (0.62 g, 62%, 97% ee
based on HPLC with a chiral column) as a viscous colorless
oil.
Melting points were determined on a Yanagimoto micro
melting point apparatus and are uncorrected. IR spectra
were recorded on a JASCO A-102 spectrometer. Mass spec-
tra (MS) were recorded on a VG-70SE spectrometer. 1H and
13C NMR spectra were run on a Hitachi R-1500, a JASCO
MY 60FT or a Varian VXR-500 spectrometer. Optical rota-
tions were measured on a JASCO DIP-1000 spectrometer.
Analytical HPLC was performed with a Shimadzu SPD-6A
instrument on a chiral phase column, Chiralcel OJ (Daicel)
or a silica gel column, Chemcosorb 5Si-U (Chemco). Merck
silica gel 60 (230±400 mesh) and Wako activated alumina
(300 mesh) were employed for column chromatography.
Extracts were dried over anhydrous MgSO4. The yeast
was used a dried yeast (super camellia, Nisshin ¯our
milling).
3.1.3. (2R,3R)-Benzyl 2-allyl-3-hydroxypiperidinecar-
boxylate (l-4). To solution of l-3 (0.42 g, 1.54 mmol,
93% ee) in MeOH (4 ml), NaBH4 (0.02 g, 0.53 mmol) was
portionwised at 08C with stirring. The mixture was stirred at
the same temperature for 1 h, made acidic with aqueous
10% HCl solution (10 ml), extracted with AcOEt
(2£20 ml). The AcOEt layer was washed with saturated
aqueous KHCO3 solution (20 ml) and brine (30 ml), dried,
and concentrated. The residue was subjected to column
chromatography (AcOEt/hexaneˆ1:3) to give l-4 (0.31 g,
3.1.1. Yeast reduction (method A). A mixture of benzyl
2-allyl-3-oxopiperidinecarboxylate (dl-3,5 1.64 g, 6.00
mmol), yeast (15 g), and sucrose (15 g) in EtOH (15 ml)
and water (300 ml) was stirred at 258C for 24 h. AcOEt
(800 ml) was added to the mixture and stirred at room
temperature for 10 min. The AcOEt layer separated by
centrifugation (£1000g, 5 min) was dried, ®ltered, and
concentrated. MeCN (50 ml) was added to the residue and
the precipitates were ®ltered through a membrane ®lter
(0.5 mm). The ®ltrate was concentrated and the residue
was subjected to column chromatography (silica gel). The
®rst eluant (AcOEt/hexaneˆ1:7) gave (R)-benzyl 2-allyl-3-
oxopiperidinecarboxylate (l-3, 0.51 g, 31%, 93% ee based
73%, 91% ee based on HPLC) as viscous oil,
1
[a]27 ˆ272.88 (c 1.00, EtOH). The H NMR spectrum of
D
this compound agreed with d-4.
3.1.4. (R)-Benzyl 2-allyl-3-[(4-methylphenylsulfonyl)-
hydrazono]piperidinecarboxylate (l-5). A mixture of
l-3 (0.98 g, 3.59 mmol, 90% ee) and TsNHNH2 (0.80 g,
4.30 mmol) in EtOH (12 ml) was stirred at the room
temperature for 24 h. The resulting precipitates were ®ltered
off, washed with EtOH, recrystallized from a mixture of
AcOEt and hexane to give l-5 (1.01 g, 64%), mp 179±
on HPLC) as a colorless oil, [a]28 ˆ240.28 (c 1.00, EtOH).
D
HPLC: column, Chiralcel OJ; column temperature, room
temperature; eluent, hexane/isopropyl alcoholˆ37:3; ¯ow
rateˆ1.5 ml/min; wavelength, 254 nm; tRˆ9.1 and
1818C. [a]29 ˆ214.68 (c 1.01, CHCl3). IR (KBr): 3220,
D
1700 cm21
.
1H NMR (500 MHz, CDCl3, rotomers) d
1
1
11.9 min. The H NMR spectrum agreed with dl-3. H
NMR (60 MHz, CDCl3, rotomers) d 1.71±2.17 (2H, m),
2.37±2.60 (4H, m), 2.98±3.44 (1H, m), 3.96±4.27 (1H,
m), 4.53±5.80 (4H, m), 5.14 (2H, s), 7.35 (5H, s). HRMS
(FAB): m/z calcd for C16H20NO3 (M111) 274.1443,
found 274.1423. The second eluant (AcOEt/hexaneˆ1:3)
gave (2S,3S)-benzyl 2-allyl-3-hydroxypiperidinecarboxyl-
ate (d-4, 0.68 g, 41%, 97% ee based on HPLC with a chiral
1.64±1.67 (1H, m), 1.83 (1H, brs), 2.08±2.55 (4H, m),
2.42 (3H, s), 3.10±3.20 (1H, m), 3.80±3.95 (1H, m),
4.70±4.86 (2H, m), 4.99±5.23 (3H, m), 5.35±5.47 (1H,
m), 7.29±7.35 (7H, m), 7.82 (2H, d, Jˆ6.5 Hz), 7.80±
8.00 (1H, m). Anal Calcd for C23H27N3O4S: C, 62.57; H,
6.16; N, 9.52. Found: C, 62.45; H, 6.24; N, 9.38.
column) as a viscous colorless oil, [a]20 ˆ178.58 (c 1.00,
3.1.5. (S)-Benzyl 2-allylpiperidinecarboxylate (l-6). To a
suspension of l-5 (0.31 g, 0.70 mmol) in MeOH (4 ml)
D
EtOH). HPLC: column, Chiralcel OJ; column temperature,

Y. Takeuchi et al. / Tetrahedron 57 (2001) 1213±1218
1217
added to the solution of NaBH3CN (0.09 g, 1.43 mmol) and
ZnCl2 (0.10 g, 0.73 mmol) in MeOH (3 ml). The mixture
was heated at re¯ux for 3 h under Ar, poured into aqueous
0.1N NaOH solution (50 ml), and extracted with AcOEt
(2£50 ml). The AcOEt layer was washed with brine
(50 ml), dried, and concentrated. The residue was subjected
to column chromatography (AcOET/hexaneˆ1:7) to give
l-6 (0.12 g, 66%, 97% ee based on HPLC) as colorless
oil, [a]27 ˆ251.58 (c 0.82, CHCl3) {lit6 [a]25 ˆ243.08
calcd for C16H19BrNO3 (M12MeOH) 352.0548, found
352.0531.
3.1.8. (3aS,7aS)-Benzyl 2-hydroxy-2-[4-oxo-3(4H)-quin-
azolinyl]-methyl]hexahydrofuro-[3,2-b]pyridine-4(2H)-
carboxylate (d-9). To a solution of d-8 (1.86 g, 4.84 mmol)
in MeCN (10 ml) added aqueous 10% HCl solution (10 ml)
and the mixture was stirred at room temperature for 1 h.
The mixture was poured into water (50 ml) and extracted
with AcOEt (2£100 ml). The combined organic layers were
washed with saturated KHCO3 solution (100 ml) and brine
(100 ml), dried, ®ltered and concentrated. A mixture of
the residue, 4(3H)-quinazolinone (0.71 g, 4.86 mmol),
anhydrous K2CO3 (0.80 g, 5.79 mmol) in dry DMF
(10 ml) was stirred at room temperature for 1 h. The
mixture was poured into brine (100 ml) and extracted with
AcOEt (2£100 ml). The combined organic layers were
washed with brine (2£100 ml), dried, ®ltered and con-
centrated. The residue was subjected to column
chromatography (Al2O3; AcOEt/isopropyl alcoholˆ4:1)
to give d-9 (1.58 g, 75%) as amorphous solid.
D
D
1
(c 4.1, CHCl3)}. The H NMR spectrum agreed with that
reported in the literature.6 1H NMR (500 MHz, CDCl3,
rotomers) d 1.40±1.42 (1H, m), 1.53±1.64 (5H, m), 2.22±
2.28 (1H, m), 2.39±2.45 (1H, m), 2.85 (1H, t, Jˆ13.3 Hz),
4.05 (1H, br d, Jˆ12.5 Hz), 4.37 (1H, br s), 5.01 (2H, dd,
Jˆ17.0, 10.0 Hz), 5.12 (2H, Abq, Jˆ12.5, 10.5 Hz), 5.70
(1H, br s), 7.29±7.35 (5H, m). HRMS (FAB): m/z calcd for
C16H22NO2 (M111) 260.1651, found 260.1648.
3.1.6. (3aS,7aS)-Benzyl 2-(bromomethyl)hexahydrofuro-
[3,2-b]pyridine-4(2H)-carboxylate (d-7). To a solution of
d-4 (1.64 g, 5.96 mmol, 97% ee) in MeCN (20 ml) was
added NBS (1.17 g, 6.57 mmol). The mixture was stirred
at room temperature for 0.5 h. The mixture was poured
into 10% aqueous Na2S2O3 solution (100 ml) and extracted
with AcOEt (2£100 ml). The combined AcOEt layers were
washed with saturated aqueous KHCO3 solution (100 ml)
and brine (100 ml), dried, ®ltered and concentrated. The
residue was subjected to column chromatography (SiO2;
hexane/AcOEtˆ3:1) to give d-7 (1.83 g, 87%, 54% de,
[a]26 ˆ140.68 (c 1.01, EtOH). The IR and 1H NMR
D
spectrum agreed with dl-9.5 1H NMR (60 MHz, CDCl3,
rotomers) d 1.37±3.12 (7H, m), 3.66±4.50 (5H, m), 5.09
(2H, s), 7.32 (5H, s), 7.46±7.90 (3H, m), 8.13±8.34 (2H, m).
HRMS (FAB): m/z calcd for C24H26N3O5 436.1872, found
436.1893.
3.1.9. d-Isofebrifugine (d-2). A mixture of d-9 (1.70 g,
3.90 mmol), 20% Pd(OH)2/C (0.09 g) in absolute MeOH
(20 ml) was stirred at room temperature for 4 h under
H2 gas. After ®ltration the solvent was removed. The
residue was crystallized from MeOH and recrystallization
from toluene to give d-2 (0.73 g, 62%) as colorless
needles, mp 129±1308C (lit1a mp 129±1308C)
[a]27 ˆ1124.38 (c 0.50, CHCl3) {lit1a [a]25 ˆ11318 (c
96% ee from HPLC) as light yellow oil, [a]27 ˆ156.38
D
(c 1.00, EtOH). HPLC: column, Chiralcel OJ; column
temperature, room temperature; eluent, hexane/isopropyl
alcoholˆ37:3; ¯ow rateˆ1.5 ml/min; wavelength, 254 nm;
tRˆ9.0, 9.9, 13.5 min. The IR and 1H NMR spectrum
agreed with dl-7.5b 1H NMR (60 MHz, CDCl3, rotomers)
d 1.25±2.60 (6H, m), 2.70±3.20 (1H, m), 3.30±3.55 (2H,
m), 3.80±5.00 (4H, m), 5.15 (2H, s), 7.36 (5H, s). HRMS
(FAB): m/z calcd for C16H21BrNO3 (M111) 354.0705,
found 354.0662.
D
D
0.35, CHCl3)}. Anal. Calcd for C16H19N3O3: C, 63.77; H,
6.36; N, 13.94. Found: C, 63.48; H, 6.43; N, 13.65. The IR
and 1H NMR spectrum agreed with the natural
compound4a,11 and dl-2.5
3.1.7. (3aS,7aS)-Benzyl 2-bromomethyl-2-methoxyhexa-
hydrofuro[3,2-b]pyridine-4(2H)-carboxylate (d-8). To a
solution of d-7 (2.23 g, 6.30 mmol) in dry THF (10 ml) was
added KOBut (1.41 g, 12.6 mmol). The mixture was stirred
at room temperature for 15 min. To the mixture was added a
solution of NBS (1.34 g, 7.53 mmol) in absolute MeOH
(20 ml). The mixture was stirred at room temperature for
1 h. The mixture was poured into 10% aqueous Na2S2O3
solution (100 ml) and extracted with AcOEt (2£100 ml).
The combined AcOEt layers were washed with saturated
aqueous KHCO3 solution (100 ml) and brine (100 ml),
dried, ®ltered and concentrated. The residue was subjected
to column chromatography (SiO2; AcOEt/hexaneˆ1:3) to
give d-8 (2.17 g, 90%, 60% de from HPLC) as colorless
3.1.10. d-Febrifugine dihydrochloride (d-1´2HCl). A
solution of d-2 (0.34 g, 1.13 mmol) in H2O (10 ml) was
heated at 808C. To the mixture added aqueous 10% HCl
solution (2 ml) and the solvent was removed. The azeotropic
treatment and the recrystallization of the residue from EtOH
gave the dihydrochloride of d-1 (0.31 g, 73%) as colorless
powder, mp 218±2198C (dec) (lit1c mp 223±2258C (dec)).
[a]29 ˆ113.38 (c 1.01, H2O) {lit1c [a]31 ˆ112.88 (c 0.85,
D
D
H2O)}. Anal. Calcd for C16H19N3O3´2HCl: C, 51.35; H,
5.66; N, 11.23. Found: C, 51.07; H, 5.77; N, 11.12. Free
base: mp 140±1418C (from AcOEt) (lit1a mp 139±1408C).
[a]26 ˆ115.78 (c 0.30, MeOH) {lit4a [a]27 ˆ113.08 (c
D
D
0.65, MeOH)}. Anal. Calcd for C16H19N3O3: C, 63.77; H,
6.36; N, 13.94. Found: C, 63.57; H, 6.64; N, 13.91. The IR
and 1H NMR spectrum of free base agreed with the natural
compound4a,11 and dl-1.5
oil, [a]24 ˆ113.88 (c 1.00, EtOH). HPLC: column,
D
Chiralcel OJ; column temperature, room temperature;
eluent, hexane/isopropyl alcoholˆ37:3; ¯ow rateˆ1.5 ml/
1
min; wavelength, 254 nm; tRˆ5.8 and 7.8 min. H NMR
(60 MHz, CDCl3, rotomers): dˆ1.37±2.45 (7H, m), 3.25
and 3.30 (total 3H, each s), 3.43±3.54 (2H, m), 3.74±5.00
(3H, m), 5.15 (2H, s), 7.35 (5H, s). MS (FAB, positive ion
mode) m/z 352 (M12MeOH, 88), 354 (M12MeOH12,
100), 384 (M1, 50), 386 (M112, 28). HRMS (FAB): m/z
3.2. Antimalarial activity
Assays and evaluation of antimalarial activities were carried
out according to the methods described previously.13

1218
Y. Takeuchi et al. / Tetrahedron 57 (2001) 1213±1218
J. Org. Chem. 1953, 18, 153. (c) Baker, B. R.; McEvoy, F. J.
J. Org. Chem. 1955, 20, 136.
Acknowledgements
4. (a) Murata, K.; Takano, F.; Fushiya, S.; Oshima, Y.
J. Nat. Prod. 1998, 61, 729. (b) Takaya, Y.; Tasaka, H.;
Chiba, T.; Uwai, K.; Tanitsu, M.; Kim, H.-S.; Wataya, Y.;
Miura, M.; Takeshita, M.; Oshima, Y. J. Med. Chem. 1999,
42, 3163.
We are grateful to the SC-NMR Laboratory of Okayama
University for the use of the facilities.
References
5. Takeuchi, Y.; Hattori, M.; Abe, H.; Harayama, T. Synthesis
1999, 1814.
1. (a) Koep¯i, J. B.; Mead, J. F.; Brockman Jr., J. A. J. Am.
Chem. Soc. 1947, 69, 1837. (b) Koep¯i, J. B.; Mead, J. F.;
Brockman Jr., J. A. J. Am. Chem. Soc. 1949, 71, 1048.
(c) Ablondi, F.; Gordon, S.; Morton, II, I. J.; Williams, J. H.
J. Org. Chem. 1952, 17, 14.
6. Takahata, H.; Kubota, M.; Takahashi, S.; Momose, T.
Tetrahedron: Asymmetry 1996, 7, 3047.
7. Sucrow, W.; Hedier, R.; Joraschek, N. Chem. Ber. 1988, 121,
1039.
8. Prelog, V. Pure Appl. Chem. 1964, 9, 119.
9. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am.
Chem. Soc. 1982, 104, 7294.
2. (a) Koep¯i, J. B.; Brockman Jr., J. A.; Moffat, J. J. Am. Chem.
Soc. 1950, 72, 3323. (b) Baker, B. R.; McEvoy, F. J.; Schaub,
R. E.; Joseph, J. P.; Williams, J. H. J. Org. Chem. 1953, 18,
178. (c) Hill, R. K.; Edwards, A. G. Chem. Ind. 1962, 858.
(d) Barringer Jr., D. F.; Berkelhammer, G.; Wayne, R. S.
J. Org. Chem. 1973, 38, 1937. (e) Kobayashi, S.; Ueno, M.;
Suzuki, R.; Ishitani, H.; Kim, H.-S.; Wataya, Y. J. Org. Chem.
1999, 64, 6833.
10. Frater, G. Helv. Chim. Acta 1980, 63, 1383.
11. Uesato, S.; Kuroda, Y.; Kato, M.; Fujiwara, Y.; Hase, Y.;
Fujita, T. Chem. Pharm. Bull. 1998, 46, 1.
12. Barringer Jr., D. F.; Berkelhammer, G.; Carter, S. D.;
Goldman, L.; Lanzilotti, A. E. J. Org. Chem. 1973, 38, 1933.
13. Takeuchi, Y.; Tokuda, S.; Takagi, T.; Koike, M.; Abe, H.;
Harayama, T.; Shibata, Y.; Kim, H.-S.; Wataya, Y.
Heterocycles 1999, 51, 1869.
3. (a) Baker, B. R.; Schaub, R. E.; McEvoy, F. J.; Williams,
J. H. J. Org. Chem. 1952, 17, 132. (b) Baker, B. R.; McEvoy,
F. J.; Schaub, R. E.; Joseph, J. P.; Williams, J. H.