Reinvestigation of the absolute stereochemistry of megastigmane glucoside, icariside B5

Article abstract of DOI:10.1248/cpb.58.1399

Icariside B₅ is one of the widely distributed megastigmane glucosides among plant sources. The absolute structure of icariside B ₅ was reinvestigated by chemical conversion from the related compound and the application of the modifie

Full text of DOI:10.1248/cpb.58.1399

October 2010
Note
Chem. Pharm. Bull. 58(10) 1399—1402 (2010)
1399
Reinvestigation of the Absolute Stereochemistry of Megastigmane
Glucoside, Icariside B5
a
c
Katsuyoshi MATSUNAMI, Hideaki OTSUKA, ,a Yoshio TAKEDA,b and Toshio MIYASE
*
a Department of Pharmacognosy, Graduate School of Biomedical Sciences, Hiroshima University; 1–2–3 Kasumi, Minami-
ku, Hiroshima 734–8553, Japan: b Faculty of Pharmacy, Yasuda Women’s University; 6–13–1 Yasuhigashi, Asaminami-ku,
Hiroshima 731–0153, Japan: and c Faculty of Pharmaceutical Sciences, University of Shizuoka; 52–1 Yata, Suruga-ku,
Shizuoka 422–8526, Japan. Received June 8, 2010; accepted July 1, 2010; published online July 27, 2010
Icariside B5 is one of the widely distributed megastigmane glucosides among plant sources. The absolute
structure of icariside B5 was reinvestigated by chemical conversion from the related compound and the applica-
tion of the modified Mosher’s method. As a result, the structure of icariside B5 was revised to be (6S,9S)-6,9-dihy-
droxymegastigman-4-en-3-one 9-O-b-D-glucopyranoside.
Key words icariside B5; dihydrovomifoliol-O-b-D-glucopyranoside; blumenol B; megastigmane glucoside
Blumenol B was isolated from the leaves of Podocarpus (cϭ0.12, MeOH)} were essentially identical to those re-
blumei in 1972.1) The absolute stereochemistry of blumenol ported for PS-1 {Ref.: [a]D20 Ϫ4.4° (cϭ0.80, MeOH)}.3) Thus
1
B was then determined as (6S,9R) by chemical conversion of 2a was elucidated to be identical to PS-1. Second, the H-
related compounds.2) In 1988, two groups independently re- and 13C-NMR spectra of 2a were remeasured in pyridine-d5
ported the isolation of b-D-glucopyranoside of blumenol B, according to the literature for EG-1.4) However, the chemical
named differently as dihydrovomifoliol-O-b-D-glucopyra- shift values in pyridine-d5 of 2a (Tables 1, 2) did not coincide
noside (PS-1) from Pinus sylvestris and icariside B5 (EG-1) with those reported for EG-1.4) These results clearly indi-
from Epimedium grandiflorum var. thunbergianum, respec- cated that PS-1 and EG-1 were not the same compounds. The
tively.3,4) (The former is designated as PS-1 and the latter detailed inspection of 13C chemical shift values measured in
EG-1 in this study for readers’ convenience.) The absolute pyridine-d5 between 2a (ϭPS-1) and the reported values for
structures of the aglycones were determined to be the same EG-14) suggested that EG-1 must be the C-9 epimer of PS-1.
as (6S,9R)-6,9-dihydroxymegastigman-4-en-3-one by com- According to the literature,3,4) the absolute structures of both
parisons of the spectral data of aglycones (PS-1a and EG-1a) PS-1 and EG-1 were determined in a similar manner, e.g., by
obtained by enzymatic hydrolysis of glucosides (PS-1 and comparison of the spectral data for aglycones with those re-
EG-1) with those reported for blumenol B.3,4) Thus com- ported for blumenol B without an adequate chiroptical
pounds PS-1 and EG-1 were concluded to be the same com- method. Therefore we next performed detailed analysis to
pound until now, although the NMR spectra were measured determine the absolute structures by chemical conversion and
in different solvents, e.g., methanol-d4 and pyridine-d5, re- the modified Mosher’s method for PS-1 and EG-1.
spectively.3,4) The modified Mosher’s method is widely used
Enzymatic hydrolysis of 2a (ϭPS-1) afforded an aglycone
for the determination of the absolute stereochemistry of chi- (2b). The (R)- and (S)-a-methoxy-a-trifluoromethylphenyl-
ral secondary alcohols.5) In this study, the absolute configura- acetic acid (MTPA) esters were prepared using the conven-
tions of these compounds (PS-1 and EG-1) were reinvesti- tional procedure (2c and 2d from 2b, see Experimental)
gated using this reliable method.
(Chart 1). The distribution pattern of DdS–R values for 2c and
First, a closely related compound, macarangioside A (2),6) 2d clearly demonstrated that 2b possessed the 9R configura-
was hydrolyzed with mild alkaline hydrolysis (Chart 1) to af- tion (Fig. 1a). In addition, the configuration of glucose was
ford degalloyl-2 (2a), of which the NMR spectra (in determined to be the D-series in HPLC analysis following
methanol-d4) and specific optical rotation value {[a]D24 Ϫ2.2° acid hydrolysis of 2 and derivatization of the liberated glu-
cose. The application of the b-D-glucosylation-induced shift-
trend rule also supported the above result (Table 1).7) The ab-
solute stereochemistry of C-6 was also confirmed based on
the circular dichroism (CD) spectra. The CD spectral data for
2a and 2b were essentially identical to those of blumenol B,
of which the absolute stereochemistry was determined by
chemical and spectroscopic analyses.2) Therefore the struc-
ture of 2a (ϭPS-1, dihydrovomifoliol-O-b-D-glucopyra-
noside) was confirmed unambiguously to be (6S,9R)-6,9-di-
hydroxymegastigman-4-en-3-one 9-O-b-D-glucopyranoside,
e.g., 9-O-b-D-glucopyranoside of blumenol B (Fig. 2).
Next, compound EG-1 was prepared from the closely re-
lated compound corchoionoside C (3), originally isolated by
Yoshikawa et al.,8) which was also isolated from Euodia
a) 0.1 M NaOMe in MeOH, rt, 1 h, b) PtO2/H2, 0 °C, 4 h, c) b-D-glucosidase,
37 °C, 12 h, d) EDC, DMAP, (R) and (S)-MTPAs in CH2Cl2, 35 °C, 12 h.
meliaefolia in our previous study.9) Partial hydrogenation of 3
afforded 3a (Chart 1). The physicochemical data, e.g., NMR
Chart 1
∗ To whom correspondence should be addressed. e-mail: hotsuka@hiroshima-u.ac.jp
© 2010 Pharmaceutical Society of Japan

1400
Vol. 58, No. 10
Table 1. 13C-NMR Spectral Data for Dihydrovomifoliol-O-b-D-glucopyranoside (2a) and Icariside B5 (3a)
2a
(ϭPS-1)
(C5D5N)
2a
(ϭPS-1)
(CD3OD)
3a
3a
(ϭEG-1)
(CD3OD)
2b
(CD3OD)
3b
(CD3OD)
C
(ϭEG-1)
(C5D5N)a)
1
2
3
4
5
6
7
8
9
10
11
12
13
42.3
50.8
197.6
126.4
168.5
78.1
34.8
33.3
75.0
20.3
24.8
24.2
21.6
102.5
75.3
78.7
72.1
78.4
63.2
42.9
51.2
201.1
126.8
171.7
79.4
34.9
33.6
76.3
20.1
24.7
24.2
21.7
102.4
75.2
78.3
71.9
77.9
63.0
43.0
51.2
200.9
126.7
171.7
79.2
35.3
42.3
50.6
198.0
126.1
168.9
78.1
33.7
32.2
76.7
22.0
24.4
24.0
21.6
104.1
75.0
78.2
71.4
78.0
62.6
43.2
51.2
201.1
126.8
171.8
79.5
34.8
33.1
77.9
22.2
24.5
24.1
22.0
104.4
75.4
78.4
71.8
78.4
62.9
43.0
51.2
200.9
126.7
171.8
79.3
35.8
35.4 (Ϫ1.8)b)
69.0 (ϩ7.3)
23.5 (Ϫ3.4)
24.7
24.1
21.8
35.4 (Ϫ2.3)c)
69.4 (ϩ8.5)
23.7 (Ϫ1.5)
24.6
24.1
21.8
1Ј
2Ј
3Ј
4Ј
5Ј
6Ј
a) Added one drop of D2O. b) d2a–2b. c) d3a–3b
.
Table 2. 1H-NMR Spectral Data for Dihydrovomifoliol-O-b-D-glucopyranoside (2a) and Icariside B5 (3a) (600 MHz)
2a (C5D5N)
(ϭPS-1)
2a (CD3OD)
(ϭPS-1)
3a (C5D5N)a)
(ϭEG-1)
3a (CD3OD)
(ϭEG-1)
C
2
2b (CD3OD)
3b (CD3OD)
2.38 (1H, d, 18)
2.77 (1H, d, 18)
6.06 (1H, br s)
2.09 (1H, m)
2.43 (1H, m)
1.71 (1H, m)
2.18 (1H, m)
2.15 (1H, dd, 18, 1) 2.16 (1H, dd, 18, 1)
2.60 (1H, d, 18) 2.58 (1H, d, 18)
5.83 (1H, dq, 1, 1) 5.83 (1H, dq, 1, 1)
1.81 (1H, m)
2.03 (1H, m)
1.49 (1H, m)
1.77 (1H, m)
2.34 (1H, dd, 18, 1)
2.82 (1H, d, 18)
5.96 (1H, br s)
2.26 (1H, ddd, 13, 13, 4)
2.38 (1H, ddd, 13, 13, 5)
2.14 (1H, dd, 18, 1)
2.65 (1H, d, 18)
5.82 (1H, br s)
1.84 (1H, ddd, 14, 13, 4)
2.06 (1H, ddd, 14, 13, 5)
2.16 (1H, dd, 18, 1)
2.59 (1H, d, 18)
5.83 (1H, dq, 1, 1)
1.79 (1H, ddd, 14, 12, 4)
1.98 (1H, ddd, 14, 12, 5)
4
7
1.77 (1H, ddd, 14, 12, 5)
1.95 (1H, ddd, 14, 12, 4)
8
9
1.43 (1H, dddd, 13, 12, 5, 5) 1.75 (1H, dddd, 13, 13, 6, 5) 1.51 (1H, dddd, 13, 13, 7, 5) 1.40 (1H, dddd, 13, 12, 8, 5)
1.65 (1H, dddd, 13, 12, 7, 4) 2.10 (1H, m) 1.78 (1H, dddd, 13, 13, 4, 4) 1.68 (1H, dddd, 13, 12, 5, 4)
4.06 (1H, dq, 6, 6) 3.81 (1H, m)
3.66 (1H, dqd, 7, 6, 5)
1.15 (3H, d, 6)
1.02 (3H, s)
1.10 (3H, s)
2.04 (3H, d, 1)
4.07 (1H, qdd, 6, 6, 6)
1.34 (3H, d, 6)
1.17 (3H, s)
1.19 (3H, s)
2.11 (3H, d, 1)
4.87 (1H, d, 8)
3.93 (1H, dd, 9, 8)
4.15 (1H, dd, 9, 9)
4.11 (1H, dd, 9, 9)
3.86 (1H, ddd, 9, 6, 2)
4.28 (1H, dd, 12, 6)
4.46 (1H, dd, 12, 2)
3.81 (1H, m)
1.24 (3H, d, 6)
1.02 (3H, s)
3.65 (1H, dqd, 8, 6, 5)
1.16 (3H, d, 6)
1.02 (3H, s)
1.09 (3H, s)
2.04 (3H, d, 1)
10 1.27 (3H, d, 6)
11 1.24 (3H, s)
12 1.20 (3H, s)
13 2.13 (3H, br s)
1Ј 4.89 (1H, d, 8)
2Ј 3.96 (1H, m)
3Ј 4.23 (1H, m)
4Ј 4.19 (1H, m)
5Ј 3.94 (1H, m)
6Ј 4.33 (1H, m)
1.17 (3H, d, 6)
1.02 (3H, s)
1.10 (3H, s)
2.04 (3H, d, 1)
4.31 (1H, d, 8)
3.13 (1H, dd, 9, 8)
3.34 (1H, dd, 9, 9)
3.27 (1H, dd, 9, 9)
3.25 (1H, m)
1.09 (3H, s)
2.04 (3H, d, 1)
4.32 (1H, d, 8)
3.14 (1H, dd, 9, 8)
3.33 (1H, m)
3.27 (1H, dd, 9, 9)
3.25 (1H, m)
3.65 (1H, dd, 12, 5)
3.65 (1H, dd, 12, 6)
3.84 (1H, dd, 12, 2)
4.53 (1H, br d, 11) 3.85 (1H, dd, 12, 2)
In parentheses, multiplicities and coupling constants (J in Hz). m: Multiplet or overlapped. Chemical shifts were determined by 1H–1H COSY and HMQC. a) Added one drop
of D2O.
Fig. 1. Modified Mosher’s Analysis
a) For 2b and b) for 3b. Dd values (dS–dR) were shown in ppm.
Fig. 2. Dihydrovomifoliol-O-b-D-glucopyranoside (PS-1) and Revised
Structure of Icariside B5 (EG-1)

October 2010
1401
spectra (in pyridine-d5) (Tables 1, 2) and specific optical rota- ranoside (PS-1), not icariside B5 (EG-1), from M. tanarius.
tion value {[a]D24 Ϫ10.5° (cϭ0.13, MeOH)} of 3a were es-
sentially identical to those reported for EG-1 {Ref.: [a]D25
Experimental
Ϫ12.9° (cϭ0.62, MeOH)},4) which indicated that 3a was
identical to EG-1. The stereochemistry of 3 was determined
by chemical conversion of roseoside, the 9-epimer of 3, e.g.,
NaBH4 reduction of the 9-ketonic functional group following
the PCC oxidation of 9-OH of roseoside.8) Thus EG-1 (ϭ3a)
must have the same stereochemistry as 3. We also performed
the modified Mosher’s analysis to confirm the above result.
The (R)- or (S)-MTPA esters were prepared similarly as de-
scribed above [3c and 3d from the aglycone (3b), see Experi-
mental]. The distribution patterns of DdS–R values for 3c and
3d clearly demonstrated that 3b possessed the 9S configura-
tion (Fig. 1b). The application of the b-D-glucosylation-in-
duced shift-trend rule7) also supported the above result (Table
1 and Experimental). The CD spectral data for 3a and 3b
were also essentially the same as those for blumenol B.
Therefore the structure of icariside B5 (EG-1) (ϭ3a) was re-
vised to be (6S,9S)-6,9-dihydroxymegastigman-4-en-3-one 9-
O-b-D-glucopyranoside, e.g., 9-O-b-D-glucoside of 9-epi-
blumenol B (Fig. 2). Finally, the NMR spectral data (CD3OD)
General Experimental Procedures HPLC was performed on octade-
sylsilanized (ODS) silica gel (Inertsil ODS-3; GL Science, Tokyo, Japan;
Fϭ6 mm, Lϭ250 mm), and the eluate was monitored with UV and refrac-
tive index detectors. Optical rotations were measured on a JASCO P-1030
polarimeter. IR spectra were recorded on a Horiba FT-710 Fourier transform
infrared spectrophotometer and UV spectra on a JASCO V-520 UV/Vis
spectrophotometer. 1H- and 13C-NMR spectra were obtained on JEOL ECA-
600 spectrometers at 600 MHz for 1H and 150 MHz for 13C, with tetram-
ethylsilane as an internal standard. Positive-ion high-resolution (HR)-elec-
trospray ionization (ESI)-time-of-flight (TOF)-MS was recorded on an Ap-
plied Biosystem QSTAR XL spectrometer. CD spectra were obtained on a
JASCO J-720 spectropolarimeter.
Mild Alkaline Hydrolysis of Macarangioside A (2) A mixture of 2
(3.0 mg) and 0.1 M NaOMe in MeOH (1.0 ml) was allowed to stand at room
temperature for 1 h under a N2 atmosphere. Liberation of degalloylated com-
pound 2a was trailed by TLC analysis (CHCl3 : MeOH : H2O, 15 : 6 : 1, Rf
values, 2: 0.32 and 2a: 0.57). The reaction mixture was neutralized with
Amberlite IR-120B (Organo) and 2a (1.2 mg) was purified by preparative
TLC [silica gel (0.25 mm thickness, applied for 9 cm and developed with
CHCl3 : MeOH : H2O, 15 : 6 : 1 for 9 cm)]. The spectral data including chi-
roptical spectra were identical to those of PS-1. 2a (ϭPS-1): Amorphous
powder; [a]D24 Ϫ2.2° (cϭ0.12, MeOH); IR nmax (film) cmϪ1: 3392, 2967,
2925, 1650, 1373, 1076, 1034; UV lmax (MeOH) nm (log e): 238 (3.77);
of icariside B5 (EG-1) isolated from E. thunbergianum var. 13C-NMR (CD3OD) dC: 201.1 (C-3), 171.7 (C-5), 126.8 (C-4), 102.4 (C-1Ј),
grandiflorum4) were found to be identical to those of
79.4 (C-6), 78.3 (C-3Ј), 78.0 (C-5Ј), 76.3 (C-9), 75.2 (C-2Ј), 71.9 (C-4Ј),
63.0 (C-6Ј), 51.2 (C-2), 42.9 (C-1), 34.9 (C-7), 33.6 (C-8), 24.7 (C-11), 24.2
3a.
(C-12), 21.7 (C-13), 20.1 (C-10); CD De (nm): ϩ1.16 (328), Ϫ5.57 (253),
The specific optical rotation value of blumenol B (2b) pre-
pared in this study was [a]D27 ϩ6.2° (cϭ0.06, CH3OH). Thus
icariside B5 (EG-1) was misidentified as a glucoside of blu-
menol B, probably due to the relatively small optical rotation
value of the aglycone [Ref: [a]D25 ϩ19.7° (cϭ0.23,
CH3OH)].4) However, 9-epi-blumenol B (3b) prepared in this
study also showed small optical rotation, [a]D27 ϩ16.6°
(cϭ0.05, CH3OH), suggesting that it was difficult to distin-
guish these epimers by optical rotation values alone at that
time. It is noteworthy that there is a slight but clear differ-
ence between blumenol B (2b) and 9-epi-blumenol B (3b)
ϩ9.76 (221) (cϭ3.09ϫ10Ϫ5 M, MeOH); HR-ESI-TOF-MS (positive-ion
mode) m/z: 411.1980 [MϩNa]ϩ (Calcd for C19H32O8Na: 411.1989).
Enzymatic Hydrolysis of 2a Compound 2a (1.2 mg) were dissolved in
1 ml of 100 mM acetate buffer (pH 5.0) and then hydrolyzed with b-D-glu-
cosidase (Oriental Yeast Co., Ltd., Japan, 10 mg) at 37 °C for 12 h by recip-
rocal shaking. The reaction mixture was extracted twice with an equal
amount of EtOAc. The aglycone (2b) (0.6 mg) was purified by preparative
TLC from the EtOAc layer (CHCl3 : MeOH, 10 : 1, Rf 0.61). (6S,9R)-6,9-
Dihydroxymegastigman-4-en-3-one (blumenol B) (2b): amorphous powder;
[a]D27 ϩ6.2° (cϭ0.06, CH3OH); IR nmax (film) cmϪ1: 3401, 2965, 2927,
1650, 1373, 1127; UV lmax (MeOH) nm (log e): 240 (3.99); 13C- and 1H-
NMR (CD3OD): Tables 1 and 2; CD De (nm): ϩ2.94 (326), Ϫ12.0 (251),
ϩ19.0 (219) (cϭ2.65ϫ10Ϫ5 M, MeOH); HR-ESI-TOF-MS (positive-ion
mode) m/z: 249.1466 [MϩNa]ϩ (Calcd for C13H22O3Na: 249.1461).
1
for H2-8 in the H-NMR spectra, i.e., chemical shifts and
Determination of Sugar Configuration The previously described
method12) was used with slight modifications. Glucosides (2 and 3, 0.2 and
0.3 mg, respectively) were hydrolyzed with 0.2 ml of 1 M HCl at 90 °C for 2 h
and the reaction mixtures were passed through the packed MB-3 ion-ex-
change resin (0.5ϫ5 cm) after washing with EtOAc (0.2 ml). After drying in
coupling patterns of H2-8 geminal protons were inverted [dH
1.43 (1H, dddd, Jϭ13, 12, 5, 5 Hz, H-8a for 2b) and dH 1.65
(1H, dddd, Jϭ13, 12, 7, 4 Hz, H-8b for 2b), dH 1.40 (1H,
dddd, Jϭ13, 12, 8, 5 Hz, H-8a for 3b) and dH 1.68 (1H,
dddd, Jϭ13, 12, 5, 4 Hz, H-8b for 3b)] (Table 2). The de- vacuo, the residues were dissolved in anhydrous pyridine (0.1 ml) and re-
acted with L-cysteine methyl ester (0.5 mg) at 60 °C for 1 h. Then o-
tailed comparison of chemical shifts and coupling patterns of
H2-8 provides important criteria to distinguish them from
each other.
tolylisothiothiocyanate (1.4 mg in 70 ml pyridine) was added to the mixtures
and further incubated at 60 °C for 1 h. The reaction mixtures were directly
analyzed using ODS HPLC [Cosmosil 5C18-ARII (Nacalai Tesque, Kyoto,
Recently, we have reported the empirical rule for the deter-
mination of the absolute stereochemistry of C-9 of the re-
Japan), 4.6ϫ250 mm, 25 °C, 25% CH3CN–50 mM H3PO4, 0.8 ml/min, UV
detection at 250 nm]. The peaks (18.1 min) were identical to the derivative of
authentic D-glucose.
lated megastigmane glucosides, simply by comparing the 13
C
Catalytic Hydrogenation of 3 Compound 3 (8.7 mg) was dissolved in
MeOH (1.0 ml), then 2 mg of Adams’ catalyst (PtO2) was added and stirred
at 0 °C for 4 h under a H2 atmosphere. Hydrogenation of 3 was monitored by
ESI-MS analysis because of the similar Rf values for reactant and product on
TLC analysis. The reaction mixture was purified using HPLC to afford 3a
(1.3 mg) [Inertsil ODS-3 (GL Science), 6ϫ250 mm, 25 °C, 15% CH3CN aq.,
1.6 ml/min, 27.9 min]. (6S,9S)-6,9-Dihydroxymegastigman-4-en-3-one 9-O-
b-D-glucopyranoside, (ϭicariside B5) (3a) (ϭEG-1): amorphous powder;
[a]D24 Ϫ10.5° (cϭ0.13, MeOH); IR nmax (film) cmϪ1: 3388, 2964, 2928,
1650, 1373, 1076, 1028; UV lmax (MeOH) nm (log e): 244 (3.97); 13C- and
1H-NMR (C5D5N and CD3OD): Tables 1 and 2; CD De (nm): ϩ1.92 (325),
Ϫ7.01 (251), ϩ11.6 (219) (cϭ2.82ϫ10Ϫ5 M, MeOH); HR-ESI-TOF-MS
(positive-ion mode) m/z: 411.1988 [MϩNa]ϩ (Calcd for C19H32O8Na:
411.1989).
chemical shift values (in methanol-d4) of C-9 and C-10 and
the anomeric carbon of the attached glucose, i.e., 13C signals
at ca. dC 76 (C-9), 20 (C-10), and 102 (C-1Ј) indicate the 9R
configuration, and ca. dC 78 (C-9), 22 (C-10), and 104 (C-1Ј)
indicate 9S for related compounds. The 13C-NMR spectral
data measured in methanol-d4 for 2a (ϭPS-1) and 3a (ϭEG-
1) also coincided with the above rule. This result further sup-
ports the usefulness of our empirical rule.10)
We reported the isolation of “icariside B5” from the leaves
of Macaranga tanarius as a known compound in a previous
paper, because PS-1 and EG-1 were considered to be the
same compound at that time.11) We should correct that here
Enzymatic Hydrolysis of 3a (؍EG-1) The aglycone (3b) (0.5 mg) was
by stating that we isolated dihydrovomifoliol-O-b-D-glucopy- prepared using a similar procedure from 3a (ϭEG-1) (1.3 mg) as described

1402
Vol. 58, No. 10
above. (TLC; CHCl3 : MeOH, 10 : 1, Rf 0.59). (6S,9S)-6,9-Dihydroxy-
megastigman-4-en-3-one (9-epi-blumenol B) (3b): amorphous powder;
[a]D27 ϩ16.6° (cϭ0.05, CH3OH); IR nmax (film) cmϪ1: 3401, 2965, 2926,
1650, 1373, 1129; UV lmax (MeOH) nm (log e): 240 (4.02); 13C- and 1H-
NMR (CD3OD): Tables 1 and 2; CD De (nm): ϩ2.91 (326), Ϫ14.4 (251),
ϩ20.0 (219) (cϭ2.21ϫ10Ϫ5 M, MeOH); HR-ESI-TOF-MS (positive-ion
mode) m/z: 249.1464 [MϩNa]ϩ (Calcd for C13H22O3Na: 249.1461).
m, H-7a), 1.56 (1H, m, H-7b), 1.55 (1H, m, H-8b), 1.36 (3H, d, Jϭ6 Hz, H3-
10), 1.01 (3H, s, H3-12), 0.96 (3H, s, H3-11); HR-ESI-TOF-MS (positive-ion
mode) m/z: 465.1869 [MϩNa]ϩ (Calcd for C23H29O5F3Na: 465.1859).
(6S,9S)-6,9-Dihydroxymegastigman-4-en-3-one 9-(S)-MTPA ester (3d):
amorphous powder; 1H-NMR (CDCl3, 600 MHz) d: 7.53—7.50 (2H, m, aro-
matic protons), 7.42—7.38 (3H, m, aromatic protons), 5.86 (1H, s, H-4),
5.15 (1H, m, H-9), 3.50 (3H, s, OMe), 2.39 (1H, d, Jϭ18 Hz, H-2a), 2.22
(1H, d, Jϭ18 Hz, H-2b), 1.99 (3H, s, H3-13), 1.86 (1H, m, H-8a), 1.78 (2H,
m, H2-7), 1.64 (1H, m, H-8b), 1.30 (3H, d, Jϭ6 Hz, H3-10), 1.044 (3H, s,
H3-12), 1.037 (3H, s, H3-11); HR-ESI-TOF-MS (positive-ion mode) m/z:
465.1866 [MϩNa]ϩ (Calcd for C23H29O5F3Na: 465.1859).
Preparation of (R)- and (S)-MTPA Esters from 2b and 3b A solution
of 2b (0.3 mg) in 1 ml of dehydrated CH2Cl2 was reacted with (R)-MTPA
(19.1 mg) in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodi-
imide hydrochloride (EDC) (19.2 mg) and 4-N,NЈ-dimethylaminopyridine
(DMAP) (10.2 mg), and stored at 35 °C for 12 h. After the addition of 1 ml
each of H2O and CHCl3, the solution was washed successively with 1 M HCl
(1 ml), NaHCO3–saturated H2O (1 ml), and saturated brine (1 ml). The or-
ganic layer was dried over Na2SO4 and then evaporated under reduced pres-
sure. The residue was purified using preparative TLC [silica gel (0.25 mm
thickness, applied for 9 cm and developed with CHCl3–(CH3)2CO (20 : 1) for
9 cm, Rf 0.50, and eluted with CHCl3–MeOH (2 : 1)] to furnish the (R)-
MTPA ester 2c (0.3 mg, 51%). Through a similar procedure, the (S)-MTPA
ester (2d) (Rf 0.49, 0.3 mg, 51%) was prepared from 2b (0.3 mg) using (S)-
MTPA (17.2 mg), EDC (16.3 mg), and DMAP (8.1 mg). The (R)-MTPA
ester 3c (Rf 0.44, 0.3 mg, 61%) was also prepared from 3b (0.25 mg) using
(R)-MTPA (17.4 mg), EDC (16.5 mg), and DMAP (8.9 mg). The (S)-MTPA
ester 3d (Rf 0.45, 0.2 mg, 41%) was also prepared from 3b (0.25 mg) using
(S)-MTPA (16.3 mg), EDC (21.2 mg), and DMAP (11.0 mg). (6S,9R)-6,9-
Dihydroxymegastigman-4-en-3-one 9-(R)-MTPA ester (2c): amorphous
Acknowledgments This research was supported by Grants-in-Aid for
Scientific Research (C) (Nos. 20590103 and 22590006) from Japan Society
for the Promotion of Science, the Research Foundation for Pharmaceutical
Sciences, and Takeda Science Foundation. We are also grateful for the use of
NMR, UV, and MS spectrometers at the Research Center for Molecular
Medicine, the Analysis Center of Life Science, and the Natural Science Cen-
ter for Basic Research and Development (N-BARD), respectively, Hi-
roshima University.
References
1) Galbraith M. N., Horn D. H. S., Chem. Commun., 1972, 113—114
(1972).
2) Weiss G., Koreeda M., Nakanishi K., Chem. Commun., 1973, 565—
566 (1973).
3) Andersson R., Lennart N. L., Phytochemistry, 27, 559—562 (1988).
4) Miyase T., Ueno A., Takizawa N., Kobayashi H., Oguchi H., Chem.
Pharm. Bull., 36, 2475—2484 (1988).
5) Kusumi T., Ooi T., Ohkubo Y., Yabuuchi T., Bull. Chem. Soc. Jpn., 79,
965—980 (2006).
6) Matsunami K., Takamori I., Shinzato T., Aramoto M., Kondo K.,
Otsuka H., Takeda Y., Chem. Pharm. Bull., 54, 1403—1407 (2006).
7) Kasai R., Suzuno M., Asakawa J., Tanaka O., Tetrahedron Lett., 18,
175—178 (1977).
8) Yoshikawa M., Shimada H., Saka M., Yoshizumi S., Yamahara J., Ma-
tsuda H., Chem. Pharm. Bull., 45, 464—469 (1997).
9) Yamamoto M., Akita T., Koyama Y., Sueyoshi E., Matsunami K.,
Otsuka H., Shinzato T., Takashima A., Aramoto M., Takeda Y., Phyto-
chemistry, 69, 1586—1596 (2008).
10) Matsunami K., Otsuka H., Takeda Y., Chem. Pharm. Bull., 58, 438—
441 (2010).
11) Matsunami K., Otsuka H., Kondo K., Shinzato K., Kawahata M., Ya-
maguchi K., Takeda Y., Phytochemistry, 70, 1277—1285 (2009).
12) Tanaka T., Nakashima T., Ueda T., Tomii K., Kouno I., Chem. Pharm.
Bull., 55, 899—901 (2007).
1
powder; H-NMR (CDCl3, 600 MHz) d: 7.54—7.50 (2H, m, aromatic pro-
tons), 7.43—7.38 (3H, m, aromatic protons), 5.86 (1H, s, H-4), 5.08 (1H, m,
H-9), 3.51 (3H, s, OMe), 2.43 (1H, d, Jϭ18 Hz, H-2a), 2.24 (1H, d,
Jϭ18 Hz, H-2b), 1.98 (3H, s, H3-13), 1.92 (1H, m, H-8a), 1.81 (1H, m, H-
7a), 1.73 (1H, m, H-8b), 1.61 (1H, m, H-7b), 1.30 (3H, d, Jϭ6 Hz, H3-10),
1.05 (3H, s, H3-12), 1.03 (3H, s, H3-11); HR-ESI-TOF-MS (positive-ion
mode) m/z: 465.1849 [MϩNa]ϩ (Calcd for C23H29O5F3Na: 465.1859).
(6S,9R)-6,9-Dihydroxymegastigman-4-en-3-one 9-(S)-MTPA ester (2d):
amorphous powder; 1H-NMR (CDCl3, 600 MHz) d: 7.54—7.50 (2H, m, aro-
matic protons), 7.41—7.37 (3H, m, aromatic protons), 5.83 (1H, s, H-4),
5.07 (1H, m, H-9), 3.57 (3H, s, OMe), 2.36 (1H, d, Jϭ18 Hz, H-2a), 2.19
(1H, d, Jϭ18 Hz, H-2b), 1.93 (3H, s, H3-13), 1.84 (1H, m, H-8a), 1.64 (1H,
m, H-8b), 1.58 (1H, m, H-7a), 1.52 (1H, m, H-7b), 1.36 (3H, d, Jϭ6 Hz, H3-
10), 1.01 (3H, s, H3-12), 0.95 (3H, s, H3-11); HR-ESI-TOF-MS (positive-ion
mode) m/z: 465.1862 [MϩNa]ϩ (Calcd for C23H29O5F3Na: 465.1859).
(6S,9S)-6,9-Dihydroxymegastigman-4-en-3-one 9-(R)-MTPA ester (3c):
amorphous powder; 1H-NMR (CDCl3, 600 MHz) d: 7.54—7.50 (2H, m, aro-
matic protons), 7.42—7.37 (3H, m, aromatic protons), 5.80 (1H, s, H-4),
5.14 (1H, m, H-9), 3.56 (3H, s, OMe), 2.27 (1H, d, Jϭ18 Hz, H-2a), 2.16
(1H, d, Jϭ18 Hz, H-2b), 1.94 (3H, s, H3-13), 1.82 (1H, m, H-8a), 1.60 (1H,