α-Glucosidase inhibition properties of cucurbitane-type triterpene glycosides from the fruits of Momordica charantia

Article abstract of DOI:10.1248/cpb.58.720

Fourteen cucurbitane-type triterpene glycosides (1-14) were isolated from a methanol extract of Momordica charantia fruits, including three new compounds, charantosides A-C (1, 5, 6). Their structures were elucidated by chemical and spectroscopic methods. All isolated compounds were evaluated for α-glucosidase inhibitory effect. Of which, 12 and 13 showed moderate inhibitory activity against α-glucosidase. Whereas, 2, 3, 6-11, and 14 showed weak inhibitory activity, and 1, 4, and 5 were inactive.

Full text of DOI:10.1248/cpb.58.720

720
Note
Chem. Pharm. Bull. 58(5) 720—724 (2010)
Vol. 58, No. 5
a
-Glucosidase Inhibition Properties of Cucurbitane-Type Triterpene
Glycosides from the Fruits of Momordica charantia
Nguyen Xuan NHIEM,a,b Phan Van KIEM,b Chau Van MINH,b Ninh Khac BAN,c Nguyen Xuan CUONG,b
Nguyen Huu TUNG,a,b Le Minh HA,b Do Thi HA,a Bui Huu TAI,a,b Tran Hong QUANG,a,b
a
d
d
,a
*
Tran Minh NGOC, Young-In KWON, Hae-Dong JANG, and Young Ho KIM
a College of Pharmacy, Chungnam National University; Daejeon 305–764, Korea: b Institute of Natural Products
Chemistry, Vietnam Academy of Science and Technology (VAST); c Institute of Ecology and Biological Resources, VAST; 18
Hoang Quoc Viet, Caugiay, Hanoi, Vietnam: and d Department of Food and Nutrition, Hannam University; Daejeon
305–811, Korea. Received November 13, 2009; accepted February 20, 2010; published online February 26, 2010
Fourteen cucurbitane-type triterpene glycosides (1—14) were isolated from a methanol extract of Mo-
mordica charantia fruits, including three new compounds, charantosides A—C (1, 5, 6). Their structures were
elucidated by chemical and spectroscopic methods. All isolated compounds were evaluated for a-glucosidase in-
hibitory effect. Of which, 12 and 13 showed moderate inhibitory activity against a-glucosidase. Whereas, 2, 3,
6—11, and 14 showed weak inhibitory activity, and 1, 4, and 5 were inactive.
Key words Momordica charantia; Cucurbitaceae; cucurbitane-type triterpene glycoside; charantoside; a-glucosidase
The plant Momordica charantia L. (Cucurbitaceae) is
widely cultivated in Asian countries, and the fruit of this
plant has been used as a bitter stomachic, a laxative, an
antidiabetic, and an anthelmintic agent for children in tra-
ditional Chinese, Indian, Vietnamese, and Indonesian medi-
cines. The alcoholic extract of M. charantia fruits has been
reported to inhibit the increase of serum glucose levels in
glucose-loaded rats.1) In addition, many cucurbitane-type
triterpenes and their glycosides were isolated from the
roots,2) fruits,3—13) seeds,14,15) leaves, and vines16—19) of this
plant.
a-Glucosidase inhibitors are oral anti-diabetic drugs used
for diabetes mellitus type 2 that work by preventing the di-
gestion of carbohydrates. Carbohydrates are normally con-
verted into simple sugars, which can be absorbed through the
intestine. Hence, inhibition of a-glucosidase enzyme can sig-
nificantly decrease the postprandial increase in blood glucose
levels after a mixed carbohydrate diet and could be a key
strategy in the control of diabetes mellitus. In fact, a potent
a-glucosidase inhibitor, acarbose, has been shown to effec-
tively reduce the intestinal absorption of sugars in humans.20)
The main drawback of acarbose is its side effects, such as
abdominal distention, flatulence, meteorism, and possibly di-
arrhea.21) Therefore, natural inhibitors from dietary plants
that have minimal side effects and have been shown to
have stronger inhibitory activity against a-glucosidase may
be useful as effective therapies for postprandial hyper-
Fig. 1. Structures of 1—14
glycemia.22) Interestingly, in our search for bioactive metabo- white powder. The absorption bands at 3407 cmꢀ1 and
lites, we have examined the constituents of the methanol ex- 1728 cmꢀ1 in the IR spectrum indicated hydroxyl and car-
tract of M. charantia fruits, and led to the isolation of four- bonyl functionalities, respectively. A [MꢁH]ꢁ pseudomolec-
teen cucurbitane-type triterpene glycosides, including three ular ion peak observed at m/z: 649.4337 (Calcd for C37H61O9:
new compounds named charantosides A—C (1, 5, 6). For 649.4316) in positive HR-electrospray ionization (ESI)-MS,
these reasons, the a-glucosidase inhibitory activity of all and the 13C-NMR spectroscopic data indicated a molecular
these compounds was tested. Herein, we report the isolation, formula of C37H60O9. The 13C-NMR spectrum of 1 showed
structure elucidation, and biological activities of compounds 37 carbon signals, of which were attributed to a methoxy
1—14 (Fig. 1).
group, a sugar moiety, and a triterpene moiety. The distor-
tionless enhancement by polarization transfer (DEPT) exper-
iment permitted differentiation of the 37 carbon signals in-
Results and Discussion
The methanol extract of M. charantia fruits yielded three cluding eight methyl, eight methylene, fifteen methine, and
1
new compounds (1, 5, 6). Compound 1 was obtained as a six quaternary carbons. The H- and 13C-NMR data assign-
© 2010 Pharmaceutical Society of Japan
∗ To whom correspondence should be addressed. e-mail: yhk@cnu.ac.kr

May 2010
721
Table 1. The NMR Spectroscopic Data for 1, 5, and 6
1
5
6
Position
a,c)
a,c)
b,c)
dC
dHa,d) (J in Hz)
dC
dHa,d) (J in Hz)
dC
dHb,d) (J in Hz)
Aglycone
1
23.4
1.41 (d, 5.4)
2.36e)
23.3
1.41 (d, 5.4)
2.38e)
19.1
1.54e)
2
3
4
5
6
7
8
9
28.7
87.4
42.6
149.2
121.5
77.0
51.1
51.2
37.6
22.8
30.2
2.10 (dd, 3.6, 13.8)
3.54 (br s)
—
28.7
87.2
42.5
147.7
123.3
66.7
51.3
50.5
37.3
23.0
30.3
1.88 (m)
3.50 (br s)
—
26.4
82.8
37.8
84.3
131.6
133.0
49.5
50.5
40.3
21.4
29.8
1.75 (m)
3.48 (br s)
—
—
—
—
5.92 (d, 4.2)
3.53 (d, 4.2)
1.55 (br s)
—
5.83 (d, 4.2)
3.98 (d, 4.2)
1.90 (br s)
—
6.20 (d, 9.6)
5.69 (dd, 3.6, 9.6)
2.04 (br s)
—
10
11
12
1.56e)
2.56 (dd, 4.8, 12.6)
2.62 (dd, 6.6, 13.2)
1.56e)
1.55e)
1.70e)
1.68 (m)
1.22 (d, 2.4)
1.67e)
1.30e)
1.69e)
13
14
15
16
17
18
19
20
21
22
46.8
48.9
35.9
28.5
52.1
15.3
209.6
37.3
19.2
40.2
—
—
1.39 (m)
46.7
48.6
35.6
28.7
52.0
15.3
210.0
33.7
19.3
44.6
—
—
1.35 (m)
45.0
47.7
33.3
27.4
50.2
14.6
181.8
36.1
18.6
39.3
—
—
1.35 (m)
2.05e)
2.05e)
1.05e)
1.52 (m)
0.96 (s)
9.72 (s)
2.60 (m)
0.95 (d, 6.6)
1.73e)
1.49 (m)
0.90 (s)
9.80 (s)
1.71 (m)
0.96 (d, 6.6)
1.61 (m)
1.49 (m)
0.92 (s)
—
1.55 (m)
0.91 (d, 6.6)
1.79e)
2.15 (dd, 3.6, 13.8)
2.18 (dd, 3.6, 13.8)
5.51 (m)
5.40 (d, 16.2)
—
1.24 (s)
1.24 (s)
0.92 (s)
1.24 (s)
0.83 (s)
23
24
25
26
27
28
29
30
7-OMe
25-OMe
3-O-All
1ꢂ
125.8
140.9
71.2
30.0
30.0
27.7
26.0
18.7
56.4
5.58e)
66.5
130.4
133.4
25.9
18.1
27.6
4.39 (dt, 2.4, 8.4)
5.13 (d, 8.4)
—
1.66 (s)
1.63 (s)
1.05 (s)
1.27 (s)
0.78 (s)
128.2
136.9
74.9
26.1
25.8
23.6
20.6
19.2
5.58e)
—
1.26 (s)
1.26 (s)
1.09 (s)
1.32 (s)
0.82 (s)
3.26 (s)
26.0
18.7
50.3
3.15 (s)
103.9
72.6
73.2
69.0
75.1
63.3
4.66 (d, 7.8)
3.29 (dd, 3.0, 7.8)
4.03 (t, 3.0)
3.46 (dd, 3.0, 9.6)
3.64 (m)
103.8
72.7
73.1
69.0
75.1
63.2
4.61 (d, 7.8)
3.22 (dd, 3.0, 7.8)
4.10 (t, 3.0)
3.42 (dd, 3.0, 9.6)
3.60 (m)
101.7
71.5
70.4
68.3
73.9
63.2
4.72 (d, 7.8)
3.42 (d, 3.0, 7.8)
4.23 (t, 3.0)
3.57 (dd, 3.0, 9.6)
3.67 (m)
3.75 (dd, 5.4, 10.8)
3.89 (dd, 4.2, 10.8)
2ꢂ
3ꢂ
4ꢂ
5ꢂ
6ꢂ
3.63 (m)
3.80 (d, 8.4)
3.61 (m)
3.78 (d, 8.4)
a) Measured in CD3OD, b) measured in CDCl3, c) 150 MHz. d) 600 MHz. e) Overlapped signals, assignments were done by HMQC, HMBC, 1H–1H COSY, and ROESY ex-
periments.
ments of 1 (Table 1) were based on the heteronuclear multi- ing frame Overhauser effect spectroscopy (ROESY) correla-
ple quantum correlation (HMQC), heteronuclear multiple tions between H-7a (dH 3.53) and biogenetically a-oriented
1
bond correlation (HMBC) (Fig. 2), H–1H correlation spec- C-30 methyl (dH 0.82) as well as OCH3 (dH 3.26) at C-7 and
troscopy (COSY) spectra, and in comparison with published b-oriented H-19 (dH 9.72). The sugar linked to C-3 con-
data.13) The NMR data of 1 were similar to that of 7b,25- firmed by the HMBC correlations between H-3 (dH 3.54) and
dimethoxycucurbita-5,23(E)-dien-19-al 3-O-b-D-allopyra- C-1ꢂ (dC 103.9) as well as those between the anomeric proton
noside7) except for the disappearance of the methoxy group H-1ꢂ (dH 4.66) and C-3 (dC 87.4). Its sugar was identified as
and instead of the hydroxyl group at C-25. The HMBC cross an allose by comparing the NMR data with that of the sugar
peaks from H-7 (dH 3.53) to OCH3 (dC 56.4), C-5 (dC moieties of charantosides II, IV, and VI,9) and by acid hydrol-
149.2), C-6 (dC 121.5), and C-9 (dC 51.2), from H-23 (dH ysis (see Experimental). Based on the above evidence, 1 was
5.58), H-24 (dH 5.58), H3-26 (dH 1.26), and H3-27 (dH 1.26) elucidated as 25-hydroxy-7b-methoxycucurbita-5,23(E)-dien-
to C-25 (dC 71.2) confirmed the methoxy and hydroxyl 19-al 3-O-b-D-allopyranoside, named charantoside A.
groups at C-7 and C-25, respectively, and the olefinic bonds
Compound 5 was also obtained as a white powder. Its
at C-5 and C-23. The methoxy group at C-7 was confirmed molecular formula was determined as C36H58O9 due to a
to be b-oriented as in 7b,25-dimethoxycucurbita-5(6),23(E)- [MꢁH]ꢁ ion in the positive HR-ESI-MS at m/z: 635.4144
dien-19-al 3-O-b-D-allopyranoside13) by the significant rotat- (Calcd for C36H59O9: 635.4159). The DEPT experiment re-

722
Vol. 58, No. 5
Table 2. Rat Intestinal a-Glucosidase Inhibitory Activity of Compounds
1—14
Compounds
% Enzyme inhibition
1
2
3
4
5.50ꢃ1.10
15.85ꢃ3.14
12.86ꢃ3.47
4.37ꢃ1.22
5
5.95ꢃ1.00
6
7
8
9
10
11
12
13
13.61ꢃ2.90
13.14ꢃ2.17
12.50ꢃ1.25
10.88ꢃ0.83
11.51ꢃ2.73
10.11ꢃ3.47
18.63ꢃ2.34
21.71ꢃ2.08
12.98ꢃ2.26
50.96ꢃ2.97
14
Acarbosea)
Percentage of enzyme inhibition at the concentration of 50 mM. a) Acarbose was
used as positive controls. Data presented is the meanꢃS.D. of samples run in triplicate.
Fig. 2. Key HMBC Correlations of 1, 5, and 6
assignments of 6 (Table 1) were based on HMQC, HMBC
1
(Fig. 2), H–1H COSY spectra, and compared to published
solved the 36 carbon signals including seven methyl, eight data.8) The sugar was located at C-3 confirmed by the HMBC
methylene, fifteen methine, and six quaternary resonances. correlations between H-3 (dH 3.48) and C-1ꢂ (dC 101.7), as
The 1H- and 13C-NMR data of 5 (Table 1) was similar to that well as between the anomeric proton H-1ꢂ (dH 4.72) and C-3
of 7b,25-dihydroxycucurbita-5,23(E)-dien-19-al 3-O-b-D-al- (dC 82.8). In addition, the methoxy protons (dH 3.15), H-23
lopyranoside,11) except for the signals due to a side chain. In (dH 5.51), H-24 (dH 5.40), H3-26 (dH 1.24), and H3-27 (dH
the HMBC spectrum, the protons of two methyl groups H-26 1.24) correlated to C-25 (dC 74.9), and proton H-10 (dH
(dH 1.66) and H-27 (dH 1.63) correlated with carbons C-24 2.62) correlated to carbon C-19 (dC 181.8) in the HMBC
(dC 130.4) and C-25 (dC 133.4); proton signal H-23 (dH spectrum, confirming that the methoxy group was at C-25,
4.39) correlated with C-20 (dC 33.7), C-22 (dC 44.6), C-24 the olefinic bond was at C-23, and the carbonyl group was at
(dC 130.4), and C-25 (dC 133.4), indicating that the double C-19. The aglycone of 6 was found to be the same as that of
bond was at C-24 and the hydroxyl group at C-23 in the side karaviloside VI,8) and its sugar moiety was also identified as
chain, similar to that of karavilagenin.8) The stereochemistry D-allose, similar to 1. Base on the above evidence, 6 was
at C-23 of 5 was deduced to be R by comparing the 13C- determined to be 25-methoxy-5b,19-epoxycucurbita-6,23-
NMR chemical shifts of side-chain [dC: 33.7 (C-20), 44.6 dien-19-on 3-O-b-D-allopyranoside, and has been named
(C-22), 66.5 (C-23), 130.4 (C-24), and 133.4 (C-25)] with charantoside C.
those of (23R)-cycloart-24-ene-3b,23-diol [dC 33.0 (C-20),
The other compounds were characterized as karaviloside
44.5 (C-22), 66.1 (C-23), 129.1 (C-24), and 133.8 (C-25)] III (2),7) 7b,25-dihydroxycucurbita-5,23(E)-dien-19-al 3-O-
and (23S)-cycloart-24-ene-3b,23-diol [dC 33.5 (C-20), 44.5 b-D-allopyranosyl (3),11) karaviloside II (4),7) momordicoside
(C-22), 67.3 (C-23), 128.4 (C-24), and 135.6 (C-25)].23) On G (7),12) momordicoside F2 (8),12) goyaglycoside-b (9),5) mo-
the other hand, the HMBC correlations observed between the mordicoside F1 (10),12) momordicoside I (11),12) momordico-
aldehydic proton H-19 (dH 9.80) and C-8 (dC 51.3), C-9 (dC side M (12),10) momordicoside A (13),14) and momordicoside
50.5), and C-11 (dC 23.0), and between H-1ꢂ (d 4.61) and C- C (14),15) (Fig. 1) by comparing their NMR spectroscopic
3 (d 87.2) confirmed that the aldehydic group was located at data with the literature values.
C-9 and the sugar was linked to C-3. Similar to 1, the sugar
The fruit of M. charantia is one of the most popular di-
was identified as D-allose. Accordingly, 5 was identified as etary botanicals for the treatment of diabetes mellitus. The
(23R) 7b,23-dihydroxycucurbita-5,24-dien-19-al 3-O-b-D-al- beneficial effects of the M. charantia extract and the mecha-
lopyranoside, a new compound named charantoside B (5).
nism with which it treats diabetes mellitus have been exten-
Compound 6 was purified as a white powder. The absorp- sively studied.24,25) However, the a-glucosidase inhibitory
tion bands at 3430 cmꢀ1 and 1749 cmꢀ1 in the IR spectrum properties of the extract and its isolated compounds have not
indicated hydroxyl and carbonyl functionalities, respective- been evaluated. Thus, we evaluated the isolated compounds
ly. The [MꢁH]ꢁ pseudomolecular ion observed at m/z: for their inhibitory activity against a-glucosidase (Table 2).
641.4179 (Calcd for C37H59O9: 647.4159) in positive HR- The most polar compound 13 which has two sugar moieties
ESI-MS, and the 13C-NMR spectroscopic data indicated a at the C-3 position and four hydroxyl groups at C-22, C-23,
molecular formula of C37H58O8. The 13C-NMR spectrum of 6 C-24, and C-25 and compound 12, which also has two sugar
exhibited 37 carbon signals that were attributed to a methoxy moieties at C-3 and C-23, showed stronger a-glucosidase in-
group, a sugar moiety, and a triterpene moiety. The DEPT hibitory activity (21.71% and 18.63% inhibition at the con-
experiment revealed differentiation of the 37 carbon signals centration of 50 mM, respectively) than the other compounds.
comprising eight methyl, eight methylene, fourteen methine, Compounds 6—11 all had a 5,19-epoxy bridge, displayed
1
and seven quaternary carbons. The H- and 13C-NMR data weak a-glucosidase inhibitory activity. Other than that,

May 2010
723
changes in the functional groups at C-3, C-19, and C-25 po- given in Table 1; ESI-MS m/z: 635 [MꢁH]ꢁ, 633 [MꢀH]ꢀ; HR-ESI-MS
m/z: 635.4144 [MꢁH]ꢁ (Calcd for C36H59O9, 635.4159).
sitions of 6—11 did not alter the degree of inhibitory activity
against a-glucosidase. Of the three new compounds, 1 and 5
Charantoside C (6): White powder; [a]D25 ꢀ30 (cꢅ0.1, MeOH); IR (KBr)
1
nmax 3430 (OH), 1749 (CꢅO), 1081 (C–O–C) cmꢀ1; H- and 13C-NMR are
displayed very weak inhibitory activity against a-glucosi-
dase. To the best of our knowledge, this is the first time a-
glucosidase inhibitory activity by cucurbitane-type triterpene
glycosides from the fruits of bitter melon has been observed.
It was also noticed that the most polar compound displayed
the strongest enzyme inhibition against a-glucosidase. This
may provide a new point for semi-synthetic chemists to cre-
ate new a-glucosidase inhibitors, which may be useful for
the development of new therapies for the treatment of dia-
betes mellitus.
given in Table 1; ESI-MS m/z: 647 [MꢁH]ꢁ, 645 [MꢀH]ꢀ; HR-ESI-MS
m/z: 647.4179 [MꢁH]ꢁ (Calcd for C37H59O9, 647.4159).
Acid Hydrolysis of 1, 5, and 6 Compounds 1, 5, and 6 (2 mg, each)
were dissolved in HCl 1 N (dioxane–H2O, 1 : 1, 1 ml) and heated for 3 h
at 80 °C in a water bath. The acidic solution was neutralized with silver
carbonate and the solvent thoroughly driven off by a stream of N2 gas
overnight. After extraction with CHCl3, the aqueous layer was concentrated
to dryness using N2 gas. The residue was dissolved in 0.1 ml of dry pyridine,
and then L-cysteine methyl ester hydrochloride in pyridine (0.06 M, 0.1 ml)
was added to the solution. The reaction mixture was heated at 60 °C for 2 h,
and 0.1 ml of trimethylsilylimidazole was added, followed by heating at
60 °C for 1.5 h. The dried product was partitioned with n-hexane and H2O
(0.1 ml of each), and the organic layer was analyzed by gas chromatography
(GC): column SPB-1 (0.25 mmꢄ30 m); detector FID, column temp. 210 °C,
injector temp. 270 °C, detector temp. 300 °C, carrier gas He (2 ml/min).
Under these conditions, standard sugars gave peaks at tR (min) 5.17 and 8.24
for D- and L-allose, respectively. A peak at tR (min) 5.17 corresponding to D-
allose was observed for all three compounds.
Glucosidase Inhibition Assay Rat intestinal acetone powders were pur-
chased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). A slightly modi-
fied version of the rat intestinal a-glucosidase assay method developed by
Kwon et al.26) was used. A total of 1 g of rat-intestinal acetone powder was
suspended in 3 ml of 0.9% saline, and the suspension was sonicated twelve
times for 30 s at 4 °C. After centrifugation (10000ꢄg, 30 min, 4 °C), the re-
sulting supernatant was used for the assay. The sample solution (50 ml) and
0.1 M phosphate buffer (pH 6.9, 100 ml) containing a-glucosidase solution
(1.0 U/ml) was incubated at 25 °C for 10 min. After pre-incubation, 5 mM p-
nitrophenyl-a-D-glucopyranoside solution (50 ml) in 0.1 M phosphate buffer
(pH 6.9) was added to each well at timed intervals. The reaction mixtures
were incubated at 25 °C for 5 min. Before and after incubation, the ab-
sorbance was read at 405 nm by a micro-plate reader Sunrise (Tecan,
Salzburg, Austria) and compared to a control which had 50 ml of buffer solu-
tion in place of the extract. The a-glucosidase inhibitory activity was ex-
pressed as % inhibition and was calculated as follows:
Experimental
General Experimental Procedures The optical rotation was deter-
mined on a Jasco DIP-370 digital polarimeter. Electrospray ionization (ESI)
mass spectra were obtained using an Agilent 1200 LC-MSD Trap spectrom-
eter. HR-ESI mass spectra were obtained using a JEOL JMS-T100LC spec-
trometer. The 1H-NMR (600 MHz) and 13C-NMR (150 MHz) spectra were
recorded on a Jeol ECA 600 spectrometer and TMS was used as an internal
standard. GC was performed on a Shidmazu-2010 instrument. Column chro-
matography (CC) was performed on silica gel (Kieselgel 60, 70—230 mesh
and 230—400 mesh, Merck) and YMC RP-18 resins.
Plant Material The fruits of M. charantia were collected in Vuthu,
Thaibinh province, Vietnam on June, 2009, and identified by Dr. Ninh Khac
Ban. A voucher specimen (INPC MC0609) was deposited at the herbarium
of Institute of Natural Products Chemistry, VAST, Vietnam.
Extraction and Isolation The dried fruits of M. charantia (5.0 kg) were
powdered and extracted with methanol (MeOH) (3ꢄ10 l) to afford the
MeOH extract (400 g), which was then suspended in water (5 l) and then ex-
tracted with chloroform (CHCl3) and ethyl acetate (EtOAc) (3ꢄ5 l each),
yielding CHCl3 (MC1, 250 g), ethyl acetate (MC2, 80 g), and water layer
(MC3) extracts. The CHCl3 fraction (250 g) was passed through a silica gel
column using a solvent mixture of n-hexane and EtOAc with increasing po-
larity as the eluent. Six fractions were collected as follows: MC1A [4 l, n-
hexane–EtOAc (40 : 1, v/v)], MC1B [4 l, n-hexane–EtOAc (20 : 1, v/v)],
MC1C [4 l, n-hexane–EtOAc (10 : 1, v/v)], MC1D [4 l, n-hexane–EtOAc
(5 : 1, v/v)], MC1E [4 l, n-hexane–EtOAc (2.5 : 1, v/v)], and MC1F [1 l,
EtOAc]. Fraction MC1C (20 g) was separated using silica gel CC eluting
with CHCl3–MeOH (20 : 1, v/v) to obtain four sub-fractions, MC1C1—
MC1C4. Sub-fraction MC1C2 (4 g) was further separated by silica gel CC
using CH2Cl2–acetone (5 : 1, v/v) as the eluent to afford four smaller frac-
tions, MC1C2A—MC1C2D. Compounds 2 (80 mg) and 4 (100 mg) were
isolated from fraction MC1C2A (850 mg) by YMC RP-18 CC eluting with
acetone–H2O (3 : 1, v/v). In a similar manner, compounds 1 (30 mg) and 3
(11 mg) were purified from fraction MC1C2C (600 mg). The fraction MC1D
(15 g) was separated using silica gel CC eluting with CHCl3–MeOH (14 : 1,
v/v) to afford four fractions, MC1D1—MC1D4. Fraction MC1D2 (2 g) af-
forded compounds 6 (25 mg) and 8 (13 mg) after subjecting it to silica gel
CC eluting with CHCl3–acetone (2 : 1, v/v) followed by YMC RP-18 CC
with acetone–H2O (2.5 : 1, v/v). Fraction MC1D3 (5 g) was separated into
five sub-fractions, MC1D3A—MC1D3E, by silica gel CC eluting with
CHCl3–n-hexane–MeOH (6 : 1 : 1, v/v). Compounds 5 (40 mg) and 9 (35 mg)
were obtained from sub-fraction MC1D3B after subjecting it to silica gel
CC eluting with CH2Cl2–acetone–H2O (1.4 : 1 : 0.2, v/v/v) followed by YMC
RP-18 CC eluting with MeOH–H2O (2 : 1, v/v). Sub-fraction MC1D3C
(1.5 g) was treated in a similar manner to MC1D3B to furnish com-
pounds 10 (35 mg) and 11 (20 mg). Fraction MC1E (19 g) was separated
into four sub-fractions, MC1E1—MC1E4, by silica gel CC eluting with
CHCl3–MeOH–H2O (5 : 1 : 0.1, v/v/v). Compounds 7 (8 mg) and 12 (15 mg)
was isolated from fraction MC1E2 (2 g) by YMC RP-18 CC eluting with
acetone–H2O (0.8 : 1, v/v). Compounds 13 (17 mg) and 14 (25 mg) were pu-
rified from sub-fraction MC1E3 (1.9 g) using YMC RP-18 CC and eluting
with MeOH–H2O (1.1 : 1, v/v).
Extract
ΔA4C0o5ntrolꢀ ΔA405
⎛
⎜
⎜
⎝
⎞
⎟
⎟
⎠
⎡
⎢
⎤
⎥
% inhibitionꢅ
ꢄ100
Control
[ΔA
]
⎢
⎣
⎥
⎦
405
Statistical Analysis All experiments were performed in triplicate. Data
is presented as the meansꢃS.D. The results were statistically analyzed by
ANOVA and Duncan’s multiple range tests. Statistical significance was ac-
cepted at a level of pꢆ0.05.
Acknowledgments This work was supported by Priority Research Cen-
ters Program through the National Research Foundation of Korea (NRF)
funded by the Ministry of Education, Science and Technology (2009-
0093815). The authors would like to thank the Korean Basic Science Insti-
tute (KBSI) for performing the NMR and MS experiments.
References and Notes
1) Karunanayake E. H., Welihinda J., Sirimanne S. R., Adorai G. S., J.
Ethnopharmacol., 11, 223—231 (1984).
2) Chen J., Tian R., Qiu M., Lu L., Zheng Y., Zhang Z., Phytochemistry,
69, 1043—1048 (2008).
3) Okabe H., Miyahara Y., Yamauchi T., Chem. Pharm. Bull., 30, 3977—
3986 (1982).
4) Okabe H., Miyahara Y., Yamauchi T., Chem. Pharm. Bull., 30, 4334—
4340 (1982).
5) Murakami T., Emoto A., Matsuda H., Yoshikawa M., Chem. Pharm.
Bull., 49, 54—63 (2001).
6) Kimura Y., Akihisa T., Yuasa N., Ukiya M., Suzuki T., Toriyama M.,
Motohashi S., Tokuda H., J. Nat. Prod., 68, 807—809 (2005).
7) Nakamura S., Murakami T., Nakamura J., Kobayashi H., Matsuda H.,
Yoshikawa M., Chem. Pharm. Bull., 54, 1545—1550 (2006).
8) Matsuda H., Nakamura S., Murakami T., Yoshikawa M., Heterocycles,
71, 331—341 (2007).
9) Akihisa T., Higo N., Tokuda H., Ukiya M., Akazawa H., Tochigi Y.,
Kimura Y., Suzuki T., Nishino H., J. Nat. Prod., 70, 1233—1239
(2007).
Charantoside A (1): White powder; [a]D25 ꢀ18 (cꢅ0.1, MeOH); IR (KBr)
1
nmax 3407 (OH), 1728 (CꢅO), 1085 (C–O–C) cmꢀ1; H- and 13C-NMR are
given in Table 1; ESI-MS m/z: 649 [MꢁH]ꢁ, 647 [MꢀH]ꢀ; HR-ESI-MS
m/z: 649.4337 [MꢁH]ꢁ (Calcd for C37H61O9, 649.4316).
Charantoside B (5): White powder; [a]D25 ꢀ14 (cꢅ0.1, MeOH); IR (KBr)
1
nmax 3411 (OH), 1727 (CꢅO), 1086 (C–O–C) cmꢀ1; H- and 13C-NMR are

724
Vol. 58, No. 5
10) Li Q. Y., Chen H. B., Liu Z. M., Wang B., Zhao Y. Y., Magn. Reson.
Chem., 45, 451—456 (2007).
11) Harinantenaina L., Tanaka M., Takaoka S., Oda M., Mogami O.,
Uchida M., Asakawa Y., Chem. Pharm. Bull., 54, 1017—1021 (2006).
12) Okabe H., Miyahara Y., Yamauchi T., Tetrahedron Lett., 23, 77—80
(1982).
C. H., J. Nat. Prod., 69, 1168—1171 (2006).
19) Chen J. C., Liu W. Q., Lu L., Qiu M. H., Zheng Y. T., Yang L. M.,
Zhang X. M., Zhou L., Li Z. R., Phytochemistry, 70, 133—140 (2009).
20) Jenkins D. J., Taylor R. H., Goff D. V., Fielden H., Misiewicz J. J., Sar-
son D. L., Bloom S. R., Alberti K. G., Diabetes, 30, 951—954 (1981).
21) Bischoff H., Puls W., Krause H. P., Schutt H., Thomas G., Diabetes
Res. Clin. Pract., 1, 53—62 (1981).
13) Liu Y., Ali Z., Khan I. A., Planta Med., 74, 1291—1294 (2008).
14) Okabe H., Miyahara Y., Yamauchi T., Miyahara K., Kawasaki T., 22) Kwon Y. I., Vattem D. A., Shetty K., Asia Pac. J. Clin. Nutr., 15, 107—
Chem. Pharm. Bull., 28, 2753—2762 (1980). 118 (2006).
15) Miyahara Y., Okabe. H., Yamauchi T., Chem. Pharm. Bull., 29, 23) Lago J. H. G., Brochini C. B., Roque N. F., Phytochemistry, 60, 333—
1561—1566 (1981). 338 (2002).
16) Yasuda M., Iwamoto M., Okabe H., Yamauchi T., Chem. Pharm. Bull., 24) Garau C., E. C., Phoenix D. A., Singh J., Int. J. Diabetes & Metabo-
32, 2044—2047 (1984). lism, 11, 45—55 (2003).
17) Fatope M. O., Takeda Y., Yamashita H., Okabe H., Yamauchi T., J. Nat. 25) Grover J. K., Yadav S. P., J. Ethnopharmacol., 93, 123—132 (2004).
Prod., 53, 1491—1497 (1990).
26) Kwon Y. I., Vattem D. A., Shetty K., Asia Pac. J. Clin. Nutr., 15, 107—
18) Chang C. I., Chen C. R., Liao Y. W., Cheng H. L., Chen Y. C., Chou
118 (2006).