Medicinal flowers. XXXIII.1 anti-hyperlipidemic and anti-hyperglycemic effects of chakasaponins I-III and structure of chakasaponin IV from flower buds of Chinese tea plant (Camellia sinensis)

Article abstract of DOI:10.1248/cpb.60.674

Effects of principal saponins, chakasaponins I-III, from the flower buds of Camellia sinensis cultivated in Fujian province, China on plasma triglyceride (TG) and glucose levels in olive oil or sucrose-loaded mice were examined. Chakasaponins I-III at 50 and 100 mg/kg significantly inhibited increases in plasma TG and glucose levels. Furthermore, they prevented gastric emptying, suggesting that the former inhibitory effect is partly dependent on the inhibition of gastric emptying. In addition, the chemical structure of a new acylated oleanane-type triterpene oligoglycoside, chakasaponin IV, was elucidated on the basis of chemical and physicochemical evidence.

Full text of DOI:10.1248/cpb.60.674

674
Note
Chem. Pharm. Bull. 60(5) 674–680 (2012)
Vol. 60, No. 5
Medicinal Flowers. XXXIII.1)
Anti-hyperlipidemic and Anti-hyperglycemic Effects of Chakasaponins
I–III and Structure of Chakasaponin IV from Flower Buds of Chinese
Tea Plant (Camellia sinensis)
’
Hisashi Matsuda, Makoto Hamao, Seikou Nakamura, Haruka Kon i, Megumi Murata, and
Masayuki Yoshikawa*
Kyoto Pharmaceutical University; Misasagi, Yamashina-ku, Kyoto 607–8412, Japan.
Received December 28, 2011; accepted February 10, 2012; published online February 27, 2012
Effects of principal saponins, chakasaponins I–III, from the flower buds of Camellia sinensis cultivated
in Fujian province, China on plasma triglyceride (TG) and glucose levels in olive oil or sucrose-loaded mice
were examined. Chakasaponins I–III at 50 and 100mg/kg significantly inhibited increases in plasma TG and
glucose levels. Furthermore, they prevented gastric emptying, suggesting that the former inhibitory effect is
partly dependent on the inhibition of gastric emptying. In addition, the chemical structure of a new acylated
oleanane-type triterpene oligoglycoside, chakasaponin IV, was elucidated on the basis of chemical and physi-
cochemical evidence.
Key words chakasaponin; Camellia sinensis; medicinal flower; tea plant; anti-hyperlipidemic effect; anti-
hyperglydemic effect
Tea flowers, the flower buds of Camellia (C.) sinensis (L.) triterpene oligoglycoside termed chakasaponin IV (1) from tea
O. KUNTZE, are used as a food garnish in Japanese-style dishes flowers cultivated in Fujian province are described.
(e.g., botebote-cha in Shimane prefecture) or drinks in Japan
Isolation of Constituents The methanolic extract (31.1%
(e.g., hanaban-cha in Shimane and Kouchi prefectures or from the dried flower buds of C. sinensis cultivated in Fujian
batabata-cha in Niigata prefecture). However, the chemical province, China) was partitioned into an EtOAc–H2O (1:1,
constituents and biological activities of tea flowers have not v/v) mixture to furnish an EtOAc-soluble fraction (3.2%)
yet been characterized. Previously, we reported the isolation and aqueous layer. The aqueous layer was further extracted
and structural elucidation of floratheasaponins A–C from with n-butanol (n-BuOH) to give n-BuOH (16.4%)- and H2O
Japanese tea flowers, the flower buds of Japanese C. sinensis.2) (11.5%)-soluble fractions as reported previously.5) The EtOAc-
Floratheasaponins A–C were found to have inhibitory effects and n-BuOH-soluble fractions were subjected to normal-phase
on serum triglyceride (TG) levels in olive oil-loaded mice,2) and reversed-phase silica gel column chromatographies and
on ethanol- and indomethacin-induced gastric mucosal le- repeated HPLC to give chakasaponin IV (1, 0.10%) together
sions in rats,3) and on serum glucose levels in sucrose-loaded with chakasaponins I (2, 0.49%),5) II (3, 0.67%),5) and III (4,
rats.3) From the flower buds of Chinese tea plants cultivated in 0.013%),5) and kaempferol 3-O-β-D-glucopyranosyl(1
Anhui province, floratheasaponins D–I were isolated together L-rhamnopyranosyl(1
6)-β-D-glucopyranoside (5, 0.018%).5,7)
with floratheasaponins A–C.4) The principal saponins, flora-
Structure of Chakasaponin IV Chakasaponin IV (1)
→3)-α-
→
°
theasaponins A–F, were found to exhibit inhibitory effects was isolated as colorless fine crystals with mp. 204.0–206.0 C
on the release of β-hexosaminidase from RBL-2H3 cells.4) In (from CHCl3–MeOH) and exhibited a positive optical rotation
the course of our studies on the saponin compositions of tea ([α]D25 +24.6 in MeOH). Its IR spectrum showed absorption
°
flowers cultivated in various provinces of China, we found bands at 1718 and 1686cm−1 ascribable to two carbonyl moi-
that floratheasaponins were not contained in the tea flowers eties and broad bands at 3450 and 1078cm−1 suggestive of an
cultivated in Fujian province, from which chakasaponins I oligoglycoside structure. Positive-ion fast atom bombardment
(2), II (3), III (4), V, and VI were isolated and their structures (FAB)-MS revealed a quasimolecular ion peak at m/z 1197
were determined.5,6) Previously, we reported inhibitory ef- (M+Na)+, high-resolution (HR) FAB-MS of which revealed
1
fects of chakasaponins I–III (2–4) against pancreatic lipase the molecular formula C57H90O25. The H-NMR (pyridine-d5)
(in vitro),5) but no anti-hyperglycemic effects and anti-hyper and 13C-NMR (Table 1) spectra of 1, which were assigned
glycemic effects have been reported. In this paper, we de- based on results of various NMR experiments,8) showed sig-
scribe the effects of principal saponins (2–4) and a principal nals assignable to an aglycone {seven methyls [δ 0.80, 1.00,
flavonol glycoside, kaempferol 3-O-β-D-glucopyranosyl(1
→3)- 1.05, 1.11, 1.24, 1.26, 1.85 (3H each, all s, H3-25, 26, 29, 24, 23,
α-L-rhamnopyranosyl(1 6)-β-D-glucopyranoside (5), on in- 30, 27)], a methylene and four methines with an oxygen func-
→
creases in plasma triglyceride and glucose levels in olive oil- tion [δ 3.25 (1H, dd, J=3.8, 13.2Hz, H-3), 3.60, 3.76 (1H each,
or sucrose-loaded mice together with their effects on gastric both d, J=10.6Hz, H2-28), 4.22 (1H, d, J=4.0Hz, H-15), 4.56
emptying in mice, rat intestinal α-glucosidase (in vitro) and (1H, d, J=4.0Hz, H-16), 6.16 (1H, dd, J=5.5, 11.8Hz, H-22)],
rat lens aldose reductase (in vitro). In addition, continuing on an olefinic proton [δ 5.44 (1H, brs, H-12)], and a tigloyl moi-
the structural elucidation of chakasaponins,5,6) the isolation ety [δ 1.41, (3H, d, J=7.0Hz, H3-22-O-Tig-4), 1.76 (3H, brs,
and structural elucidation of a new acylated oleanane-type H3-22-O-Tig-5), 6.83 (1H, dq-like, H-22-O-Tig-3)] and four
glycopyranosyl parts [δ 4.92 (1H, d, J=7.2Hz, H-1′), 5.01 (1H,
d, J=7.4Hz, H-1″″), 5.74 (1H, d, J=7.6Hz, H-1″), 5.77 (1H, d,
The authors declare no conflict of interest.
*To whom correspondence should be addressed. e-mail: myoshika@mb.kyoto-phu.ac.jp
© 2012 The Pharmaceutical Society of Japan

May 2012
675
Table 1. 13C-NMR (125MHz) Data for Chakasaponins IV (1) and 1a
C-
1
1a
C-
1
1a
1
2
3
4
5
6
7
8
9
38.9
26.5
89.4
39.5
55.5
18.7
36.9
41.4
47.1
36.7
23.9
124.7
144.5
47.7
67.5
73.8
45.4
41.6
46.9
31.9
41.4
74.0
27.9
16.8
15.8
17.5
21.3
62.9
33.5
25.1
38.9
26.5
89.5
39.5
55.5
18.7
36.4
41.4
47.1
36.9
23.9
124.8
145.0
48.1
67.3
73.7
44.9
41.4
47.3
31.7
45.5
74.2
27.9
16.8
15.8
17.4
21.3
69.4
33.7
25.4
22-O-Tig
1
167.8
129.8
136.6
14.0
2
3
4
5
12.2
GlcA
1′
2′
3′
4′
5′
6′
Gal
105.6
79.1
84.0
71.1
76.4
172.1
105.5
79.0
83.9
71.1
76.4
172.2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
1″
2″
3″
4″
5″
6″
Ara
1‴
2‴
3‴
4‴
5‴
Xyl
1″″
2″″
3″″
4″″
5″″
103.5
73.3
75.1
70.0
75.7
61.9
103.4
73.3
75.1
70.0
75.7
61.9
101.7
81.9
72.7
68.2
65.8
101.7
82.0
72.7
68.3
65.8
106.8
75.7
78.2
70.7
67.4
106.8
75.7
78.2
70.7
67.4
Measured in pyridine-d5.
Chart 1. Chemical Structures of 1–5
J=5.4Hz H-1‴)]. Treatment of 1 with 10% aqueous KOH–50% signals assignable to an aglycone [δ: 0.80, 1.01, 1.06, 1.12,
aqueous 1,4-dioxane (1:1) liberated desacyl-chakasaponin 1.13, 1.25, 1.85 (3H each, all s, H3-25, 26, 29, 24, 30, 23, 27),
IV (1a) and an organic acid, tiglic acid, which was identified 3.23 (1H, m, H-3), 3.71, 4.13 (1H each, both d, J=11.0Hz,
by HPLC analysis of the p-nitrobenzyl derivative.9,10) The H2-28), 4.45 (1H, m, H-15), 4.47 (1H, m, H-16), 4.62 (1H, m,
molecular formula of 1a was determined as C52H84O24 from H-22), 5.42 (1H, brs, H-12)] and four glycopyranosyl parts.
positive-ion FAB-MS [m/z 1115 (M+Na)+]. The 1H-NMR The position of the tigloyl group in 1 was characterized on the
(pyridine-δ5) and 13C-NMR (Table 1) spectra of 1a,8) showed basis of an heteronuclear multiple bond connectivity (HMBC)

676
Vol. 60, No. 5
experiment, in which a long-range correlation was observed 3-O-[β-D-galactopyranosyl(1
between the 22-proton and tigloyl carbonyl carbon. In addi- arabinopyranosyl(1 3)]-β-D-glucopyranosiduronic acid (1).
tion, comparison of the 13C-NMR data for 1 with those for
Effects of 2–5 on TG Levels after Loading of Olive Oil
→2)][β-D-xylopyranosyl(1→2)-α-L-
→
1a revealed acylation shifts around the 21- and 22-positions. in Mice Effects of chakasaponins (2–4) and a flavonol gly-
The oligoglycoside structure at the 3-position in 1 and 1a coside (5) on the increase in plasma triglyceride (TG) levels
was elucidated by a HMBC experiment, which showed long- in olive oil-loaded mice were examined. As shown in Table
range correlations between the following proton and carbons: 2, the chakasaponins significantly inhibited the increase at 25
H-1′ and 3-C; H-1″ and C-2′; H-1‴ and C-3′; H-1″″ and C-2‴. and/or 50mg/kg (per os (p.o.)), but flavonol glycoside (5) did
Finally, acid hydrolysis of 1a with 5% aqueous H2SO4–1,4- not. Previously, escins from the seeds of Aesculus turbinata
dioxane (1:1, v/v) yielded a known triterpene, A1-barrigenol,11) were found to have anti-hyperlipidemic effects in lipid emul-
as an aglycone, together with L-arabinose, D-galactose, D- sion-loaded mice.14) The inhibitory effect of the chakasaponins
glucuronic acid, and D-xylose, which were identified by GLC in the present study was similar to that of escin IIa.
analysis of their trimethylsilyl thiazolidine derivatives.12,13)
Effects of 2–5 on Glucose Levels after Loading of Su-
On the basis of those findings, the structure of chakasa- crose in Mice Effects of 2–5 on the increase in plasma glu-
ponin IV was determined to be 22-O-tigloyl-A1-barrigenol cose levels in sucrose-loaded mice were examined. As shown
in Table 3, the chakasaponins significantly inhibited the in-
crease at 50 and 100mg/kg. (p.o.), but the flavonol glycoside
(5) did not. We previously reported that escin IIa inhibited
the increase in glucose levels in glucose-loaded rats.15,16) In
the present study, the inhibitory effect of the chakasaponins
tended to be stronger than that of escin IIa.
Effects of 2–5 on Gastric Emptying (GE) in Mice Pre-
viously, we reported that the inhibition of gastric emptying
markedly reduced increase in blood glucose and TG levels
in sugar and olive oil-loaded rats and mice.16–20) Additionally,
inhibitory effects on GE by chakasaponins I (2) and II (3),21)
but not 4 and 5, on GE were reported. Therefore, effects of
2–5 on GE in mice were compared. As shown in Table 4, the
chakasaponins significantly prevented GE at 50–100mg/kg
Fig. 1. DQF COSY and HMBC Correlations of 1
(p.o.) 30min after the loading of a carboxymethyl cellulose
Table 2. Effects of Chakasaponins I–III (2–4) and Kaempferol 3-O-Glc(1
Mice
→
3)-Rha(1
→
6)-Glc (5) on Increases in Plasma TG Levels in Olive Oil-Loaded
Plasma TG (mg/100mL)
Dose
(mg/kg, p.o.)
Treatment
Normal
n
2.0h
4.0h
6.0h
—
—
25
50
—
—
25
50
—
—
25
50
—
—
25
50
—
—
5
6
8
6
6
6
8
6
6
6
8
6
6
6
8
6
6
7
9
7
7
6
8
8
8
115.5 12.4**
440.5 45.2
435.7 67.4
284.2 9.6**
152.4 13.5**
553.8 49.8
431.8 49.8**
249.5 31.1**
124.0 8.4**
407.2 73.0
394.1 81.4
214.4 62.7*
164.8 20.1**
418.1 54.5
469.5 86.6
383.9 41.8
91.9 9.4**
440.3 60.2
371.3 41.5
203.8 52.1**
114.9 18.1**
546.7 59.4
464.2 45.3
308.9 61.4*
126.7 9.0**
359.4 43.2
296.8 45.2
347.5 27.2
149.8 15.0**
522.7 44.0
436.9 63.3
416.7 71.2
94.7 8.3**
385.1 71.4
300.5 67.2
314.3 88.2
128.0 13.5**
398.6 42.1
395.5 114.9
444.6 46.7
97.3 7.4**
393.2 60.1
297.0 67.4
160.4 47.7**
110.4 12.0**
375.9 66.2
244.5 38.5
285.8 46.9
122.9 10.4**
267.8 37.8
197.7 26.6
257.7 34.4
118.5 14.7*
259.8 50.3
240.2 24.1
390.0 73.9
87.0 11.3**
207.8 36.1
184.8 36.8
255.5 60.0
103.2 12.0*
231.6 44.0
213.6 64.1
191.4 33.3
90.6 9.4**
263.3 45.0
171.9 24.9
129.1 16.6**
134.8 7.7*
271.5 62.5
310.9 91.5
270.2 44.4
Control
Chakasaponin I (2)
Normal
Control
Chakasaponin II (3)
Normal
Control
Chakasaponin III (4)
Normal
Control
5
Normal
Control
Orlistat
10
—
—
50
100
Normal
Control
Escin IIa
Values represent the mean S.E.M. Significantly different from the control group, *p<0.05, **p<0.01.

May 2012
677
Table 3. Effects of Chakasaponins I–III (2–4) and Kaempferol 3-O- Glc(1
Loaded Mice
→
3)-Rha(1
→
6)-Glc (5) on Increases in Plasma Glucose Levels in Sucrose-
Plasma glucose (mg/100mL)
Dose
(mg/kg, p.o.)
Treatment
Normal
n
0.5h
1.0h
2.0h
—
—
6
8
6
6
6
8
6
5
6
9
6
6
6
8
6
6
6
9
6
6
6
9
6
6
128.5 5.3**
226.9 8.2
149.4 2.6**
194.2 6.2
121.7 5.3**
144.3 3.0
161.7 6.5
161.6 6.6
117.5 8.0**
138.2 4.6
148.4 2.4
150.9 3.6
124.2 3.9**
141.2 3.5
157.4 5.9*
160.0 3.6**
125.9 6.6**
167.2 3.9
157.9 7.8
159.5 7.9
131.8 6.4**
163.7 3.7
151.5 6.3
152.8 3.8
112.9 5.1**
168.9 7.6
166.6 7.2
175.9 11.1
Control
Chakasaponin I (2)
50
199.6 8.7**
196.8 8.2**
116.5 7.9**
221.5 14.4
185.7 9.9**
178.5 8.3**
126.5 4.0**
238.7 5.9
213.7 9.4
100
—
200.1 5.9
Normal
132.7 10.2**
200.3 10.5
179.3 5.3
—
Control
Chakasaponin II (3)
50
100
—
187.3 10.7
135.4 3.9**
207.3 9.7
Normal
—
Control
Chakasaponin III (4)
50
215.9 10.6**
189.0 10.2**
124.1 8.7**
241.7 13.4
270.5 16.1
243.7 18.0
124.8 7.3**
218.7 4.0
198.0 6.3
100
—
187.6 6.8*
133.8 7.2**
217.5 10.3
207.9 11.4
225.0 16.7
143.0 5.4**
208.9 6.8
Normal
Control
5
—
50
100
—
Normal
Control
Acarbose
—
10
162.4 11.7**
153.8 10.2**
122.2 6.3**
263.3 16.2
254.2 14.0
206.4 12.9**
183.8 3.8*
185.4 8.1*
123.2 4.9**
235.1 10.7
224.2 10.4
205.0 7.5**
20
—
Normal
Control
Escin IIa
—
50
100
Values represent the mean S.E.M. Significantly different from the control group, *p<0.05, **p<0.01.
Table 4. Effects of Chakasaponins I–III (2–4) and Kaempferol 3-O-
Glc(1 3)-Rha(1 6)-Glc (5) on Gastric Emptying in Mice
sodium salt (CMC-Na) solution, but the flavonol glycoside (5)
had no such effect, similar to the results obtained with olive
oil- and sucrose-loaded mice. We previously reported that
escin IIa inhibited GE in mice and this contributed to the anti-
hyperglycemic effect in glucose-loaded rats and mice.16,22–24)
In the present study, the inhibitory effect of the chakasaponins
tended to be stronger than that of escin IIa. These findings
suggest that the inhibition of GE is involved in the anti-hyper-
glycemic and anti-hyperlipidemic effects.
Effects of 2–5 on the Activities of Rat Intestinal
α-Glucosidase and Rat Lens Aldose Reductase Next,
inhibitory effects of 2–5 on the activities of rat intestinal
α-glucosidase (maltase and sucrase) were examined, since
2–4 significantly reduced plasma glucose levels in sucrose-
loaded mice. Compounds 2–5 did not inhibit the enzymatic
activity less than 14.6% at 400µM. Aldose reductase has been
reported to catalyze the reduction of glucose to sorbitol as a
key enzyme in the polyol pathway. Sorbitol does not readily
diffuse across cell membranes, and its intracellular accumula-
tion has been implicated in chronic complications of diabetes
such as cataracts.25) Therefore, the inhibitory effects of 2–5 on
the activity of rat lens aldose reductase were examined. The
chakasaponins (2–4) had little effect (inhibition at 100µM:
3.0%, 5.1%, 18.0%, respectively), but 5 had a moderate effect
with an IC50 value of 20.8µM.
→
→
Dose
(mg/kg, p.o.)
Gastric emptying
Treatment
Control
n
(%)
—
25
9
75.0 4.1
Chakasaponin I (2)
7
7
68.5 5.0
50
50.2 3.0**
37.9 1.4**
73.1 2.1
100
—
7
Control
10
7
Chakasaponin II (3)
25
65.2 3.9
50
7
56.9 2.7**
40.5 6.7**
76.8 2.4
100
—
7
Control
10
7
Chakasaponin III (4)
25
70.7 1.9
50
7
52.0 2.7**
39.8 1.7**
81.4 4.7
100
—
7
Control
7
5
50
6
90.5 1.3
100
6
87.6 2.0
Control
8
86.2 1.3
Escin IIa
25
50
6
81.7 6.5
6
69.6 3.0*
53.8 4.0**
100
6
Each value represents the mean S.E.M. Significantly different from the control
group, *p<0.05, **p<0.01.
In conclusion, principal saponins, chakasaponins I–III
(2–4) from the flower buds of Camellia sinensis cultivated rat intestinal α-glucosidase or rat lens aldose reductase. Fur-
in Fujian province, China significantly inhibited increases in thermore, inhibition of GE was involved in the anti-hypergly-
plasma TG and glucose levels in olive oil or sucrose-loaded cemic and anti-hyperlipidemic effects of 1–3. In addition, the
mice at 50 and 100mg/kg, but did not effect the activity of chemical structure of a new acylated oleanane-type triterpene

678
Vol. 60, No. 5
oligoglycoside, chakasaponin IV, was elucidated on the basis
of chemical and physicochemical evidence.
The procedure used to isolate compounds 2–5 is described
in ref. 5.
Chakasaponin IV (1): Colorless fine crystals from CHCl3–
25
°
°
MeOH; mp 204.0–206.0 C; [α]D +24.6 (c=0.89, MeOH);
Experimental
General Experimental Procedures The following instru- IR (KBr) νmax 3450, 2962, 1718, 1686, 1078cm−1; H-NMR
ments were used to obtain physical data: specific rotations, (500MHz, pyridine-d5) δ: 0.80, 1.00, 1.05, 1.11, 1.24, 1.26,
a Horiba SEPA-300 digital polarimeter (l=5cm); IR spectra, 1.85 (3H each, all s, H3-25, 26, 29, 24, 23, 30, 27), 1.41 (3H,
Shimadzu FTIR-8100 spectrometer; FAB-MS and high resolu- d, J=7.0Hz, H3-22-O-Tig-4), 1.76 (3H, brs, H3-22-O-Tig-5),
tion (HR)-FAB-MS, a JEOL JMS-SX 102A mass spectrom- 3.25 (1H, dd, J=3.8, 13.2Hz, H-3), 3.60, 3.76 (1H each, both
1
1
eter; H-NMR spectra, JEOL JNM-LA 500 (500MHz) and d, J=10.6Hz, H2-28), 4.22 (1H, d, J=4.0Hz, H-15), 4.56 (1H,
JEOL JNM-ECA 600 (600MHz) spectrometers; 13C-NMR d, J=4.0Hz, H-16), 4.92 (1H, d, J=7.2Hz, H-1′), 5.01 (1H, d,
spectra, JEOL JNM-LA (125MHz) and JEOL JNM-ECA J=7.4Hz, H-1″″), 5.44 (1H, brs, H-12), 5.74 (1H, d, J=7.6Hz,
600 (150MHz) spectrometers with tetramethylsilane as an H-1″), 5.77 (1H, d, J=5.4Hz H-1‴), 6.16 (1H, dd, J=5.5,
internal standard; and HPLC detector, Shimadzu RID-6A 11.8Hz, H-22), 6.83 (1H, dq-like, H-22-O-Tig-3); 13C-NMR
refractive index and SPD-10Avp UV-VIS detectors. HPLC (125MHz, pyridine-d5) δC: given in Table 1; positive-ion FAB-
column, COSMOSIL 5C18-MS-II {[250×4.6mm i.d. (5µm) MS: m/z 1197 (M+Na)+; HR-FAB-MS: m/z 1197.5676 [Calcd
for analytical purposes] and [250×20mm i.d. (5µm) for pre- for C57H90O25Na (M+Na)+: 1197.5669].
parative purposes], Nacalai Tesque, Japan} and Develosil
Alkaline Hydrolysis of Chakasaponin IV (1) A solution
C30-UG-5 {[250×4.6mm i.d. (5µm) for analytical purposes] of chakasaponin IV (1, 12mg) in 50% aqueous 1,4-dioxane
and [250×20mm i.d. (5µm) for preparative purposes], Nomura (0.5mL) was treated with 10% aqueous KOH (0.5mL) and
°
stirred at 37 C for 1h. The reaction mixture was neutralized
Chemical, Japan} columns were used.
The following materials were used for chromatography: with Dowex HCR W2 (H+ form) and the resin was removed
ordinary-phase silica gel column chromatography, Silica gel by filtration. Evaporation of the solvent from the filtrate under
BW-200 (Fuji Silysia Chemical, Ltd., Japan, 150–350 mesh); reduced pressure yielded a reaction product. A part of the re-
reverse-phase silica gel column chromatography, Chromatorex action product was dissolved in (CH2)2Cl2 (2.0mL) and the so-
ODS DM1020T (Fuji Silysia Chemical, Ltd., Japan, 100–200 lution was treated with p-nitrobenzyl-N,N′-diisopropylisourea
°
mesh); TLC, precoated TLC plates with Silica gel 60F254 (10mg), then stirred at 80 C for 1h. The reaction solution was
(Merck, 0.25mm) (ordinary phase) and Silica gel RP-18 F254S subjected to HPLC [column: YMC-Pack ODS-A, 250×4.6mm
(Merck, 0.25mm) (reverse phase); reversed-phase HPTLC, i.d.; mobile phase: MeCN–H2O (50:50, v/v); detection: UV
precoated TLC plates with Silica gel RP-18 WF254S (Merck, (254nm); flow rate: 1.0mL/min] to identify the p-nitrobenzyl
0.25mm). Detection was achieved by spraying with 1% ester of tiglic acid (tR 35.1min) from 1. The rest of the reac-
Ce(SO4)2–10% aqueous H2SO4 followed by heating.
tion product was subjected to normal-phase silica gel column
Plant Material The flower buds of Camellia sinensis, cul- chromatography [50mg, CHCl3–MeOH–H2O (10:3:1 lower
tivated in Fujian province, China, were collected in 2006. The layer
→
6:4:1)] to give desacyl-chakasaponin IV (1a, 10.0mg).
botanical identification was undertaken by one of the authors
Desacyl-chakasaponin IV (1a): Colorless fine crystals from
26
°
°
(M. Y.). A voucher of the plant is on file in our laboratory CHCl3–MeOH; mp 211.0–213.0 C; [α]D +38.7 (c=0.33,
1
(2006. China-06F).
MeOH); H-NMR (600MHz, pyridine-d5) δ: 0.80, 1.01, 1.06,
Isolation of Chakasaponin IV The dried flower buds of 1.12, 1.13, 1.25, 1.85 (3H each, all s, H3-25, 26, 29, 24, 30, 23,
C. sinensis (1.5kg, Fujian province, China) were extracted 27), 3.23 (1H, m, H-3), 3.71, 4.13 (1H each, both d, J=11.0Hz,
three times with MeOH (15L×3) for three hours under reflux. H2-28), 4.45 (1H, m, H-15), 4.47 (1H, m, H-16), 4.62 (1H, m,
Evaporation of the solvent under reduced pressure provided a H-22), 4.90 (1H, d, J=7.2Hz, H-1′), 5.00 (1H, d, J=7.4Hz,
MeOH extract (467g, 31.1%). A portion (450g) of the MeOH H-1″″), 5.42 (1H, brs, H-12), 5.70 (1H, d, J=7.4Hz, H-1′), 5.77
extract was partitioned into an EtOAc (2L×3)/H2O (2L) mix- (1H, d, J=5.2Hz H-1‴); 13C-NMR (150MHz, pyridine-d5) δC:
ture to furnish an EtOAc-soluble fraction (46g, 3.2%) and given in Table 1; positive-ion FAB-MS: m/z 1115 (M+Na)+;
aqueous phase, which was extracted with n-BuOH (2L×3) to HR-FAB-MS: m/z 1115.5245 [Calcd for C52H84O24Na
give n-BuOH (237g, 16.4%)- and H2O (166g, 11.5%)-soluble (M+Na)+: 1115.5250].
fractions as reported previously.5) A part of the n-BuOH-sol-
uble fraction (142g) was subjected to normal-phase silica gel solution of 1a (2.0mg) in 5% aq. H2SO4–1,4-dioxane (1:1, v/v,
column chromatography [3kg, CHCl3 CHCl3 :MeOH:H2O 1.0mL) was heated under reflux for 2h. After cooling, the re-
(30:10:1 10:3:1 7:3:1 6:4:1)
MeOH] to give five frac- action mixture was neutralized with Amberlite IRA-400 (OH−
Acid Hydrolysis of Desacyl-Chakasaponin IV (1a) A
→
→
→
→
→
tions [Fr. 1, Fr. 2, Fr. 3, Fr. 4 (115.0g), Fr. 5]. Fr. 4 (115g) was form) and the resin was removed by filtration. Evaporation
subjected to reversed-phase silica gel column chromatogra- of the solvent from the filtrate under reduced pressure gave a
phy [530g, MeOH:H2O (10:90→30:70→50:50→70:30→Me product, which was applied to a Sep-Pack C18 cartridge with
OH)] to give nine fractions [Fr. 4-1, Fr. 4-2, Fr. 4-3, Fr. 4-4, H2O and MeOH. The MeOH eluate was identified as A1-bar-
Fr. 4-5, Fr. 4-6 (12.5g), Fr. 4-7, Fr. 4-8, and Fr. 4-9]. Fraction rigenol. The H2O eluate was concentrated and the residue was
4-6 (12.5g) was subjected to reversed-phase silica gel column treated with L-cysteine methylester hydrochloride (1.0mg) in
°
→60:40→MeOH) pyridine (0.1mL) at 60 C for 1h. After the reaction, the solu-
chromatography [370g, MeOH:H2O (50:50
to give five fractions [Fr. 4-6-1, Fr. 4-6-2, Fr. 4-6-3 (6.2g), Fr. tion was treated with N,O-bis(trimethylsilyl)trefluoroacetamide
°
4-6-4, Fr. 4-6-5]. A part of Fr. 4-6-3 (1.0g) was further sepa- (0.1mL) at 60 for 1h. The supernatant was then subjected
rated by HPLC [MeOH–H2O (65:35, v/v)] to give chakasapo- to GLC to identify the derivatives of D-glucuronic acid (i),
nin IV (1, 14.2mg).
D-galactose (ii), L-arabinose (iii) and D-xylose (iv). (Supelco

May 2012
679
STB-1 [30m×0.25mm i.d.] capillary column; 230 C column were performed as described in our previous report.20,26) The
temperature; N2 carrier gas; tR, (i) 26.5min, (ii) 25.6min, (iii) rat small intestinal brush border membrane fraction was pre-
°
15.1min, (iv) 19.3min).
pared and its suspension in 0.1M maleate buffer (pH 6.0) was
Bioassay Methods. Animals Male ddY mice weighing used as the small intestinal α-glucosidase of maltase, sucrase,
about 25–30g were purchased from Kiwa Laboratory Animal isomaltase, and trehalase. The enzyme suspension was diluted
Co., Ltd., Wakayama, Japan. The animals were housed at a to hydrolyse maltose and sucrose to produce ca. 0.30 and ca.
°
constant temperature of 23 2 C and fed a standard labora- 0.15µmol/tube of D-glucose in the following reaction. The
tory chow (MF, Oriental Yeast Co., Ltd., Tokyo, Japan). The substrate (maltose: 37mM, sucrose: 37mM), test compound
animals were fasted for 20–24h prior to the beginning of the and enzyme in 0.1M maleate buffer (pH 6.0, 0.1mL) were
°
experiment, but were allowed free access to tap water. All of incubated together at 37 C. After 30min, 0.4mL of water
the experiments were performed with conscious mice unless was added to the test tube, and the tube was immediately im-
otherwise noted. The experimental protocol was approved by mersed in boiling water for 2min to stop the reaction, then
the Experimental Animal Research Committee at Kyoto Phar- cooled with water. The glucose concentration was determined
maceutical University.
by an enzymatic method. Each test sample was dissolved in
Effect on Plasma Triglyceride (TG) Levels in Olive dimethyl sulfoxide (DMSO). Measurements were performed
Oil-Treated Mice The experiments were performed as de- in duplicate, and the IC50 was determined graphically. An
scribed in our previous report.20) Each test sample was admin- intestinal α-glucosidase inhibitor, acarbose (IC50 for maltase
istered orally to fasted mice and olive oil (5mL/kg) was ad- and sucrase of 2.0 and 1.7µM, respectively),20) was used as a
ministered p.o. 30min thereafter. Blood samples (ca. 0.1mL) reference compound.
were collected into a polyethylene tube (1.5mL) containing
Effect on Rat Lens Aldose Reductase The experiments
heparin (5 units/tube) from the infraorbital venosus plexus, 2, were performed as described previously.20,26) The supernatant
4, and 6h after the olive oil treatment. Plasma TG levels were fluid of rat lens homogenate was used as a crude enzyme. The
determined by an enzymatic method using Triglyceride E test enzyme suspension was diluted to produce ca. 10nmol/tube of
Wako (Wako Pure Chemical Industries Ltd., Osaka, Japan).
nicotinamide adenine dinucleotide phosphate (NADP) in the
Effect on Plasma Glucose Levels in Sucrose-Loaded following reaction. The incubation mixture contained 135mM
Mice The experiments were performed as described previ- phosphate buffer (pH 7.0), 100mM Li2SO4, 0.03mM reduced
ously.26) Each test sample was administered orally to fasted nicotinamide adenine dinucleotide phosphate (NADPH),
mice, and a 20 (w/v) % sucrose solution (10mL/kg) was 1mM DL-glyceraldehyde as a substrate, and 100µL of enzyme
administered p.o. 30min thereafter. Blood samples (ca. fraction, with 25µL of sample solution, in a total volume of
0.1mL) were collected into a polyethylene tube (1.5mL) con- 0.5mL. The reaction was initiated by the addition of NADPH
°
taining heparin (5 units/tube) from the infraorbital venosus at 30 C. After 30min, the reaction was stopped by the ad-
plexus, 0.5, 1, and 2h after the sucrose loading. After the dition of 150µL of 0.5M HCl. Then, 0.5mL of 6M NaOH
centrifugation of blood samples, plasma glucose levels were containing 10mM imidazole was added, and the solution was
°
determined enzymatically using Glucose CII test Wako (Wako heated at 60 C for 20min to convert NADP to a fluorescent
Pure Chemical Industries Ltd., Osaka, Japan). An intestinal product. Fluorescence was measured using a fluorescence mi-
α-glucosidase inhibitor, acarbose, was used as a reference croplate reader (FLUOstar OPTIMA, BMG Labtechnologies)
compound.
at an excitation wavelength of 360nm and an emission wave-
Gastric Emptying (GE) in Mice The rate of GE was length of 460nm. Each test sample was dissolved in DMSO.
determined using phenol red.20) Briefly, a solution of 1.5% car- Measurements were performed in duplicate, and the IC50 was
boxymethyl cellulose sodium salt (CMC-Na) containing 0.05% determined graphically. An aldose reductase inhibitor, epalr-
phenol red as a marker was given intragastrically (0.3mL/ estat (IC50=0.072µM),20) was used as a reference compound.
mouse) to 20–24h-fasted mice. Thirty minutes later, the mice
Statistics Values are expressed as mean S.E.M. For the
were sacrificed by cervical dislocation under ether anesthesia. statistical analysis, a one-way analysis of variance followed by
’
The abdominal cavity was opened, the gastroesophageal junc- William s test for dose-dependent effects was used.
tion and pylorus were clamped, and the stomach was removed,
weighed, placed in 10mL of 0.1M NaOH, and homogenized.
Acknowledgements This research was supported by the
The suspension was allowed to settle for 1h at room tem- Academic Frontier Project and a Grant-in-Aid for Scientific
perature, and 5mL of the supernatant was added to 0.5mL Research from the Ministry of Education, Culture, Sports,
of 20% trichloroacetic acid (w/v) and centrifuged at 3000rpm Science and Technology (MEXT) of Japan, and Hoh-ansha
for 20min. Next, 4mL of supernatant was mixed with 4mL Foundation, and Japan-China Medical Association.
of 0.5M NaOH, and the amount of phenol red was determined
from the absorbance at 560nm. The test sample was given
orally 30min prior to the administration of the CMC-Na solu-
tion. Gastric emptying (%) in the 30-min period was calcu-
lated according to the following equation:
References and Notes
1) Part XXXII: Morikawa T., Li X., Nishida E., Nakamura S.,
Ninomiya K., Matsuda H., Hamao M., Muraoka O., Hayakawa T.,
Yoshikawa M., Chem. Pharm. Bull., 59, 889–895 (2011).
2) Yoshikawa M., Morikawa T., Yamamoto K., Kato Y., Nagatomo A.,
Matsuda H., J. Nat. Prod., 68, 1360–1365 (2005).
gastric emptying (%)
3) Yoshikawa M., Wang T., Sugimoto S., Nakamura S., Nagatomo A.,
Matsuda H., Harima S., Yakugaku Zasshi, 128, 141–151 (2008).
4) Yoshikawa M., Nakamura S., Kato Y., Matsuhira K., Matsuda H.,
Chem. Pharm. Bull., 55, 598–605 (2007).
amount of phenol red in test sample-treated group
=1
×100
amount of phenol red administered
Effect on Rat Intestinal α-Glucosidase The experiments

680
5) Yoshikawa M., Sugimoto S., Kato Y., Nakamura S., Wang T., Ya-
mashita C., Matsuda H., Chem. Biodivers., 5, 903–915 (2009).
6) Yoshikawa M., Sugimoto S., Nakamura S., Matsuda H., Chem.
Pharm. Bull., 56, 1297–1303 (2008).
Vol. 60, No. 5
16) Matsuda H., Murakami T., Li. Y., Yamahara J., Yoshikawa M.,
Bioorg. Med. Chem., 6, 1019–1023 (1998).
17) Yoshikawa M., Matsuda H., Biofactors, 13, 231–237 (2000).
18) Yoshikawa M., Nakamura S., Ozaki K., Kumahara A., Morikawa T.,
Matsuda H., J. Nat. Prod., 70, 210–214 (2007).
19) Shimoda H., Ninomiya K., Nishida N., Yoshino T., Morikawa T.,
Matsuda H., Yoshikawa M., Bioorg. Med. Chem. Lett., 13, 223–228
(2003).
20) Matsuda H., Asao Y., Nakamura S., Hamao M., Sugimoto S., Hongo
M., Pongpiriyadacha Y., Yoshikawa M., Chem. Pharm. Bull., 57,
487–494 (2009).
21) Hamao M., Matsuda H., Nakamura S., Nakashima S., Semura S.,
Maekubo S., Wakasugi S., Yoshikawa M., Bioorg. Med. Chem., 19,
6033–6041 (2011).
22) Matsuda H., Li Y., Murakami T., Yamahara J., Yoshikawa M., Eur.
J. Pharmacol., 368, 237–243 (1999).
23) Matsuda H., Li Y., Yoshikawa M., Life Sci., 66, PL41–PL46 (2000).
24) Matsuda H., Li Y., Yoshikawa M., Life Sci., 67, 2921–2927 (2000).
25) Terashima H., Hama K., Yamamoto R., Tsuboshima M., Kikkawa
R., Hatanaka I., Shigeta Y., J. Pharmacol. Exp. Ther., 229, 226–230
(1984).
7) Wada S., He P., Hashimoto I., Watanabe N., Sugiyama K., Biosci.
Biotechnol. Biochem., 64, 2262–2265 (2000).
1
8) The H- and 13C-NMR spectra of 1 were assigned with the aid of
distortionless enhancement by polarization transfer (DEPT), DQF
COSY, heteronuclear multiple quantum coherence spectroscopy
(HMQC), and HMBC experiments.
9) Yoshikawa M., Murakami T., Yoshizumi S., Murakami N., Yama-
hara J., Matsuda H., Chem. Pharm. Bull., 44, 1899–1907 (1996).
10) Matsuda H., Nakamura S., Fujimoto K., Moriuchi R., Kimura Y.,
Ikoma N., Hata Y., Muraoka O., Yoshikawa M., Chem. Pharm.
Bull., 58, 1617–1621 (2010).
11) Ito S., Ogino T., Sugiyama H., Kodama M., Tetrahedron Lett., 24,
2289–2294 (1967).
12) Hara S., Okabe H., Mihashi K., Chem. Pharm. Bull., 34, 1843–1845
(1986).
13) Nakamura S., Murakami T., Nakamura J., Kobayashi H., Matsuda
H., Yoshikawa M., Chem. Pharm. Bull., 54, 1545–1550 (2006).
14) Hu J. N., Zhu X. M., Han L. K., Saito M., Sun Y. S., Yoshikawa M., 26) Morikawa T., Chaipech S., Matsuda H., Hamao M., Umeda Y., Sato
’
Kimura Y., Zheng Y. N., Chem. Pharm. Bull., 56, 12–16 (2008).
15) Yoshikawa M., Murakami T., Matsuda H., Yamahara J., Murakami
N., Kitagawa I., Chem. Pharm. Bull., 44, 1454–1464 (1996).
H., Tamura H., Kon i H., Ninomiya K., Yoshikawa M., Pongpiri-
yadacha Y., Hayakawa T., Muraoka O., Bioorg. Med. Chem., 20,
832–840 (2012).