Acylated phenylethanoid oligoglycosides with hepatoprotective activity from the desert plant Cistanche tubulosa1

Article abstract of DOI:10.1016/j.bmc.2010.01.047

The methanolic extract from fresh stems of Cistanche tubulosa (Orobanchaceae) was found to show hepatoprotective effects against d-galactosamine (d-GalN)/lipopolysaccharide (LPS)-induced liver injury in mice. From the extract, three new phenylethanoid oligoglycosides, kankanosides H₁ (1), H₂ (2), and I (3), were isolated together with 16 phenylethanoid glycosides (4-19) and two acylated oligosugars (20, 21). The structures of 1-3 were determined on the basis of spectroscopic properties as well as of chemical evidence. Among the isolates, echinacoside (4, IC₅₀ = 10.2 μM), acteoside (5, 4.6 μM), isoacteoside (6, 5.3 μM), 2′-acetylacteoside (8, 4.8 μM), and tubuloside A (10, 8.6 μM) inhibited d-GalN-induced death of hepatocytes. These five isolates, 4 (31.1 μM), 5 (17.8 μM), 6 (22.7 μM), 8 (25.7 μM), and 10 (23.2 μM), and cistantubuloside B₁ (11, 21.4 μM) also reduced TNF-α-induced cytotoxicity in L929 cells. Moreover, principal constituents (4-6) exhibited in vivo hepatoprotective effects at doses of 25-100 mg/kg, po.

Full text of DOI:10.1016/j.bmc.2010.01.047

Bioorganic & Medicinal Chemistry 18 (2010) 1882–1890
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Acylated phenylethanoid oligoglycosides with hepatoprotective activity
from the desert plant Cistanche tubulosa1
Toshio Morikawa a, Yingni Pan a,b, Kiyofumi Ninomiya a, Katsuya Imura a, Hisashi Matsuda c,
b
a,
*
Masayuki Yoshikawa a,c, , Dan Yuan , Osamu Muraoka
*
a Pharmaceutical Research and Technology Institute, Kinki University, 3-4-1 Kowakae, Higashi-osaka, Osaka 577-8502, Japan
b School of Traditional Chinese Medicines, Shenyang Pharmaceutical University, 103 Wenhua Rd., Shenyang 110016, People’s Republic of China
c Kyoto Pharmaceutical University, Misasagi, Yamashina-ku, Kyoto 607-8412, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The methanolic extract from fresh stems of Cistanche tubulosa (Orobanchaceae) was found to show
Received 28 November 2009
Revised 15 January 2010
Accepted 16 January 2010
Available online 25 January 2010
hepatoprotective effects against D-galactosamine (D-GalN)/lipopolysaccharide (LPS)-induced liver injury
in mice. From the extract, three new phenylethanoid oligoglycosides, kankanosides H1 (1), H2 (2), and I
(3), were isolated together with 16 phenylethanoid glycosides (4–19) and two acylated oligosugars (20,
21). The structures of 1–3 were determined on the basis of spectroscopic properties as well as of chemical
evidence. Among the isolates, echinacoside (4, IC50 = 10.2
5.3 M), and tubuloside A (10, 8.6
M), 20-acetylacteoside (8, 4.8
hepatocytes. These five isolates, 4 (31.1 M), 5 (17.8 M), 6 (22.7
and cistantubuloside B1 (11, 21.4 M) also reduced TNF- -induced cytotoxicity in L929 cells. Moreover,
principal constituents (4–6) exhibited in vivo hepatoprotective effects at doses of 25–100 mg/kg, po.
Ó 2010 Elsevier Ltd. All rights reserved.
l
M), acteoside (5, 4.6
M) inhibited -GalN-induced death of
M), 8 (25.7 M), and 10 (23.2 M),
lM), isoacteoside (6,
Keywords:
Hepatoprotective activity
Kankanoside
Cistanche tubulosa
Phenylethanoid glycoside
Orobanchaceae
l
l
l
D
l
l
l
l
l
l
a
Anti-TNF-a activity
1. Introduction
in isolated rat thoracic aorta.11 In our continuing studies on constit-
uents in the plant, a methanolic extract of fresh stems of C. tubulosa
Cistanche tubulosa (SCHRENK) R. WIGHT (Orobanchaceae) is a
perennial parasitic plant growing on roots of Salvadora or Calotropis
species, and distributed in North Africa, Arabia, and Asian countries.2
Stems of C. tubulosaas well asCistanche salsa andCistanche deserticola
have traditionally been used for treatment of impotence, sterility,
lumbago, and body weakness as well as a promoting agent of blood
circulation.2,3 Previously, several phenylethanoids, iridoids, mono-
terpenes, and lignans were isolated from Chinese and Pakistan
C. tubulosa.2,4–9 In the course of our characterization studies on bio-
active constituents in this natural medicine, we previously reported
that five iridoids, kankanosides A–D and kankanol, a monoterpene
glycoside, kankanoside E, two phenylethanoid oligoglycosides,
kankanosides F and G (23), and an acylated oligosugar, kankanose
(21), were isolated from the methanolic extract of dried stems of
C. tubulosa.10,11 In addition, the methanolic extract and several iso-
lates such as echinacoside (4), acteoside (5), cistanoside F (20), 21,
and kankanoside F were found to show vasorelaxant effects, that
is, inhibitory effects against contractions induced by noradrenaline
was found to show protective effect on liver injury induced by
-galactosamine ( -GalN)/lipopolysaccharide (LPS) in mice. From
D
D
the methanolic extract, we have isolated three new phenylethanoid
oligoglycosides named kankanosides H1 (1), H2 (2), and I (3) together
with 16phenylethanoidglycosides (4–19) and two acylatedoligosu-
gars (20, 21). This paper deals with the isolation and structural elu-
cidation of these new phenylethanoid oligoglycosides (1–3).
Structural requirements of the phenylethanoid glycosides for the
hepatoprotective activity are also discussed (Fig. 1).
2. Results and discussion
2.1. Protective effect of methanolic extract from stems of
C. tubulosa on liver injury induced by D-GalN/LPS in mice
Fresh stems of C. tubulosa (cultivated in Urumuqi, Xinjiang
Province, China) were extracted with methanol under reflux to
yield a methanolic extract (8.36% from the fresh stems). As shown
in Table 1, at doses of 250–500 mg/kg, po, the methanolic extract
showed inhibitory effect on the increase in serum aspartate amino-
transaminase (sAST) and alanine aminotransaminase (sALT), mark-
* Corresponding authors. Tel.: +81 75 595 4633; fax: +81 75 595 4768 (M.Y.); tel.:
+81 6 6721 2332; fax: +81 6 6729 3577 (O.M.).
E-mail addresses: myoshika@mb.kyoto-phu.ac.jp (M. Yoshikawa), muraoka@
phar.kindai.ac.jp (O. Muraoka).
ers of liver injury, induced by D-GalN/LPS in mice.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.01.047

T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
1883
HO
HO
O
HO
O
O
H
5
2
O
O
O
OH
OH
OH
OH
6
4
OH
O
OH
OH
O
1'''
7
1
3
8
HO
H
OH
HO
HO
H
OH
HO
HO
O
O
H
O
H
O
O
O
OH
O
O
O
1'
HO
4'
O
2'
1''
H
OAc
OAc
OH
H
H
H
OHOH
OHOH
OHOH
HO
HO
HO
O
O
O
O
HO
O
O
kankanoside H1 (1)
kankanoside H2 (2)
kankanoside I (3)
OH
OH
OH
R3O
R3O
O
HO
O
H
OH
R1
O
O
O
O
R2
O
O
O
OH
HO
HO
O
R2O
H
HO
H
OH
OR1
OH
R1
H
R2
R3
Glc
H
H
H
OHOH
OHOH
salidroside (19)
echinacoside (4):
acteoside (5):
isoacteoside (6):
cis-acteoside (7):
trans-Caf
trans-Caf
H
HO
HO
H
RO
O
H
H
trans-Caf
O
O
cis-Caf
H
H
H
OH
R1 R2 R3
HO
O
O
2'-acetylacteoside (8):
decaffeoylacteoside (9):
tubuloside A (10):
cistantubuloside B1 (11):
cistantubuloside B2 (12):
arenarioside (13):
wiedemanninoside C (14):
tubuloside B (22):
Ac trans-Caf
cistantubuloside A (15):
H
H
H
H
Glc
H
H
H
H
H
OH
H
syringalide A 3'-O-Rha (16):
campneoside I (17):
Ac trans-Caf
Glc
OHOH
OH OCH3
OH OH
HO
HO
H
H
H
H
trans-p-Cou
cis-p-Cou
trans-Caf
trans-Fer
H
Glc
campneoside II (18):
Glc
O
OH
O
Xyl
O
HO
HO
Glc
O
O
H
cistanoside F (20):
kankanose (21):
R = H
R = Glc
Ac
trans-Caf
O
O
HO
O
trans-p-Cou: trans-p-coumaroyl Glc: β-D-glucopyranosyl
HO
cis-p-Cou: cis-p-coumaroyl
trans-Caf: trans-caffeoyl
cis-Caf: cis-caffeoyl
Xyl: β-D-xylopyranosyl
Rha: α-L-rhamnopyranosyl
OH
H
OHOH
kankanoside G (23)
trans-Fer: trans-feruloyl
Figure 1. Chemical structures of compounds 1–23 from stems of Cistanche tubulosa.
Table 1
Inhibitory effects of the methanolic extract from stems of C. tubulosa on D-GalN/LPS-induced liver injury in mice
Treatment
Dose (mg/kg, po)
n
sAST
sALT
(Karmen unit)
Inhibition (%)
(Karmen unit)
Inhibition (%)
Normal (vehicle)
—
—
7
11
86 5b
10,714 1520
—
—
28 6b
6823 1011
—
—
Control (D-GalN/LPS)
MeOH extract
250
500
1000
8
8
8
4653 1698b
2049 556b
904 272b
56.6
80.9
91.6
3632 1527
1318 397b
701 226b
46.8
80.7
89.7
Each value represents the mean S.E.M.
Significantly different from the control, ap <0.05, bp <0.01.
2.2. Chemical constituents from stems of C. tubulosa
rhamnopyranoside4 (16, 0.0017%), campneosides I17,18 (17,
0.015%) and II17,18 (18, 0.0005%), and salidroside11 (19, 0.0027%),
and two acylated oligosugars, cistanoside F11 (20, 0.0004%) and kan-
kanose11 (21, 0.0004%).
A methanolic extract from the fresh stems of C. tubulosa was sub-
jected to Diaion HP-20 column chromatography (H2O?MeOH) to
give H2O- and MeOH-eluted fractions (5.63% and 2.73%, respec-
tively). The MeOH-eluted fraction was subjected to SiO2 and ODS
column chromatographies and finally HPLC to furnish three new
phenylethanoid oligoglycosides, kankanosides H1 (1, 0.0008%), H2
(2, 0.0001%), and I (3, 0.0007%) together with 16 phenylethanoid gly-
cosides, echinacoside2,11 (4, 0.45%), acteoside2,11 (5, 0.28%), isoacteo-
side2,11 (6, 0.041%), cis-acteoside12,13 (7, 0.0007%), 20-acetylac
teoside2,11 (8, 0.0038%), decaffeoylacteoside14 (9, 0.0002%), tubulo-
side A2,11 (10, 0.013%), cistantubulosides B19 (11, 0.0020%), B29 (12,
0.0003%), arenarioside15 (13, 0.0006%), wiedemanninoside C16 (14,
2.3. Structures of kankanosides H1 (1), H2 (2), and I (3)
Kankanoside H1 (1) was obtained as a white powder with neg-
ative optical rotation (½a D24
ꢁ45.3 in MeOH). Its IR spectrum
ꢀ
showed absorption bands at 3416, 1736, 1638, 1605, 1518, 1070,
and 1040 cmꢁ1 ascribable to hydroxyls, ester carbonyls, ether func-
tions, and aromatic rings. The positive- and negative-ion FABMS
spectra of
1 showed quasimolecular ion peaks at m/z 835
(M+Na)+ and m/z 811 (MꢁH)ꢁ, respectively, and the molecular for-
0.0008%), cistantubuloside A9 (15, 0.0009%), syringalide A 30-O-
a
-L
-
mula was determined as C37H48O20 by high-resolution positive-ion

1884
T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
FABMS measurement. The 1H and 13C NMR spectra of 1 (CD3OD,
Tables 2 and 3), which were assigned by various NMR experi-
ments,19 showed signals assignable to two methylenes [d 2.70
(2H, m, H2-7), 3.66, 4.07 (1H each, both m, H2-8)], ortho- and
meta-coupled ABC-type aromatic protons [d 6.52 (1H, dd, J = 1.8,
7.8 Hz, H-6), 6.65 (1H, d, J = 1.8 Hz, H-2), 6.67 (1H, d, J = 7.8 Hz,
ner-Glc-C-3 (dC 80.5); and terminal-Glc-H-1 and inner-Glc-C-6 (dC
69.3) (Fig. 2). Finally, alkaline hydrolysis of 1 with 5% potassium
hydroxide (KOH) liberated trans-p-coumaric acid, which was iden-
tified by HPLC analysis, together with a deacylated product. The
deacylated product was successively treated with 1.0 M hydro-
chloric acid (HCl) to liberate L-rhamnose and D-glucose, which
H-5)], two b-
terminal-Glc-H-1), 4.54 (1H, d, J = 8.2 Hz, inner-Glc-H-1)], and a
-rhamnopyranosyl moiety [d 1.06 (3H, d, J = 6.0 Hz, Rha-H3-6),
D
-glucopyranosyl moieties [d 4.29 (1H, d, J = 7.8 Hz,
were identified by HPLC analysis using an optical rotation detec-
tor.10,11 Thus, the structure of kankanoside H1 was elucidated to
a-
L
be 2-(3,4-dihydroxyphenyl)ethyl O-
[b- -glucopyranosyl-(1?6)]-2-O-acetyl-4-O-trans-p-coumaroyl-b-
-glucopyranoside (1).
a-L-rhamnopyranosyl-(1?3)-
4.80 (1H, br s, Rha-H-1)] together with an acetyl group [d 1.98
(3H, s)] and a trans-p-coumaroyl group {an trans-olefin [d 6.34,
7.67 (1H each, both d, J = 16.0 Hz, H-8 and 7)] and ortho-coupled
A2B2-type aromatic protons [d 6.80, 7.46 (2H each, both d,
J = 8.7 Hz, H-3,5 and 2,6)]}. Connectivities between the oligoglyco-
side and acyl moieties in 1 were characterized by a HMBC experi-
ment, which showed long-range correlations between the
following proton and carbon pairs: inner-Glc-H-1 and C-8 (dC
71.9); inner-Glc-H-2 [d 4.88 (1H, dd-like)] and the acetyl carbonyl
carbon (dC 171.4); inner-Glc-H-4 [d 5.08 (1H, dd, J = 9.6, 9.6 Hz)]
and the p-coumaroyl carbonyl carbon (dC 168.2); Rha-H-1 and in-
D
D
Kankanoside H2 (2) was isolated as a white powder with nega-
tive optical rotation (½a D25
ꢁ52.4 in MeOH). Its molecular formula
ꢀ
C37H48O20 was found to be the same as that of 1 by high-resolution
positive-ion FABMS measurement. The spectroscopic properties of
2 were very similar to those of 1, except for signals due to the cis-p-
coumaroyl group [d 5.79, 6.95 (1H each, both d, J = 12.8 Hz, H-8 and
7), 6.76, 7.73 (2H each, both d, J = 8.7 Hz, H-3,5 and 2,6)]. Further-
more, HMBC experiment on 2 also revealed similar modes of long-
range correlations as those detected for 1 between the following
proton and carbon pairs: inner-Glc-H-1 [d 4.51 (1H, d, J = 7.8 Hz)]
and C-8 (dC 72.0); inner-Glc-H-2 [d 4.88 (1H, m)] and the acetyl car-
bonyl carbon (dC 171.4); inner-Glc-H-4 [d 4.98 (1H, dd, J = 9.6,
9.7 Hz)] and the p-coumaroyl carbonyl carbon (dC 166.7); Rha-H-
1 [d 4.77 (1H, d, J = 1.4 Hz, Rha-H-1)] and inner-Glc-C-3 (dC 80.7);
and terminal-Glc-H-1 [d 4.28 (1H, d, J = 7.8 Hz)] and inner-Glc-C-6
(dC 69.4) (Fig. 2). By the degradation study, 2 gave cis-p-coumaric
acid and two same sugars as those obtained in the case of 1.
Consequently, the structure of kankanoside H2 was elucidated to
Table 2
1H NMR (600 MHz, CD3OD) data for kankanosides H1 (1), H2 (2), and I (3)
Position
1
2
3
dH (J Hz)
6.65 (d, 1.8)
dH (J Hz)
6.63 (d, 1.8)
dH (J Hz)
2
3
4
5
6
7
7.26–7.28 (m)
7.26–7.28 (m)
7.18 (m)
7.26–7.28 (m)
7.26–7.28 (m)
2.95 (2H, dd, 7.3,
7.8)
be 2-(3,4-dihydroxyphenyl)ethyl O-
[b- -glucopyranosyl-(1?6)]-2-O-acetyl-4-O-cis-p-coumaroyl-b-
glucopyranoside (2).
a-L-rhamnopyranosyl-(1?3)-
6.67 (d, 7.8)
6.52 (dd, 1.8, 7.8) 6.52 (dd, 1.8, 8.2)
2.70 (2H, m)
6.65 (d, 8.2)
D
D-
2.69 (2H, m)
Kankanoside I (3) was also isolated as a white powder with neg-
8
3.66 (m)
4.07 (m)
3.62 (m)
4.04 (m)
3.79 (m)
4.10 (dt, 16.9, 7.3)
ative optical rotation (½a D25
ꢁ62.2 in MeOH). Its molecular formula,
ꢀ
C35H46O18, was determined by positive- and negative-FABMS and
8-O-Glc
HRFABMS measurements. The spectroscopic properties of
3
10
20
30
40
50
4.54 (d, 8.2)
4.88 (dd-like)
4.01 (m)
5.08 (dd, 9.6, 9.6) 4.98 (dd, 9.6, 9.7)
3.84 (m)
4.51 (d, 7.8)
4.88 (m)
3.94 (dd, 9.2, 9.6)
4.40 (d, 8.2)
3.99 (dd, 8.2, 9.2)
3.81 (m)
5.01 (dd, 9.6, 10.1)
3.76 (m)
(CD3OD, Tables 2 and 3) were superimposable on those of echina-
coside (4), except for signals due to the aglycone part {{two meth-
ylenes {d 2.95 (2H, dd, J = 7.3, 7.8 Hz, H2-7), [3.79 (1H, m), 4.10 (1H,
dt, J = 16.9, 7.3 Hz), H2-8]}, monosubstituted aromatic protons [d
7.18 (1H, m, H-4), 7.26–7.28 (4H, m, H-2,6, 3,5)]}}. Alkaline
hydrolysis of 3 with 5% KOH liberated trans-caffeic acid, and the
deacylated product was successively treated with 1.0 M HCl to lib-
3.75 (ddd, 2.3, 5.9,
9.7)
3.62 (m)
60
3.66 (dd, 6.0,
11.9)
3.56 (m)
3.95 (dd, 1.8,
11.9)
3.91 (dd, 2.3, 12.0)
3.94 (dd, 1.8, 11.8)
erate L-rhamnose and D-glucose. Finally, connectivities of the acyl
30-O-Rha
group and the sugar moieties in 3 were clarified by the HMBC
experiment, as shown in Figure 2. On the basis of above-mentioned
evidence, the structure of kankanoside I was elucidated to be
100
200
300
400
500
600
4.80 (br s)
3.50 (m)
3.62 (m)
3.25 (m)
3.50 (m)
4.77 (d, 1.4)
3.46 (m)
3.62 (m)
3.24 (m)
3.53 (m)
5.18 (d, 1.8)
3.53 (m)
3.92 (m)
3.26 (m)
3.50 (m)
2-phenylethyl O-
a
-L-rhamnopyranosyl-(1?3)-[b-
D-glucopyrano-
syl-(1?6)]-4-O-trans-caffeoyl-b-
D
-glucopyranoside (3).
1.06 (3H, d, 6.0)
1.11 (3H, d, 6.0)
1.08 (3H, d, 6.0)
60-O-Glc
1000
2.4. Effects of chemical constituents from the stems of
C. tubulosa on -GalN-induced cytotoxicity in primary cultured
mouse hepatocytes
4.29 (d, 7.8)
3.24 (m)
3.34 (m)
3.24 (m)
3.22 (m)
3.64 (dd, 5.1,
12.0)
4.28 (d, 7.8)
3.18 (m)
3.34 (m)
3.20 (m)
3.18 (m)
3.62 (m)
4.28 (d, 7.4)
3.20 (m)
3.31 (m)
3.24 (m)
3.24 (m)
D
2000
3000
4000
5000
D
-GalN/LPS-induced liver injury is recognized to develop via
immunological responses.19 This type of liver injury is reported
to occur in two forms. First, sensitivity to TNF- by depletion of
uridine triphosphate in hepatocytes is increased by -GalN. Second,
pro-inflammatory mediators, such as NO and TNF- , are released
from LPS-activated macrophages (Kupffer’s cells). Apoptosis of
hepatocytes by TNF- is reported to have an important role in
GalN/LPS-induced liver injury.20
6000
3.60 (dd, 5.0, 11.9)
a
3.83 (br d, ca. 12) 3.80 (dd, 2.3, 12.3)
3.80 (m)
D
20-O-Ac
a
1.98 (3H, s)
1.97 (3H, s)
40-O-Acyl
a
D-
2
3
5
6
7
8
7.46 (d, 8.7)
6.80 (d, 8.7)
6.80 (d, 8.7)
7.46 (d, 8.7)
7.67 (d, 16.0)
6.34 (d, 16.0)
7.73 (d, 8.7)
6.76 (d, 8.7)
6.76 (d, 8.7)
7.73 (d, 8.7)
6.95 (d, 12.8)
5.79 (d, 12.8)
7.05 (d, 2.3)
In our previous studies on hepatoprotective compounds from
natural medicines, we reported that several constituents from Hove-
nia dulcis,21 Bupleurum scorzonerifolium,22,23 Curcuma zedoaria,24–26
Angelica furcijuga,27,28 Betula platyphylla var. japonica,29 Pisum
6.77 (d, 8.2)
6.95 (dd, 2.3, 8.2)
7.59 (d, 16.0)
6.27 (d, 16.0)

T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
1885
Table 3
inhibitory effect on
D
-GalN-induced cytotoxicity in primary cultured
13C NMR (150 MHz, CD3OD) data for kankanosides H1 (1), H2 (2), and I (3)
hepatocytes. Since the methanolic extract from the stems of C.
tubulosa showed hepatoprotective effects on -GalN/LPS-induced li-
ver injury in mice (vide ante), inhibitory effect of the constituents on
-GalN-induced cytotoxicity in primary cultured mouse hepato-
D
Position
1
2
3
dC
dC
dC
D
1
2
3
4
5
6
7
8
131.7
117.2
146.0
144.6
116.3
121.3
36.3
131.8
117.2
146.0
144.6
116.3
121.3
36.3
140.0
130.0
129.4
127.2
129.4
130.0
37.2
cytes was examined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assay. As shown in Table 4,
echinacoside(4, IC50 = 10.2
(6, 5.3
M), 20-acetylacteoside (8, 4.8
8.6
M) and B11 (22, 14.6 M), and kankanoside G11 (23, 14.8
showed strong activity. Their activities were greater than that of
commercial silybin (38.8
M),33–43 as a positive control.44,45 The
l
M), acteoside(5, 4.6
M), tubulosides A (10,
M)
lM), isoaceteoside
l
l
l
l
l
71.9
72.0
72.1
l
8-O-Glc
10
20
30
40
50
60
101.7
75.1
80.5
70.7
74.8
69.3
101.7
75.1
80.7
70.8
74.9
69.4
104.2
76.1
81.6
70.4
74.7
69.4
structural requirements of the phenylethanoid glycosides for the
activity were as followings; (1) the aglycone part was essential for
the activity [4 >> kankanose (21, >100
>100 M)];(2)theaglyconehavingthe3,4-dihydroxygroupshowed
stronger activity than that having the 4-hydroxy group (6 > 23); (3)
the 60-O-b-
-glucopyranosyl moiety reduced the activity (4 < 5,
10 < 8); (4) the 8-O-b-
lM), 5 >> cistanoside F (20,
l
D
30-O-Rha
D
-glucopyranosyl part having the 40-O-caf-
100
200
300
400
500
600
103.3
71.8
72.6
73.6
70.8
18.4
103.5
71.9
72.6
73.7
70.8
18.2
103.1
72.1
72.3
73.8
70.6
18.5
feoyl group showed stronger activity than that having the 60-O-caf-
feoryl group (5 = 6, 8 > 22); and (5) introduction of the 20-O-acetyl
moiety reduced the activity (8 5 5, 22 < 6).
Next, effects of the principal constituents (4–6) on NO and TNF-
a
productions, as markers of macrophage activation in LPS-acti-
60-O-Glc
1000
vated mouse peritoneal macrophages were examined.43,46,47 As
104.7
74.9
77.8
71.5
77.9
62.6
104.8
74.9
77.9
71.5
78.0
62.6
104.7
75.1
77.8
71.5
77.9
62.6
2000
the result, these constituents (4–6) showed neither NO nor TNF-
3000
a
production inhibitory activities (IC50 >100
Thus, these compounds were found not to affect the overproduc-
tions of NO and TNF- from LPS-activated macrophages.
lM, data not shown).
4000
5000
a
6000
20-O-Ac
1
2
2.5. Effects of the chemical constituents on TNF-
cytotoxicity in L929 cells
a-induced
20.9
171.4
20.9
171.4
40-O-Acyl
In order to clarify the effects of the constituents on the sensitiv-
ity of hepatocytes to TNF- , TNF- -induced decrease in viability of
L929 cells, a TNF-
-sensitive cell line,48 was examined using the
MTT assay. Without test samples, the cell viability was reduced
to ca. 60% after incubation of the cells with 20 ng/mL TNF- for
44 h compared to the case without TNF- . As shown in Table 5,
echinacoside (4, IC50 = 31.1 M), acteoside (5, 17.8 M), isoaceteo-
side (6, 22.7 M), tubuloside A (10,
M), 20-acetylacteoside (8, 25.7
23.2 M), and cistantubuloside B1 (11, 21.4 M) inhibited the de-
1
2
3
4
5
6
7
8
9
127.0
131.4
116.9
161.5
116.9
131.4
148.0
114.7
168.2
127.5
134.4
115.9
160.5
115.9
134.4
147.6
115.7
166.7
127.6
115.3
146.8
149.8
116.5
123.3
148.2
114.7
168.5
a
a
a
a
a
l
l
l
l
l
l
crease of the cell viability, and their activities were found greater
than those of piperine42,43 and silybin. The structural requirements
of the phenylethanoid glycosides for the activity were as follow-
ings; (1) the aglycone part was essential for the activity [4 >> kan-
kanose (21, >100 lM)]; (2) the aglycone having the 3,4-dihydroxy
group showed stronger activity than that having the 4-hydroxy
sativum,30 Salacia reticulata,31 Tilia argentea,32 Anastatica hierochun-
tica,33 Panax notoginseng,34 Cyperus longus,35 Erycibe expansa,36
Camellia sinensis,37 Sedum sarmentosum,38,39 Sinocrassula indica,40
Hedychium coronarium,41 and Piper chaba42,43 showed hepatoprotec-
tive effects on the liver injury induced by
D-GalN/LPS in mice and/or
HO
O
HO
HO
O
O
H
5
O
O
O
OH
OH
OH
OH
6
1
4
OH
OH
O
OH
1'''
8
3
HO
HO
H
OH
HO
HO
HO
HO
H
OH
O
O
H
O
H
7
2
O
2'
O
O
O
OH
O
O
O
1'
O
O
4'
H
1''
O
OH
H
H
H
OHOH
OHOH
OHOH
O
O
HO
HO
HO
O
O
O
O
HO
O
O
1
2
3
1
1H–H COSY:
HMBC :
Figure 2. 1H–1H COSY and HMBC correlations for 1–3.

1886
T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
Table 4
Inhibitory effects of the methanolic extract and its constituents from stems of C. tubulosa on D-GalN-induced cytotoxicity in primary cultured mouse hepatocytes
Inhibition (%)
0
l
g/mL
3
l
g/mL
10
l
g/mL
30
l
g/mL
100
l
g/mL
IC50
(
l
l
g/mL)
M)
MeOH extract
MeOH-eluted fraction
H2O-eluted fraction
0.0 1.8
0.0 1.5
0.0 1.5
9.1 2.9a
17.3 1.9b
26.6 4.1b
0.0 2.8
29.2 1.4b
40.2 3.6b
ꢁ2.1 2.1
53.0 2.4b
73.5 0.9b
ꢁ8.7 1.5
97.3
34.9
13.6 2.5a
ꢁ1.8 2.9
0
lM
3
lM
10
lM
30
lM
100
lM
IC50
(
Kankanoside H1 (1)
Kankanoside H2 (2)
Kankanoside I (3)
Echinacoside (4)
Acteoside (5)
Isoacteoside (6)
20-Acetylacteoside (8)
Tubuloside A (10)
Cistantubuloside B1 (11)
Wiedemanninoside C (14)
Cistantubuloside A (15)
Syringalide A 30-O-Rha (16)
Salidroside (19)
Cistanoside F (20)
Kankanose (21)
Tubuloside B (22)
Kankanoside G (23)
Silybinc
0.0 1.8
0.0 0.6
0.0 0.6
0.0 2.1
0.0 2.4
0.0 4.4
0.0 1.9
0.0 3.7
0.0 1.0
0.0 0.5
0.0 1.9
0.0 1.3
0.0 1.8
0.0 1.5
0.0 2.8
0.0 4.4
0.0 3.0
0.0 0.3
8.7 3.2
4.4 1.1
3.9 0.6
16.4 4.2a
11.6 1.3b
13.6 0.3b
46.7 4.3b
71.8 2.3b
57.3 2.2b
58.4 5.3b
50.2 4.6b
10.3 1.7b
11.5 0.9b
8.2 3.4
20.4 2.2b
18.2 1.9b
25.9 1.7b
67.7 1.7b
119.2 5.4b
101.2 5.9b
95.2 3.2b
74.6 0.9b
18.5 1.6b
20.6 2.6b
17.0 4.1b
35.7 4.0b
ꢁ0.7 1.8
7.7 3.9
34.0 2.4b
26.3 0.9b
27.7 2.5b
32.8 1.4b
40.9 1.3b
43.7 2.1b
41.9 3.2b
31.1 1.6b
3.1 1.2
10.2
4.6
5.3
4.8
8.6
31.2 2.7b
39.4 2.8b
15.3 3.4b
55.7 6.1b
0.2 1.3
4.5 1.7
3.0 1.5
9.7 0.7
0.9 0.6
21.4 1.5b
1.4 1.4
4.0 2.6
ꢁ1.3 2.9
33.6 4.5b
33.3 3.3b
7.7 0.7
71.2
2.0 0.7
21.2 0.8a
ꢁ2.8 2.8
ꢁ4.9 1.3
ꢁ7.9 2.1
75.4 2.8b
72.7 4.1b
45.2 8.8b
8.6 2.3
14.6
14.8
38.8
12.6 3.6a
4.8 1.1
77.0 5.5b
Each value represents the mean S.E.M. (N = 4).
Significantly different from the control, ap <0.05, bp <0.01.
c
Commercial silybin was purchased from Funakoshi Co., Ltd (Tokyo, Japan).
group [4 > cistantubuloside A (15, >100
30-O-
-rhamnopyranoside (16, >100 M), 6 > kankanoside
(23, >100 -glucopyranosyl moiety reduced
M)]; (3) the 60-O-b-
the activity (4 < 5); (4) the 8-O-b- -glucopyranosyl part having
the 40-O-caffeoyl group showed stronger activity than that having
l
M), 5 > syringalide A
ments for their protective effects on hepatocyte injury induced
by -GalN, and cytoprotective effects on L929 cells caused by
TNF- were as followings; (1) the aglycone part was essential for
a-
L
l
G
D
l
D
a
D
the activity; (2) the aglycone having the 3,4-dihydroxy group
showed stronger activity than that having the 4-hydroxy group;
the 60-O-caffeoryl group [5 > 6, 8 > tubuloside B (22, >100
lM)];
(3) the 60-O-b-
the 8-O-b-
D
-glucopyranosyl moiety reduced the activity; (4)
and (5) introduction of the 20-O-acetyl moiety reduced the activity
(8 < 5, 22 < 6). These requirements were found similar to those ob-
served in the previous chapter concerning the inhibitory effects on
D
-glucopyranosyl part having the 40-O-caffeoyl group
showed stronger activity than that having the 60-O-caffeoryl group;
and (5) introduction of the 20-O-acetyl moiety reduced the activity.
Furthermore, principal phenylethanoid glycosides, echinacoside
(4), acteoside (5), and isoacteoside (6), inhibited the increase in
D-GalN-induced cytotoxicity in primary cultured mouse hepatocytes.
These in vitro findings suggest the following possible mecha-
nisms of action for the hepatoprotective effect of the constituents
of stems of C. tubulosa: (i) decreasing of -GalN-induced cytotoxic-
sAST and sALT at doses of 25–100 mg/kg po in
mice, and those inhibitory effects were suggested to be dependent
on the reduced sensitivity of hepatocytes to TNF- . The detail
D-GalN/LPS-treated
D
ity (4–6, 8, 10, 22, and 23), and (ii) decreasing of TNF-
a-induced
a
cytotoxicity (4–6, 8, 10, and 11).
mechanisms of action of the phenylethanoid glycosides should
be studied further.
2.6. Protective effects of echinacoside (4), acteoside (5), and
3. Experimental
3.1. General
isoacteoside (6) against liver injury induced by
mice and their mode of action
D-GalN/LPS in
Finally, we examined the effect of principal phenylethanoid gly-
cosides, echinacoside (4), acteoside (5), and isoacteoside (6) on the
Thefollowinginstrumentswereusedto obtainspectralandphys-
ical data: specific rotations, Horiba SEPA-300 digital polarimeter
(l = 5 cm); UV spectra, Shimadzu UV-1600 spectrometer; IR spectra,
Shimadzu FTIR-8100 spectrometer; 1H and 13C NMR spectra, JEOL
JNM-ECA600 (600 and 150 MHz) and JEOL JNM-ECS400 (400 and
100 HMz) spectrometers with tetramethylsilane as an internal stan-
dard; FABMS and high-resolution FABMS, JEOL JMS-SX 102A mass
spectrometer; HPLC detector, Shimadzu RID-10A refractive index,
Shimadzu SPD-10A UV–vis, and Shodex OR-2 optical rotation detec-
D
-GalN/LPS-induced liver injury in mice. As shown in Table 6 and
4–6 significantly inhibited the increase of sAST and sALT induced
by -GalN/LPS in mice at doses of 25–100 mg/kg, po. On the other
hand, 4–6 did not affect the overproduction of TNF- from LPS-
activated macrophages. Furthermore, 4–6 significantly inhibited
the death of L929 cells caused by TNF- , indicating a decrease in
the sensitivity of L929 cells to TNF- (vide ante). These findings
suggest that 4–6 reduce the sensitivity of these cells to the TNF-
-induced cytotoxicity. Many compounds that inhibit cell death
induced by -GalN and production of TNF- have been re-
ported,24,26,32 but there reported few compounds that selectively
reduce the sensitivity of hepatocytes to TNF-a 42,43,49,50
D
a
a
a
tors. HPLC column, Cosmosil 5C18-MS-II and pNAP (Nacalai Tesque
a
Inc., 250 ꢂ 4.6 mm id) and (250 ꢂ 20 mm id) columns were used
D
a
for analytical and preparative purposes, respectively.
The following experimental conditions were used for chroma-
tography: normal-phase silica gel column chromatography (CC),
silica gel 60 N (Kanto Chemical Co., Ltd, 63–210 mesh, spherical,
neutral); reversed-phase silica gel CC, Diaion HP-20 (Nippon
Rensui) and Chromatorex ODS DM1020T (Fuji Silysia Chemical,
.
In conclusion, from fresh stems of C. tubulosa, three new phe-
nylethanoid glycosides, kankanosides H1 (1), H2 (2), and I (3) to-
gether with 16 known phenylethanoid glycosides (4–19) and two
acylated oligosugars (20, 21) were isolated. Structural require-

T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
1887
Table 5
Inhibitory effects of the methanolic extract and its constituents from stems of C. tubulosa on TNF-
a-induced cytotoxicity in L929 cells
Inhibition (%)
0
l
g/mL
3
3
l
g/mL
10
l
g/mL
30
l
g/mL
100
l
g/mL
IC50
(
(
l
g/mL)
MeOH extract
MeOH-eluted fraction
H2O-eluted fraction
0.0 1.4
0.0 7.5
0.0 0.7
17.6 8.1
21.5 8.3
9.1 2.0
40.5 5.3b
58.4 4.9b
18.6 6.2
58.3 4.6b
76.0 2.3b
11.7 3.7
47.9 4.4b
77.4 15.1b
ꢁ31.1 4.5
18.4
9.0
0
lM
lM
10
lM
30
lM
100
l
M
IC50
lM)
Echinacoside (4)
Acteoside (5)
Isoacteoside (6)
20-Acetylacteoside (8)
Tubuloside A (10)
Cistantubuloside B1 (11)
Cistantubuloside A (15)
Syringalide A 30-O-Rha (16)
Campneoside I (17)
Salidroside (19)
0.0 4.8
0.0 1.1
0.0 1.2
0.0 3.1
0.0 2.4
0.0 3.9
0.0 2.3
0.0 2.9
0.0 2.0
0.0 6.1
0.0 1.9
0.0 4.9
0.0 2.8
0.0 1.3
0.0 2.6
5.2 3.5
16.4 1.3a
ꢁ4.6 3.5
2.3 5.0
22.5 1.6b
24.1 4.6b
19.0 2.6
8.9 6.6
45.7 6.0b
58.4 2.5b
61.9 5.9b
64.1 4.9b
55.2 2.8b
32.8 10.8b
4.6 1.6
80.4 4.5b
91.9 5.3b
102.4 8.7b
107.3 10.4b
101.9 2.2b
122.7 13.7b
11.2 1.1a
22.2 6.4b
7.5 3.1
31.1
17.8
22.7
25.7
23.2
21.4
14.7 4.6a
36.2 4.8b
31.0 4.4b
3.6 0.5
ꢁ14.7 17.2
2.8 1.2
4.5 1.0
7.7 2.9
4.6 1.4
13.3 3.3
1.9 5.8
ꢁ8.8 8.5
ꢁ1.2 7.9
ꢁ1.1 1.2
10.7 4.7
1.3 0.9
ꢁ8.3 10.5
2.2 1.8
ꢁ5.4 5.1
ꢁ1.0 4.8
Kankanose (21)
1.3 1.8
0.8 0.1
Tubuloside B (22)
Kankanoside G (23)
Piperine42,43
13.4 4.7
4.7 0.5
36.4 13.3a
3.1 2.6
39.2 6.3b
2.9 1.4
5.5 1.6a
5.3 2.8
5.3 1.4a
22.0 3.8b
10.6 0.9b
48.0 4.1b
41.8 1.4b
50.8 3.9b
Silybinc
60.4
Each value represents the mean S.E.M. (N = 4).
Significantly different from the control, ap<0.05, bp<0.01.
c
Commercial silybin was purchased from Funakoshi Co., Ltd (Tokyo, Japan).
Table 6
Inhibitory effects of principal constituents (4–6) from stems of C. tubulosa on D-GalN/LPS-induced liver injury in mice
Treatment
Dose (mg/kg, po)
n
sAST
sALT
(Karmen unit)
Inhibition (%)
(Karmen unit)
Inhibition (%)
Normal (vehicle)
—
—
5
12
8
8
8
8
8
8
5
58 6b
11,768 1621
4562 1413a
3914 1181b
5736 3048a
3703 1594b
6339 1950
3425 848b
95 5b
—
—
25 2b
—
—
Control (D-GalN/LPS)
5484 666
3084 1117
2634 920
3047 1462
2220 1045a
3278 1021
2265 567a
19 1b
Echinacoside (4)
25
100
25
100
25
100
—
61.2
66.7
51.3
68.5
46.1
70.9
—
43.8
52.0
44.4
59.5
40.2
58.7
—
Acteoside (5)
Isoacteoside (6)
Normal (vehicle)
Control (
Hydrocortisone
D
-GalN/LPS)
—
10
8
7
9126 1477
—
94.2
9830 1605
—
97.7
627 262b
247 123b
Each value represents the mean S.E.M.
Significantly different from the control, ap<0.05, bp<0.01.
Ltd, 100–200 mesh); normal-phase TLC, pre-coated TLC plates with
silica gel 60F254 (Merck, 0.25 mm); reversed-phase TLC, pre-coated
TLC plates with silica gel RP-18 F254S (Merck, 0.25 mm); reversed-
phase HPTLC, pre-coated TLC plates with silica gel RP-18 WF254S
(Merck, 0.25 mm), detection was achieved by spraying with 1%
Ce(SO4)2–10% aqueous H2SO4, followed by heating.
(167.84 g, 5.63% and 81.21 g, 2.73%, respectively). The MeOH-eluted
fraction (61.00 g) was subjected to normal-phase silica gel CC
[1.8 kg, CHCl3–MeOH–H2O (15:3:0.4?10:3:0.5?6:4:1, v/v/v)?M
eOH] to give seven fractions [Fr. 1 (1.12 g), 2 (9.56 g), 3 (0.89 g), 4
(10.69 g), 5 (8.84 g), 6 (12.52 g), and 7 (4.60 g)]. The fraction 2
(9.56 g) was separated by reversed-phase silica gel CC [400 g,
MeOH–H2O (10:90, v/v)?MeOH?acetone] to give five fractions
[Fr. 2–1 (55.9 mg), 2–2 (4.48 g), 2–3 (3.42 g), 2–4 (1.16 g), and 2–5
(31.9 mg)]. The fraction 2–3 (1.00 g) was subjected to HPLC [Cosmo-
sil 5C18-MS-II, CH3CN–1% aqueous AcOH (7:93, v/v)] to give seven
fractions [Fr. 2–3–1 (66.5 mg), 2–3–2 (20.4 mg), 2–3–3 (26.0 mg),
2–3–4 (136.6 mg), 2–3–5 (380.6 mg), 2–3–7 (84.9 mg)]. The fraction
3.2. Plant material
Fresh stems of C. tubulosa cultivated at Urumuqi, Xinjiang Prov-
ince, China were collected in September 2007. The plant material
was identified by one of the authors (M.Y.). A voucher of the plant
material is on file in our laboratory.
2–3–4 (106.6 mg) was further purified by HPLC [Cosmosil
pNAP,
CH3CN–1% aqueous AcOH (5:95, v/v)] to give salidroside (19,
13.7 mg, 0.0027%). The fraction 4 (10.69 g) was separated by re-
versed-phase silica gel CC [500 g, MeOH–H2O (30:70,
v/v)?MeOH?acetone] to give four fractions [Fr. 4–1 (878.2 mg),
4–2 (7.06 g), 4–3 (1.57 g), and 4–4 (792.8 mg)]. The fraction 4–2
(1.50 g) was subjected to HPLC [Cosmosil 5C18-MS-II, CH3CN–1%
aqueous AcOH (20:80, v/v)] to give five fractions {Fr. 4–2–1
(42.9 mg), 4–2–2 [= campneoside I (17, 67.0 mg, 0.014%)], 4–2–3
3.3. Extraction and isolation
The fresh stems of C. tubulosa (2.98 kg) were finely cut and ex-
tracted three times with methanol under reflux for 3 h. Evaporation
of the solvent under reduced pressure provided a methanolic extract
(249.1 g, 8.36%). The methanolic extract was subjected to Diaion HP-
20 CC (5.0 kg, H2O?MeOH) to give H2O- and MeOH-eluted fractions

1888
T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
[= acteoside (5, 1.21 g, 0.26%)], 4–2–4 (41.6 mg), and 4–2–5
(181.3 mg)}. The fraction 4–2–4 (41.6 mg) was further purified by
CD3OD) dC: given in Table 3. Positive-ion FABMS: m/z 835 (M+Na)+.
Negative-ion FABMS: m/z 811 (MꢁH)ꢁ.
HPLC [Cosmosil
pNAP, CH3CN–1% aqueous AcOH (18:82, v/v)] to
give isoacteoside (6, 19.1 mg, 0.0040%) and cis-acteoside (7,
3.4 mg, 0.0007%). The fraction 4–3 (1.57 g) was purified by HPLC
[Cosmosil 5C18-MS-II, CH3CN–1% aqueous AcOH (20:80, v/v)] to give
11 fractions [Fr. 4–3–1 (30.4 mg), 4–3–2 (55.2 mg), 4–3–3 (= 17,
22.1 mg, 0.0010%), 4–3–4 (= 5, 224.6 mg, 0.010%), 4–3–5 (27.4 mg),
4–3–6 (43.6 mg), 4–3–7 (= 6, 825.0 mg, 0.037%), 4–3–8 [= syringa-
3.3.3. Kankanoside I (3)
A white powder, ½a D25
ꢁ62.2 (c 1.00, MeOH). High-resolution po-
ꢀ
sitive-ion FABMS: Calcd for C35H46O18Na (M+Na)+: 777.2581.
Found: 777.2585. UV [MeOH, nm (log
4.04), 334 (4.23). IR (KBr): 3415, 1717, 1647, 1603, 1508, 1070,
1046 cmꢁ1 1H NMR (600 MHz, CD3OD) d: given in Table 2. 13C
e)]: 246 (3.97), 295 (sh,
.
lide A 30-O-
a
-
L
-rhamnopyranoside (16, 37.6 mg, 0.0017%)], 4–3–9
NMR (150 MHz, CD3OD) dC: given in Table 3. Positive-ion FABMS:
(39.8 mg), 4–3–10 [= 20-acetylacteoside (8, 85.4 mg, 0.0038%)], and
m/z 777 (M+Na)+. Negative-ion FABMS: m/z 753 (MꢁH)ꢁ.
4–3–11 (64.6 mg)]. The fraction 4–3–6 (43.6 mg) was further puri-
fied by HPLC [Cosmosil
pNAP, CH3CN–1% aqueous AcOH (18:82,
3.4. Alkaline and acid hydrolysis of 1–3
v/v)] to give kankanosides H1 (1, 17.0 mg, 0.0008%) and H2 (2,
3.3 mg, 0.0001%). The fraction 4–3–9 (39.8 mg) was further purified
by HPLC [Cosmosil pNAP, CH3CN–1% aqueous AcOH (18:82, v/v)] to
Solutions of 1, 2, and 3 (each 1.5 mg) in 5% aqueous potassium
hydroxide (KOH, 0.5 mL) were stirred at 40 °C for 1 h. Each solution
was neutralized with Dowex HCR W2 (H+ form), and the resins
were removed by filtration. Evaporation of the solvent under re-
duced pressure yielded the corresponding deacylated products,
give kankanoside I (3, 15.4 mg, 0.0007%) and 6 (3.1 mg, 0.0001%).
The fraction 5 (8.84 g) was separated by reversed-phase silica gel
CC [400 g, MeOH–H2O (20:80?30:70, v/v)?MeOH?acetone] to
give seven fractions [Fr. 5–1 (870.2 mg), 5–2 (478.9 mg), 5–3
(3.72 g), 5–4 (979.9 mg), 5–5 (1.19 g), 5–6 (1.27 g), and 5–7
(130.1 mg)]. The fraction 5–1 (870.2 mg) was purified by HPLC [Cos-
which were subjected to HPLC analysis [column: Cosmosil
pNAP,
250 ꢂ 4.6 mm id; mobile phase: CH3CN–1% aqueous AcOH
(15:85, v/v); detection: UV (254 nm); flow rate: 1.0 mL/min] to
be identified as trans-caffeic acid (tR 9.9 min from 3), trans-p-cou-
maric acid (tR 17.1 min from 1), and cis-p-coumaric acid (tR
17.8 min from 2), respectively. Then each was dissolved in 1.0 M
HCl (1.0 mL) and heated at 80 °C for 3 h. After being cooled, the
reaction mixture was neutralized with Amberlite IRA-400 (OHꢁ
form), and the resins were removed by filtration. After removal
of the solvent under reduced pressure, the residue was separated
by Sep-Pak C18 cartridge column (H2O?MeOH). The H2O-eluted
fraction was subjected to HPLC analysis under the following condi-
tions: HPLC column, Kaseisorb LC NH2-60-5, 4.6 mm id ꢂ 250 mm
(Tokyo Kasei Co., Ltd, Tokyo, Japan); detection, optical rotation
[Shodex OR-2 (Showa Denko Co., Ltd, Tokyo, Japan); mobile phase,
CH3CN–H2O (85:15, v/v); flow rate 0.8 mL/min]. Identification of
mosil pNAP, CH3CN–1% aqueous AcOH (5:95, v/v)] to give decaffeoy-
lacteoside (9, 5.1 mg, 0.0002%) and cistanoside F (20, 8.1 mg,
0.0004%). The fraction 5–3 (3.72 g) was purified by HPLC [Cosmosil
5C18-MS-II, CH3CN–1% aqueous AcOH (15:85, v/v)] to give six frac-
tions [Fr. 5–3–1 (128.6 mg), 5–3–2 [= kankanose (21, 9.4 mg,
0.0004%)], 5–3–3 (64.6 mg), 5–3–4 (72.3 mg), 5–3–5 [= echinaco-
side (4, 2.54 g, 0.11%), and 5–3–6 (318.5 mg)]. The fraction 5–3–4
(72.3 mg) was further purified by HPLC [Cosmosil
pNAP, CH3CN–
1% aqueous AcOH (10:90, v/v)] to give campneoside II (18,
10.6 mg, 0.0005%). Thefraction 5–4 (979.9 mg) was purified by HPLC
[Cosmosil 5C18-MS-II, CH3CN–1% aqueous AcOH (15:85, v/v)] to give
nine fractions [Fr. 5–4–1 (28.6 mg), 5–4–2 (90.5 mg), 5–4–3
(99.7 mg), 5–4–4 (= 4, 25.7 mg, 0.0012%), 5–4–5 (16.4 mg), 5–4–6
(318.5 mg), 5–4–7 (21.6 mg), 5–4–8 (= 5, 204.4 mg, 0.0091%), and
5–4–9 (134.7 mg)]. The fraction 5–4–6 (318.5 mg) was purified by
L-rhamnose (i) and D-glucose (ii) liberated from 1, 2, or 3 in the
H2O-eluted fraction were carried out by comparison of their reten-
tion times and optical rotation with those of authentic samples [i,
tR 9.9 min (negative)] and [ii, tR 17.9 min (positive)].
HPLC [Cosmosil
pNAP, CH3CN–1% aqueous AcOH (15:85, v/v)] to
give cistantubulosides B1 (11, 44.9 mg, 0.0020%), B2 (12, 7.6 mg,
0.0003%), and A (15, 21.1 mg, 0.0009%) and wiedemanninoside C
(14, 17.2 mg, 0.0008%). The fraction 5–5 (1.19 g) was purified by
HPLC [Cosmosil 5C18-MS-II, CH3CN–1% aqueous AcOH (18:82, v/v)]
to give tubuloside A (10, 300.3 mg, 0.013%) and arenarioside (13,
5.8 mg, 0.0006%). The fraction 6 (12.52 g) was separated by re-
versed-phase silica gel CC [600 g, MeOH–H2O (20:80?30:70,
v/v)?MeOH?acetone] to give four fractions [Fr. 6–1 (553.8 mg),
6–2 (10.69 g), 6–3 (1.40 g), and 6–4 (17.8 mg)]. The fraction 6–2
(500.0 mg) was purified by HPLC [Cosmosil 5C18-MS-II, CH3CN–1%
aqueous AcOH (25:75, v/v)] to give 4 (353.1 mg, 0.34%).
3.5. Bioassay
3.5.1. Reagents
LPS (from Salmonella enteritidis), minimum essential medium
(MEM), and William’s E medium were purchased from Sigma–Al-
drich Chemical (St. Louis, MO, USA); fetal bovine serum (FBS)
was from Invitrogen (Carlsbad, CA, USA); and other chemicals were
from Wako Pure Chemical Industries, Co., Ltd (Osaka, Japan). 96-
Well microplates were purchased from Sumitomo Bakelite Co.,
Ltd (Tokyo, Japan).
3.3.1. Kankanoside H1 (1)
3.5.2. Animals
A white powder, ½a D24
ꢁ45.3 (c 1.06, MeOH). High-resolution po-
ꢀ
Male ddY mice were purchased from Kiwa Laboratory Animal
Co., Ltd, (Wakayama, Japan). The animals were housed at a con-
stant temperature of 23 2 °C and were fed a standard laboratory
chow (MF, Oriental Yeast Co., Ltd, Tokyo, Japan). All of the experi-
ments were performed with conscious mice unless otherwise
noted. The experimental protocol was approved by the Experimen-
tal Animal Research Committee of Kinki University.
sitive-ion FABMS: Calcd for C37H48O20Na (M+Na)+: 835.2637.
Found: 835.2633. UV [MeOH, nm (log
4.36), 317 (4.52). IR (KBr): 3416, 1736, 1638, 1605, 1518, 1070,
1044 cmꢁ1 1H NMR (600 MHz, CD3OD) d: given in Table 2. 13C
e)]: 226 (4.33), 293 (sh,
.
NMR (150 MHz, CD3OD) dC: given in Table 3. Positive-ion FABMS:
m/z 835 (M+Na)+. Negative-ion FABMS: m/z 811 (MꢁH)ꢁ.
3.3.2. Kankanoside H2 (2)
3.5.3. Protective effects on
mice
D-GalN/LPS-induced liver injury in
A white powder, ½a D25
ꢁ52.4 (c 0.27, MeOH). High-resolution po-
ꢀ
sitive-ionFABMS:Calcdfor C37H48O20Na (M+Na)+: 835.2637. Found:
The method described by Tiegs et al.51 was modified and used for
this experiment. Briefly, male ddY mice weighing about 25–30 g
835.2644. UV [MeOH, nm (log
(4.37). IR (KBr): 3420, 1717, 1638, 1605, 1508, 1159, 1067 cmꢁ1
1H NMR (600 MHz, CD3OD) d: given in Table 2. 13C NMR (150 MHz,
e)]: 228 (4.26), 290 (sh, 4.22), 316
.
were fasted for 20 h before the experiment.
D-GalN (350 mg/kg)
and LPS (10 g/kg) dissolved in saline were injected intraperitone-
l

T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
1889
ally to produce liver injury. Each test sample was given orally 1 h be-
fore the -GalN/LPS injection. Blood samples were collected from the
infraorbital venous plexus 10 h after -GalN/LPS injection. sAST and
(100 lg/mL) at 37 °C under a 5% CO2 atmosphere. The cells were
D
inoculated in 96-well tissue culture plate [3 ꢂ 104 cells/well in
D
100
lL/well in MEM]. After 44 h of incubation in the medium
sALT levels were determined using the Reitman–Frankel method
(commercial kit, Transaminase CII-Test Wako, Wako Pure Chemical
Industries, Co., Ltd). Hydrocortisone was used as a reference com-
pound. Test samples were suspended with 5% Arabic gum solution,
and the solution was administered orally at 10 mL/kg in each exper-
iment, while the vehicle was given orally at 10 mL/kg in the corre-
sponding control group.
TNF-
a (20 ng/mL) with or without the test sample, the viability
of the cells was assessed by the MTT colorimetric assay (vide
ante).49,50 Each test compound was dissolved in DMSO, and the
solution was added to the medium (final concentration in DMSO
0.5%).
3.6. Statistics
3.5.4. Protective effects on cytotoxicity induced by
primary cultured mouse hepatocytes
D
-GalN in
Values are expressed as means S.E.M. One-way analysis of
variance (ANOVA) followed by Dunnett’s test was used for statisti-
cal analysis. Probability (p) values less than 0.05 were considered
significant.
The hepatoprotective effect of the constituents were deter-
mined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoli-
um bromide (MTT) colorimetric assay using primary cultured
mouse hepatocytes.33–43 Hepatocytes were isolated from male
ddY mice (30–35 g) by collagenase perfusion method. A cell sus-
Acknowledgments
pension at 4 ꢂ 104 cells in 100
l
L William’s E medium containing
T.M., K.N., and O.M. were supported by a Grant-in Aid for Scien-
tific Research from ‘High-tech Research Center’ Project for Private
Universities: matching fund subsidy from a Grant-in Aid for Scien-
tific Research from the Ministry of Education, Culture, Sports,
Science and Technology (MEXT) of Japan, 2007–2011 and also sup-
ported by a Grant-in Aid for Scientific Research from MEXT. M.Y.
and H.M. were supported by the 21st COE program, Academic Fron-
tier Project, and a Grant-in Aid for Scientific Research from MEXT.
FBS (10%), penicillin G (100 units/mL), and streptomycin (100
l
g/mL) was inoculated in a 96-well microplate and pre-cultured
for 4 h at 37 °C under a 5% CO2 atmosphere. The medium was
added with 100 of the fresh medium containing -GalN
(2 mM) with or without the test sample and the hepatocytes were
cultured for 44 h. The medium was exchanged with 100 L of the
fresh medium, and 10 L of MTT [5 mg/mL in phosphate buffered
saline (PBS)] solution was added to the medium. After 4 h of incu-
bation, the medium was removed, and 100 L of isopropanol con-
lL
D
l
l
l
References and notes
taining 0.04 M HCl was then added to dissolve the formazan
produced in the cells. The optical density (O.D.) of the formazan
solution was measured by microplate reader at 570 nm (reference:
655 nm). Inhibition (%) was obtained by following formula.
1. This paper is number 35 in the series ‘Bioactive constituents from Chinese
natural medicines’. The number 34: Muraoka, O.; Morikawa, T.; Zhang, Y.;
Ninomiya, K.; Nakamura, S.; Matsuda, H.; Yoshikawa, M. Tetrahedron 2009, 65,
4142.
2. Kobayashi, H.; Oguchi, H.; Takizawa, N.; Miyase, T.; Ueno, A.; Usmanghani, K.;
Ahmad, M. Chem. Pharm. Bull. 1987, 35, 3309.
3. Xinjiang Science and Technology Press, ‘Culture Techniques of Xinjiang Staple
Medicinal Plants’, Xinjiang Institute of Traditional Chinese and Ethnologic
Medicines Ed. 2004, pp. 84–88.
Inhibition ð%Þ ¼ ½ðO:D:ðsampleÞ ꢁ O:D:ðcontrolÞÞ=ðO:D:ðnormalÞ
ꢁ O:D:ðcontrolÞÞꢀ ꢂ 100
4. Yoshizawa, F.; Deyama, T.; Takizawa, N.; Usmanghani, K.; Ahmad, M. Chem.
Pharm. Bull. 1990, 38, 1927.
5. Du, N.; Zhou, P.; Wang, J.; Liu, C.; Li, W. Zhongguo Yaoke Daxue Xuebao 1993, 24,
46.
6. Xue, D. Zhongguo Zhongyao Zazhi 1997, 22, 170.
7. Song, Z.; Mo, S.; Chen, Y.; Tu, P.; Li, W.; Zhao, Y.; Zheng, J. Zhongguo Zhongyao
Zazhi 2000, 25, 728.
3.5.5. Effects on production of NO in LPS-stimulated mouse
peritoneal macrophages
Screening test for NO production using TGC-induced mouse per-
itoneal macrophages was performed as described previously.43,46,47
8. Song, Z.; Tu, P.; Zhao, Y.; Zheng, J. Zhongcaoyao 2000, 31, 808.
9. Tu, P.; Song, Z.; Shi, H.; Jiang, Y.; Zhao, Y. Helv. Chim. Acta 2006, 89, 927.
10. Xie, H.; Morikawa, T.; Matsuda, H.; Nakamura, S.; Muraoka, O.; Yoshikawa, M.
Chem. Pharm. Bull. 2006, 54, 669.
11. Yoshikawa, M.; Matsuda, H.; Morikawa, T.; Xie, H.; Nakamura, S.; Muraoka, O.
Bioorg. Med. Chem. 2006, 14, 7468.
12. Kikuchi, M.; Yamauchi, Y. Yakugaku Zasshi 1985, 105, 442.
13. Budzianowski, J.; Skrzypczak, L. Phytochemistry 1995, 38, 997.
14. Karasawa, H.; Kobayashi, H.; Takizawa, N.; Miyase, T.; Fukushima, S. Yakugaku
Zasshi 1986, 106, 562.
3.5.6. Effects on production of TNF-
peritoneal macrophages
a in LPS-stimulated mouse
TNF-
a released in the medium was determined as described
previously28 with
a
slight modification. Briefly, TGC-induced
mouse peritoneal macrophages (5 ꢂ 105 cells/well) were collected
from the peritoneal cavities of male ddY mice, suspended in 100 lL
15. Andary, C.; Privat, G.; Wylde, R.; Heitz, A. J. Nat. Prod. 1985, 48, 778.
16. Abougazar, H.; Bedir, E.; Khan, I. A.; Çalis, Í. Planta Med. 2003, 69, 814.
17. Imakura, Y.; Kobayashi, S.; Mima, A. Phytochemistry 1985, 24, 139.
18. Wu, J.; Huang, J.; Xiao, Q.; Zhang, S.; Xiao, Z.; Li, Q.; Long, L.; Huang, L. Magn.
Reson. Chem. 2004, 42, 659.
19. Freudenberg, M. A.; Galanos, C. Infect. Immun. 1991, 59, 2110.
20. Josephs, M. D.; Bahjat, F. R.; Fukuzuka, K.; Ksontini, R.; Solorzano, C. C.;
Edwards, C. K., III; Tannahill, C. L.; Mackay, S. L. D.; Copeland, E. M., III;
Moldawer, L. L. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2000, 278, R1196.
21. Yoshikawa, M.; Murakami, T.; Ueda, T.; Yoshizumi, S.; Ninomiya, K.; Murakami,
N.; Matsuda, H.; Saito, M.; Fujii, W.; Tanaka, T.; Yamahara, J. Yakugaku Zasshi
1997, 117, 108.
of RPMI 1640 supplemented with 5% FBS, penicillin (100 units/mL),
and streptomycin (100 g/mL), and pre-cultured in 96-well micro-
plates at 37 °C in 5% CO2 in air for 1 h. Nonadherent cells were re-
moved by washing with PBS, and 100 of fresh medium
containing various concentrations of test compound was added.
After 10 min, 100 L of the medium containing 10 g/mL LPS
was added and incubated for 4 h. TNF- production in each well
was determined using an ELISA kit (Mouse TNF- ELISA kit, Invit-
l
lL
l
l
a
a
rogen). Each test compound was dissolved in DMSO, and the solu-
tion was added to the medium (final DMSO concentration 0.5%).
22. Matsuda, H.; Murakami, T.; Ninomiya, K.; Inadzuki, M.; Yoshikawa, M. Bioorg.
Med. Chem. Lett. 1997, 7, 2193.
23. Yoshikawa, M.; Murakami, T.; Hirano, K.; Inadzuki, M.; Ninomiya, K.; Matsuda,
H. Tetrahedron Lett. 1997, 38, 7395.
24. Matsuda, H.; Ninomiya, K.; Morikawa, T.; Yoshikawa, M. Bioorg. Med. Chem.
Lett. 1998, 8, 339.
25. Matsuda, H.; Morikawa, T.; Ninomiya, K.; Yoshikawa, M. Bioorg. Med. Chem.
2001, 9, 909.
26. Morikawa, T.; Matsuda, H.; Ninomiya, K.; Yoshikawa, M. Biol. Pharm. Bull. 2002,
25, 627.
3.5.7. Inhibitory effects on TNF-
cells
L929 cells (Dainippon Pharmaceuticals, Osaka, Japan) were
maintained in Minimum Essential Medium Eagle (MEM, Sigma–Al-
drich) containing 10% FBS, 1% MEM Non-Essential Amino acids
a-induced cell death in L929
(Invitrogen), penicillin
G
(100 units/mL), and streptomycin

1890
T. Morikawa et al. / Bioorg. Med. Chem. 18 (2010) 1882–1890
27. Matsuda, H.; Murakami, T.; Kageura, T.; Ninomiya, K.; Toguchida, I.; Nishida,
N.; Yoshikawa, M. Bioorg. Med. Chem. Lett. 1998, 8, 2191.
28. Yoshikawa, M.; Nishida, N.; Ninomiya, K.; Ohgushi, T.; Kubo, M.; Morikawa, T.;
Matsuda, H. Bioorg. Med. Chem. 2006, 14, 456.
39. Zhang, Y.; Morikawa, T.; Nakamura, S.; Ninomiya, K.; Matsuda, H.; Muraoka, O.;
Yoshikawa, M. Heterocycles 2007, 71, 1565.
40. Ninomiya, K.; Morikawa, T.; Xie, H.; Matsuda, H.; Yoshikawa, M. Heterocycles
2008, 75, 1983.
29. Matsuda, H.; Ishikado, A.; Nishida, N.; Ninomiya, K.; Fujiwara, H.; Kobayashi,
Y.; Yoshikawa, M. Bioorg. Med. Chem. Lett. 1998, 8, 2939.
30. Murakami, T.; Kohno, K.; Ninomiya, K.; Matsuda, H.; Yoshikawa, M. Chem.
Pharm. Bull. 2001, 49, 1003.
41. Nakamura, S.; Okazaki, Y.; Ninomiya, K.; Morikawa, T.; Matsuda, H.;
Yoshikawa, M. Chem. Pharm. Bull. 2008, 56, 1704.
42. Matsuda, H.; Ninomiya, K.; Morikawa, T.; Yasuda, D.; Yamaguchi, I.; Yoshikawa,
M. Bioorg. Med. Chem. Lett. 2008, 18, 2038.
31. Yoshikawa, M.; Ninomiya, K.; Shimoda, H.; Nishida, N.; Matsuda, H. Biol. Pharm.
Bull. 2002, 25, 72.
43. Matsuda, H.; Ninomiya, K.; Morikawa, T.; Yasuda, D.; Yamaguchi, I.; Yoshikawa,
M. Bioorg. Med. Chem. 2009, 17, 7313.
32. Matsuda, H.; Ninomiya, K.; Shimoda, H.; Yoshikawa, M. Bioorg. Med. Chem.
2002, 10, 707.
44. Feher, J.; Deak, G.; Muzes, G.; Lang, I.; Niederland, V.; Nekam, K.; Karteszi, M.
Orv. Hetil. 1989, 130, 2723.
33. Yoshikawa, M.; Xu, F.; Morikawa, T.; Ninomiya, K.; Matsuda, H. Bioorg. Med.
Chem. Lett. 2003, 13, 1045.
34. Yoshikawa, M.; Morikawa, T.; Kashima, Y.; Ninomiya, K.; Matsuda, H. J. Nat.
Prod. 2003, 66, 922.
35. Xu, F.; Morikawa, T.; Matsuda, H.; Ninomiya, K.; Yoshikawa, M. J. Nat. Prod.
2004, 67, 569.
36. Matsuda, H.; Morikawa, T.; Xu, F.; Ninomiya, K.; Yoshikawa, M. Planta Med.
2004, 70, 1201.
37. Li, N.; Morikawa, T.; Matsuda, H.; Ninomiya, K.; Li, X.; Yoshikawa, M.
Heterocycles 2007, 71, 1193.
38. Ninomiya, K.; Morikawa, T.; Zhang, Y.; Nakamura, S.; Matsuda, H.; Muraoka, O.;
Yoshikawa, M. Chem. Pharm. Bull. 2007, 55, 1185.
45. Skottova, N.; Krecman, V. Physiol. Res. 1998, 47, 1.
46. Matsuda, H.; Kiyohara, S.; Sugimoto, S.; Ando, S.; Nakamura, S.; Yoshikawa, M.
Biol. Pharm. Bull. 2009, 32, 147.
47. Yoshikawa, M.; Morikawa, T.; Oominami, H.; Matsuda, H. Chem. Pharm. Bull.
2009, 57, 957.
48. Kouroku, Y.; Fujita, E.; Jimbo, A.; Mukasa, T.; Tsuru, T.; Momoi, M. Y.; Momoi, T.
Biochem. Biophys. Res. Commun. 2000, 270, 972.
49. Xie, Y.; Morikawa, M.; Nakamura, S.; Ninomiya, K.; Imura, K.; Muraoka, O.;
Yuan, D.; Yoshikawa, M. Chem. Pharm. Bull. 2008, 56, 1628.
50. Morikawa, T.; Wang, L.-B.; Ninomiya, K.; Yokoyama, E.; Matsuda, H.; Muraoka,
O.; Wu, L.-J.; Yoshikawa, M. Chem. Pharm. Bull. 2009, 57, 361.
51. Tiegs, G.; Wolter, M.; Wendel, A. Biochem. Pharmacol. 1989, 38, 627.