Estrogenic and anti-estrogenic activities of Cassia tora phenolic constituents

Article abstract of DOI:10.1248/cpb.55.1476

Through an estrogenic activity bioassay-guided fractionation of the 70% ethanolic extract of Cassia tora seeds two new phenolic triglucosides, torachrysone 8-O-[β-D-glucopyranosyl(1→3)-O-β-D- glucopyranosyl(1→6)-O-β-D-glucopyranoside] (1) and toralactone 9-O-[β-D-glucopyranosyl-(1→3)-O-β-D-glucopyranosyl-(1→6) -O-β-D-glucopyranoside] (2), along with seven known compounds were isolated. The structures of the new compounds were elucidated on the basis of spectroscopic and chemical evidence. The estrogenic activity of the fractions and the isolated compounds were investigated using the estrogen-dependent proliferation of MCF-7 cells. In addition, the yeast two hybrid assay expressing estrogen receptor α(ERα) and β (ERβ) and the ERα competitor screening assay (ligand binding screen) were used to verify the binding affinities of the isolated compounds to ER. Furthermore, a naringinase pre-treatment of the 70% alcoholic extract of Cassia tora seeds resulted in a significant increase in its estrogenic activity. From the naringinase pre-treated extract six compounds were isolated, among which 6-hydroxymusizin and aurantio-obtusin showed the most potent estrogenic activity, while torachrysone, rubrofusarin and toralactone showed a significant anti-estrogenic activity. Finally, the structure requirements responsible for the estrogenic activity of the isolated compounds were studied by investigating the activity of several synthetic compounds and chemically modifying the isolated compounds. The basic nucleus 1,3,8-trihyroxynaphthalene (T₃HN) was found to play a principal role in the binding affinity of these compounds to ER.

Full text of DOI:10.1248/cpb.55.1476

1476
Chem. Pharm. Bull. 55(10) 1476—1482 (2007)
Vol. 55, No. 10
Estrogenic and Anti-estrogenic Activities of Cassia tora Phenolic
Constituents
Ali Mahmoud EL-HALAWANY,a Mi Hwa CHUNG,a Norio NAKAMURA,b Chao-Mei MA,a
Tsutomu NISHIHARA,c and Masao HATTORI*,a
a Institute of Natural Medicine, University of Toyama; 2630 Sugitani, Toyama 930–0194, Japan: b Faculty of
Pharmaceutical Sciences, Doshisha Women’s College of Liberal Arts; Kodo, Kyotanabe, Kyoto 610–0395, Japan: and
c Faculty of Pharmaceutical Sciences, Hyogo College of Medicine; 1–3–6 Minatojima, Chuo-ku, Kobe 650–0045, Japan.
Received May 16, 2007; accepted July 17, 2007
Through an estrogenic activity bioassay-guided fractionation of the 70% ethanolic extract of Cassia tora
seeds two new phenolic triglucosides, torachrysone 8-O-[b-D-glucopyranosyl(1→3)-O-b-D-glucopyranosyl(1→6)-
O-b-D-glucopyranoside] (1) and toralactone 9-O-[b-D-glucopyranosyl-(1→3)-O-b-D-glucopyranosyl-(1→6)-O-b-D-
glucopyranoside] (2), along with seven known compounds were isolated. The structures of the new compounds
were elucidated on the basis of spectroscopic and chemical evidence. The estrogenic activity of the fractions and
the isolated compounds were investigated using the estrogen-dependent proliferation of MCF-7 cells. In addition,
the yeast two hybrid assay expressing estrogen receptor a (ERa) and b (ERb) and the ERa competitor screening
assay (ligand binding screen) were used to verify the binding affinities of the isolated compounds to ER. Further-
more, a naringinase pre-treatment of the 70% alcoholic extract of Cassia tora seeds resulted in a significant in-
crease in its estrogenic activity. From the naringinase pre-treated extract six compounds were isolated, among
which 6-hydroxymusizin and aurantio-obtusin showed the most potent estrogenic activity, while torachrysone,
rubrofusarin and toralactone showed a significant anti-estrogenic activity. Finally, the structure requirements re-
sponsible for the estrogenic activity of the isolated compounds were studied by investigating the activity of sev-
eral synthetic compounds and chemically modifying the isolated compounds. The basic nucleus 1,3,8-trihyroxy-
naphthalene (T3HN) was found to play a principal role in the binding affinity of these compounds to ER.
Key words Cassia tora; 6-hydroxymusizin; 1,3,8-trihyroxynaphthalene; estrogenic activity; rubrofusarin; toralactone
Estrogen is a key regulator of the cellular processes in- ried out for the ethanolic extract of Cassia tora, and the bio-
volved in the development and maintenance of the reproduc- logically active principles of this extract were also investi-
tive function.1) Phytoestrogens are polyphenolic non-steroidal gated.
plant compounds with estrogen-like biological activity. Re-
cently, an increasing number of epidemiological and experi- Results
mental studies has suggested that the consumption of phytoe-
The 70% ethanolic extract of Cassia tora seeds signifi-
strogens may have a protective effect on estrogen-related cantly stimulated the proliferation of estrogen dependent
conditions like menopause and estrogen related diseases such MCF-7 cells indicating its possible estrogenic activity, which
as prostate and breast cancers, osteoporosis and cardiovascu- is significantly increased after being treated with naringinase
lar diseases.2) In our search for new phytoestrogens, the (Table 1). The 70% ethanolic extract was fractionated on
ethanolic extract of Cassia tora exhibited a significant stimu- DIAON HP-20 to obtain 3 fractions (25% methanol, 50%
lation of estrogen dependent MCF-7 cells, suggesting its methanol, 100% methanol). Estrogenic activity of these frac-
estrogenic activity. Cassia tora L. (Fabaceae) is widely dis- tions was evaluated using MCF-7 cells before and after
tributed in tropical Asian countries and its seeds have been naringinase treatment (Table 1). The 50% methanol fraction
used as a traditional medicine for constipation, asthenia, eye and the naringinase pre-treated 70% ethanolic extract were
disease, hepatitis, diuretic, hemoglobin disorders and as an
antidysenteric.3) Hypotensive activity of the seed extract has
Table 1. Estrogenic Effects of Fractions of Cassia tora on the Proliferation
of MCF-7 Cells
also been reported.4,5) In India the plant is used for the treat-
ment of snakebites and scorpion stings.6) The seeds were
MCF-7 cell proliferation (% of control)
Fraction
found to contain a diversity of phenolic constituents, i.e.
anthraquinones, naphthopyrones, naphthalenes, hydroan-
thracenes and their glucosides.7—11) These phenolic con-
stituents were reported to have various biological and phar-
macological effects such as antibacterial,11) antifungal,12) an-
timutagenic,13) hypolipidemic14) and anti-allergic activities.11)
In addition, the aqueous extract of the seed15) and naphthopy-
rones9) and anthraquinones10) isolated from the seeds exhib-
ited hepatoprotective activity.
In this study, the isolation of two new phenolic trigluco-
sides from a 70% ethanol extract of Cassia tora seeds is
reported. Furthermore, a naringinase pretreatment, as a mimic
of the metabolic activity of gastro-intestinal flora, was car-
1 mg/ml
10 mg/ml
Chloroform-soluble
70% EtOH extract
100% MeOH eluate
50% MeOH eluate
147.8ꢀ12.8*
126.8ꢀ4.1*
145.3ꢀ16.1*
152.5ꢀ17.4**
153.8ꢀ1.7**
206.8ꢀ11.8**
142.4ꢀ7.8*
86.9ꢀ10.6
82.5ꢀ2.1
98.6ꢀ2.7
115.2ꢀ12.7
169.3ꢀ9.4**
110.8ꢀ10.9
157.6ꢀ18.9**
137.5ꢀ11.3*
140.1ꢀ14.4*
85.7ꢀ7.9
25% MeOH eluate
NT 70% EtOH extract
NT 100% MeOH eluate
NT 50% MeOH eluate
NT 25% MeOH eluate
118.9ꢀ11.3
17b-Estradiol showed maximum activity (202.2ꢀ3.2%) at
a concentration of
10ꢁ12 M. Each value represents the meanꢀS.E. (nꢂ6). Asterisks indicate significant dif-
ference from the control at pꢃ0.05 (∗), pꢃ0.01 (∗∗). NT, naringinase treated.
∗ To whom correspondence should be addressed. e-mail: saibo421@inm.u-toyama.ac.jp
© 2007 Pharmaceutical Society of Japan

October 2007
1477
selected for further purification due to their relatively high
estrogenic activity. The 50% methanol fraction was subjected
to intensive chromatographic purifications using normal and
reversed phase silica gel and Sephadex LH-20 to get two new
compounds (1, 2) along with seven known compounds [au-
rantio-obtusin 6-O-b-D-glucoside (3),7) torachrysone 8-O-
b-D-gentiobioside (4),11) toralactone 9-O-b-D-gentiobioside
(5),8) 6-hydroxymusizin 8-O-b-D-glucoside (6),16) torachrysone
tetraglucoside (7), rubrofusarin triglucoside (8)11) and
chrysophanol triglucoside (9)].9) From the naringinase pre-
treated total extract six known compounds were isolated and
Fig. 1. New Compounds Isolated from Cassia tora Seeds
identified as [chrysophanol (10), physcion (11), 9-methoxy- der. The ESI-MS spectrum showed an [MꢄH]ꢄ at m/z 759.
chrysophanol (12),17) aurantio-obtusin (13),7) toralactone On the basis of its ESI-MS datum and elemental analysis
(14)18) and rubrofusarin (15)].(11) In addition, torachrysone data, compound 2 was assigned the molecular formula
(16)19,20) and 6-hydroxymusizin (17)16) were obtained through C33H42O20. The UV spectrum showed absorption maxima at
acid hydrolysis of their corresponding isolated glycosides 6 237, 247, 270, 277 and 383 nm, suggesting the naphtha-a-
and 7. Finally, nor-rubrofusarin (18) and nor-toralactone (19) pyrone nucleus of this compound.3,18) The 1H-NMR spectrum
were obtained by the demethylation of compounds 14 and showed two aromatic protons at dH 6.50 (1H, s, H-4) and
15. The known compounds were identified by comparison 7.13 (1H, s, H-5), two doublet protons at dH 6.86 (1H, d,
with reported data, and the structures of the new compounds Jꢂ2.2 Hz, H-8) and 6.92 (1H, d, Jꢂ2.2 Hz, H-6), aromatic
were determined as follows.
methyl protons at dH 2.22 (3H, s, CH3-3) and a methoxyl
Compound 1 was obtained as yellow needles. The electron proton at dH 3.88 (3H, s, OCH3). These data confirmed the
spray ionization mass (ESI-MS) spectrum showed an aglycone moiety to be toralactone.3,18) The three anomeric
[MꢄH]ꢄ ion peak at m/z 733. On the basis of its ESI-MS protons at dH 5.09, 4.31 and 4.28 revealed the presence of
and elemental analysis data, compound 1 was assigned the three sugar moieties. The 13C-NMR chemical shifts of the
molecular formula C32H44O19. The UV spectrum showed sugar moieties, and the comparison of the sugar resulting
absorption maxima at 240, 265, 310, 324 and 340 nm, sug- from the acid hydrolysis of 2 with authentic samples,
1
gesting the torachrysone nucleus.11,19,20) The H-NMR spec- revealed that the three sugar moieties were D-glucose.3,11) The
trum showed an aromatic singlet at dH 7.09 (1H, s, H-4), two site of attachment of the three glucose units to the aglycone
meta-coupled doublets at dH 7.06 (1H, d, Jꢂ2.1 Hz, H-7) and was found to be at OH-9 as indicated from the HMBC corre-
6.89 (1H, d, Jꢂ2.1 Hz, H-5), one aromatic methyl at dH 2.21 lation between the proton at dH 5.09 (H-1ꢅ) and the carbon at
(3H, s, CH3-3), a methoxyl group at dH 3.82 (3H, s, OCH3) dC 157.6 (C-9) of the aglycone. Similar to compound 1, the
and an acetyl group at dH 2.48 (3H, s, COCH3), in addition downfield shifted carbons at dC 68.8 and 88.1 corresponding
to the chelated hydroxyl group signal at dH 9.46 (1H, s, to C-6ꢅ and C-3ꢆ, respectively, revealed the presence of 1→6
OH-1). By comparison with the previously reported data, and 1→3 linkages among the three glucose units.3) In the
1
these H-NMR data confirmed the torachrysone nucleus of HMBC spectrum, long-range correlations were found be-
1.11,19,20) Moreover, the presence of three anomeric protons at tween the anomeric proton at dH 4.31 (H-1ꢆ) and the carbon
dH 5.09 (1H, d, Jꢂ8.0 Hz, H-1ꢅ) , 4.32 (1H, d, Jꢂ8.0 Hz, H- at dC 68.8 (C-6ꢅ), and between the anomeric proton at dH
1ꢆ) and 4.30 (1H, d, Jꢂ8.0 Hz, H-1ꢇ) indicated the presence 4.28 (H-1ꢇ) and the carbon at dC 88.1 (C-3ꢆ) which con-
of three sugar moieties. The attachment of the sugar moieties firmed the 1→6 and 1→3 linkages. All the glucose moieties
to C-8 of the aglycone was confirmed by the heteronuclear were found to be in the b configuration as indicated by the J
multiple bond coherence (HMBC) correlation between the values of the anomeric protons (Jꢂ8 Hz). Based on the for-
anomeric proton of glucose-I at dH 5.09 and C-8 of the agly- merly mentioned data and by comparison with reported data
cone at dC 155.2. All the glucose moieties were found to be of rubrofusarin triglucoside and cassiacide C2,3,11) compound
in the b configuration as indicated by the J values of their 2 was determined to be toralactone 9-O-[b-D-glucopyra-
anomeric protons (Jꢂ8.0 Hz). The 13C-NMR chemical shifts nosyl-(1→3)-O-b-D-glucopyranosyl-(1→6)-O-b-D-glucopy-
of the sugar moieties, and the comparison of the sugar result- ranoside].
ing from the acid hydrolysis of 1 with that of authentic sam-
Estrogenic Activity of the Isolated Compounds Using
ples, revealed that the three sugar moieties were D-glucose. estrogen-dependent MCF-7 cell-proliferation, yeast two hy-
In the 13C-NMR spectrum, the downfield shifted C-6ꢅ and C- brid assay expressing ERa and ERb and the ligand binding
3ꢆ at dC 68.8 and at dC 88.1, respectively, revealed the pres- screen, almost all the isolated compounds from the 50%
ence of 1→6 and 1→3 linkages among the three glucose methanol fraction and the naringinase-pretreated 70%
1
molecules.3,11) In the HMBC spectrum, H–13C long-range ethanolic extract of Cassia tora were tested for their estro-
correlations were observed between the anomeric proton at genic activities at various concentrations. In addition, a
dH 4.32 (H-1ꢆ) and the carbon at dC 68.8 (C-6ꢅ), and between group of synthetic compounds (20—24), torachrysone (16),
the anomeric proton at dH 4.30 (H-1ꢇ) and the carbon at dC 6-hydroxymusizin (17), nor-rubrofusarin (18) and nor-
88.1 (C-3ꢆ), which confirmed the 1→6 and 1→3 linkages. toralactone (19) (Fig. 2) were also investigated in order to
From these findings the structure of compound 1 was deter- study the structure activity requirements of these compounds.
mined to be torachrysone 8-O-[b-D-glucopyranosyl(1→3)-O-
Estrogen-Dependent MCF-7 Cells The positive control
17b-estradiol showed maximum cell proliferation (212.8% of
b-D-glucopyranosyl(1→6)-O-b-D-glucopyranoside] (Fig. 1).
Compound 2 was obtained as yellowish amorphous pow- the control) at a concentration of 10ꢁ12 M (Table 2). Aurantio-

1478
Vol. 55, No. 10
obtusin 6-O-b-D-glucoside (3) showed a weak estrogenic shaped manner versus their concentrations, while 1,6-dihy-
activity at 10ꢁ5 M. The aglycones, aurantio-obtusin (12), droxynaphthalene (23) showed a significant stimulation of
6-hydroxymusizin (17) and 1,3,8-trihydroxynaphtahlene (24) cell proliferation at 10ꢁ6 M.
showed significant increase in their estrogenic activity upon
Yeast Two-Hybrid Assay In the yeast expressing ERa
decreasing their concentrations with a maximum activity at (Table 3), 6-hydroxymusizin 8-O-b-D-glucoside (6) showed
10ꢁ7 M. Nor-rubrofusarin (18) and nor-toralactone (19) sig- estrogenic activity at 10ꢁ4 M. In addition, 6-hydroxymusizin
nificantly stimulated the MCF-7 cell proliferation in a bell- (17), nor-rubrofusarin (18) and nor-toralactone (19) showed a
significant estrogenic activity at 10ꢁ4 M. On the other hand, in
the yeast expressing ERb, 6-hydroxymusizin 8-O-b-D-gluco-
side (6) showed estrogenic activity at 10ꢁ4 M. An appreciable
induction of b-galactosidase activity was observed for auran-
tio-obtusin (12), 6-hydroxymusizin (17), nor-rubrofusarin
(18) and nor-toralactone (19) at a concentration of 10ꢁ4 M.
Ligand Binding Screen 6-Hydroxymusizin (17), nor-
rubrofusarin (18), nor-toralactone (19) and T3HN (24)
showed concentration dependent increases in their binding to
ERa with IC50 values of 2.5ꢈ10ꢁ4 M, 3.2ꢈ10ꢁ3 M, 3.9ꢈ
10ꢁ4 M and 3.9ꢈ10ꢁ4 M, respectively, while physcion (11),
aurantio-obtusin (12), toralactone (14), rubrofusarin (15),
torachrysone (16), 2,6-dihydroxynaphthalene (20) and 1,6-
dihydroxynaphthalene (23) showed IC50 values higher than
5ꢈ10ꢁ3 M (Table 4).
Anti-estrogenic Activity Using the estrogen-dependent
MCF-7 cell proliferation and yeast two-hybrid assays, all the
isolated compounds were tested for their anti-estrogenic
activities at various concentrations. Inhibition of 17b-estra-
diol (E2)-induced proliferation of MCF-7 cells and E2-in-
duced b-galactosidase activity by the isolated compounds
were determined under the conditions at which these com-
pounds did not inhibit the proliferation/growth of the MCF-7
and the yeast cells. Tamoxifen, an estrogen receptor antago-
nist, was used as the positive control.
Inhibition of 17b-Estradiol (E2)-Induced Stimulation of
Fig. 2. Chemical Structures of Isolated Compounds (3—15) from the
Seeds of Cassia tora and of Synthetic Compounds (18—24)
Table 2. Estrogenic and Anti-estrogenic Activities of Isolated and Synthetic Compounds on MCF-7 Cells
Relative cell-proliferation inhibition
(% of 10ꢁ12 M E2 proliferation value)
Relative cell-proliferation (% of vehicle)
10ꢁ6
Comp.
10ꢁ5
M
M
10ꢁ7
M
10ꢁ5
M
10ꢁ6
M
10ꢁ7
M
3
5
6
7
8
127.2ꢀ6.0*
110.4ꢀ4.3
94.7ꢀ2.3
116.3ꢀ6.4
104.9ꢀ2.8
85.1ꢀ9.7
93.1ꢀ12.2
101.1ꢀ8.8
96.6ꢀ17.8
90.5ꢀ4.8
93.2ꢀ3.3
96.9ꢀ3.2
99.6ꢀ5.8
126.6ꢀ21.5
170.7ꢀ5.8**
157.7ꢀ5.9**
97.6ꢀ8.3
93.5ꢀ9.2
90.0ꢀ4.9
110.2ꢀ4.6
116.2ꢀ4.4
103.3ꢀ6.5
125.3ꢀ9.6
103.8ꢀ9.8
106.7ꢀ4.6
99.8ꢀ8.6
94.5ꢀ6.3
112.8ꢀ9.4
93.1ꢀ2.9
110.3ꢀ5.1
84.1ꢀ12.7
96.9ꢀ1.2
106.4ꢀ3.9
106.8ꢀ2.6
99.8ꢀ4.6
95.9ꢀ8.8
94.4ꢀ2.8
147.5ꢀ7.5**
90.9ꢀ2.5
90.5ꢀ3.2
104.1ꢀ2.9
96.7ꢀ13.2
122.8ꢀ2.9
106.1ꢀ2.3
104.0ꢀ2.4
105.1ꢀ5.3
96.2ꢀ2.8
89.7ꢀ2.1*
95.8ꢀ3.5
98.3ꢀ1.9
94.6ꢀ1.8
98.7ꢀ4.6
97.8ꢀ2.6
95.6ꢀ3.3
94.2ꢀ2.5
92.7ꢀ3.4
93.3ꢀ2.0
107.7ꢀ2.9
99.7ꢀ4.0
94.3ꢀ3.2
95.0ꢀ2.7
74.0ꢀ2.9**
94.1ꢀ1.9
93.0ꢀ5.3
104.8ꢀ3.8
98.1ꢀ6.0
104.6ꢀ2.5
103.1ꢀ2.1
93.6ꢀ2.3
87.6ꢀ1.5**
95.3ꢀ1.4
101.4ꢀ1.9
96.5ꢀ1.6
91.5ꢀ4.9
91.2ꢀ1.9
91.9ꢀ2.9
102.9ꢀ7.0
92.1ꢀ2.5
102.2ꢀ4.4
111.2ꢀ3.2
92.7ꢀ2.9
96.5ꢀ4.0
98.3ꢀ2.5
104.1ꢀ1.9
98.3ꢀ4.4
104.1ꢀ3.2
106.9ꢀ5.8
93.8ꢀ1.2
103.9ꢀ2.5
106.0ꢀ3.3
91.4ꢀ4.3
90.4ꢀ2.7
91.1ꢀ2.8
97.5ꢀ3.8
97.7ꢀ1.4
92.9ꢀ2.7
92.1ꢀ1.0
90.9ꢀ3.0
91.7ꢀ2.6
91.9ꢀ5.0
106.7ꢀ4.0
113.2ꢀ1.9
88.9ꢀ4.4*
91.3ꢀ2.5
97.1ꢀ3.7
101.2ꢀ4.4
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Tam.
86.5ꢀ3.3
95.4ꢀ4.7
97.1ꢀ3.9
96.0ꢀ9.2
115.3ꢀ5.2
157.7ꢀ8.5**
130.2ꢀ13.4
144.9ꢀ9.3**
96.8ꢀ3.4
105.2ꢀ8.9
100.7ꢀ8.6
110.5ꢀ6.1
142.6ꢀ21.8**
137.7ꢀ16.3*
179.2ꢀ24.7**
157.7ꢀ3.5**
95.1ꢀ6.0
106.4ꢀ9.7
106.7ꢀ4.4
140.5ꢀ14.7**
140.8ꢀ8.5**
72.3ꢀ4.2
125.3ꢀ16.5
Each value represents the meanꢀS.E. (nꢂ6). Asterisks denote significant differences from the control at pꢃ0.05 (∗), pꢃ0.01 (∗∗). The estrogenic activity of 17b-estradiol
was 212.8ꢀ19.6% at a concentration of 10ꢁ12 M (100%). The relative cell-proliferation inhibition was expressed as a percentage of the control (100%), observed in the presence of
10ꢁ12 M 17b-estradiol alone. Tam. denotes (tamoxifen).

October 2007
1479
Table 3. Induction of b-Galactosidase in the Yeast Two-Hybrid Assay Expressing ERa and ERb by Isolated and Synthetic Compounds
b-Galactosidase activity (U)
Comp.
ERa
ERb
10ꢁ4
M
10ꢁ5
M
10ꢁ6
M
10ꢁ4
M
10ꢁ5
M
10ꢁ6
M
3
5
6
7
8
39.9ꢀ12.4
50.7ꢀ1.5
81.0ꢀ5.9*
32.8ꢀ7.2
59.5ꢀ4.8
32.2ꢀ3.7
34.6ꢀ3.7
56.4ꢀ11.1
47.0ꢀ12.4
38.1ꢀ10.2
54.2ꢀ1.5
48.7ꢀ4.8
45.2ꢀ7.3
93.2ꢀ7.8*
128.4ꢀ12.1**
190.3ꢀ19.1**
52.2ꢀ8.5
36.3ꢀ2.8
42.7ꢀ4.6
41.6ꢀ0.5
41.5ꢀ6.3
47.6ꢀ2.4
41.9ꢀ2.7
49.5ꢀ2.8
39.8ꢀ6.3
51.4ꢀ2.4
35.8ꢀ3.2
40.7ꢀ3.2
48.8ꢀ5.1
38.7ꢀ2.2
55.8ꢀ7.9
51.3ꢀ2.8
41.1ꢀ2.3
25.7ꢀ6.3
31.2ꢀ10.9
67.5ꢀ7.6
67.5ꢀ7.8
42.3ꢀ3.1
36.9ꢀ6.2
44.5ꢀ0.4
40.4ꢀ1.8
35.2ꢀ6.8
36.5ꢀ0.6
41.2ꢀ2.6
41.8ꢀ2.9
31.2ꢀ9.7
38.1ꢀ4.5
39.9ꢀ2.2
39.1ꢀ2.2
48.5ꢀ4.5
46.5ꢀ0.6
43.2ꢀ2.9
35.6ꢀ2.7
38.1ꢀ4.5
36.1ꢀ9.7
44.2ꢀ4.1
50.7ꢀ7.3
49.7ꢀ6.9
78.6ꢀ5.4
53.0ꢀ5.4
55.6ꢀ1.4
42.6ꢀ1.9
24.0ꢀ1.3
70.1ꢀ12.4
66.0ꢀ9.4
109.8ꢀ5.8**
52.0ꢀ7.5
66.4ꢀ13.2
41.0ꢀ1.6
54.4ꢀ1.6
56.6ꢀ1.4
321.4ꢀ12.4**
74.4ꢀ6.0
73.1ꢀ9.4
59.0ꢀ13.2
57.5ꢀ7.4
348.4ꢀ23.6**
141.4ꢀ0.10**
187.2ꢀ20.9*
52.2ꢀ8.5
36.3ꢀ2.8
42.7ꢀ4.6
54.4ꢀ2.2
48.4ꢀ2.1
60.0ꢀ10.4
39.9ꢀ11.6
53.5ꢀ4.9
44.1ꢀ4.4
46.2ꢀ4.4
65.9ꢀ4.3
67.5ꢀ2.2
45.8ꢀ2.4
51.6ꢀ2.1
55.6ꢀ4.9
47.4ꢀ11.6
87.1ꢀ1.4**
54.8ꢀ5.8
66.0ꢀ1.5
42.3ꢀ3.1
36.9ꢀ6.2
44.5ꢀ0.4
40.4ꢀ1.8
54.7ꢀ1.5
48.1ꢀ0.6
53.4ꢀ1.9
48.9ꢀ2.2
42.4ꢀ9.1
50.2ꢀ5.6
42.0ꢀ1.8
55.1ꢀ10.8
72.6ꢀ2.2
51.8ꢀ0.6
30.9ꢀ7.9
52.4ꢀ1.9
48.9ꢀ5.6
31.9ꢀ9.1
55.5ꢀ2.2
61.3ꢀ4.1
66.7ꢀ7.6
78.6ꢀ5.4
53.0ꢀ4.4
55.6ꢀ1.4
42.6ꢀ1.9
53.9ꢀ9.5
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
41.6ꢀ0.5
55.0ꢀ10.6
Each value represents the meanꢀS.E. of three independent experiments (nꢂ3). Asterisks denote significant differences from the control at pꢃ0.05 (∗), pꢃ0.01 (∗∗). DMSO
(control) was 40.8ꢀ2.7 and 37.2ꢀ7.3 in the yeast expressing ERa and ERb, respectively. 17b-Estradiol was 1553.7ꢀ13.5 and 2126.0ꢀ79.3 at a concentration of 10ꢁ7 M in the
yeast expressing ERa and ERb, respectively.
Table 4. ERa Binding Affinity of Isolated Compounds from Cassia tora
nificant anti-estrogenic activity in concentration dependent
manner. Torachrysone (16), 6-hydroxymusizin (17), 1,3-di-
hydroxynaphthalene (22) and 1,6-trihydroxynaphthalene (23)
showed activity at a 10ꢁ4 M concentration. On the other hand,
in the yeast expressing ERb, 9-methoxychrysophanol (13),
toralactone (14), rubrofusarin (15), torachrysone (16), 6-hy-
droxymusizin (17), and T3HN (24) showed activity at 10ꢁ4 M,
while nor-rubrofusarin (18) and nor-toralactone (19) at 10ꢁ5
and 10ꢁ6 M.
According to the aforementioned data, among the isolated
glycosides, aurantio-obtusin 6-O-b-D-glucoside (3) and 6-hy-
droxymusizin 8-O-b-D-glucoside (6), exhibited weak or no
estrogenic activity. Regarding the aglycones and the synthetic
compounds, aurantio-obtusin (12), 6-hydroxymusizin (17),
nor-rubrofusarin (18), nor-toralactone (19) and 1,3,8-trihy-
droxynaphthalene (24) exhibited the most potent estrogenic
and anti-estrogenic activities in the in vitro assays. On the
other hand, toralactone (14), rubrofusarin (15) and
torachrysone (16) showed the most significant anti-estrogenic
activity, when evaluated by the yeast two-hybrid assay.
and Synthetic Compounds
% Binding to ERa
Compound
IC50
5ꢈ10ꢁ3
M
10ꢁ4
M
10ꢁ5
M
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
ꢁ2.5ꢀ1.7
24.2ꢀ0.1**
62.9ꢀ1.1**
0.8ꢀ0.3
20.1ꢀ0.2*
27.1ꢀ5.0**
35.6ꢀ2.3**
80.4ꢀ2.1**
52.8ꢀ0.1**
66.6ꢀ0.1**
30.8ꢀ0.2** ꢁ16.3ꢀ0.7
ꢁ10.7ꢀ3.1
3.5ꢀ0.8
32.8ꢀ1.5**
86.6ꢀ1.2**
ꢁ3.0ꢀ3.4
ꢁ1.1ꢀ1.9
10.9ꢀ4.3
ꢁ8.5ꢀ0.2
5.1ꢀ5.6
19.5ꢀ0.9
12.8ꢀ0.8
37.8ꢀ0.5**
24.7ꢀ0.9**
42.1ꢀ9.1**
ꢁ4.1ꢀ0.2
ꢁ12.0ꢀ0.5
ꢁ1.9ꢀ0.8
ꢁ5.6ꢀ1.1
0.9ꢀ13.3
2.3ꢀ2.1
ꢁ0.9ꢀ0.5
13.1ꢀ2.1
5.5ꢀ1.1
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
1.6ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
2.5ꢈ10ꢁ4
3.2ꢈ10ꢁ3
3.9ꢈ10ꢁ4
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
ꢉ5ꢈ10ꢁ3
3.9ꢈ10ꢁ4
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
2.4ꢀ2.2
ꢁ7.4ꢀ0.6
ꢁ4.2ꢀ0.3
ꢁ8.5ꢀ1.3
ꢁ32.2ꢀ28.5 ꢁ14.9ꢀ1.2
12.1ꢀ1.4 5.9ꢀ0.8
22.9ꢀ1.5** ꢁ2.9ꢀ0.1
A percentage of 17b-estradiol (positive control) bound to ERa was 52.5ꢀ1.7% at a
concentration of 10ꢁ7 M (IC50ꢂ8ꢈ10ꢁ8 M). Asterisks indicate significant difference
from the control at pꢃ0.05 (∗), pꢃ0.01 (∗∗), (nꢂ4).
Discussion
Several phenolic compounds such as isoflavonoids,
MCF-7 Cells Tamoxifen inhibited the 17b-estradiol-medi- coumestans and lignans were reported to have estrogenic
ated proliferation of MCF-7 cells by 16% at 10ꢁ5 M (Table 2). and/or anti-estrogenic activities due to their structural simi-
Physcion (11) showed a significant anti-estrogenic activity at larity to the mammalian estrogens. The main structural fea-
10ꢁ5 and 10ꢁ6 M concentrations, while 1,3-dihydroxynaph- ture of these compounds was reported to be the presence of
thalene (22) showed significant anti-estrogenic activity at two phenolic groups separated by a planar core of about
10ꢁ7 M.
Inhibition of 17b-Estradiol-Induced b-Galactosidase
12.8 Å.21)
Despite lacking the formerly mentioned characteristic fea-
Activity in the Yeast Two-Hybrid Assay In the yeast ture, the naphthopyrones, toralactone (14) and rubrofusarin
expressing ERa (Table 5), rubrofusarin (15), nor-rubro- (15), and the acetyl naphthalenes, torachrysone (16) and
fusarin (18), nor-toralactone (19) and T3HN (24) showed sig- 6-hydroxymusizin (17), exhibited a significant estrogenic

1480
Vol. 55, No. 10
Table 5. Inhibitory Effects of Isolated and Synthetic Compounds on the Induction of b-Galactosidase Activity by 17b-Estradiol in the Yeast Two-Hybrid
Assay (ERa and ERb)
b-Galactosidase activity (% of control)
Comp.
ERa
ERb
10ꢁ4
M
10ꢁ5
M
10ꢁ6
M
10ꢁ4
M
10ꢁ5
M
10ꢁ6
M
3
5
6
7
8
92.8ꢀ9.0
97.3ꢀ7.4
94.0ꢀ6.5
103.1ꢀ2.3
118.7ꢀ8.3
95.7ꢀ1.4
95.7ꢀ3.7
82.4ꢀ6.7
101.7ꢀ7.3
82.9ꢀ0.9
79.1ꢀ1.3*
63.6ꢀ0.8**
66.4ꢀ2.4**
52.4ꢀ3.9**
—
93.4ꢀ1.9
104.5ꢀ1.1
98.8ꢀ18.6
116.5ꢀ4.9
110.1ꢀ5.9
104.0ꢀ1.4
110.4ꢀ5.9
89.4ꢀ3.1
107.1ꢀ6.2
105.8ꢀ2.0
120.1ꢀ5.6
104.2ꢀ4.8
91.8ꢀ1.9
106.9ꢀ3.2
110.1ꢀ3.9
97.7ꢀ2.2
103.4ꢀ5.3
111.8ꢀ7.5
92.0ꢀ1.9
80.4ꢀ5.6*
108.2ꢀ5.4
95.8ꢀ5.3
76.8ꢀ3.8*
73.7ꢀ2.2**
105.6ꢀ1.3
91.0ꢀ1.9
105.7ꢀ5.9
112.0ꢀ8.9
90.1ꢀ1.9
97.1ꢀ2.7
90.4ꢀ2.9
99.6ꢀ2.8
121.8ꢀ1.8
95.8ꢀ5.0
90.7ꢀ0.7
90.9ꢀ6.7
82.6ꢀ1.2
129.1ꢀ11.7
78.8ꢀ1.3**
78.5ꢀ3.5**
50.7ꢀ1.6**
66.3ꢀ1.8**
31.7ꢀ0.8**
—
102.4ꢀ3.6
91.9ꢀ4.6
110.5ꢀ3.1
102.9ꢀ2.7
101.5ꢀ12.8
97.3ꢀ6.5
103.3ꢀ9.2
89.0ꢀ3.2
117.2ꢀ3.3
102.5ꢀ4.3
88.2ꢀ4.8
82.2ꢀ3.2
88.2ꢀ8.9
99.8ꢀ4.7
63.6ꢀ2.1**
43.3ꢀ3.1**
104.9ꢀ1.5
106.9ꢀ5.5
125.4ꢀ8.01
93.8ꢀ6.7
95.9ꢀ3.4
100.1ꢀ1.1
110.3ꢀ2.5
105.8ꢀ1.7
111.1ꢀ3.3
103.7ꢀ4.2
110.1ꢀ7.8
98.1ꢀ6.7
123.3ꢀ3.9
117.9ꢀ2.1
124.2ꢀ4.3
99.8ꢀ9.6
99.7ꢀ11.0
117.7ꢀ7.9
94.0ꢀ1.5
75.8ꢀ4.3**
103.4ꢀ3.3
101.8ꢀ9.2
117.9ꢀ6.6
107.8ꢀ2.1
94.1ꢀ1.9
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Tam.
90.3ꢀ6.4
102.5ꢀ2.8
78.9ꢀ3.2*
67.3ꢀ0.8**
82.2ꢀ4.8
86.8ꢀ2.2
57.4ꢀ3.9**
38.5ꢀ1.9**
95.7ꢀ2.5
92.3ꢀ5.5
90.6ꢀ4.7
97.2ꢀ9.5
76.4ꢀ3.4**
114.2ꢀ4.2
—
—
103.8ꢀ4.3
96.1ꢀ9.5
67.3ꢀ3.5**
76.1ꢀ3.2**
57.9ꢀ3.7**
11.9ꢀ1.1**
110.6ꢀ19.2
94.6ꢀ5.4
81.7ꢀ5.8
86.0ꢀ1.6
55.3ꢀ2.0**
22.4ꢀ2.5**
85.7ꢀ1.4
98.7ꢀ10.9
117.9ꢀ1.5
103.3ꢀ7.7
Each value represents the meanꢀS.E. of three independent experiments (nꢂ3). Asterisks denote significant differences from the control at pꢃ0.05 (∗), pꢃ0.01 (∗∗). b-Galac-
tosidase activity (U) of 17b-estradiol was 641.2ꢀ22.5 and 2333.4ꢀ112.5 in yeast expressing ERa and ERb, respectively, at a concentration of 10ꢁ7 M (100%). b-Galactosidase
activity (U) of the tested compounds was calculated as a percentage of the 17b-estradiol activity. Compounds 18 and 19 were toxic at a concentration of 10ꢁ4 M. Tam. denotes
(tamoxifen).
and/or anti-estrogenic activities (Tables 4, 5). This finding over, physcion (11), which has a methyl group at the 6 posi-
has led us to investigate the main structure requirements tion, showed only anti-estrogenic activity.
responsible for the estrogenic activity of these compounds.
To the best of our knowledge, this is the first report about
Therefore, a group of synthetic compounds (20—24) pos- the estrogenic and anti-estrogenic activities of these naturally
sessing a partial similarity to the previously mentioned active occurring phenolic compounds based on naphthalene and
compounds were investigated for their estrogenic activity. naphthopyrone nucleuses. Finally, the increase in estrogenic
Moreover, demethylation of toralactone (14) and rubro- activity of the naringinase pretreated extract of Cassia tora
fusarin (15) was carried out to study the effect of the methyl suggests the activation of this extract by intestinal flora.
group at position 7 in these compounds on the activity.
In conclusion, the current study introduced the 1,3,8-trihy-
Among the tested synthetic and chemically modified com- droxynaphthalene (T3HN) as a new basic nucleus responsible
pounds, nor-rubrofusrain (18), nor-toralactone (19) and for estrogenic activity, which might be of interest in the dis-
T3HN (24) showed almost similar estrogenic and anti-estro- covery of new estrogenic and anti-estrogenic compounds
genic activities to those of 6-hydroxymusizin (17) (Tables based on this nucleus.
2—5). By comparing the structures of the formerly men-
tioned compounds it was found that trihydroxynaphthalene
Experimental
Chemicals and Reagents Naringinase was purchased from Sigma Co.
(T3HN) is a common feature in these compounds, indicating (St. Louis, MO, U.S.A.). Dulbecco’s modified Eagle’s medium (DMEM) was
obtained from Gibco BRL (Grand Island, NY, U.S.A.). Fetal bovine serum
(FBS) was purchased from ICN Biomedicals, Inc. (Aurora, OH, U.S.A.).
Streptomycin, 0.25% trypsin, and O-nitrophenyl b-D-galactoside (ONPG)
were purchased from Nacalai Tesque Co. (Kyoto, Japan). 17b-Estradiol was
its principal role in the estrogenic activity. It was also found
that methylation of a free hydroxy group at position 3 of the
T3HN nucleus as in toralactone (14), rubrofusarin (15) and
torachrysone (16) resulted in reduction of estrogenic activity
and binding affinity to the estrogenic receptor ERa, but
retained anti-estrogenic activity. On the other hand, 1,3- (22)
and 1,6-dihydroxylated naphthalene (23) showed weaker
estrogenic activities, which confirmed the significance of the
trihydroxylated system for effective binding to estrogen
receptors.
The estrogenic activity of aurantio-obtusin (13) is in
accordance with the previously reported data about the
importance of the phenolic group at 2 and 6 position in the
antharquinone nucleus for their estrogenic activity.22) More-
purchased from Calbiochem Co. (Darmstadt, Germany). Human serum was
obtained from Bio-Whittakar (Walkersville, MD, U.S.A.). 3-(4,5-Dimethyl-
2-thiazolyl)-2,3-diphenyl-2H-tetrazolium bromide (MTT), penicillin, norit
SX-II charcoal, 2,6-dihydroxynaphthalene, 6-hydroxy-2-naphthoic acid, 1,3-
dihydroxynaphthalene, 1,6-dihydroxnaphthalene and tamoxifen were pur-
chased from Wako Chem. Co. (Osaka, Japan). 20T-Zymolyase from Seika-
gaku Kogyo Co. (Tokyo, Japan). Dextran 70T was obtained from Amersham
Pharmacia Biotech AB. (Uppsala, Sweden). 1,3,8-Trihydroxynaphthalene
(T3HN) was kindly provided by Professor Yutaka Ebizuka (Graduate School
of Pharmaceutical Sciences, The University of Tokyo).
General Experimental Procedures The melting point was measured
on a Yanagimoto microhot stage melting point apparatus. Optical rotations
were measured with a DIP-360 automatic polarimeter (Jasco Co., Tokyo).

October 2007
1481
int.): 733 [MꢄH]ꢄ (100), 571 [MꢁGlc.ꢄH]ꢄ (69), 409 [Mꢁ2ꢈGlc.ꢄH]ꢄ
(50), 247 [agl.ꢄH]ꢄ (77). Anal. Found: C, 52.4; H, 6.2. C32H44O19 requires:
C, 52.4; H, 6.0%.
UV spectra were measured with a UV-2200 UV–VIS recording spectropho-
tometer (Shimadzu Co., Kyoto). IR spectra were measured with a Jasco
1
FT/IR-230 infrared spectrometer. H- and 13C-NMR were measured with a
JHA-LAA 400 WB-FT (1H, 400 MHz; 13C, 100 MHz; Jeol, Tokyo) spec-
trometer, the chemical shifts being represented as ppm with tetramethylsi-
lane as an internal standard. ESI-MS was carried out on an Esquire 3000
mass spectrometer (Bruker Daltanik GmbH, Bremen, Germany) system with
ESI ionization source. Elemental analysis was carried out on a Perkin Elmer
analyzer 2400 II. TLC was carried out on precoated silica gel 60 F254
(0.25 mm, Merck) and RP-18 F254S (0.25 mm, Merck Co.). Column chro-
matography (CC) was carried out on (BW-820MH silica gel), ODS DM
1020T (ODS, Fuji Silysia, Nagoya, Japan), Diaion HP-20 (Mitsubishi Kasei,
Tokyo) and Sephadex LH-20 (Pharmacia Co.). Medium pressure liquid
chromatography (MPLC) was performed on LiChroprep Rp-18 (size A,
Merck Co.). Preparative HPLC was performed on a Tosoh CCPM-CCPM-II
system (Tosoh Co., Tokyo) equipped with a UV 8020 detector and TSK gel
ODS-80Ts column (21.5ꢈ300 mm, Tosoh Co.).
Plant Material Cassia tora seeds were purchased at the Japanese mar-
ket and identified by a specialist in botanical identification of Suntory Com-
pany, Japan (Lot No. T-260).
Extraction and Isolation The seeds of Cassia tora were extracted with
70% ethanol and the extract was concentrated in vacuum. The ethanolic
extract (100 g) was suspended in water and fractionated with CHCl3 (3ꢈ1 l)
to obtain a chloroform fraction (7 g). The remaining water solution was
applied to a Diaion HP-20 column (60 cmꢈ6 cm) and stepwisely eluted with
H2O, 25% methanol, 50% methanol and 100% methanol. The 50% methanol
Toralactone 9-O-[b-D-Glucopyranosyl-(1→3)-O-b-D-glucopyranosyl-
(1→6)-O-b-D-glucopyranoside] (2) Yellowish amorphous powder. [a]D25
ꢁ38.7° [cꢂ0.15, MeOH]. UV (MeOH) nm lmax (log e): 237 (4.26), 247
(4.14), 270 (4.68), 277 (4.7), 383 (3.78). IR (KBr) cm-1: 3360, 1681, 1650,
1
1624, 1579, 1407, 1370, 1231, 1165, 1040, 868, 812. H-NMR (DMSO-d6,
400 MHz): d 2.22 (3H, s, CH3-3), 3.88 (3H, s, OCH3), 4.00 (1H, d, Jꢂ
12.0 Hz, H-6ꢅ), 4.28 (1H, d, Jꢂ8.0 Hz, H-1ꢇ), 4.31 (1H, d, Jꢂ8.0 Hz, H-1ꢆ),
5.09 (1H, d, Jꢂ8.0 Hz, H-1ꢅ), 6.50 (1H, s, H-4), 6.86 (1H, d, Jꢂ2.2 Hz, H-
8), 6.92 (1H, d, Jꢂ2.2 Hz, H-6), 7.13 (1H, s, H-5), 12.5 (1H, s, OH-9). 13C-
NMR (DMSO-d6, 100 MHz): d 18.8 (CH3-3), 55.6 (OCH3), 60.5 (C-6ꢆ),
61.0 (C-6ꢇ), 68.6 (C-4ꢆ), 68.8 (C-6ꢅ), 70.1 (C-4ꢅ), 70.1 (C-4ꢇ), 72.2 (C-2ꢆ),
73.2 (C-2ꢅ), 73.8 (C-2ꢇ), 75.7 (C-5ꢇ), 75.7 (C-5ꢅ), 76.1 (C-5ꢆ), 76.3 (C-3ꢅ),
76.9 (C-3ꢇ), 88.1 (C-3ꢆ), 98.5 (C-10a), 100.5 (C-6), 100.9 (C-1ꢅ), 101.9 (C-
8), 102.7 (C-1ꢆ), 104.1 (C-1ꢇ), 104.1 (C-4), 109.2 (C-9a), 111.6 (C-5), 132.4
(C-4a), 141.6 (C-5a), 152.7 (C-3), 157.6 (C-9), 161.4 (C-7), 162.8 (C-10),
166.8 (C-1). ESI-MS m/z (rel. int.): 781 [MꢄNa]ꢄ (50), 759 [MꢄH]ꢄ (100),
597 [MꢁGlc.ꢄH]ꢄ (24), 273 [aglꢄH]ꢄ (43). Anal. Found: C, 52.03; H,
5.77. C33H42O20 requires: C, 52.28; H, 5.59%.
Preparation of the Naringinase-Treated Extracts and Fractions23)
The 70% alcoholic extract and 25%, 50% and 100% MeOH fractions of C.
tora (40 mg, each) were incubated with naringinase enzyme (20 mg) in 2 ml
of 0.2 M acetate buffer (pHꢂ4.7) at 37 °C for 4 h. The resulting solution was
then extracted with BuOH (10 mlꢈ3) and the combined BuOH extracts were
evaporated under vacuum to get the naringinase-treated extract and frac-
tions. These, in turn, were dissolved in DMSO (10 mg/ml) as stock solutions
to test their estrogenic activity.
fraction (60 g) was subjected to
a silica gel column eluted with
CHCl3–MeOH–H2O (9 : 1 : 0.1→5 : 5 : 0.5 v/v/v) to yield 8 fractions. Com-
pound 3 (13 mg) was obtained from fraction 2 after purification on Sephadex
LH-20 using methanol as eluent followed by a medium pressure liquid chro-
matography (MPLC) Rp-18 column using MeOH–H2O (1 : 1 v/v). Fraction 4
was subjected to an ODS column to obtain compounds 5 (25 mg) and 6
(14 mg). Compound 4 (5 mg) was obtained from fraction 5 by purification
on a MPLC Rp-18 column eluted with MeOH–H2O (4 : 6 v/v). Fraction 8
was further purified using a MPLC Rp-18 column gradiently eluted with
MeOH–H2O to obtain compounds 8 (70 mg) and 9 (14 mg) in addition to
two other fractions 8-1 and 8-2. Fraction 8-1 was purified on Sephadex LH-
20 using MeOH–H2O (1 : 1 v/v) to get compounds 1 (20 mg) and 7 (75 mg).
Finally, compound 2 (6 mg) was obtained from fraction 8-2 by prep HPLC
(acetonitrile: 0.1% TFA/H2O (30% v/v) 5 ml/min, monitored at 285 nm).
Naringinase Treatment of the Extract and Isolation of the Agly-
cones23) The 70% ethanolic extract (25 g) was incubated with naringinase
enzyme (12.5 g) in 1000 ml of 0.2 M acetate buffer (pHꢂ4.7) at 37 °C for
24 h. The solution was extracted with EtOAc three times (500 mlꢈ3). The
pooled EtOAc extract was evaporated under vacuum to obtain the naringi-
nase treated extract (15 g). This extract was applied to a silica gel column
(30 cmꢈ5 cm) eluted with hexane–ethyl acetate (9.5 : 0.5—5 : 5 v/v). Five
fractions were obtained and screened for their estrogenic activity using
MCF-7 cells. Because of their relatively high estrogenic activity fractions 1
and 2 were selected for further purification. Fraction 1 was applied to silica
gel column (15 cmꢈ2 cm) using silica gel 60 and gradiently eluted with
hexane–ethyl acetate (9.5 : 0.5—9 : 1 v/v) to obtain compounds 10 (300 mg),
11 (25 mg) and 12 (10 mg). Fraction 2 was injected to an MPLC Rp-18 col-
umn eluted with MeOH–H2O (8 : 2 v/v) affording compound 13 (30 mg) and
a mixture of compounds 14 and 15. The isolation of compounds 14 and 15
was achieved by repeated purification using prep. HPLC (acetonitrile: 0.1%
TFA/H2O (75% v/v) 5 ml/min, monitored at 285 nm) to afford compounds
14 (30 mg) and 15 (130 mg).
Acid Hydrolysis of Torachrysone Tetraglucoside (4) and 6-Hydroxy-
musizin 8-O-b-D-Glucoside (6)11) A solution of 4 and 6 (10 mg of each) in
1% H2SO4 (0.5 ml) in a sealed tube was heated separately on a boiling water
bath for 1 h. The solution was then extracted with EtOAC, and concentrated
under vacuum to obtain torachrysone (17) (2 mg) and 6-hydroxymusizin
(16) (5 mg). Both compounds were identified by 1H-NMR and EI-MS.
Acid Hydrolysis of Compounds 1 and 211) A solution of 1 and 2 (4 mg
of each) in a 1% H2SO4 (0.2 ml) in a sealed tube was heated on a boiling
water bath for 1 h. The solution was then extracted with EtOAc, and concen-
trated under vacuum to get torachrysone and toralactone, respectively, and
1
the identity of the aglycones was verified by H-NMR, EI-MS, TLC and by
comparison with the reported data.11,13,18—20) The aqueous layer was neutral-
ized with sodium carbonate and freeze dried. The sugar component of the
residue was detected by TLC through comparison with authentic samples to
be glucose. The absolute configuration of the sugar was determined as
D-glucose according to the method described by Hara et al.25)
Demethylation of Toralactone (14) and Rubrofusrin (15)24) Toralac-
tone and rubrofusarin (5 mg each) were separately dissolved in acetic anhy-
dride (5 ml), treated gradually with HI (dꢂ1.7) and refluxed gently for 2 h
with occasional shaking. At the end of the reaction, a saturated solution of
sodium thiosulfate (5 ml) was added. The resulting mixture was extracted by
EtOAc and evaporated under vacuum. The resulting demethylated com-
pounds were purified using prep HPLC (acetonitrile: 0.1% TFA/H2O (80%
v/v), 5 ml/min, monitored at 285 nm) to afford nor rubrofusrin (18) (2 mg)
and nor toralactone (19) (1.5 mg).
MCF-7 Cells Proliferation Assay Estrogen-sensitive human MCF-7
breast cells were grown in DMEM supplemented with 5% FBS, penicillin,
and streptomycin. The cells were harvested by trypsinization (0.25% trypsin)
and plated at a concentration of 5ꢈ103 cells/well in DMEM supplemented
with 5% FBS in 96-well tissue culture plates (Iwaki Co., Chiba, Japan) and
allowed to attach for 24 h. Then the culture medium was replaced with phe-
nol red-free DMEM containing 10% heat-inactivated dextran/charcoal-
stripped (DC) human serum prior to the addition of compounds and 17b-
estradiol. Stock solutions of test compounds in DMSO were diluted with DC
medium. The final DMSO concentration in culture medium did not exceed
0.1%, and this concentration did not affect cell viability. After 4 d in a hu-
midified incubator with 5% CO2 at 37 °C, the proliferation of the cells was
measured using the MTT method.26)
Torachrysone 8-O-[b-Glucopyranosyl(1→3)-O-b- -glucopyranosyl(1→
D
6)-O-b-D-glucopyranoside] (1) Yellow needles (MeOH). mp 249—
251 °C. [a]D25 ꢁ38.5° [cꢂ0.32, MeOH]. UV (MeOH) nm lmax (log e): 240
(4.95), 265 (4.61), 310 (4.12), 324 (4.11), 340 (4.12). IR (KBr) cmꢁ1: 3390,
1629, 1398, 1260, 1165. 1H-NMR (DMSO-d6, 400 MHz): d 2.21 (3H, s,
CH3-3), 2.48 (3H, s, COCH3), 3.82 (3H, s, OCH3), 4.00 (1H, d, Jꢂ12.0 Hz,
H-6ꢅ), 4.30 (1H, d, Jꢂ8.0 Hz, H-1ꢇ), 4.32 (1H, d, Jꢂ8.0 Hz, H-1ꢆ), 5.09
(1H, d, Jꢂ8.0 Hz, H-1ꢅ), 6.89 (1H, d, Jꢂ2.1 Hz, H-5), 7.06 (1H, d,
Jꢂ2.1 Hz, H-7), 7.09 (1H, s, H-4), 9.46 (1H, s, OH-1). 13C-NMR (DMSO-
d6, 100 MHz): d 19.5 (CH3-3), 32.2 (COCH3), 55.4 (OCH3), 60.7 (C-6ꢆ),
61.0 (C-6ꢇ), 68.3 (C-4ꢆ), 68.8 (C-6ꢅ), 70.0 (C-4ꢅ), 70.1 (C-4ꢇ), 72.2 (C-2ꢆ),
73.3 (C-2ꢅ), 73.8 (C-2ꢇ), 76.0 (C-5ꢇ), 76.1 (C-5ꢅ), 76.4 (C-5ꢆ), 76.9 (C-3ꢅ,
C-3ꢇ), 88.1 (C-3ꢆ), 101.1 (C-5), 102.4 (C-1ꢅ), 102.8 (C-1ꢆ), 103.3 (C-7),
104.1 (C-1ꢇ), 108.6 (C-4a), 118.8 (C-4), 123.2 (C-2), 133.6 (C-3), 136.8 (C-
8a), 151.0 (C-1), 155.2 (C-8), 158.4 (C-6), 204.5 (CO). ESI-MS m/z (rel.
Yeast Two-Hybrid Assay The yeast two-hybrid assay was carried out
according to the method of Nishikawa and Kanayama.27,28) Briefly, yeast
cells expressing ERa and ERb were separately grown overnight at 30 °C
with shaking in a synthetic defined medium (SD) lacking tryptophan and
leucine. Yeast cells were treated with 17b-estradiol and the isolated com-
pounds for 4 h at 30 °C, and b-galactosidase activity was determined as fol-

1482
Vol. 55, No. 10
lows. The growth of the yeast cells was monitored by measuring the turbid-
ity at 600 nm. The treated yeast cells were collected by centrifugation
(8000ꢈg, 5 min) and re-suspended in 200 ml of Z-buffer (0.1 M sodium
phosphate, pH 7.0, 10 mM KCl, and 1 mM MgSO4) containing 1 mg/ml of
zymolyase at 37 °C for 15 min. The reaction was started by the addition of
40 ml of 4 mg/ml O-nitrophenol b-D-galactopyranoside (ONPG) as a sub-
strate. When a yellow color developed (incubation time: t), 100 ml of 1 M
Na2CO3 was added to stop the reaction. The absorbance of the solution
(150 ml) was measured at 420 and 550 nm. The b-galactosidase activity was
determined using the following formula:
Pharm. Bull., 46, 1650—1652 (1998).
4) Koo A., Wang J. S., Li K. M., Amer. J. Chin. Med., 4, 245—248
(1976).
5) Chan S. H., Koo A., Li K. M., Amer. J. Chin. Med., 4, 383—389
(1976).
6) Chopra R. N., Chopra I. C., Handa K. L., Kapur L. D., “Chopra’s In-
digenous Drugs of India,” 2nd ed., Academic Publishers, Calcutta,
1958, pp. 524, 601.
7) Kitanaka S., Takido M., Chem. Pharm. Bull., 32, 860—864 (1984).
8) Kitanaka S., Takido M., Chem. Pharm. Bull., 36, 3980—3984 (1988).
9) Wong S., Wong M., Seligmann O., Wagner H., Phytochemistry, 28,
211—214 (1989).
Uꢂ1000ꢈ(A420ꢁ1.75ꢈA550)/(tꢈ0.05ꢈA600
)
10) Wong S., Wong M., Seligmann O., Wagner H., Planta Med., 55, 276—
280 (1989).
11) Hatano T., Uebayashi H., Ito H., Shiota S., Tsuchiya T., Yoshida T.,
Chem. Pharm. Bull., 47, 1121—1127 (1999).
12) Kim Y., Lee C., Kim H., Lee H., J. Agric. Food Chem., 52, 6096—
6100 (2004).
13) Choi J., Lee H., Park Y., Ha J., Kang S., Planta Med., 63, 11—14
(1997).
14) Patil U., Saraf S., Dixit V. K., J. Ethnopharm., 90, 249—252 (2004).
15) Wu C., Hseih C., Song T., Yen G., J. Agric. Food Chem., 49, 2579—
2586 (2001).
16) Brown K., Cameron D., Weiss U., Tetrahedron Lett., 6, 471—476
(1969).
17) Savard J., Brassard P., Tetrahedron, 4, 3455—3464 (1984).
18) Takahashi S., Takido M., Yakugaku Zasshi, 93, 261—267 (1972).
19) Shibata S., Morishita E., Kaneda M., Kimura Y., Takido M., Takahashi
S., Chem. Pharm. Bull., 17, 454—457 (1969).
20) Wei B., J. Nat. Prod., 55, 967—969 (1992).
21) Klopman G., Chakravarti S., Chemosphere, 51, 445—459 (2003).
22) Matsuda H., Shimoda H., Morikawa T., Yoshikawa M., Bioorg. Med.
Chem. Lett., 2001, 1839—1842 (2001).
23) Ahn E., Akao T., Nakamura N., Komatsu K., Nishihara T., Hattori M.,
J. Trad. Med., 21, 81—86 (2004).
24) Suleimanov T., Chem. Nat. Compd., 40, 13—15 (2004).
25) Hara S., Okabe H., Mihashi K., Chem. Pharm. Bull., 35, 501—506
(1987).
26) Carmichael J., DeGraph W. G., Grazer A. F., Minn J. D., Michell J. B.,
Cancer Res., 47, 936—942 (1987).
27) Nishikawa J., Saito K., Goto J., Dakeyama F., Matsuo M., Nishihara T.,
Toxicol. Appl. Pharmacol., 154, 76—83 (1999).
Ligand Binding Screen An estrogen-R (a) competitor screening kit
was purchased from Wako Chemicals Japan Inc. The assay depends on the
competition between the samples applied in different concentrations and the
labeled estrogen mixture. The amount of the ligand that binds to the ERa
coated on the microplate well is determined by the dynamic equilibrium
among all the ligand concentrations in the mixture, the difference of their
binding affinities to the receptor and the incubation time. Therefore, the re-
duction in fluorescence intensities from the labeled estrogen retained is an
indication of the affinity of the added compounds to the estrogen receptor.
The isolated compounds were tested in 10ꢁ5, 10ꢁ4 and 5ꢈ10ꢁ3 M concentra-
tions. Estradiol was used as a positive control and a labeled estrogen mixture
was used as a negative control. The results were calculated as percentages of
the negative control.
Anti-estrogenic Assay To examine the antagonistic activity of the test
compounds, the inhibition of b-galactosidase activity and the proliferation
of MCF-7 cells, which had been induced by 10ꢁ7 M and 10ꢁ12 M 17b-estra-
diol, respectively, were measured at various concentrations of test com-
pounds.
Statistical Analysis Each set of experiments was repeated at least three
times. Values are expressed as meanꢀS.E.M. One-way analysis of variance
followed by Dunnett’s test was used for statistical analysis.
Acknowledgements Part of this study was financially supported by
Suntory Ltd., Osaka, Japan. The authors are also grateful to Prof. Yutaka
Ebizuka (Graduate School of Pharmaceutical Sciences, The University of
Tokyo) for kindly providing the T3HN sample.
References
1) McDonnell D., Trends Endocrinol. Metab., 10, 310—311 (1999).
2) Cos P., Bruyne T., Apers S., Berghe D., Pieters L., Vlietinik A., Planta
Med., 69, 589—599 (2003).
28) Kanayama T., Mamiya S., Nishihara T., Nishikawa J., J. Biochem.
(Tokyo), 133, 791—797 (2003).
3) Kitanaka S., Nakayama T., Shibano T., Ohkoshi E., Takido M., Chem.