New cardenolide glycosides from the seeds of Digitalis purpurea and their cytotoxic activity

Article abstract of DOI:10.1271/bbb.120922

A chemical investigation of Digitalis purpurea seeds led to the isolation of three new cardenolide glycosides (1, 8 and 11), together with 12 known cardenolide glycosides (2-7, 9, 10 and 12-15). The structures of 1, 8 and 11 were determined by 1D and 2D NMR spectroscopic analyses and the results of an acid or enzymatic hydrolysis. The cytotoxic activity of the isolated compounds (1-15) against HL-60 leukemia cells was examined. Compounds 2, 9, 11 and 12 showed potent cytotoxicity against HL-60 cells with respective 50% inhibition concentration (IC₅₀) values of 0.060, 0.069, 0.038, and 0.034μM. Compounds 2, 9 and 11 also exhibited potent cytotoxic activity against HepG2 human liver cancer cells with respective IC₅₀ values of 0.38, 0.79, and 0.71 μM. An investigation of the structure-activity relationship showed that the cytotoxic activity was reduced by the introduction of a hydroxy group at C-16 of the digitoxigenin aglycone, methylation of the C-3′ hydroxy group at the fucopyranosyl moiety, and acetylation of the C-3′ hydroxy group at the digitoxopyranoyl moiety.

Full text of DOI:10.1271/bbb.120922

Biosci. Biotechnol. Biochem., 77 (6), 1186–1192, 2013
New Cardenolide Glycosides from the Seeds of Digitalis purpurea
and Their Cytotoxic Activity
y
Minpei KURODA,1; Satoshi KUBO,1 Yukiko MATSUO,1 Tomomi ATOU,1 Junichi SATOH,2
1
Tomofumi FUJINO,2 Makio HAYAKAWA,2 and Yoshihiro MIMAKI
1Department of Medicinal Pharmacognosy, Tokyo University of Pharmacy and Life Sciences, School of Pharmacy,
1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan
2Department of Hygiene and Health Sciences, Tokyo University of Pharmacy and Life Sciences, School of Pharmacy,
1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan
Received December 3, 2012; Accepted March 5, 2013; Online Publication, June 7, 2013
[doi:10.1271/bbb.120922]
A chemical investigation of Digitalis purpurea seeds
led to the isolation of three new cardenolide glycosides
(1, 8 and 11), together with 12 known cardenolide
glycosides (2–7, 9, 10 and 12–15). The structures of 1, 8
and 11 were determined by 1D and 2D NMR spectro-
scopic analyses and the results of an acid or enzymatic
hydrolysis. The cytotoxic activity of the isolated com-
pounds (1–15) against HL-60 leukemia cells was exam-
ined. Compounds 2, 9, 11 and 12 showed potent
cytotoxicity against HL-60 cells with respective 50%
inhibition concentration (IC50) values of 0.060, 0.069,
0.038, and 0.034 ꢀM. Compounds 2, 9 and 11 also
exhibited potent cytotoxic activity against HepG2
human liver cancer cells with respective IC50 values of
0.38, 0.79, and 0.71 ꢀM. An investigation of the struc-
ture-activity relationship showed that the cytotoxic
activity was reduced by the introduction of a hydroxy
group at C-16 of the digitoxigenin aglycone, methylation
of the C-30 hydroxy group at the fucopyranosyl moiety,
and acetylation of the C-30 hydroxy group at the
digitoxopyranoyl moiety.
examination of D. purpurea seeds which resulted in the
isolation of 15 cardenolide glycosides (1–15) (Fig. 1),
including three new compounds (1, 8 and 11). We report
here the structures of the new compounds determined by
extensive spectroscopic analyses, including two-dimen-
sional (2D) NMR, as well as the results of an acid or
enzymatic hydrolysis. The cytotoxic activity of 1–15
against HL-60 human promyelocytic leukemia cells and
the structure-activity relationship were also examined.
Compounds 2, 9 and 11, which each showed potent
cytotoxic activity against HL-60 cells, were also
evaluated for their cytotoxicity against HepG2 human
liver cancer cells.
Results and Discussion
The MeOH extract of D. purpurea seeds was passed
through a porous-polymer polystyrene resin (Diaion HP-
20) column. The EtOH fraction was purified by silica gel
and octadecylsilanized (ODS) silica gel column chro-
matography (CC), and by reversed-phase preparative
HPLC, giving cardenolide glycosides (1–15). Known
compounds 2–7, 9, 10 and 12–15 were respectively
identified as digitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!
4)-ꢀ-D-fucopyranoside (glucodigifucoside, 2),14) digitox-
igenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-digitalopy-
ranoside (odorobioside G, 3),7) gitoxigenin 3-O-ꢀ-D-
glucopyranosyl-(1!4)-ꢀ-D-fucopyranoside (glucogito-
fucoside, 4),16) gitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!
4)-ꢀ-D-digitalopyranoside (digitalin, 5),16) gitoxigenin
Key words: Digitalis purpurea; Scrophulariazeae; car-
denolide glycoside; HL-60 cells; cytotoxic
activity
Cardenolides are a group of naturally occurring
products found in a limited number of plant families
such as Scrophulariazeae, Apocynaceae, Liliaceae, and
Ranunculaceae. We have previously reported the iso-
lation and structural characterization of the cardenolides
from rhizomes of Convallaria majalis (Liliaceae)1) and
seeds of Adonis aestivalis (Ranunculaceae),2) as well as
their cytotoxic activities against cultured tumor cells and
normal cells. In the 1950s and 1960s, phytochemical
studies were carried out on Digitalis purpurea seeds
(Scrophulariaceae), and more than 10 cardenolide
glycosides were isolated and identified.3–17) As part of
our continuous search for cardenolides with cytotoxic
activity,1,2) we performed a systematic phytochemical
3-O-ꢀ-
4)-ꢀ-D-digitalopyranoside (neogitostin, 6),13) gitoxigenin
3-O-ꢀ- -glucopyranosyl-(1!
-glucopyranosyl-(1!4)-ꢀ-
4)-ꢀ-D-digitalopyranoside (gitostin, 7),15) digitoxigenin
3-O-ꢀ- -digitoxo-
D-glucopyranosyl-(1!6)-ꢀ-D-glucopyranosyl-(1!
D
D
D
-glucopyranosyl-(1!4)-3-O-acetyl-ꢀ-
D
pyranoside (30-O-acetylglucoevatromonoside, 9),18)
gitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-3-O-acetyl-
ꢀ-D-digitoxopyranoside (acetylglucogitoroside, 10),4)
digitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-digi-
toxopyranosyl-(1!4)-ꢀ-D-digitoxopyranosyl-(1!4)-ꢀ-
y
To whom correspondence should be addressed. Fax: +81-42-676-4579; E-mail: kurodam@toyaku.ac.jp
Abbreviations: COSY, correlation spectroscopy; DMEM, Dulbecco’s modified Eagle’s medium; DMSO, dimethyl sulfoxide; FBS, fetal bovine
serum; FCS, fetal calf serum; HMBC, heteronuclear multiple-bond connectivity; HMQC, heteronuclear multiple quantum correlation; HPLC, high-
performance liquid chromatography; HR-ESI-TOF-MS, high-resolution-electrospray ionization-time of flight-mass spectrometry; IC50, 50%
inhibition concentration; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide; NMR, nuclear magnetic resonance; PBS,
phosphate-buffered saline

Cytotoxic Cardenolides from Digitalis purpurea Seeds
1187
23
O
cardenolide skeleton. In addition, signals for three
anomeric protons were present at ꢁH 4.93 (d, J ¼
7:8 Hz), 5.35 (dd, J ¼ 9:6, 1.8 Hz), and 5.15 (dd, J ¼
9:5, 1.8 Hz). The methyl carbon signals at ꢁC 18.60 and
18.39, and proton signals at ꢁH 1.59 (d, J ¼ 6:3 Hz) and
1.41 (d, J ¼ 6:2 Hz) suggested that 1 possessed two
21
O
18
20
22
19
H
HO
HO
14
O
H
OH
O
HO
O
O
3
O
O
HO 1'''
1''
H
1'
OH
OAc
1
deoxy sugars. The H-NMR [ꢁH 2.11 (3H, s)] and 13C-
1
O
O
NMR [ꢁC 170.14 (C=O) and 21.26 (Me)] peaks indicated
the presence of an acetyl group. Acid hydrolysis of 1 with
0.025 M HCl gave 3ꢀ,14ꢀ-dihydroxycard-20(22)-enolide
(digitoxigenin), and the carbohydrate moieties D-digi-
toxose and D-glucose. Identification of D-digitoxose and
D-glucose, including their absolute configurations, was
carried out by a direct HPLC analysis of the hydrolysate,
and 1 was identified as a monoacetate of digitoxigenin
16
R1
R5O
R3O
H
O
HO 1''
O
R4O
H
OH
O
O
R2O
H
HO 1'
R1
H
H
R2
H
Me
H
R3 R4 R5
HO
2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Glc
H
O
Glc:
Ac:
HO
HO
3
4
5
6
7
8
HO
OH
1
triglycoside. An analysis of the H-1H COSY spectra
OH Me
OH Me
OH Me
O
and HMQC spectra allowed the exact sugar sequences of
the monosaccharides to be determined. The sequential
assignments from H-1 to H2-6 or Me-6 of each mono-
Glc
H
OH Me Glc
H
O
O
1
saccharide were established by a H-1H COSY experi-
R1
H
ment. The signal multiplet patterns and coupling
constants (Table 1) indicated the presence of two ꢀ-D-
digitoxopyranosyl (4C1) units (Dgt0 and Dgt00) and a ꢀ-D-
glucopyranosyl (4C1) unit (Glc000). The HMQC spectrum
showed that these proton resonances were correlated
with those of the one-bond coupled carbons (Table 2).
The anomeric configurations of the Dgt and Glc moieties
were ascertained by the relatively large J values of their
anomeric protons (7.8–9.6 Hz). The HMBC spectrum of
1 showed long-range correlations between H-1 of Glc000
at ꢁH 4.93 and C-4 of Dgt00 at ꢁC 83.39, H-1 of Dgt00 at ꢁH
5.35 and C-4 of Dgt0 at ꢁC 79.84, H-1 of Dgt0 at ꢁH 5.15
and C-3 of the aglycone at ꢁC 73.83, and between H-3 of
Dgt0 at ꢁH 5.81 and the acetyl carbonyl carbon at ꢁC
170.14. The structure of 1 accordingly was established as
digitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-digi-
HO
H
OH
O
R2O
O
O
1'
O
HO
HO 1''
H
OAc
R1 R2
9
H
H
H
Glc
10 OH
11
H
O
O
R1
H
HO
O
HO
O
HO
O
H
OH
O
O
HO 1''''
O
O
1'''
O
1''
H
OH
1'
OH
OR2
R1 R2
12
H
H
H
Ac
13 OH
14
H
15 OH Ac
Fig. 1. Chemical Structures of 1–15.
toxopyranosyl-(1!4)-3-O-acetyl-ꢀ-D-digitoxopyranoside.
Compound 8 was shown to have a molecular formula
of C42H66O19 by the HR-ESI-TOF-MS data (m=z
897:4096 þ Na]þ). Enzymatic hydrolysis of 8 with ꢀ-
D-glucosidase produced a partial hydrolysate, identified
as 5, together with D-glucose. The deduced molecular
formula for 8 was the same as those of concomitantly
isolated cardenolide glycosides 6 and 7. Furthermore,
the spectral features of 8 were essentially analogous to
those of 6 and 7. The 13C-NMR spectrum suggested that
8 was an isomer of 6 and 7 and that the terminal ꢀ-D-
glucopyranosyl (Glc000) unit was attached to C-3 of the
inner Glc00 unit (Table 2). In the HMBC spectrum of 8,
long-range correlations were observed between H-1 of
Glc000 and C-3 of Glc00, H-1 of Glc00 and C-4 of ꢀ-D-
digitalopyranse (Dtl0), and between H-1 of Dtl0 and C-3
of the aglycone. Compound 8 was thus formulated as
gitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!3)-ꢀ-D-gluco-
pyranosyl-(1!4)-ꢀ-D-digitalopyranoside.
D-digitoxopyranoside (purpurea glycoside A, 12),3)
gitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-digi-
toxopyranosyl-(1!4)-ꢀ-D-digitoxopyranosyl-(1!4)-ꢀ-
D-digitoxopyranoside (purpurea glycoside B, 13),3)
digitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-digi-
toxopyranosyl-(1!4)-ꢀ-D-digitoxopyranosyl-(1!4)-3-
O-acetyl-ꢀ-D-digitoxopyranoside (purlanoside A, 14),6)
and gitoxigenin 3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-
digitoxopyranosyl-(1!4)-ꢀ-D-digitoxopyranosyl-(1!4)-
3-O-acetyl-ꢀ-D-digitoxopyranoside (purlanoside B, 15).6)
This is the first report on the isolation of 9 from
D. purpurea seeds. Detailed spectroscopic data were
almost unavailable for identifying 2–7, 9, 10 and 12–15,
so a structural determination of these compounds was
carried out based on a spectroscopic analysis, including
2D NMR data.
Compound 1 was obtained as an amorphous solid and
its molecular formula was determined as C43H66O16,
based on the HR-ESI-TOF-MS data (m=z 861.4305
½M þ Naꢀþ). The IR spectrum of 1 showed absorption
bands for hydroxyl (3378 cmꢁ1) and carbonyl (1740
cmꢁ1) groups. The 1H-NMR spectrum contained two
methyl group singlets at ꢁH 1.01 and 0.88 (Me-18 and
Me-19), an oxymethylene group at ꢁH 5.32 and 5.04
(each dd, J ¼ 18:1, 1.5 Hz, H2-21), and an olefinic group
at ꢁH 6.14 (br s, H-22), these being characteristic of the
The HR-ESI-TOF-MS data (m=z: 893.4169 ½M þ
Naꢀþ) showed 11 to have a molecular formula of
C43H66O18. The 1H- and 13C-NMR spectral data
suggested that the structure of the aglycone moiety of
11 was identical to that of 9. The molecular formula for
1
11 differed from that for 9 by C6H10O5, and the H-
NMR spectrum showed signals for three anomeric
protons (Table 1). Enzymatic hydrolysis of 11 gave 9
and D-glucose. When the 13C-NMR spectrum of 11 was

1188
M. KURODA et al.
Table 1. 1H-NMR Data for the Sugar Moieties of 1, 8 and 11 in C5D5N
1
8
11
Positions
ꢁH
J (Hz)
9.5, 1.8
ꢁH
J (Hz)
7.6
9.7, 7.6
ꢁH
J (Hz)
9.5
10
20
5.15
2.41
2.10
5.81
3.63
4.17
1.41
dd
ddd
dd
ddd
dd
dq
d
4.68
4.43
—
d
dd
5.16
2.36
2.06
5.90
3.77
4.26
1.68
br d
m
a
13.9, 3.2, 1.8
13.9, 9.5, 3.2
3.2, 3.2, 3.1
9.4, 3.1
b
m
30
40
50
60
3.53
4.32
3.71
1.54
3.67
dd
m
m
d
9.7, 2.8
br d
dd
m
3.0
9.5, 3.1
9.4, 6.2
6.2
6.4
d
6.2
Methoxy
Acetyl
s
2.11
s
2.09
s
100
5.35
2.31
1.96
4.70
3.59
4.31
1.59
—
dd
9.6, 1.8
5.15
3.97
—
d
dd
7.8
8.8, 7.8
4.90
3.97
—
d
dd
7.8
8.7, 7.8
200
a
ddd
ddd
ddd
dd
13.4, 2.9, 1.8
13.4, 9.6, 2.9
2.9, 2.9, 2.8
9.6, 2.8
b
300
400
500
600
4.20
4.04
3.88
4.50
4.25
dd
dd
m
8.8, 8.7
9.4, 8.7
4.22
4.18
4.01
4.53
4.29
dd
dd
m
8.8, 8.7
8.8, 8.8
dd
d
9.6, 6.3
6.3
a
dd
dd
10.8, 2.2
10.8, 4.8
dd
dd
11.6, 2.2
11.6, 5.9
b
1000
2000
3000
4000
5000
6000
4.93
3.91
4.22
4.21
3.95
4.50
4.37
d
7.8
5.28
4.05
4.26
4.22
4.00
4.32
d
7.8
5.13
4.08
4.20
4.19
3.88
4.45
d
7.9
8.3, 7.9
dd
dd
dd
m
7.9, 7.8
7.9, 7.9
7.9, 7.9
dd
dd
dd
m
m
8.5, 7.8
8.7, 8.5
8.7, 8.7
dd
m
m
ddd
m
9.0, 3.3, 3.3
a
dd
dd
11.8, 2.6
11.8, 5.1
b
compared with that of 9, a set of six additional signals
corresponding to a terminal ꢀ-D-glucopyranosyl unit
(Glc000) was observed (Table 2). The signal from C-4 of
the Glc00 moiety and its neighboring carbons varied,
whereas all the other signals remained mainly unaffected.
In the HMBC spectrum of 11, H-1 of Glc000 exhibited a
long-range correlation with C-4 of Glc00. HMBC
correlations were also observed between H-1 of Glc00
and C-4 of Dgt0, H-1 of Dgt0 and C-3 of the aglycone,
and between H-3 of Dgt0 and the acetyl carbonyl carbon.
The structure of 11 thus was determined as digitoxigenin
3-O-ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-glucopyranosyl-
(1!4)-3-O-acetyl-ꢀ-D-digitoxopyranoside.
noside, showed higher activity than 3 and 5, these being
the corresponding 3-O-methyl derivatives of the inner
fucopyranosyl moiety of 2 and 4. Compounds 14 and 15
were respectively the 3-O-acetyl derivatives of the
innermost digitoxopyranosyl unit of 12 and 13, and were
less cytotoxic than 12 and 13. These results imply that a
slight modification of the sugar moieties affected the
cytotoxicity against tumor cells. In contrast, glycosyla-
tion of the terminal glucosyl moiety of 5 and 9 had little
effect on the activity. Compounds 2, 9, 11 and 12, which
showed potent cytotoxic activity against HL-60 cells,
were also cytotoxic to HepG2 human liver cancer cells
with mean respective IC50 values of 0.38, 0.79, 0.71, and
3.2 mM (Table 4). These activities are as potent as, or
more potent than that of actinomycin D which was used
as the positive control (IC50 1.2 mM).
The cardenolide glycosides (1–15) isolated in this
study were simultaneously evaluated for their cytotoxic
activity against HL-60 leukemia cells (Table 3). The
positive controls were etoposide and cisplatin, which
had respective IC50 values of 0.46 mM and 1.6 mM against
HL-60 cells. Compounds 2, 9, 11 and 12 showed potent
cytotoxic activity against HL-60 cells, with mean
respective IC50 values of 0.060, 0.069, 0.038, and
0.034 mM. The considerable cytotoxicity of 1, 3, 4, 10, 13
and 14 was comparable to that of etoposide, with mean
IC50 values ranging from 0.19 to 0.56 mM, whereas 5–8
and 15 were moderately cytotoxic with mean IC50
values from 1.9 to 3.7 mM. When the cytotoxic activity of
the digitoxigenin glycosides (2, 3, 9, 12 and 14) was
compared with that of the corresponding gitoxigenin
glycosides (4, 5, 10, 13 and 15), which contained the
same sugar moiety at C-3 of the digitoxigenin aglycone
moiety, the digitoxigenin glycosides displayed higher
cytotoxic activity than the gitoxigenin glycosides,
indicating that the introduction of a hydroxy group at
the C-16 position of the aglycone reduced the cytotoxic
activity. On the other hand, 2 and 4, which contained the
diglycoside ꢀ-D-glucopyranosyl-(1!4)-ꢀ-D-fucopyra-
Experimental
General information. Optical rotation data were measured with a P-
1030 (Jasco, Tokyo, Japan) automatic digital polarimeter. IR spectra
were recorded with a Jasco FT-IR 620 spectrophotometer, and UV
spectra with a Jasco V-630 spectrophotometer. NMR spectra were
acquired with a DRX-500 spectrometer (500 MHz for 1H-NMR;
Bruker, Karlsruhe, Germany), using standard Bruker pulse programs.
Chemical shifts are given as ꢁ-values with reference to TMS as an
internal standard. ESI-TOF-MS data were obtained with a Micromass
LCT mass spectrometer (Waters, Manchester, UK). Diaion HP-20
porous-polymer polystyrene resin (Mitsubishi-Chemicals, Tokyo,
Japan), 300 mesh silica gel (Fuji-Silysia Chemicals, Aichi, Japan),
75 mm ODS silica gel (Nacalai-Tesque, Kyoto, Japan), and Sephadex
LH-20 (GE Healthcare Japan, Tokyo, Japan) were used for column
chromatography (CC). TLC was carried out on precoated 0.25-mm-
thick silica gel 60 F254 (Merck, Darmstadt, Germany) and 0.25-mm-
thick RP-18 F254S (Merck) plates, and spots were visualized by
spraying with 10% H2SO4 aq. solution and subsequent heating. HPLC
was performed by using a system comprising a CCPM pump (Tosoh,
Tokyo, Japan), a CCP PX-8010 controller (Tosoh), an RI-8010

Cytotoxic Cardenolides from Digitalis purpurea Seeds
Table 2. 13C-NMR Data for 1–15 in C5D5N
1189
Positions
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
2
30.77 30.56 30.45 30.56 30.40 30.43 30.41 30.40 30.78 30.71 30.78 30.38 30.46 30.82 30.80
27.16 27.19 27.20 27.06 27.01 27.02 27.01 27.14 27.18 27.01 27.18 26.51 27.08 27.17 27.05
73.83 74.50 74.46 74.47 74.42 74.39 74.40 74.35 73.91 73.83 73.89 73.50 75.15 73.83 73.79
30.81 30.70 30.68 30.68 30.61 30.63 30.62 30.66 30.88 30.79 30.85 30.81 30.79 30.82 30.80
37.05 36.66 36.59 36.65 36.54 36.56 36.55 36.57 37.07 37.00 37.07 36.97 37.01 37.06 36.99
27.32 27.30 27.33 27.12 27.11 27.12 27.11 27.14 27.33 27.07 27.33 27.31 27.08 27.33 27.11
21.55 21.52 21.53 21.26 21.25 21.26 21.26 21.27 21.56 21.28 21.56 21.57 21.29 21.56 21.26
41.90 41.89 41.91 42.10 42.09 42.10 42.10 42.12 41.92 42.07 41.91 41.71 42.11 41.90 42.12
35.88 35.88 35.90 35.84 35.81 35.83 35.83 35.84 35.90 35.80 35.90 35.88 35.82 35.89 35.84
35.51 35.48 35.46 35.42 35.38 35.38 35.39 35.41 35.53 35.42 35.52 35.33 35.46 35.52 35.46
22.00 21.99 22.00 21.93 21.92 21.93 21.93 21.95 22.01 21.91 22.01 22.01 21.94 22.00 21.95
39.86 39.85 39.88 40.04 40.00 40.04 40.05 40.06 39.88 39.99 39.87 39.98 40.06 39.87 40.05
50.13 50.11 50.13 50.51 50.51 50.51 50.51 50.52 50.14 50.49 50.13 50.07 50.51 50.13 50.52
84.64 84.64 84.66 84.16 84.12 84.16 84.17 84.17 84.65 84.11 84.64 84.45 84.18 84.65 84.15
33.16 33.16 33.18 43.98 44.00 43.99 43.98 44.00 33.18 43.93 33.17 33.24 43.99 33.17 44.00
27.04 27.06 27.70 72.31 72.28 72.31 72.31 72.33 27.06 72.28 27.05 27.06 72.32 27.05 72.32
51.48 51.46 51.49 59.41 59.42 59.40 59.42 59.44 51.49 59.39 51.49 51.44 59.43 51.48 59.45
16.20 16.19 16.20 16.97 16.95 16.96 16.97 16.98 16.21 16.95 16.21 16.17 16.98 16.20 16.98
23.92 23.73 23.70 23.75 23.70 23.71 23.70 23.73 23.94 23.92 23.93 23.90 23.95 23.93 23.95
176.02 176.06 176.02 172.65 172.69 172.64 172.63 172.63 175.99 172.65 176.00 176.04 172.64 176.01 172.62
73.74 73.73 73.73 76.73 76.74 76.72 76.72 76.73 73.73 76.73 73.73 73.71 76.74 73.73 76.73
117.65 117.63 117.65 120.14 120.15 120.14 120.15 120.17 117.68 120.14 117.67 117.61 120.16 117.66 120.19
174.56 174.56 174.54 174.68 174.67 174.68 174.66 174.66 174.53 174.68 174.53 174.52 174.66 174.54 174.65
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
10
96.87 103.51 103.37 103.48 103.34 103.34 103.30 103.28 96.85 96.78 96.83 96.51 96.53 96.88 96.85
37.00 72.98 70.85 72.98 71.36 71.27 71.37 71.33 37.07 37.00 37.07 39.14 39.12 36.99 36.99
71.08 75.58 85.56 75.58 85.52 85.63 85.46 85.60 71.16 71.12 70.98 67.55 67.58 71.03 71.02
79.84 83.44 76.68 83.39 76.58 75.65 76.85 76.28 80.97 80.89 80.86 83.49 83.47 79.91 79.91
69.57 70.52 70.49 70.51 70.46 70.54 70.38 70.41 69.84 69.80 69.80 68.98 68.98 69.58 69.58
18.60 17.67 17.74 17.65 17.73 17.97 17.66 17.69 18.88 18.84 18.82 18.66 18.68 18.23 18.23
20
30
40
50
60
Methoxy
Acetyl 170.14
21.26
58.98
58.95 58.86 58.99 58.98
170.89 170.88 170.71
21.32 21.28 21.27
170.13 170.12
21.27 21.26
100
200
300
400
500
600
99.82 107.05 105.53 107.00 105.48 105.04 105.21 104.83 106.11 106.02 105.54 99.80 99.82 99.85 99.85
38.74 76.25 76.09 76.23 76.07 75.72 75.51 74.67 75.21 75.16 74.71 38.81 38.79 38.76 38.76
67.65 78.54 78.34 78.53 78.31 78.32 76.63 88.26 78.25 78.20 76.64 67.57 67.57 67.61 67.62
83.39 71.57 71.85 71.59 71.80 71.82 81.35 70.03 71.48 71.43 81.35 83.46 83.47 83.51 83.87
68.97 78.63 78.62 78.60 78.62 77.71 76.41 78.24 78.39 78.34 76.25 68.64 68.66 68.92 68.93
18.39 62.82 63.10 62.83 63.06 70.39 62.43 62.69 62.94 62.86 62.46 18.61 18.60 18.62 18.62
1000
2000
3000
4000
5000
6000
106.10
75.13
78.34
71.57
78.28
62.63
105.59 104.85 105.93
75.20 74.73 75.71
78.51 78.22 78.24
71.69 71.45 71.61
78.51 78.45 78.61
62.76 62.38 62.54
105.05 99.89 99.88 99.85 99.82
74.78 38.64 38.62 38.59 38.59
78.23 67.43 67.43 67.48 67.48
71.54 83.04 83.04 82.87 82.87
78.49 68.69 68.66 68.62 68.62
62.31 18.41 18.43 18.59 18.59
10000
20000
30000
40000
50000
60000
106.15 106.13 106.13 106.13
75.14 75.14 75.14 75.15
78.42 78.39 78.38 78.38
71.61 71.59 71.59 71.59
78.41 78.39 78.38 78.38
62.70 62.64 62.63 62.64
detector (Tosoh) or a Shodex OR-2 detector (Showa-Denko, Tokyo,
Japan), and a Rheodyne injection port. A Tosoh TSK-gel ODS-100Z
(10 mm i.d. ꢂ 250 mm, 5 mm) was used for preparative HPLC. The
following materials and reagents were used for the cell culture and
assay of cytotoxic activity: Spectra Classic microplate reader (Tecan,
Salzburg, Austria); 96-well flat-bottom plate (Iwaki Glass, Chiba,
Japan); JCRB 0085 HL-60 cells (Human Science Research Resources
Bank, Osaka, Japan); FBS and FCS (Bio-Whittaker, Walkersville, MD,
USA); the RPMI 1640 medium, DMEM, etoposide, cisplatin, and
MTT (Sigma, St. Louis, MO, USA); actinomycin D (Wako Pure
Chemical Industries, Osaka, Japan); and penicillin G sodium salt and
streptomycin sulfate (Meiji-Seika, Tokyo, Japan). All other chemicals
used were of biochemical reagent grade.
Plant material. The seeds of D. purpurea were obtained in 2008
from a wholesale firm in Richters, Ontario, Canada. A few seeds were
germinated and identified by one of the authors (Y.M.). A voucher
specimen was deposited in our laboratory.
Extraction and isolation. The seeds of D. purpurea (2.0 kg) were
extracted twice with hot MeOH (6 L). The MeOH extract was
concentrated under reduced pressure, and the viscous concentrate
(140 g) was applied to Diaion HP-20 CC and successively eluted with
20% MeOH, EtOH, and EtOAc. The EtOH-eluted portion (fraction B,
80 g) was separated by silica gel CC with a CHCl3–MeOH gradient
(9:1, 4:1, 2:1), and then with MeOH, to afford 11 fractions (fractions
B-I to XI). Fraction B-II was applied to an ODS silica gel column, and
was eluted with MeCN–H2O (2:3) and MeOH to afford 17 subfractions

1190
M. KURODA et al.
Table 3. Cytotoxic Activities of 1–15, Etoposide, and Cisplatin
against HL-60 Cells
passing through an Amberlite IRA-96SB column and then chromato-
graphed on silica gel eluted with EtOAc to give an aglycone
(digitoxigenin, 1.1 mg) and a sugar fraction (0.5 mg). The sugar
fraction was analyzed by HPLC under the following conditions:
Shodex Sugar SC1011 column (8:0 mm i.d. ꢂ 300 mm, 5 mm,
Showa-Denko); H2O solvent; flow rate, 1.0 mL/min flow rate; 80 ꢃC
column temperature; RI and OR detection. D-Glucose and D-digitoxose
were identified by comparing their retention times and optical rotation
values with those of authentic samples; tR (min) D-glucose (7.77,
positive optical rotation) and D-digitoxose (9.63, positive optical
rotation).
Compound
IC50 (mM)a
Compound
IC50 (mM)a
1
2
3
4
5
6
7
8
9
0:56 ꢄ 0:0028
0:060 ꢄ 0:0013
0:19 ꢄ 0:018
0:49 ꢄ 0:077
1:9 ꢄ 0:21
10
11
12
13
14
15
0:40 ꢄ 0:041
0:038 ꢄ 0:00040
0:034 ꢄ 0:0019
0:26 ꢄ 0:013
0:32 ꢄ 0:054
2:4 ꢄ 0:064
3:0 ꢄ 0:19
3:7 ꢄ 0:10
3:5 ꢄ 0:30
Etoposide
Cisplatin
0:46 ꢄ 0:0017
1:6 ꢄ 0:029
25
0:069 ꢄ 0:00040
Compound 2. Amorphous solid; ½ꢂꢀ
þ9.53ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 705.3447 ½M þ Naꢀþ (calcd. for C35H54NaO13
,
aData are presented as the mean values ꢄ SEM of three experiments
705.3462); UV ꢃmax (MeOH) nm (log "): 218 (3.14); IR ꢄmax
(film) cmꢁ1: 3427 (OH), 1740 (C=O); 1H-NMR (C5D5N) ꢁ: 6.14
(1H, br s, H-22), 5.32 (1H, br d, J ¼ 17:9 Hz, H-21a), 5.21 (1H, d,
J ¼ 7:8 Hz, H-100), 5.04 (1H, br d, J ¼ 17:9 Hz, H-21b), 4.72 (1H, d,
J ¼ 7:7 Hz, H-10), 4.35 (1H, br s, H-3), 2.79 (1H, dd, J ¼ 8:0, 5.8 Hz,
H-17), 1.64 (3H, d, J ¼ 6:4 Hz, H-60), 1.00 (3H, s, Me-18), 0.82 (3H, s,
Me-19); see Table 2 for the 13C-NMR (C5D5N) data.
performed in triplicate.
Table 4. Cytotoxic Activities of 2, 9, 11, 12, and Actinomycin D
against HepG2 Cells
Compound
IC50 (mM)a
Compound
IC50 (mM)a
2
9
0:38 ꢄ 0:09
0:79 ꢄ 0:17
0:71 ꢄ 0:16
12
3:20 ꢄ 0:50
25
Compound 3. Amorphous solid; ½ꢂꢀ
ꢁ2.65ꢃ (c 0.11, MeOH);
D
HR-ESI-TOF-MS m=z: 719.3550 ½M þ Naꢀþ (calcd. for C36H56NaO13
,
11
Actinomycin D
1:2 ꢄ 0:19
719.3619); UV ꢃmax (MeOH) nm (log "): 217 (3.16); IR ꢄmax
(film) cmꢁ1: 3460 (OH), 1736 (C=O); 1H-NMR (C5D5N) ꢁ: 6.14
(1H, br s, H-22), 5.32 (1H, dd, J ¼ 18:2, 1.5 Hz, H-21a), 5.04 (1H, dd,
J ¼ 18:2, 1.5 Hz, H-21b), 5.15 (1H, d, J ¼ 7:7 Hz, H-100), 4.69 (1H, d,
J ¼ 7:7 Hz, H-10), 4.34 (1H, br s, H-3), 3.67 (3H, s, methoxyl), 2.79
(1H, dd, J ¼ 8:4, 5.7 Hz, H-17), 1.59 (3H, d, J ¼ 6:4 Hz, H-60), 1.01
(3H, s, Me-18), 0.79 (3H, s, Me-19); see Table 2 for the 13C-NMR
(C5D5N) data.
aData are presented as the mean values ꢄ SEM of three experiments
performed in triplicate.
(fractions B-II-A to Q). Fraction B-II-E was separated by silica gel CC
with CHCl3–MeOH–H2O (60:10:1), ODS silica gel CC with MeCN–
H2O (1:2), and preparative HPLC using MeCN–H2O (3:2, 1:2) to yield
3 (88.0 mg) and 10 (9.2 mg). Fraction B-II-H was purified by silica gel
CC with CHCl3–MeOH (9:1) and ODS silica gel CC with MeCN–H2O
(1:2) to afford 13 (14.4 mg). Fraction B-II-J was separated by
preparative HPLC, using MeCN–H2O (2:3) to give 9 (35.6 mg) and
15 (92.6 mg). Fraction B-II-L was subjected to chromatography on
silica gel eluted with CHCl3–MeOH–H2O (40:10:1) to afford 1
(17.0 mg) as a pure compound, and 12 with several impurities.
Compound 12 (6.9 mg) was purified by silica gel CC eluted with
CHCl3–MeOH (19:1). Compound 14 (23.5 mg) was isolated from
fraction B-II-N via preparative HPLC, using MeCN–H2O (2:3).
Fraction B-IV was subjected to ODS silica gel CC eluted with
MeCN–H2O (1:2) to give 2 (102 mg) and 5 (230 mg) as pure
compounds, and 11 with some impurities. Compound 11 (6.1 mg)
was purified by ODS silica gel CC eluted with MeCN–H2O (2:3) and
by preparative HPLC using MeCN–H2O (2:3). Fraction B-VI was
subjected to ODS silica gel CC eluted with MeCN–H2O (1:2) and then
with MeOH which produced eight subfractions (fractions B-VI-A to
H). Fraction B-VI-A was then applied to an ODS silica gel column and
eluted with MeCN–H2O (1:5) and then with MeCN to afford 10
subfractions (fractions B-VI-A-1 to 10). Fraction B-VI-A-9 was
purified by silica gel CC with CHCl3–MeOH–H2O (20:10:1) and then
with MeOH, and nine subfractions (fractions B-VI-A-9-i to ix) were
collected. Fraction B-VI-A-9-ii was purified by preparative HPLC,
using MeCN–H2O (2:5), to give 4 (5.7 mg). Fraction B-VI-A-9-iv was
subjected to ODS silica gel CC eluted with MeCN–H2O (1:3) and to
Sephadex LH-20 CC eluted with MeOH to afford 7 (23.2 mg) and 8
(4.0 mg). Fraction B-VI-A-9-vi was purified by ODS silica gel CC
eluted with MeCN–H2O (1:3, 3:2) to give 6 (14.0 mg).
25
Compound 4. Amorphous solid; ½ꢂꢀ
ꢁ4.18ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 721.3371 ½M þ Naꢀþ (calcd. for C35H54NaO14
,
721.3411); UV ꢃmax (MeOH) nm (log "): 218 (3.18); IR ꢄmax
(film) cmꢁ1: 3426 (OH), 1738 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25
(1H, br s, H-22), 5.68 (1H, br d, J ¼ 18:2 Hz, H-21a), 5.57 (1H, br d,
J ¼ 18:2 Hz, H-21b), 5.20 (1H, d, J ¼ 7:8 Hz, H-100), 4.99 (1H, t-like,
J ¼ 8:0 Hz, H-16), 4.71 (1H, d, J ¼ 7:6 Hz, H-10), 4.32 (1H, br s, H-3),
3.28 (1H, d, J ¼ 7:9 Hz, H-17), 2.72 (1H, dd, J ¼ 14:5, 8.4 Hz, H-15a),
1.63 (3H, d, J ¼ 6:4 Hz, H-60), 1.11 (3H, s, Me-18), 0.83 (3H, s,
Me-19); see Table 2 for the 13C-NMR (C5D5N) data.
25
Compound 5. Amorphous solid; ½ꢂꢀ
ꢁ12.6ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 735.3875 ½M þ Naꢀþ (calcd, for C36H56NaO14
,
735.3932); UV ꢃmax (MeOH) nm (log "): 218 (3.22); IR ꢄmax
(film) cmꢁ1: 3409 (OH), 1735 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25
(1H, br s, H-22), 5.71 (1H, br d, J ¼ 18:2 Hz, H-21a), 5.57 (1H, br d,
J ¼ 18:2 Hz, H-21b), 5.15 (1H, d, J ¼ 7:7 Hz, H-100), 4.98 (1H, t-like,
J ¼ 7:7 Hz, H-16), 4.68 (1H, d, J ¼ 7:6 Hz, H-10), 4.32 (1H, br s, H-3),
3.66 (3H, s, methoxyl), 3.28 (1H, d, J ¼ 7:9 Hz, H-17), 2.73 (1H, dd,
J ¼ 14:6, 8.4 Hz, H-15a), 1.59 (3H, d, J ¼ 6:3 Hz, H-60), 1.11 (3H, s,
Me-18), 0.80 (3H, s, Me-19); see Table 2 for the 13C-NMR (C5D5N)
data.
25
Compound 6. Amorphous solid; ½ꢂꢀ
þ1.15ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 897.4096 ½M þ Naꢀþ (calcd. for C42H66NaO19
,
897.4096); UV ꢃmax (MeOH) nm (log "): 217 (3.17); IR ꢄmax
(film) cmꢁ1: 3421 (OH), 1735 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25
(1H, br s, H-22), 5.70 (1H, dd, J ¼ 18:3, 1.6 Hz, H-21a), 5.57 (1H, dd,
J ¼ 18:3, 1.6 Hz, H-21b), 5.20 (1H, d, J ¼ 7:8 Hz, H-1000), 5.14 (1H, d,
J ¼ 7:8 Hz, H-100), 4.99 (1H, t-like, J ¼ 8:0 Hz, H-16), 4.63 (1H, d,
J ¼ 7:7 Hz, H-10), 3.64 (3H, s, methoxy), 3.28 (1H, d, J ¼ 7:9 Hz,
H-17), 2.73 (1H, dd, J ¼ 14:6, 8.4 Hz, H-15a), 1.70 (3H, d, J ¼ 6:3 Hz,
H-60), 1.11 (3H, s, Me-18), 0.80 (3H, s, Me-19); see Table 2 for the
13C-NMR (C5D5N) data.
26
Compound 1. Amorphous solid; ½ꢂꢀ
þ14.46ꢃ (c 0.145, MeOH);
D
HR-ESI-TOF-MS m=z: 861.4305 ½M þ Naꢀþ (calcd. for C43H66NaO16
,
861.4249); UV ꢃmax (MeOH) nm (log "): 217 (3.67); IR ꢄmax
(film) cmꢁ1: 3378 (OH), 1740 (C=O); 1H-NMR (C5D5N) ꢁ: 6.14
(1H, br s, H-22), 5.32 (1H, dd, J ¼ 18:1, 1.5 Hz, H-21a), 5.04 (1H, dd,
J ¼ 18:1, 1.5 Hz, H-21b), 4.25 (1H, br s, W1=2 ¼ 8:3 Hz, H-3), 2.79
(1H, dd, J ¼ 8:8, 5.4 Hz, H-17), 1.01 (3H, s, Me-18), 0.88 (3H, s,
Me-19); see Table 1 for signals of the sugar moieties; see Table 2 for
the 13C NMR (C5D5N) data.
25
Compound 7. Amorphous solid; ½ꢂꢀ
þ0.13ꢃ (c 0.10, MeOH);
D
Acid hydrolysis of 1. A solution of 1 (2.9 mg) in 0.025 M HCl
(dioxane-H2O, 1:1, 2.0 mL) was heated at 95 ꢃC for 60 min in an Ar
atmosphere. After cooling, the reaction mixture was neutralized by
HR-ESI-TOF-MS m=z: 897.4006 ½M þ Naꢀþ (calcd. for C42H66NaO19
,
897.4096); UV ꢃmax (MeOH) nm (log "): 218 (3.12); IR ꢄmax
(film) cmꢁ1: 3423 (OH), 1732 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25

Cytotoxic Cardenolides from Digitalis purpurea Seeds
1191
(1H, br s, H-22), 5.69 (1H, br d, J ¼ 18:2, H-21a), 5.56 (1H, br d,
J ¼ 18:2 Hz, H-21b), 5.17 (1H, d, J ¼ 7:8 Hz, H-1000), 5.08 (1H, d,
J ¼ 7:8 Hz, H-100), 4.98 (1H, t-like, J ¼ 7:8 Hz, H-16), 4.67 (1H, d,
J ¼ 7:6 Hz, H-10), 3.64 (3H, s, methoxy), 3.28 (1H, d, J ¼ 7:9 Hz,
H-17), 2.72 (1H, dd, J ¼ 14:6, 8.4 Hz, H-15a), 1.53 (3H, d, J ¼ 6:3 Hz,
H-60), 1.11 (3H, s, Me-18), 0.80 (3H, s, Me-19); see Table 2 for the
13C-NMR (C5D5N) data.
(1H, br s, H-22), 5.39 (1H, br d, J ¼ 10:0 Hz, H-100), 5.37 (1H, br d,
J ¼ 9:5 Hz, H-1000), 5.36 (1H, br d, J ¼ 9:5 Hz, H-10), 5.32 (1H, br d,
J ¼ 18:1 Hz, H-21a), 5.04 (1H, br d, J ¼ 18:1 Hz, H-21b), 4.99 (1H, d,
J ¼ 7:9 Hz, H-10000), 2.78 (1H, dd, J ¼ 8:3, 5.7 Hz, H-17), 1.57 (3H, d,
J ¼ 6:2 Hz, H-6000), 1.41 (3H, d, J ¼ 6:2 Hz, H-60), 1.28 (3H, d,
J ¼ 6:3 Hz, H-600), 1.00 (3H, s, Me-18), 0.87 (3H, s, Me-19); see
Table 2 for the 13C-NMR (C5D5N) data.
26
25
Compound 8. Amorphous solid; ½ꢂꢀ
ꢁ3.48ꢃ (c 0.10, MeOH);
Compound 13. Amorphous solid; ½ꢂꢀ
ꢁ5.4ꢃ (c 0.10, MeOH);
D
D
HR-ESI-TOF-MS m=z: 897.4096 ½M þ Naꢀþ (calcd. for C42H66NaO19
,
HR-ESI-TOF-MS m=z: 965.4713 ½M þ Naꢀþ (calcd. for C47H74NaO19
,
897.4096); UV ꢃmax (MeOH) nm (log "): 217 (4.30); IR ꢄmax
(film) cmꢁ1: 3395 (OH), 1731 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25
(1H, br s, H-22), 5.70 (1H, br d, J ¼ 18:2 Hz, H-21a), 5.57 (1H, br d,
J ¼ 18:2, H-21b), 4.98 (1H, dd, J ¼ 8:4, 7.9 Hz, H-16), 4.31 (1H, br s,
W1=2 ¼ 8:1 Hz, H-3), 3.28 (1H, d, J ¼ 7:9 Hz, H-17), 2.73 (1H, dd,
J ¼ 14:6, 8.4 Hz, H-15a), 1.11 (3H, s, Me-18), 0.81 (3H, s, Me-19); see
Table 1 for the signals for the sugar moieties; see Table 2 for the
13C-NMR (C5D5N) data.
965.4722); UV ꢃmax (MeOH) nm (log "): 217 (3.11); IR ꢄmax
(film) cmꢁ1: 3411 (OH), 1737 (C=O); 1H-NMR (C5D5N) ꢁ: 6.24
(1H, br s, H-22), 5.70 (1H, br d, J ¼ 18:0 Hz, H-21a), 5.57 (1H, br d,
J ¼ 18:2 Hz, H-21b), 5.39 (1H, br d, J ¼ 9:8 Hz, H-100), 5.37 (1H, br d,
J ¼ 9:7 Hz, H-1000), 5.35 (1H, br d, J ¼ 9:2 Hz, H-10), 4.99 (1H, d,
J ¼ 7:6 Hz, H-10000), 4.97 (1H, t-like, J ¼ 7:3 Hz, H-16), 3.27 (1H, d,
J ¼ 7:8 Hz, H-17), 2.72 (1H, dd, J ¼ 14:5, 8.5 Hz, H-15a), 1.57 (3H, d,
J ¼ 6:2 Hz, H-6000), 1.41 (3H, d, J ¼ 6:2 Hz, H-60), 1.29 (3H, d,
J ¼ 6:2 Hz, H-600), 1.11 (3H, s, Me-18), 0.89 (3H, s, Me-19); see
Table 2 for the 13C-NMR (C5D5N) data.
Enzymatic hydrolysis of 8. Compound 8 (2.5 mg) was treated with
ꢀ-D-glucosidase (Sigma-Aldrich, 7.0 mg) in a 0.2 M AcOH–AcONa
buffer (pH 5.0, 2.0 mL) at room temperature for 24 h. The reaction
mixture was subjected to preparative HPLC using MeCN–H2O (1:3) to
yield 5 (1.8 mg) and a sugar fraction (0.4 mg). An HPLC analysis of the
sugar fraction under the same conditions as those used for 1 showed the
presence of D-glucose.
25
Compound 14. Amorphous solid; ½ꢂꢀ
þ22.2ꢃ (c 0.11, MeOH);
HR-ESI-TOF-MS m=z: 991.4893 ½M þ NaꢀDþ (calcd. for C49H76NaO19
,
991.4879); UV ꢃmax (MeOH) nm (log "): 214 (3.12); IR ꢄmax
(film) cmꢁ1: 3381 (OH), 1738 (C=O); 1H-NMR (C5D5N) ꢁ: 6.14
(1H, br s, H-22), 5.82 (1H, br d, J ¼ 3:0 Hz, H-30), 5.33 (1H, br d,
J ¼ 10:8 Hz, H-100), 5.32 (1H, br d, J ¼ 17:6 Hz, H-21a), 5.31 (1H, br
d, J ¼ 8:1 Hz, H-1000), 5.15 (1H, br d, J ¼ 9:3 Hz, H-10), 5.08 (1H, br d,
J ¼ 17:6 Hz, H-21b), 4.99 (1H, d, J ¼ 7:7 Hz, H-10000), 4.26 (1H, br s,
H-3), 2.79 (1H, dd, J ¼ 8:3, 5.7 Hz, H-17), 2.11 (3H, s, acetyl), 1.57
(3H, d, J ¼ 6:2 Hz, H-6000), 1.42 (3H, d, J ¼ 6:2 Hz, H-60), 1.33 (3H, d,
J ¼ 6:1 Hz, H-600), 1.01 (3H, s, Me-18), 0.88 (3H, s, Me-19); see
Table 2 for the 13C-NMR (C5D5N) data.
25
Compound 9. Amorphous solid; ½ꢂꢀ
þ29.0ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 731.3578 ½M þ Naꢀþ (calcd. for C37H56NaO13
,
731.3619); UV ꢃmax (MeOH) nm (log "): 220 (3.21); IR ꢄmax
(film) cmꢁ1: 3301 (OH), 1750 (C=O); 1H-NMR (C5D5N) ꢁ: 6.14
(1H, br s, H-22), 5.98 (1H, br d, J ¼ 3:1 Hz, H-30), 5.32 (1H, br d,
J ¼ 18:0 Hz, H-21a), 5.18 (1H, br d, J ¼ 9:4 Hz, H-10), 5.04 (1H, br d,
J ¼ 18:0 Hz, H-21b), 4.96 (1H, d, J ¼ 7:8 Hz, H-100), 2.79 (1H, dd,
J ¼ 8:3, 5.8 Hz, H-17), 2.08 (3H, s, acetyl), 1.70 (3H, d, J ¼ 6:2 Hz,
H-60), 1.01 (3H, s, Me-18), 0.89 (3H, s, Me-19); see the Table 2 for the
13C-NMR (C5D5N) data.
25
Compound 15. Amorphous solid; ½ꢂꢀ
þ33.6ꢃ (c 0.11, MeOH);
D
HR-ESI-TOF-MS m=z: 1007.4802 ½M þ Naꢀþ (calcd. for
C
49H76NaO20, 1007.4828); UV ꢃmax (MeOH) nm (log "): 218 (3.14);
IR ꢄmax (film) cmꢁ1: 3399 (OH), 1733 (C=O); 1H-NMR (C5D5N) ꢁ:
6.25 (1H, br s, H-22), 5.82 (1H, br d, J ¼ 2:8 Hz, H-30), 5.70 (1H, br d,
J ¼ 18:2 Hz, H-21a), 5.57 (1H, br d, J ¼ 18:2 Hz, H-21b), 5.33 (1H, br
d, J ¼ 11:3 Hz, H-100), 5.31 (1H, br d, J ¼ 10:5 Hz, H-1000), 5.15 (1H, br
d, J ¼ 9:4 Hz, H-10), 4.99 (1H, d, J ¼ 7:8 Hz, H-10000), 4.98 (1H, t-like,
J ¼ 8:5 Hz, H-16), 3.28 (1H, d, J ¼ 7:9 Hz, H-17), 2.72 (1H, dd,
J ¼ 14:4, 8.2 Hz, H-15a), 2.11 (3H, s, acetyl), 1.57 (3H, d, J ¼ 6:2 Hz,
H-6000), 1.41 (3H, d, J ¼ 6:2 Hz, H-60), 1.33 (3H, d, J ¼ 6:1 Hz, H-600),
1.12 (3H, s, Me-18), 0.90 (3H, s, Me-19); see Table 2 for the 13C-NMR
(C5D5N) data.
25
Compound 10. Amorphous solid; ½ꢂꢀ
þ39.0ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 747.3567 ½M þ Naꢀþ (calcd. for C37H56NaO14
,
747.3568); UV ꢃmax (MeOH) nm (log "): 219 (3.04); IR ꢄmax
(film) cmꢁ1: 3275 (OH), 1756 (C=O); 1H-NMR (C5D5N) ꢁ: 6.24
(1H, br s, H-22), 5.96 (1H, br d, J ¼ 2:8 Hz, H-30), 5.70 (1H, br d,
J ¼ 18:2 Hz, H-21a), 5.57 (1H, br d, J ¼ 18:2 Hz, H-21b), 5.16 (1H, br
d, J ¼ 9:4 Hz, H-10), 4.98 (1H, t-like, J ¼ 8:3 Hz, H-16), 4.95 (1H, d,
J ¼ 7:7 Hz, H-100), 4.23 (1H, br s, H-3), 3.27 (1H, d, J ¼ 7:9 Hz,
H-17), 2.71 (1H, dd, J ¼ 14:7, 8.5 Hz, H-15a), 2.07 (3H, s, acetyl),
1.68 (3H, d, J ¼ 6:2 Hz, H-60), 1.11 (3H, s, Me-18), 0.89 (3H, s,
Me-19); see Table 2 for the 13C-NMR (C5D5N) data.
HL-60 cell culture assay. HL-60 cells were maintained in an RPMI
1640 medium containing 10% FBS supplemented with 2 mM of
L-glutamine, 100 units/mL of penicillin G sodium salt, and 100 mg/mL
of streptomycin sulfate. The leukemia cells were washed and
resuspended in this medium to 4 ꢂ 104 cells/mL, and 196 mL of this
cell suspension was placed in each well of a 96-well flat-bottom plate.
The cells were incubated in 5% CO2/air for 24 h at 37 ꢃC. After this
incubation, 4 mL of an EtOH–H2O (1:1) solution containing the sample
was added to give a final concentrations of 0.1–10 mg/mL, and 4 mL of
EtOH–H2O (1:1) was added into the control wells. The cells were
incubated for a further 72 h in the presence of each agent, and then the
cell growth was evaluated by a modified MTT reduction assay.19)
Briefly, after terminating the cell culture, 10 mL of 5 mg/mL of MTT in
PBS was added to each well, and the plate was reincubated in 5%
CO2/air for 4 h at 37 ꢃC. The plate was then centrifuged at 1500 g for
5 min to precipitate the cells and MTT formazan. An aliquot of 150 mL
of the resulting supernatant was removed from each well, and 175 mL
of DMSO was added to dissolve the MTT formazan crystals. The plate
was mixed on a microshaker for 10 min, and then read on a microplate
reader at 550 nm.
26
Compound 11. Amorphous solid; ½ꢂꢀ
þ17.08ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 893.4169 ½M þ Naꢀþ (calcd. for C43H66NaO18
,
893.4147); UV ꢃmax (MeOH) nm (log "): 217 (4.30); IR ꢄmax
(film) cmꢁ1: 3395 (OH), 1731 (C=O); 1H-NMR (C5D5N) ꢁ: 6.25
(1H, br s, H-22), 5.32 (1H, br d, J ¼ 18:2 Hz, H-21a), 5.04 (1H, br d,
J ¼ 18:2, H-21b), 4.26 (1H, br s, W1=2 ¼ 8:1 Hz, H-3), 2.79 (1H, dd,
J ¼ 8:4, 5.7 Hz, H-17), 1.01 (3H, s, Me-18), 0.89 (3H, s, Me-19); see
Table 1 for the sugar moieties; see Table 2 for the 13C-NMR (C5D5N)
data.
Enzymatic hydrolysis of 11. Compound 11 (3.0 mg) was treated with
ꢀ-D-glucosidase (Sigma-Aldrich, 7.0 mg) in a 0.2 M AcOH–AcONa
buffer (pH 5.0, 2.0 mL) at room temperature for 24 h. The reaction
mixture was applied to a silica gel column eluted with CHCl3–MeOH–
H2O (40:10:1) and then by MeOH to yield 9 (1.7 mg) and a sugar
fraction (0.2 mg). The HPLC analysis of the sugar fraction under the
same conditions as those used for 1 showed the presence of D-glucose.
26
Compound 12. Amorphous solid; ½ꢂꢀ
ꢁ6.8ꢃ (c 0.10, MeOH);
D
HR-ESI-TOF-MS m=z: 949.4727 ½M þ Naꢀþ (calcd. for C47H74NaO18
,
HepG2 cell culture assay. HepG2 cells were cultured in DMEM
containing 10% FCS, 50 units/mL of penicillin G sodium salt, and
50 mg/mL of streptomycin sulfate in a humidified atmosphere of 8.5%
949.4773); UV ꢃmax (MeOH) nm (log "): 216 (3.17); IR ꢄmax
(film) cmꢁ1: 3410 (OH), 1739 (C=O); 1H-NMR (C5D5N) ꢁ: 6.13

1192
M. KURODA et al.
CO2/air at 37 ꢃC. The cells were washed and resuspended in the
medium at a concentration of 1:7 ꢂ 104 cells/mL, and 6 mL of the cell
suspension was placed in a 60-mm dish. The cells were incubated in
8.5% CO2/air for 24 h at 37 ꢃC. After this incubation, 6 mL of DMSO
containing the sample was added to give final concentrations of 0.08–
5.0 mM, and 6 mL of DMSO was added to the control dishes. The cells
were incubated for 48 h in the presence of each agent, and then the cell
number was counted.
8) Hoji K, Miki T, Sakashita A, Okano A, and Miyataka K, Chem.
Pharm. Bull., 9, 276–288 (1961).
9) Miyatake K, Okano A, Hoji K, Miki T, and Sakashita A, Chem.
Pharm. Bull., 9, 375–378 (1961).
10) Okano A, Hoji K, Miki T, and Sakashita A, Chem. Pharm. Bull.,
7, 226–232 (1959).
11) Okano A, Hoji K, Miki T, and Sakashita A, Chem. Pharm. Bull.,
7, 222–225 (1959).
12) Okano A, Hoji K, Miki T, and Sakashita A, Chem. Pharm. Bull.,
7, 212–221 (1959).
References
13) Okano A, Pharm. Bull., 6, 173–177 (1958).
14) Okano A, Pharm. Bull., 5, 272–276 (1957).
15) Miyatake K, Okano A, Hoji K, and Miki T, Pharm. Bull., 5,
163–167 (1957).
1) Higano T, Kuroda M, Sakagami H, and Mimaki Y, Chem.
Pharm. Bull., 55, 337–339 (2007).
2) Kubo S, Kuroda M, Matsuo Y, Masatani D, Sakagami H, and
Mimaki Y, Chem. Pharm. Bull., 60, 1275–1282 (2012).
3) Hoji K, Chem. Pharm. Bull., 9, 576–578 (1961).
4) Hoji K, Chem. Pharm. Bull., 9, 571–575 (1961).
5) Miyatake K, Okano A, Hoji K, Miki T, and Sakashita A, Chem.
Pharm. Bull., 9, 519–523 (1961).
16) Miyatake K, Okano A, Hoji K, and Miki T, Pharm. Bull., 5,
157–163 (1957).
17) Sato D, Ishii H, and Nishimura Y, Yakugaku Zasshi, 74, 560
(1954).
18) Imre Z and Yurdun T, Planta Med., 54, 529–531 (1988).
19) Sargent JM and Taylor CG, Br. J. Cancer, 60, 206–210 (1989).
6) Hoji K, Chem. Pharm. Bull., 9, 291–295 (1961).
7) Hoji K, Chem. Pharm. Bull., 9, 289–291 (1961).