Murine metabolism and absorption of lancemaside A, an active compound in the roots of Codonopsis lanceolata

Article abstract of DOI:10.1007/s11418-010-0415-8

Lancemaside A, a triterpenoid saponin isolated from the roots of Codonopsis lanceolata, has been reported to ameliorate the reduction of blood testosterone levels induced by immobilization stress in mice. In the present study, we investigated the metabolism and absorption of lancemaside A in mice. After oral administration of lancemaside A at 100 mg/kg body weight, the unmetabolized compound appeared rapidly in plasma (t_max = 0.5 h). Lancemaside A has a low bioavailability (1.1%) because of its metabolism by intestinal bacteria and its poor absorption in the gastrointestinal tract. Furthermore, we identified four metabolites from the cecum of mice after oral administration of lancemaside A: codonolaside II, echinocystic acid, echinocystic acid 28-O-β-D-xylopyranosyl-(1 → 4)- α-L-rhamnopyranosyl-(1 → 2)-α-L-arabinopyranosyl ester, and echinocystic acid 28-O-α-L- rhamnopyranosyl-(1 → 2)- α-L-arabinopyranosyl ester. Among these metabolites, codonolaside II and echinocystic acid were detected in plasma, and their t_max values were 4 and 8 h, respectively. These findings should be helpful for understanding the mechanism of the biological effect of lancemaside A. The Japanese Society of Pharmacognosy and Springer 2010.

Full text of DOI:10.1007/s11418-010-0415-8

J Nat Med (2010) 64:321–329
DOI 10.1007/s11418-010-0415-8
ORIGINAL PAPER
Murine metabolism and absorption of lancemaside A,
an active compound in the roots of Codonopsis lanceolata
Noriko Komoto
Takeshi Nishihama
Minoru Hayama
•
Makoto Ichikawa
•
•
Sanae Ohta
Mitsuyasu Ushijima Yukihiro Kodera
Setsuko Sekita Masanori Kuroyanagi
•
Daisuke Nakano
•
•
•
•
Osamu Shirota
•
•
Received: 6 November 2009 / Accepted: 1 March 2010 / Published online: 10 April 2010
Ó The Japanese Society of Pharmacognosy and Springer 2010
Abstract Lancemaside A, a triterpenoid saponin isolated
from the roots of Codonopsis lanceolata, has been reported
to ameliorate the reduction of blood testosterone levels
induced by immobilization stress in mice. In the present
study, we investigated the metabolism and absorption of
lancemaside A in mice. After oral administration of lanc-
emaside A at 100 mg/kg body weight, the unmetabolized
compound appeared rapidly in plasma (tmax = 0.5 h).
Lancemaside A has a low bioavailability (1.1%) because of
its metabolism by intestinal bacteria and its poor absorption
in the gastrointestinal tract. Furthermore, we identified four
metabolites from the cecum of mice after oral adminis-
tration of lancemaside A: codonolaside II, echinocystic
acid, echinocystic acid 28-O-b-D-xylopyranosyl-(1 ? 4)-
a-L-rhamnopyranosyl-(1 ? 2)-a-L-arabinopyranosyl ester,
and echinocystic acid 28-O-a-L-rhamnopyranosyl-(1 ? 2)-
a-L-arabinopyranosyl ester. Among these metabolites,
codonolaside II and echinocystic acid were detected in
plasma, and their tmax values were 4 and 8 h, respectively.
These findings should be helpful for understanding the
mechanism of the biological effect of lancemaside A.
Keywords Lancemaside A Á Metabolism Á
Codonopsis lanceolata Á Codonolaside II Á
Echinocystic acid Á Bioavailability
Introduction
The roots of Codonopsis lanceolata Trautv. (Campanula-
ceae) have been used as a traditional medicine for their
antitussive, expectorant, and antidotal effects. Recently, the
root extracts have been found to inhibit the reduction of
blood testosterone levels induced by immobilization stress
in mice [1]. Moreover, a supplement containing the root
extracts has been reported to ameliorate partial androgen
deficiency of the aging male (PADAM)-like symptoms as
well as the total testosterone levels in middle-aged men
with PADAM [2]. It has been previously reported that the
components in the roots of C. lanceolata include saponins
[
3–7] and phenylpropanoid glycosides [8]. Among them,
N. Komoto (&) Á M. Ichikawa Á S. Ohta Á D. Nakano Á
T. Nishihama Á M. Ushijima Á Y. Kodera Á M. Hayama
Healthcare Research Institute, Wakunaga Pharmaceutical
Company, Ltd., 1624 Shimokotachi, Kodacho, Akitakata,
Hiroshima 739-1195, Japan
lancemaside A (1, Fig. 1), a 3,28-bidesmosidic triterpenoid
saponin, has been reported to ameliorate the reduction of
blood testosterone levels induced by immobilization stress
in mice [8]. Therefore, 1 is expected to be useful for the
prevention of and as a complementary therapy for PADAM
symptoms associated with the lowering of blood testos-
terone levels.
e-mail: komoto_n@wakunaga.co.jp
O. Shirota Á S. Sekita
Laboratory of Pharmacognosy and Natural Product Chemistry,
Faculty of Pharmaceutical Sciences at Kagawa Campus,
Tokushima Bunri University, 1314-1 Shido, Sanuki,
Kagawa 769-2193, Japan
It is known that the bioavailability of saponins is low
because of their poor oral absorption and gastrointestinal
metabolism [9, 10]. Some of saponins are transformed to
their deglycosylated compounds by intestinal bacteria, and
the resulting compounds exhibit biological activities
[11, 12]. This suggests that 1 may not itself be responsible
M. Kuroyanagi
Department of Life Sciences, Faculty of Life and Environmental
Sciences, Prefectural University of Hiroshima, 562 Nanatsuka,
Shobara, Hiroshima 727-0023, Japan
123

3
22
J Nat Med (2010) 64:321–329
Milli-Q Labo system (Millipore, Bedford, MA, USA) was
used for mobile phase preparation.
Preparation of standard compounds 2–5
Codonolaside II (2)
Compound 1 (50 mg) was dissolved in 40 ml of 0.5 mol/l
sodium acetate buffer (pH 5.0), and b-glucuronidase
(21 mg) was then added to the mixture, followed by
incubation for 6 h at 37°C. After incubation, the mixture
was applied to a Diaion HP-20 (Mitsubishi Chemical Co.,
Tokyo, Japan) column, and the column was washed with
10% methanol, followed by elution with methanol. The
methanol eluate was concentrated in vacuo. The residue
was purified by silica gel column chromatography with
chloroform–methanol–water (40:10:1). Finally, 25 mg of
codonolaside II (2) was isolated and identified by com-
parison of the NMR spectroscopic data with those reported
Fig. 1 Chemical structures of lancemaside A (1) and its metabolites
2–5). b-GlcA, b-D-glucuronopyranosyl; a-Ara, a-L-arabinopyranosyl;
a-Rha, a-L-rhamnopyranosyl; b-Xyl, b-D-xylopyranosyl
(
for its biological activities. However, no data are available
on the absorption and metabolism of 1.
In the present study, we investigated the pharmacokinetics
of 1 in mice. The metabolism of 1 in the gastrointestinal tract
was evaluated by using liquid chromatography–mass spec-
trometry (LC–MS) with an electrospray ionization (ESI)
source. In addition, 1 and its metabolites were identified by
1
3
in the literature [5]. The C-NMR data of 2 are shown in
Table 1.
Echinocystic acid (3)
n
using LC–MS with multiple-stage mass spectrometry (MS ),
which is extensively applied to the online analysis of saponins
Compound 1 (80 mg) was dissolved in 80 ml of acetate
buffer (pH 5.0), and b-glucuronidase (40 mg) was then
added to the mixture, followed by incubation for 6 h at
37°C. After incubation, the mixture was applied to a Diaion
HP-20 column, and the column was washed with 30%
methanol, followed by elution with methanol. The metha-
nol eluate was concentrated in vacuo. The residue was
dissolved in 3 ml of methanol, and 12 ml of 3% potassium
hydroxide was then added to the mixture, followed by
heating at 80°C for 2 h. The mixture was added to 50%
methanol (40 ml) and applied to a Diaion HP-20 column,
and the column was washed with 40% methanol, followed
by elution with methanol. The methanol eluate was con-
centrated in vacuo. The residue was purified by silica gel
column chromatography with chloroform–methanol (20:1).
Finally, 9 mg of echinocystic acid (3) was isolated and
identified by comparison of the NMR spectroscopic data
[13–15]. As shown in Fig. 1, four metabolites, namely,
codonolaside II (2), echinocystic acid (3), echinocystic
acid 28-O-b-D-xylopyranosyl-(1 ? 4)-a-L-rhamnopyrano-
syl-(1 ? 2)-a-L-arabinopyranosyl ester (4), and echinocystic
acid 28-O-a-L-rhamnopyranosyl-(1 ? 2)-a-L-arabinopyr-
anosyl ester (5), were identified from the cecum of mice after
oral administration of 1. We also established an LC–MS
method to determine the concentration of 1 and its gastroin-
testinal metabolites, 2 and 3, in mouse plasma, and then
investigated the pharmacokinetics of 1 in mice.
Materials and methods
Chemicals and reagents
1
3
Lancemaside A (1), lancemaside E (used as the internal
standard, IS), and foetidissimoside A were isolated from
the roots of C. lanceolata [6, 16]. The IS solution was
prepared at a concentration of 0.05 lg/ml in ethanol. Saline
with those reported in the literature [17]. The C-NMR
data of 3 are shown in Table 1.
Compounds 4 and 5
(
0.9% NaCl injectable solution) was obtained from Fuso
Pharmaceutical Co., Ltd. (Osaka, Japan). Acetonitrile for
LC–MS was purchased from Wako Pure Chemical Indus-
tries, Ltd. (Osaka, Japan). b-Glucuronidase (type H-5,
Foetidissimoside A (47 mg) was dissolved in 40 ml of
acetate buffer (pH 5.0), and b-glucuronidase (21 mg) was
then added to the mixture, followed by incubation for 6 h
at 37°C. After the incubation, the mixture was applied to a
Diaion HP-20 column, and the column was washed with
30% methanol, followed by elution with methanol. The
methanol eluate was concentrated in vacuo. The residue
6
54200 U/g) was obtained from Sigma (St. Louis, MO,
USA). All other chemicals of reagent grade were purchased
from Tokyo Kasei Kogyo Co., Ltd. (Tokyo, Japan) and
Wako Pure Chemical Industries, Ltd. Water purified with a
1
23

J Nat Med (2010) 64:321–329
323
1
3
Table 1 C-NMR chemical shifts of lancemaside A (1) and its metabolites (2–5) in pyridine-d
5
Carbon
1
2
3
4
5
Carbon
1
2
4
5
1
2
3
4
5
6
7
8
9
38.7
26.6
89.0
39.5
55.8
18.3
33.4
39.9
47.1
36.9
23.7
39.1
28.1
78.2
39.4
55.9
18.8
33.5
40.0
47.3
37.4
23.8
39.0
28.2
78.1
39.4
55.9
18.8
33.6
39.9
47.3
37.4
23.8
39.1
28.1
78.1
39.4
55.9
18.9
33.5
40.0
47.3
37.4
23.8
39.1
28.1
78.1
39.4
55.9
18.8
33.5
40.0
47.3
37.4
23.8
Glucuronic acid
1
107.2
2
75.5
78.2
73.4
77.9
172.8
3
4
5
6
Arabinose
1
93.4
75.1
69.4
65.8
62.7
93.4
93.7
93.5
75.2
70.1
66.1
63.1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
75.3
69.5
65.8
62.7
75.0
70.3
66.3
63.4
3
122.8
144.3
42.0
36.1
74.0
49.5
41.2
47.0
30.9
35.9
32.1
28.1
16.9
15.6
17.5
27.1
175.9
33.2
24.7
122.8
144.4
42.1
36.1
74.0
49.5
41.2
47.1
30.9
36.0
32.1
28.8
16.6
15.7
17.6
27.1
175.9
33.2
24.8
122.4
145.1
42.1
36.2
74.8
48.9
41.5
47.3
31.0
36.1
32.9
28.7
16.6
15.7
17.5
27.2
180.0
33.3
24.7
122.8
144.4
42.1
36.2
74.0
49.5
41.3
47.2
30.9
36.0
32.1
28.8
16.6
15.8
17.6
27.1
175.9
33.2
24.7
122.8
144.5
42.1
36.2
73.9
49.6
41.3
47.2
30.9
36.0
32.1
28.7
16.6
15.8
17.6
27.1
175.9
33.2
24.7
4
5
Rhamnose
a
1
100.9
71.8
72.7
83.3
68.4
18.3
101.0
71.8
72.7
83.4
68.5
18.4
101.1
71.9
72.7
83.6
68.6
18.4
101.5
72.4
72.6
73.9
70.5
18.5
2
3
4
5
6
a
Xylose (inner)
1
2
3
4
5
106.1
74.9
87.0
68.9
66.8
106.1
74.9
87.1
69.0
66.8
Xylose (terminal)
1
2
3
4
5
106.0
106.0
75.3
78.2
70.9
67.3
106.9
76.0
78.6
71.0
67.4
75.3
78.1
70.9
67.3
Chemical shifts are given in d values (ppm), with tetramethylsilane as an internal standard
a
Values may be interchanged
was purified by silica gel column chromatography with
chloroform–methanol–water (40:10:1). Finally, 4 (25 mg)
and 5 (6 mg) were isolated and identified by high-resolu-
tion (HR)-ESI–MS (Micromass Q-Tof micromass spec-
trometer, Waters Corp., Milford, MA, USA) and NMR
humidity) with a 12-h light/dark cycle and free access to
food and water, before the experiments were performed.
All animal treatments in this study were approved by the
Institutional Animal Care and Use Committee of Waku-
naga Pharmaceutical Co., Ltd.
(
JNM-ECP500 spectrometer, JEOL, Tokyo, Japan)
1 13
including H-NMR, C-NMR, HMQC, and HMBC.
LC–MS conditions
Animals
LC–MS analysis was carried out by coupling a Shimadzu
HPLC system (LC-20A Prominence, Shimadzu, Kyoto,
Japan) to a ThermoElectron LTQ linear ion trap mass
spectrometer (ThermoElectron, San Jose, CA, USA) fitted
with an ESI source. HPLC separation was performed on a
100 9 2 mm (3 lm) YMC-Pack Pro C18 RS column
Six-week-old male ddY mice (SLC, Shizuoka, Japan) were
purchased. The mice were acclimated for 1 week in a
laboratory animal facility maintained at constant tempera-
ture (23 ± 3°C) and humidity (50 ± 20% relative
123

3
24
J Nat Med (2010) 64:321–329
(
YMC, Kyoto, Japan). The mobile phase consisted of (A)
.1% formic acid in water and (B) 0.1% formic acid in
blood was centrifuged at 1000g for 10 min, and plasma
was collected and stored at -35°C until analysis.
0
acetonitrile at a flow rate of 0.2 ml/min. Gradient elution
was used, with the initial mobile phase at 34% B, ramped
to 100% B at 3.2 min, and held until 4.2 min. At the end of
the gradient program, the eluent composition was brought
to the initial condition in 0.1 min, and the column was
equilibrated for 5.7 min before the next injection. The
column temperature was maintained at 40°C, and the
injection volume was 5 ll. The mass spectrometric con-
ditions were as follows: electrospray voltage, 4.5 kV;
heated capillary temperature, 350°C; sheath gas (nitrogen
gas) pressure, 35 arbitrary units; auxiliary gas (nitrogen
gas) pressure, 5 arbitrary units. Identification of metabo-
lites of 1 in the cecum was carried out by using LC–MS
Plasma sample preparation for LC–MS analysis
A plasma sample (40 ll) was transferred to a 1.5-ml glass
tube, and 900 ll of the IS solution was added. The mixture
was shaken for 20 s and then centrifuged at 11000g for
5 min. The supernatant was dried, and the residue was
dissolved in 100 ll methanol and then centrifuged at
11000g for 5 min. The supernatant was analyzed by LC–
MS to determine 1–3.
Validation of the analytical method
for the determination of lancemaside A
and its metabolite in plasma
n
with MS at a collision energy of 30%. LC–MS analysis for
determination of 1–3 in plasma was performed in selective
reaction monitoring (SRM). Three different time segments
were used: (1) from 0 to 3.5 min for 1 and IS, (2) from 3.5
to 5.2 min for 2, and (3) from 5.2 to 7 min for 3, respec-
tively. The SRM detection setup for 1 was m/z
Specificity was assessed by analyzing three different
sources of plasma. Chromatograms were visually
inspected for interfering chromatographic peaks from
endogenous substances. To examine the linearity of 1–3,
calibration standards with ten concentrations were pre-
pared and assayed. The final concentration ranges were
11–11363 ng/ml for 1, 8–7819 ng/ml for 2, and 6–
5513 ng/ml for 3. A calibration curve was derived from
the peak area ratio (analyte/IS) versus the known con-
centrations of the calibration samples. A weighting of 1/
1
189 ? 647; for IS it was m/z 1351 ? 809; for 2 it was
m/z 1013 ? 469, 471, 541, and 923; and for 3 it was m/z
71 ? 407, 409, and 425. The isolation width of the pre-
cursor ion was set at 3 m/z for 1, 2, and IS, and at 2 m/z for
, respectively, and the collision energy was set at 30%.
4
3
2
Determination of lancemaside A
and its metabolites in cecum
x (where x is the concentration of a given standard) was
used for curve fitting. To evaluate the accuracy and
precision of the method, recovery experiments (n = 3)
were carried out by adding different concentrations of 1–
3 to plasma. The added concentrations were 11–5681 ng/
ml for 1, 8–3909 ng/ml for 2, and 6–2756 ng/ml for 3.
The accuracy was expressed in terms of relative error,
and the precision was evaluated in terms of the relative
standard deviation (RSD). The limit of detection (LOD)
was determined as the analyte concentration that yielded
a signal equal to three times the noise level.
Compound 1 (100 mg/kg body weight) was administered
orally to three mice that had been fasted for 16 h. At 8 h
after administration, the animals were exsanguinated by
decapitation, and their ceca were rapidly extirpated and
frozen in liquid nitrogen. The frozen cecum samples were
homogenized with a 15-fold volume of methanol by using
a pestle and mortar. The mixture was centrifuged at 11000g
for 5 min, and the supernatant was analyzed by LC–MS to
determine 1 and its metabolites.
Pharmacokinetic analysis
Administration of lancemaside A
and collection of plasma samples
Pharmacokinetic parameters were calculated by using the
MOMENT computer program [18]. The peak plasma
concentration (Cmax) and time to reach Cmax (tmax) after
oral administration were read directly from the experi-
mental data. The total area under the plasma concen-
Compound 1 was administered to mice in a single dose of
1
1
00 mg/kg body weight for oral administration (p.o.) and
mg/kg body weight for intravenous administration (i.v.).
Compound 1 was dissolved in distilled water at 5 mg/ml
for oral administration and in saline at 0.2 mg/ml for
intravenous administration. Mice were fasted for 16 h prior
to administration. Blood samples were collected by
decapitation in a heparin-containing tube at 0.5, 1, 2, 4, 8,
and 24 h after oral administration and at 2, 5, 10, 20, 30,
tration–time curve from time zero to infinity (AUC0–inf)
was calculated by the trapezoidal rule extrapolation
method. The elimination rate constant (K ) was calcu-
el
lated from the slope of the line by regression analysis of
the log plasma concentration versus time plot, and the
half-life (t1/2b) was calculated as 0.693/K . The bio-
el
6
0, and 120 min after intravenous administration. The
availability (F) was calculated from the following
1
23

J Nat Med (2010) 64:321–329
325
equation: F(%) = {[AUC0–inf (p.o.)][dose (i.v.)]}/{[AUC0–inf
(
from peaks 1–5 produced the same product ions, at m/z
425, 409, and 407, with similar intensities, and this frag-
mentation pattern was similar to that reported previously
for echinocystic acid moiety [19]. The data suggest the
presence of an echinocystic acid moiety in the structure of
these compounds.
i.v.)][dose (p.o.)]} 9 100%.
Results and discussion
Identification of lancemaside A metabolites in cecum
On the basis of these results, the presumed metabolites 2–
, corresponding to peaks 2–5 (Fig. 2), were synthesized
5
Intestinal flora plays an important role in the metabolism of
glycosides in humans and animals. Metabolic reactions
such as deglycosylation and deglucuronidation occur dur-
ing passage through the gastrointestinal tract, especially in
the small intestine. To study the intestinal metabolism of
lancemaside A (1), the cecum sample from mice at 8 h
after oral administration of 1 was analyzed by LC–MS in
the negative ion mode. Five peaks (peaks 1–5) were
enzymatically and chemically and identified by NMR and
MS. The structures of 2 and 3 were determined to be
codonolaside II and echinocystic acid, respectively, by
comparison with the NMR spectroscopic data from previous
reports [5, 17]. The structures of the new compounds 4 and 5
were determined based on their spectroscopic data, includ-
ing MS, 1D- and 2D-NMR analyses. The molecular formula
of 4 was assigned as C H O , determined from its
4
6 74 16
-
observed in the chromatogram of the cecum sample
n
Fig. 2). The MS data of their peaks are summarized in
pseudomolecular ion [M–H] at m/z 881.4925 (calcd for
C H O : 881.4899) in the HR-ESI–MS spectrum. The
(
4
6 73 16
1
Table 2. Peak 1 was identified as lancemaside A (1) on the
basis of its retention time and mass fragmentation pattern
in comparison with those of authentic compounds. The
mass spectra of peaks 2 (4.2 min), 4 (4.3 min), and 5
H-NMR spectrum of 4 showed the presence of seven singlet
methyl groups [d 0.95 (3H, s), 1.03 (3H, s), 1.04 (3H, s),
1.15 (6H, s), 1.22 (3H, s), 1.81 (3H, s)], one doublet methyl
H
group [dH 1.78 (3H, d, J = 5.5 Hz)], and three anomeric
(
5.3 min) showed the ions to be at m/z 1013, 881, and 749,
respectively, corresponding to the loss of 176 Da (glucu-
ronic acid), 308 Da (glucuronic acid and xylose), and
protons [d 5.20 (1H, d, J = 7.5 Hz), 5.85 (1H, s), 6.48 (1H,
H
1
d, J = 3.5 Hz)]. The C-NMR spectrum of 4 showed the
3
presence of three anomeric carbons (d 93.7, 101.1, 106.9),
c
4
40 Da (glucuronic acid and two xyloses) from 1. The
mass spectrum of peak 3 (5.8 min) showed the ions to be at
as shown in Table 1. These data, together with correlations
observed in the HMQC and HMBC spectra, indicated the
absence of a xylose moiety in 2. Thus the structure of 4 was
determined to be echinocystic acid 28-O-b-D-xylopyrano-
syl-(1 ? 4)-a-L-rhamnopyranosyl-(1 ? 2)-a-L-arabinopyr-
anosyl ester. The molecular formula of 5 was assigned as
C H O , which was determined from its pseudomolecu-
m/z 471, corresponding to the aglycone (echinocystic acid)
n
moiety of 1. In MS spectra (Table 2), the m/z 471 ions
4
1 66 12
-
lar ion [M–H] at m/z 749.4486 (calcd for C H O :
4
1 65 12
1
49.4476) in the HR-ESI–MS spectrum. The H-NMR
7
spectrum of 5 showed the presence of seven singlet methyl
groups [d 0.97 (3H, s), 1.03 (6H, s), 1.14 (3H, s), 1.16 (3H,
H
s), 1.22 (3H, s), 1.81 (3H, s)], one doublet methyl group [dH
1
.72 (3H, d, J = 6.0 Hz)], and two anomeric protons [d
H
1
.84 (1H, s), 6.55 (1H, d, J = 2.5 Hz)]. The C-NMR
3
5
spectrum of 5 showed the presence of two anomeric carbons
dc 93.5, 101.5), as shown in Table 1. These data, together
(
with correlations observed in the HMQC and HMBC spec-
tra, indicated the absence of one xylose moiety present in 4.
Thus, the structure of 5 was determined to be echinocystic
acid 28-O-a-L-rhamnopyranosyl-(1 ? 2)-a-L-arabinopyr-
anosyl ester. In the negative ion LC–MS analysis, the
metabolite peaks (peaks 2–5) observed in the cecum sample
after oral administration of 1 had retention times and mass
spectra that were identical to those of the synthetic com-
pounds 2–5, and thus the structures of the metabolites were
determined to be codonolaside II (2), echinocystic acid (3),
echinocystic acid 28-O-b-D-xylopyranosyl-(1 ? 4)-a-L-
rhamnopyranosyl-(1 ? 2)-a-L-arabinopyranosyl ester (4),
Fig. 2 Representative chromatograms of a cecum sample at 8 h after
oral administration of lancemaside A. a Total ion chromatogram,
mass chromatograms at b m/z 1189, c m/z 1013, d m/z 881, e m/z
749 ? 795, f m/z 471. Peak 1, lancemaside A; peaks 2–5, metabolites
of lancemaside A
123

3
26
J Nat Med (2010) 64:321–329
n
Table 2 MS spectra in the negative ion mode of lancemaside A and its metabolites detected in the cecum at 8 h after oral administration of
lancemaside A
a
n
MS type
Peak
Retention time (min)
3.0
Isolated m/z
Major product ions (m/z)
-
2
MS
b
1
2
(Lancemaside A)
(Codonolaside II)
(Echinocystic acid)
1189 [M–H]
647 (100)
3
MS
647
471
407
629 (43), 571 (57), 525 (25), 471 (100), 407 (30)
425 (21), 409 (30), 407 (100)
391 (100), 377 (46)
4
MS
5
MS
-
2
MS
4.2
1013 [M–H]
995 (31), 923 (72), 541 (36), 471 (53), 469 (100)
337 (100)
3
MS
4
4
4
69
71
07
3
MS
425 (25), 409 (35), 407 (100)
391 (100), 377 (35)
4
MS
-
2
MS
3
4
5.8
4.3
471 [M–H]
07
881 [M–H]
425 (26), 409 (44), 407 (100)
391 (100), 377 (46)
3
MS
4
-
-
2
MS
471 (100)
3
MS
4
4
71
07
425 (33), 409 (45), 407 (100)
391 (100), 377 (45)
4
MS
2
MS
5
5.3
749 [M–H]
471 (100)
3
4
4
7
71
07
MS
425 (22), 409 (37), 407 (100)
391 (100), 377 (39)
4
MS
-
2
MS
95 [M?HCOO]
749 (15), 539 (100)
a
Peak numbers are the same as in Fig. 2
b
Relative abundances (%) are given in parentheses
and echinocystic acid 28-O-a-L-rhamnopyranosyl-(1 ? 2)-
a-L-arabinopyranosyl ester (5), as shown in Fig. 1.
It is known that saponins are metabolized in the gas-
trointestinal tract into their deglycosylated form. Akao
et al. [20] reported that ginsenoside Rb is enzymatically
1
metabolized into compound K as a result of the hydrolysis
of the glycosidic linkage by rat intestinal bacteria. Gly-
cyrrhizin has been shown to be hydrolyzed by glucuroni-
dase of intestinal bacteria to glycyrrhetinic acid, an active
principal aglycone [11, 21]. Therefore, our data suggest
that orally administered 1 was metabolized into its degly-
cosylated metabolites 2–5 by the digestive enzymes in the
gastrointestinal tract. In this metabolic pathway, the ester
linkage at C-28 of 1 was hydrolyzed to a carboxylic acid.
To our knowledge, this is the first report of the hydrolysis
of the carboxyl ester bond of saponin occurring in the
gastrointestinal tract.
Fig. 3 Selective reaction monitoring chromatograms of rat plasma
sample spiked with lancemaside A, codonolaside II, echinocystic
acid, and lancemaside E (internal standard). The product ions were
LC–MS method to quantify lancemaside A and its
metabolites in plasma
monitored at
a
m/z 1189 ? 647 for lancemaside A,
b m/z
1
4
1
013 ? 469, 471, 541, and 923 for codonolaside II, c m/z
71 ? 407, 409, and 425 for echinocystic acid, and d m/z
351 ? 809 for lancemaside E
Lancemaside A (1) and its metabolites, 2 and 3, in plasma
were quantified by LC–MS analysis, which has high
specificity and selectivity. The ESI spectra mainly showed
-
the deprotonated molecule ([M–H] ) of 1 (m/z 1189), 2 (m/
z 1013), 3 (m/z 471), and IS (m/z 1351). The product ions at
m/z 647 for 1, m/z 469, 471, 541 and 923 for 2, m/z 407,
in SRM mode in order to quantify them. Typical SRM
chromatograms of a blank plasma sample spiked with 1 at
284 ng/ml, 2 at 195 ng/ml, 3 at 138 ng/ml, and IS at
1203 ng/ml are shown in Fig. 3. The retention times of
1–3, and IS were 3.1, 4.2, 5.8, and 2.6 min, respectively.
409 and 425 for 3, and m/z 890 for IS in the MS/MS
experiments of the deprotonated molecule were monitored
1
23

J Nat Med (2010) 64:321–329
327
Fig. 4 Plasma concentration–time profiles of a lancemaside A, b c
codonolaside II, and c echinocystic acid following oral administration
of lancemaside A at 100 mg/kg to mice. Each point represents the
mean ± SE (n = 9 or 10)
Validation of the LC–MS method to quantify
lancemaside A and its metabolites in plasma
Specificity was assessed by analyzing blank plasma from
three different sources. No endogenous interfering sources
were observed on the chromatograms obtained from blank
plasma at the retention times of 1–3 and IS. The calibration
curves of 1–3 in the respective spiked samples had corre-
lation coefficients of 0.999, 0.990, and 0.997, respectively.
Accuracy and precision were determined by using three
replicates of the spiked samples. The relative error ranged
from -8 to 3% for 1, from -13 to 12% for 2, and from
-
16 to 7% for 3, whereas intraday RSD did not exceed
4% for 1, 15% for 2, and 13% for 3. The LOD was
ng/ml for 1 and 3, and 2 ng/ml for 2.
1
3
Study of lancemaside A pharmacokinetics
The plasma concentration–time profiles after oral admin-
istration of lancemaside A (1) at 100 mg/kg are shown in
Fig. 4a, and some relevant pharmacokinetic parameters are
listed in Table 3. Compound 1 was rapidly absorbed after
oral administration, and the concentration in plasma
showed peaks at 0.5 h. The values of Cmax, t1/2b, and
AUC0–inf were 1063 ng/ml, 5.02 h, and 2634 ng h/ml,
respectively. After intravenous administration at 1 mg/kg,
the plasma level of 1 declined rapidly (Fig. 5), and the
concentration–time curve exhibited a biphasic decline. The
t1/2b and AUC0–inf values were 0.54 h and 2359 ng h/ml,
respectively. The bioavailability (F) of 1 was 1.1%. It is
known that saponins are very poorly absorbed following
oral administration to animals. Liang et al. [10] reported
the bioavailability of tubeimoside I was only 0.23% in rats,
and Li et al. [22] reported the bioavailability of ginsenoside
Rg1, Rb1, and Rd to be 15.62, 0.28, and 0.34%, respec-
tively, in rats. Moreover, after oral administration of gly-
cyrrhizin to rats, the unmetabolized compound was not
detected in plasma, but glycyrrhetic acid, which is con-
verted by intestinal bacteria, was detected [21]. Therefore,
our data indicated that the poor bioavailability of 1 might
be due to its low absorption from the gastrointestinal tract
and/or its metabolism by intestinal bacteria. After oral
administration of 1 to mice, deglycosylated metabolites 2
and 3 were detected in the plasma after 0.5 and 2 h,
respectively, and the concentration in plasma showed peaks
123

3
28
J Nat Med (2010) 64:321–329
Table 3 Pharmacokinetic parameters of lancemaside A, codonola-
side II, and echinocystic acid in mice following oral (p.o., 100 mg/kg)
and intravenous (i.v., 1 mg/kg) administration of lancemaside A
four deglycosylated metabolites, 2–5, in the cecum were
identified, and metabolites 2 and 3 were detected in plasma.
We speculate that metabolites 2 and 3 may contribute to
the biological effects following oral administration of 1.
Thus, our findings should be helpful for understanding the
mechanism of the biological effects of 1.
Lancemaside A
Codonolaside II Echinocystic
acid
p.o.
0.5
i.v.
p.o.
p.o.
t
max (h)
4
8
C
max (ng/ml) 1063
61
42
–
t
1/2b (h)
5.02
0.54
7.83
590
References
AUC0–inf
2634
2359
494
(
ng h/ml)
1
2
. Sekita S, Kuroyanagi M, Yasuda K, Mizuno I, Ushijima M,
Hayama M, Ichikawa M, Sumihiro M (2005) Drug for amelio-
rating male climacteric disorder. Patent WO200515426
. Ushijima M, Mizuno I, Suzuki E, Amayasu R, Ishii S, Nishihama
T, Morihara N, Kashimoto N, Mouri Y, Sumioka I, Kuroyanagi
M, Sekita S, Hayama M (2007) Improvement of PADAM-like
symptoms in middle-aged men by a designer food containing
Codonopsis lanceolata. Pharmacometrics 72:23–30
F (%)
1.1
3
. Lee KT, Choi J, Jung WT, Nam JH, Jung HJ, Park HJ (2002)
Structure of a new echinocystic acid bisdesmoside isolated from
Codonopsis lanceolata roots and the cytotoxic activity of pros-
apogenins. J Agric Food Chem 50:4190–4193
4
. Lee KW, Jung HJ, Park HJ, Kim DG, Lee JY, Lee KT (2005) b-
D-Xylopyranosyl-(1 ? 3)-b-D-glucuronopyranosyl echinocystic
acid isolated from the roots of Codonopsis lanceolata induces
caspase-dependent apoptosis in human acute promyelocytic leu-
kemia HL-60 cells. Biol Pharm Bull 28:854–859
5
6
. Li JP, Liang ZM, Yuan Z (2007) Triterpenoid saponins and anti-
inflammatory activity of Codonopsis lanceolata. Pharmazie
6
2:463–466
. Ushijima M, Komoto N, Sugizono M, Mizuno I, Sumihiro M,
Ichikawa M, Hayama M, Kawahara N, Nakane T, Shirota O,
Sekita S, Kuroyanagi M (2008) Triterpene glycosides from the
roots of Codonopsis lanceolata. Chem Pharm Bull 56:308–314
. Yuan Z, Liang ZM (2006) A new triterpenoid saponin from
Codonopsis lanceolata. Chin Chem Lett 17:1460–1462
. Ushijima M, Mizuno I, Ohta S, Sumihiro M, Mouri Y, Ichikawa
M, Hayama M, Sekita S, Kuroyanagi M (2008) Codonopsis
lanceolata ameliorate plasma testosterone reduction induced by
immobilization stress in middle-aged mice. Pharmacometrics
7
8
Fig. 5 Plasma concentration–time profiles of lancemaside A follow-
ing intravenous administration of lancemaside A at 1 mg/kg body
weight to mice. Each point represents mean ± SE (n = 4)
at 4 and 8 h, respectively (Table 3). Our results suggest
that metabolites 2 and 3 may be involved in the biological
effects when compound 1 was administered orally to mice.
In addition, metabolite 2 was detected in both the small
intestine and cecum, while 3 was detected only in the
cecum (data not shown). Our data indicated that 1 was
metabolized to 2 in the upper gastrointestinal tract and then
further metabolized to 3 in the cecum. On the other hand,
metabolites 4 and 5, identified in the cecum, were not
detected in plasma at any time.
7
5:69–75
9. Han M, Sha X, Wu Y, Fang X (2005) Oral absorption of ginse-
noside Rb using in vitro and in vivo models. Planta Med 71:398–
04
1
4
1
1
0. Liang MJ, Zhang WD, Zhang C, Liu RH, Shen YH, Li HL, Wang
XL, Wang XW, Zhu JB, Chen CL (2007) Quantitative determi-
nation of the anticancer agent tubeimoside I in rat plasma by
liquid chromatography coupled with mass spectrometry. J
Chromatogr B 845:84–89
1. Hattori M, Sakamoto T, Kobashi K, Namba T (1983) Metabolism
of glycyrrhizin by human intestinal flora. Planta Med 48:38–42
12. Paek IB, Moon Y, Kim J, Ji HY, Kim SA, Sohn DH, Kim JB, Lee
HS (2006) Pharmacokinetics of a ginseng saponin metabolite
compound K in rats. Biopharm Drug Dispos 27:39–45
3. Li B, Abliz Z, Tang M, Fu G, Yu S (2006) Rapid structural
characterization of triterpenoid saponins in crude extract from
Symplocos chinensis using liquid chromatography combined with
electrospray ionization tandem mass spectrometry. J Chromatogr
A 1101:53–62
4. Ha YW, Na Y-C, Seo J-J, Kim S-N, Linhardt RJ, Kim YS (2006)
Qualitative and quantitative determination of ten major saponins
in Platycodi radix by high performance liquid chromatography
In conclusion, the present study investigated the
metabolism and absorption of lancemaside A (1), which is
considered to be one of the biologically active compounds
present in the roots of C. lanceolata. After oral adminis-
tration of 1 to mice, the unmetabolized compound appeared
rapidly in plasma, and its low bioavailability (1.1%) was
due to deglycosylation by the intestinal bacteria and to poor
absorption in the upper gastrointestinal tract. Furthermore,
1
1
1
23

J Nat Med (2010) 64:321–329
329
with evaporative light scattering detection and mass spectrome-
try. J Chromatogr A 1135:27–35
of triterpenoid saponins in the roots of Codonopsis lanceolata by
liquid chromatography-mass spectrometry. J Nat Med 62:423–
429
1
5. Madl T, Sterk H, Mittelbach M (2006) Tandem mass spectro-
metric analysis of a complex triterpene saponin mixture of
Chenopodium quinoa. J Am Soc Mass Spectrom 17:795–806
6. Shirota O, Nagamatsu K, Sekita S, Komoto N, Kuroyanagi M,
Ichikawa M, Ohta S, Ushijima M (2008) Preparative separation
of the saponin lancemaside A from Codonopsis lanceolata by
centrifugal partition chromatography. Phytochem Anal 19:403–
20. Akao T, Kanaoka M, Kobashi K (1998) Appearance of com-
1
pound K, a major metabolite of ginsenoside Rb by intestinal
1
bacteria, in rat plasma after oral administration. Measurement of
compound K by enzyme immunoassay. Biol Pharm Bull 21:245–
249
21. Takeda S, Ishihara K, Wakui Y, Amagaya S, Maruno M, Akao T,
Kobashi K (1996) Bioavailability study of glycyrrhetic acid after
oral administration of glycyrrhizin in rats; relevance to the
intestinal bacterial hydrolysis. J Pharm Pharmacol 48:902–905
22. Li L, Sheng Y, Zhang J, Wang C, Guo D (2004) HPLC deter-
mination of four active saponins from Panax notoginseng in rat
serum and its application to pharmacokinetic studies. Biomed
Chromatogr 18:849–856
4
10
1
1
1
7. Zhang Z, Koike K, Jia Z, Nikaido T, Guo D, Zheng J (1999)
Triterpenoidal saponins acylated with two monoterpenic acids
from Gleditsia sinensis. Chem Pham Bull 47:388–393
8. Tabata K, Yamaoka K, Kaibara A, Suzuki S, Terakawa M, Hata
T (1999) Moment analysis program available on Microsoft Excel.
Xenobio Metab Dispos 14:286–293
9. Ichikawa M, Ohta S, Komoto N, Ushijima M, Kodera U, Hayama
M, Shirota O, Sekita S, Kuroyanagi M (2008) Rapid identification
123