The influence of the sennosides on absorption of glycyrrhetic acid in rats

Article abstract of DOI:10.1248/bpb.28.1897

In the course of our clinical studies of Kampo medicine (traditional Japanese medicines), we observed the pharmacokinetic interactions between two herbs. When Onpito (TJ-8117, Kampo medicine) containing licorice and rhubarb was administered orally to human subjects, we observed that the AUC _(0-lim) and C_max of glycyrrhetic acid (GA) in plasma were lower than those treated with other Kampo medicines containing licorice. In this study, we demonstrate the pharmacokinetic interactions of GA derived from glycyrrhizinic acid (GL) in licorice and anthraquinones derived from rhubarb. To our knowledge, this is the first report to investigate the pharmacokinetic interactions between two herbs. When GL was orally co-administrated to rats with a non-effective dose of sennoside A having purgative activity, the AUC _(0-lim) and C_max of GA decreased. In addition, sennoside A did not affect the metabolism of GL by the intestinal bacteria in vitro. In the examination using an in situ loop of rat colon, the remaining ratio of GA rose drastically by the co-administration of sennoside A, sennidin A and rhein. Observed inhibition activity of these anthraquinones on GA absorption depended on the concentration of the components added. The maximum inhibition ratio was approximately 75% by rhein, 60% by sennoside A and 25% by sennidin A. We conclude that the decrease of the pharmacokinetic parameters of GA in human plasma observed in the clinical study of TJ-8117 is attributable to an interactive action of absorption from the intestinal tract by anthraquinones contained in or derived from rhubarb.

Full text of DOI:10.1248/bpb.28.1897

October 2005
Biol. Pharm. Bull. 28(10) 1897—1902 (2005)
1897
The Influence of the Sennosides on Absorption of Glycyrrhetic Acid in
Rats
,
a,b
a,b
a
a
Yasuharu MIZUHARA,* Yukiho TAKIZAWA, Kazuhisa ISHIHARA, Takayuki ASANO,
a
a
a
a,c
a,c
Hirotaka KUSHIDA, Takashi MOROTA, Yo s h io KASE, Shuichi TAKEDA, Masaki ABURADA,
Masaaki NOMURA, and Koichi YOKOGAWA
d
d
a
Medicinal Evaluation Laboratory, Tsumura & Co.; 3586 Yoshiwara, Ami-machi, Inasiki-gun, Ibaraki 300–1192, Japan:
b
Department of Medicinal Informatics, Graduate School of Medical Science, Kanazawa University; 13–1 Takara-machi,
c
Kanazawa, Ishikawa 920–8641, Japan: Faculty of Pharmacy, Musashino University; 1–1–20 Shinmachi, Nishitokyo,
Tokyo 202–8585, Japan: and Department of Hospital Pharmacy, School of Medicine, Kanazawa University; 13–1
d
Takara-machi, Kanazawa, Ishikawa 920–8641, Japan. Received February 28, 2005; accepted June 27, 2005
In the course of our clinical studies of Kampo medicine (traditional Japanese medicines), we observed the
pharmacokinetic interactions between two herbs. When Onpito (TJ-8117, Kampo medicine) containing licorice
and rhubarb was administered orally to human subjects, we observed that the AUC(0—lim) and Cmax of glycyr-
rhetic acid (GA) in plasma were lower than those treated with other Kampo medicines containing licorice. In
this study, we demonstrate the pharmacokinetic interactions of GA derived from glycyrrhizinic acid (GL) in
licorice and anthraquinones derived from rhubarb. To our knowledge, this is the first report to investigate the
pharmacokinetic interactions between two herbs. When GL was orally co-administrated to rats with a non-effec-
tive dose of sennoside A having purgative activity, the AUC(0—lim) and Cmax of GA decreased. In addition, senno-
side A did not affect the metabolism of GL by the intestinal bacteria in vitro. In the examination using an in situ
loop of rat colon, the remaining ratio of GA rose drastically by the co-administration of sennoside A, sennidin A
and rhein. Observed inhibition activity of these anthraquinones on GA absorption depended on the concentra-
tion of the components added. The maximum inhibition ratio was approximately 75% by rhein, 60% by senno-
side A and 25% by sennidin A. We conclude that the decrease of the pharmacokinetic parameters of GA in
human plasma observed in the clinical study of TJ-8117 is attributable to an interactive action of absorption
from the intestinal tract by anthraquinones contained in or derived from rhubarb.
Key words glycyrrhetic acid; interaction; absorption; rhein; licorice; rhubarb
Herbal medicines are increasingly used in the world, and
Kampo medicines are orally administered so that the
herb–drug interactions are recognized as a clinically impor- chemical components contained in the Kampo medicines
1—5)
tant problem.
Recently, various clinical and non-clinical may be manipulated by the intestinal flora before being ab-
studies including in vitro studies related to pharmacokinetic sorbed into the body. For example, the following hydrolysis,
interactions between herbs and new drugs have been re- glycosides are transformed by intestinal flora to the corre-
6—11)
12)
15,16)
ported.
Durr et al. reported that the administration of sponding aglycones.
St. John’s Wort (Hypericum perforatum) extract to 8 healthy
In a previous paper, we reported that only GA was de-
male volunteers during 14 d resulted in an 18% decrease of tected instead of GL in the plasma when GL was orally ad-
digoxin exposure after a single digoxin dose (0.5 mg), in 1.4- ministered to rats. In addition, we clarified that GA was not
and 1.5-fold increased expressions of duodenal P-glycopro- detected in the plasma after oral administration of GL to
17)
tein/MDR1 and CYP3A4, respectively, and in a 1.4-fold in- germ-free rats having no intestinal bacteria. These results
crease in the functional activity of hepatic CYP3A4 in clini- suggest that GL is hydrolyzed to GA by intestinal bacteria
cal study. However, to our knowledge, there has been no re- after oral administration of GL to rats, and is absorbed as GA
port dealing with the pharmacokinetic interactions between from the intestinal tract. Absorption of GL might be smaller
two herbs. In the course of our clinical studies of Kampo
medicine, we observed changes in various pharmacokinetic
parameters of a marker component of Kampo medicine. In
this article, we demonstrate a pharmacokinetic interaction
between two herbs by various non-clinical investigations of a
representative marker component of a herb.
Glycyrrhizinic acid (GL), one of the main components of
licorice root (Glycyrrhizae radix), shows various biological
13)
activities such as having an anti-inflammatory and an anti-
14)
hepatitis effect and is contained in many types of Kampo
medicines. GL is a glycoside having two glucuronic acid
molecules attached to the hydroxyl group at C-3 position of
glycyrrhetic acid (GA) (Fig. 1). Since GA is derived from
GL via metabolism as mentioned below, GA has been fre-
quently used as a pharmacokinetic marker of Kampo medi-
cines including licorice.
Fig. 1. The Metabolism of GL and Sennosides by Intestinal Bacteria
To whom correspondence should be addressed. e-mail: mizuhara_yasuharu@mail.tsumura.co.jp © 2005 Pharmaceutical Society of Japan
∗

1
898
Vol. 28, No. 10
18)
than that of GA due to its hydrophilic property. In contrast,
Akao et al. reported that the elevation of the glutamic-ox-
alacetic transaminase (GOT) in serum induced by CCl was
4
inhibited in normal rats after oral administration of GL, and
19)
not observed in germ-free rats. These results suggest that
the hydrolysis of GL to GA by intestinal bacteria is deter-
mined by the pharmacological effect of orally administered
GL.
Onpito (TJ-8117) is one Kampo medicine composed of
five herbs, Rhei rhizoma (rhubarb), Glycyrrhizae radix
(licorice), Ginseng radix, Zingiberis rhizoma and Aconiti
tuber. In a clinical study of TJ-8117 in healthy human sub-
jects, we observed that the pharmacokinetics parameters (the
plasma concentration–time curve, AUC(0—lim) and the maxi- Fig. 2. Correlation of AUC of GA in Human Plasma and Dosage after Ad-
mum peak plasma concentration, Cmax) of GA, a marker sub- ministration of Various Kampo Medicines to Human Subjects
Kampo A—J (not including rhubarb).
stance, in plasma after oral administration of TJ-8117 were
much lower than those after administration of the other
Kampo medicines including licorice (Fig. 2). A detailed re- water freely available. The rats were fasted overnight prior to
port of the clinical study of TJ-8117 is now in preparation, the experiment after acclimation for one week.
we will publish the report including pharmacokinetics of the
other components soon after the data analyses.
The Purgative Activity of Sennoside A in Rats Senno-
side A suspended in 0.5% sodium carboxymethylcellulose
He et al. reported a similar result, whereby the co-admin- (CMC) was orally administered to rats at doses of 2.5, 5, 10,
istration of antibiotics to rats such as amoxicillin and metron- and 20 mg/kg/10 ml. The frequency of diarrhea was recorded
idazole decreased the AUC of GA after the administration of after administration. The condition of the feces was classified
20)
GL. Thus, the influence on intestinal bacteria must be one into four grades, normal feces (0), mucous feces (0.25),
of the reasons of decreased GA absorption. Moreover, diar- loose stool (0.5), and diarrhea (1). A probit plot of the corre-
rhea was also speculated to be for other reasons, because di- sponding feces score calculated the ED50 value of purgative
arrhea affected the condition of intestinal bacteria and might activity.
shorten the retention time of intestinal content. Considering
The Simultaneous Administration of GL and Senno-
this point, diarrhea caused by anthraquinones (sennoside A, side A Suspensions of GL at a dose of 5 mg/kg and senno-
sennidin A, rhein and rhein-9-anthrone) contained in rhubarb side A at a dose of 0, 1.6, 5.6 and 19.6 mg/kg were prepared
were thought to be involved in the lowering of GA absorp- with 0.5% CMC, and these suspensions were orally adminis-
tion in the clinical study of TJ-8117.
trated to rats (7 weeks, each nꢁ6). A dosage of 1.6 mg/kg of
To elucidate the phenomenon in this clinical study, TJ- sennoside A is a non-effective dose, 5.6 mg/kg is equivalent
117, we investigated the effect of the anthraquinones con- to ED50 of purgative activity and 19.6 mg/kg causes diarrhea
8
cerned as active principles for the purgative effect of rhubarb in rats. Approximately 300 ml of blood was collected from
to the pharmacokinetics of GL and GA in rats.
the femoral artery cannula treated in heparin at 1, 2, 4, 6, 9,
2, 18, 24 and 32 h after administration. The plasma was ob-
1
MATERIALS AND METHODS
tained by centrifugation (3000 rpm, 10 min, 4 °C) and was
stored at ꢂ20 °C until analysis.
Chemicals GL was purchased from Wako Pure Chem.
Elucidation of the Absorption of GA in Rats That Have
Industries (Osaka, Japan). GA and rhein purchased from Undergone Ileostomy Ileostomy was conducted in rats by
Tokyo Kasei Co. (Tokyo, Japan) were purified by silica gel the insertion of cannula (polyethylene tube O.D. 3 mm, Hi-
column chromatography and recrystallization. Sennoside A biki) at the ileocecum junction under anesthetic. GL sus-
and sennidin A were purchased from EXTRASYNTHESE. pended in 0.5% CMC was orally administered at a dose of
Rhein-9-anthrone was synthesized from rhein according to 5 mg/kg to these operated on rats (nꢁ5—6). Approximately
21)
the method reported by Witte et al. Namely, 100 mg of 300 ml of blood was collected at 1, 2, 4, 6, 9, 12, 18, 24 and
rhein was suspended in 20 ml of acetic acid, and was re- 32 h after administration from heparin-treated cannula in-
fluxed. 3.5 ml of 40% SnCl -hydrochloric acid solution was serted in the femoral artery. The plasma was obtained by
2
dropped to this suspension under warm conditions (140 °C), centrifugation and stored at ꢂ20 °C until analysis.
and stirred for 2 h. This reaction mixture was left to stand
The cannula (polyethylene tube O.D. 0.61 mm, Natsume
overnight at 4 °C, and the insoluble precipitate was collected factory) was inserted into the cecum from the ileocecum
washed with water and crystallized from dichloromethane/ junction for direct administration to cecum. GL and senno-
methanol. Glycyrrhetic acid acetate was synthesized from side A suspended in 0.5% CMC were administered in the
GA by an ordinal method. b-galactosidase was purchased cecum from this cannula at a dose of 5 mg/kg with or without
from Boehringer Mannheim GmbH.
1.6 mg/kg of sennoside A to these treated rats (nꢁ5—6).
Animals Male Sprague Dawley rats, weighing 235— Blood samples were collected at the same time points and by
3
06 g, were purchased from Charles River Japan (Atsugi, the same method described above. The plasma obtained by
Japan). The rats were bred within one day, during a 12 h pe- centrifugation was stored at ꢂ20 °C until analysis.
riod, at a constant temperature (23ꢀ2 °C) and humidity
GA Analysis of Rat Plasma by Enzyme Immuno Assay
55ꢀ10%). The rats were fed standard laboratory chow with The plasma concentration of GA was measured by Enzyme
(

October 2005
1899
Immuno Assay (EIA), as reported by Kanaoka et al. with tions were as follows: column was Cosmosil Packed Column
22)
®
some modification. The antiserum of GA was obtained by 5C18-300 (4.6ꢃ150 mm); the mobile phase was 75%
repeated administration of the synthesized N-glycyrrhetyl- methanol including 2% acetic acid, 1 ml/min of flow rate,
glycine–BSA complex to rabbits. The cross reaction of anti- column temperature at 45 °C, and detection wavelength at
serum for 11 analogs, following GL, 18a-GA, b-estradiol, 254 nm. Ethanol solutions of GA (60—600 ng/ml) for the
aldosterone, sodium deoxycholate, hydrocortisone, pro- calibration curve were prepared separately. A total of 50 ml of
gesterone, 5a-androstan-17b-ol-3-one, etiocholan-17b-ol-3- i.s. solution and 100 ml of physiological saline were added to
one, cholesterol and carbenoxolone was 1% or less. As a la- 50 ml of the obtained solutions. The lower limit of quantita-
beled antigen, the g-glycyrrhetylaminobutylic acid b-galac- tion in this analysis was 150 ng/ml.
tosidase complex was synthesized.
The Metabolism of GL by Intestinal Bacteria (in Vitro)
The obtained plasma samples were diluted 16 times with The cecum of the rats (nꢁ5) were promptly taken out after
buffer A (0.02 mol/l sodium phosphate buffer (pH 7.0) con- removing all blood and the cecum contents were moved into
taining 0.1% of BSA, 0.1% of sodium azide and 1 mmol/l centrifuge tubes filled with N gas. These samples were di-
2
magnesium chloride). These samples (each 100 ml) were in- luted 4 times with GAM broth. In order to extract several en-
cubated with 100 ml of the GA antiserum solution diluted zymes produced by intestinal microflora, the ultrasonic
6
0000 times with buffer A for 30 min at room temperature. crusher treated the cecum contents for 3 min under cooling
Subsequently, 50 ml of the labeled antigen diluted 80000 by ice.
times with buffer A were added to the above reaction mix-
These cecum contents (1.9 ml) in the tubes filled with N2
tures, and incubated for 90 min at room temperature. Further- gas were pre-incubated for several minutes at 37 °C. After
®
more, 100 ml of MARCELLA10 (Non-dissolved secondary the pre-incubation, 50 ml of the GL solution (final concentra-
antibody: DAINIPPON PHARMACEUTICAL CO., LTD) tion; 0.0823 mg/ml) was added to the mixture with 50 ml
diluted 5 times with buffer A was added to the reaction mix- of sennoside A solution (final concentration; 0.0086 and
tures, and incubated for 40 min at room temperature. These 0.0863 mg/ml) or 50 ml of sennidin A solution (final concen-
reaction mixtures were washed with 1 ml of buffer A by tration; 0.0054 and 0.0538 mg/ml). These mixtures were in-
removing the supernatant after centrifugation (3000 rpm, cubated for 3 h at 37 °C. Then, 0.5 ml of 1 mol/l HCl was
15 min, 4 °C). This procedure was repeated twice. After added to stop the reaction. The reaction mixtures were cen-
washing the reaction mixtures, 100 ml of 4-methylumbellif- trifuged (3000 rpm, 10 min, 4 °C), and then 0.5 ml of the su-
ꢂ4
eryl-b-D-galactoside aqueous solution (1.5ꢃ10 mol/l) was pernatant was collected. Then, 20 ml of the internal standard
added to the precipitate and it was incubated for 30 min at solution (glycyrrhetic acid acetate: 100 mg/ml ethanol solu-
3
0 °C. Subsequently, the reaction was stopped by the addition tion) and 2 ml of ethyl acetate were added to the collected su-
of 2 ml of 0.1 mol/l glycine–NaOH (pH 10.3). The fluores- pernatant and the mixtures were agitated for 10 min. After
cence intensity of the sample was measured by the fluo- centrifugation (3000 rpm, 10 min, 4 °C), 1 ml of the upper
rophotometer (F-4500 types a spectrum a fluorophotometer, layer was evaporated under a N gas stream and dissolved in
2
Hitachi, Ltd.) at 365 nm for the excitation wavelength, and at 200 ml of HPLC mobile phase. A total of 30 ml of this solu-
448 nm for the fluorescence wavelength. The lower limit of tion was injected into the HPLC (Shimadzu, Japan; pump:
quantity in this method was 4 ng/ml.
LC-6A, auto sampler: SIL-6A, column oven: CTO-6A, and
Gastro Intestinal Absorption at the in Situ Loop UV detecter: SPD-6A). HPLC conditions were as follows:
®
Method The rats were used after fasting overnight. Each column was Cosmosil Packed Column 5C18-300 (4.6ꢃ
rat was anesthetized with Nembutal (1 ml/kg, i.p.). The loop 250 mm); the mobile phase was methanol: 2% acetic acid
of the colon (2 cm), duodenum (1 cm), jejunum (5 cm), ileum (v:vꢁ80:20), 1.1 ml/min flow rate, column temperature was
(5 cm), and cecum were prepared by the authorized method. at 45 °C and detection wavelength at 254 nm. The concentra-
The test compounds (GA (only), GAꢄsennoside A, GAꢄ tion of GA was calculated by the internal standard method
sennidin A, GAꢄrhein, GAꢄrhein-9-anthrone) were injected (the ratio of the peak height of GA to glycyrrhetic acid ac-
in a 0.5 ml volume into each loop under several combination etate). The range of the calibration curve was from 200 ng/ml
patterns. At 30 min after injection, the intestinal loop was to 20 mg/ml.
incised after blood collection with heparinized syringes
Statistical Analysis The plasma concentration of GA
from the inferior vena cava. The content of the each loop was calculated by using a calibration curve program
was recovered, and then the residual amount of GA was NEWCRA3J (Wakabayashi). The pharmacokinetic param-
determined by HPLC.
eters were calculated by WinNonlin (non-compartment
The Determination of GA Content in the Intestines model; Pharsight Inc.). The peak plasma concentration
HPLC) The content of GA in the each intestine was deter- (Cmax), and time to reach the maximum plasma concentration
(
mined by HPLC. Each intestinal content was made uniform (tmax) were calculated from the actual measurements. The
with physiological saline at 20 ml and homogenized by Poly- area under the plasma concentration–time curve (AUC(0—lim)
tron (NITI-ON, Japan). A total of 50 ml of ethanol and gly- was computed by the trapezoid method.
)
cyrrhetic acid acetate solution (internal standard; i.s.:
Microsoft Excel calculated the mean value. Data was ex-
1
0 mg/ml in methanol) were added to the obtained ho- pressed as the meanꢀS.D. (nꢁ5—6). The significant differ-
mogenates (100 ml). After the samples were stirred and cen- ences were evaluated using Dunnett’s multiple comparison
trifuged (15000 rpm, 5 min, 4 °C), 10 ml of the supernatant test or Student’s t-test.
was injected into HPLC (Shimadzu, Japan; pump: LC-10AD,
degasser: DGU-12A, auto sampler: SIL-10AXL, column
oven: COT-10A, and UV detector: SPD-10A). HPLC condi-

1
900
Vol. 28, No. 10
RESULTS
the group treated with GL (5 mg/kg) alone, no statistical sig-
nificant difference was observed in tmax. Moreover, there was
The Purgative Activity of Sennoside A in Rats The in- no other parameter showing a significant change among the
cidence of diarrhea after oral administration of sennoside A co-administration groups.
at doses of 2.5, 5, 10 and 20 mg/kg to rats was observed.
Examination of Absorption in Rats That Have Under-
After administration of sennoside A, diarrhea was observed gone Ileostomy In order to specify the part where the inter-
after 3—4 h and reached a peak at 9 h in all groups. The fre- action of GL and sennoside A occurred, GL was intracecally
quency of diarrhea appearance at 9 h was 15, 50, 90, and administrated with or without sennoside A to normal rats and
100%, respectively (Fig. 3). The grade of diarrhea was indi- orally administrated to rats that had undergone ileostomy.
cated as follows. In the 2.5 mg/kg group, mucous feces were The plasma concentration of GA was determined by EIA.
observed in one, and loose stool was observed in one of the The plasma concentration profile of GA after intracecal or
five rats. In the 5 mg/kg group, mucous feces were observed oral administration of GL to rats is shown in Fig. 4B. The
in two, loose stool was observed in two, and diarrhea was ob- pharmacokinetic parameters are shown in Table 1. The area
served in one of the five rats. In the 10 mg/kg group, loose under the curve (AUC(0—lim)) of GA after intracecally admin-
stool was observed in one rat, and diarrhea was observed in istration of GL to rats was significantly reduced by sennoside
four of the five rats. In the 20 mg/kg group, all rats showed A co-administration, similar to oral administration in rats. In
diarrhea. From these results, the ED50 value was computed as addition, Cmax also showed a tendency to decrease. Moreover,
5
.6 mg/kg.
The Combined Treatment of Sennoside A and GL to of GL to rats that had undergone ileostomy.
Rats Based on the results of the purgative activity of sen- Loop Absorption Test in Situ In order to elucidate the
GA was not detected in the plasma after oral administration
noside A in rats, the optimal dose of sennoside A was deter- interaction of GL and sennosides or its derivatives directly,
mined. At doses of 1.6, 5.6, or 19.6 mg/kg (ratio: 3.5) of sen- the absorption of GA from in situ loops under various condi-
noside A orally administrated with 5 mg/kg of GL to rats. tions was examined. The amount of remaining GA in the
Then, the plasma concentration profile of GA was investi- loops after co-administration of GA and sennosides or its de-
gated (Fig. 4A). The pharmacokinetic parameters are shown rivatives is shown in Fig. 5.
in Table 1.
Although the amount of remaining GA significantly in-
Although AUC(0—lim) and Cmax of GA decreased signifi-
cantly by co-administration with sennoside A compared to
Table 1. Pharmacokinetic Parameters of GA after Administration of GL to
Rats
AUC(0—lim)
ng·h/ml)
Cmax
tmax
Study
Dosing
(
(ng/ml)
(h)
In vivo study 1 GL (control)
GLꢄSA 1.6 mg/kg
2401ꢀ1020 312ꢀ89
1201ꢀ802* 175ꢀ92*
1014ꢀ700** 174ꢀ106*
742ꢀ429** 124ꢀ75**
3652ꢀ2131 310ꢀ240
8.2ꢀ2.0
7.7ꢀ2.9
7.2ꢀ2.1
6.7ꢀ2.0
7.0ꢀ5.9
5.2ꢀ3.6
—
GLꢄSA 5.6 mg/kg
GLꢄSA 19.6 mg/kg
In vivo study 2 Normal rat GL (i.c.)
Normal rat GLꢄSA (i.c.) 1400ꢀ880* 176ꢀ138
Ileostomized rat GL (p.o.)
—
—
GL: glycyrrhizinic acid, SA: sennoside A. i.c.: intracecal administration, p.o.: oral
administration. In vivo study 1: oral administration of GL (5 mg/kg) with or without
sennoside A to male rats (meanꢀS.D., nꢁ5—6) ∗ pꢅ0.05 vs. control, ∗∗ pꢅ0.01 vs.
control (Dunnet’s multiple comparison test). In vivo study 2: oral administration of GL
Fig. 3. Time Course of Diarrhea Appearance after Administration of Sen-
noside A (n=5)
(5 mg/kg) to male ileostomized rats and after intracecal administration of GL with or
Frequency of diarrhea appearance (%): total diarrhea score/maximum score
5)ꢃ100.
without sennoside A (1.6 mg/kg) to male rats (meanꢀS.D., nꢁ5—6) ∗ pꢅ0.05 vs. nor-
male rat (Student’s t-test).
(
Fig. 4. (A) Effect of Sennoside A on Plasma Concentration of GA after Oral Administration of GL with or without Sennoside A to Male Rats
(
(
meanꢀS.D., n=6): ꢀ, GL (5 mg/kg); ꢁ, GL (5 mg/kg)ꢄSennoside A (1.6 mg/kg); ꢂ, GL (5 mg/kg)ꢄSennoside A (5.6 mg/kg); ꢃ, GLꢄSennoside A
19.6 mg/kg) and (B) Plasma Concentration of Glycyrrhetic Acid after Oral Administration of GL to Male Ileostomized Rats (ꢃ) and after Intracecal Admin-
istration of GL with (ꢀ) or without (ꢁ) Sennoside A to Male Normal Rats (MeanꢀS.D., nꢁ5—6)

October 2005
1901
Fig. 5. Effect of Sennoside A, Sennidin A, Rhein-9-anthrone and Rhein
on GA Absorption from Rat in Situ Loop at Colon (MeanꢀS.D., nꢁ5)
Fig. 7. Regional Differences of the Effect of Rhein on GA Absorption
from Rat in Situ Loop (MeanꢀS.D., nꢁ5)
A: Control (GA 0.3 mg/ml), B: GA 0.3 mg/mlꢄrhein-9-anthrone 0.095 mg/ml, C:
GA 0.3 mg/mlꢄsennoside A 0.15 mg/ml, D: GA 0.3 mg/mlꢄsennidin A 0.095 mg/ml,
E: GA 0.3 mg/mlꢄrhein 0.1 mg/ml. ∗∗ pꢅ0.01 vs. control (Dunnett’s multiple compari-
son test) (meanꢀS.D., nꢁ5).
Control: GA 0.3 mg/ml. ∗∗ pꢅ0.01 vs. control, ∗ pꢅ0.05 vs. control (Student’s t-
test).
Fig. 8. The Metabolic Activity of GA in Rat Intestinal Bacteria after GL
(0.0823 mg/ml) Incubated with Sennoside A or Sennidin A (MeanꢀS.D.,
nꢁ5)
Fig. 6. Dose Dependency of Rhein (ꢄ), Sennoside A (ꢅ) and Sennidin A
ꢆ: without sonication, ꢂ: sonication. A: GL 0.0823 mg/ml (0.1 mmol/l), B: GL
.0823 mg/mlꢄsennoside A 0.0086 mg/ml (0.01 mmol/l), C: GL 0.0823 mg/mlꢄsenno-
side A 0.0863 mg/ml (0.1 mmol/l), D: GL 0.0823 mg/mlꢄsennidin A 0.0054 mg/ml
0.01 mmol/l), E: GL 0.0823 mg/mlꢄsennidin A 0.0538 mg/ml (0.1 mmol/l).
(
ꢃ) on GA Absorption from Rat in Situ Loop at Colon (MeanꢀS.D., nꢁ5)
0
GA: 0.3 mg/ml. Dashed line: prediction curve (Langmuir isotherm).
(
creased by co-administration with sennoside A, sennidin A
and rhein, it was not affected by the co-administration with testinal flora must be highly involved in the effectiveness of
rhein-9-anthrone.
the herbal medicines. In our clinical study of TJ-8117, we
When dosage dependency of sennoside A, sennidin A and observed a decrease in various pharmacokinetic parameters
rhein on the interaction was examined, the absorption (AUC(0—lim) and Cmax) of GA derived from GL in licorice
(ꢁdoseꢂthe amount of remaining GA) of GA decreased in a (Fig. 2). In this case, we first speculated that the cause of this
dosage-dependent manner in all compounds.
phenomenon might be due to diarrhea caused by rhubarb, be-
These inhibition activities were saturated at 75% in rhein, cause GA has been considered to be transformed from GL by
at 60% in sennoside A and at 25% in sennidin A. Figure 6 intestinal bacteria.
shows the observed difference of the maximum inhibition ac-
In this study, we discovered that the plasma concentration
tivity among each compound. Significant inhibition activity of GA in rats decreased by concomitant oral administration
by rhein was observed only in the ileum and colon (Fig. 7). of GL with sennoside A, one of the active principles of the
The Metabolism of GL by Intestinal Bacteria (in Vitro) purgative action of rhubarb. The frequency of diarrhea ap-
The metabolic activity of GL to GA incubated with or with- pearance after administration of sennoside A to rats was
out sennoside A or sennidin A is shown in Fig. 8. Without ul- dose-dependent from 2.5 to 10 mg/kg. However, when GL
trasonication treatment, the metabolic activity was constant was orally co-administrated to rats with a non-effective dose
in all treatments. In the experiments using ultrasonicated of sennoside A (1.6 mg/kg), the AUC(0—lim) and Cmax of GA
cecum content for the extraction of the internal enzymes con- decreased when compared to those treated with GL alone.
tained in the intestinal bacteria, the metabolic activity of GA Moreover, there was not a significant change of these phar-
increased to five times that seen without sonication. How- macokinetics parameters of GA among the co-administration
ever, no significant change was observed in each treatment groups. These results suggest that the onset of diarrhea does
group.
not direct cause lowering the pharmacokinetic parameters of
GA.
DISCUSSION
When GL was orally administrated to rats that had under-
gone ileostomy, GA was not detected in the plasma suggest-
Since herbal medicines are generally prepared by extrac- ing the involvement of intestinal flora in the hydrolysis of
tion with hot water, they contain many water-soluble compo- GL. Moreover, the AUC(0—lim) and Cmax of GA decreased
nents such as glycosides. The intestinal bacteria often partici- when GL was intracecally administrated to rats with senno-
pate in the metabolism and absorption of these glycosides. side A, when compared to those after administration of GL
Thus, the condition of the gastro-intestinal system and in- alone. These results indicate that the main absorption part of

1
902
Vol. 28, No. 10
GA is in the colon after transformation of GL to GA by in- CONCLUSION
testinal bacteria, and that GA absorption in the colon is af-
fected by sennoside A or its derivatives.
Based on this investigation, the decrease in GA absorption
In this study, the purgative action was observed after the in human subjects in the clinical study of TJ-8117 is strongly
administration of sennoside A to the rats. However, senno- suggested to be a direct interaction of GA absorption with
side A itself has been reported to have no purgative activity, sennosides and its derivatives. However, the mechanism of
with metabolites produced by intestinal bacteria causing the this interaction is still ambiguous. Thus, further studies to
23,24)
diarrhea.
Sennoside A is hydrolyzed by intestinal bacte- elucidate the mechanism are required.
ria, and is transformed into sennidin A. Furthermore, it is re-
duced to rhein-9-anthrone under anaerobic conditions in the REFERENCES
25—27)
intestine and is oxidized to rhein.
Rhein-9-anthrone has
1
)
Fugh-Berman A., Lancet, 355, 134—138 (2000).
2) Ruschitzka F., Meier P. J., Turina M., Luscher T. F., Noll G., Lancet,
55, 548—549 (2000).
Izzo A. A., Ernst E., Drugs, 61, 2163—2175 (2001).
been reported to be one of the main components related to
the purgative activity caused by rhubarb.
In order to elucidate the interaction of GL and sennosides
or its derivatives directly, we investigated their direct influ-
ence on the absorption of GA using an in situ loop. Inhibi-
tion of GA absorption was observed in rhein (approximately
28)
3
3)
4) Ioannides C., Xenobiotica, 32, 451—478 (2002).
5) Zhou S., Gao Y., Jiang W., Huang M., Xu A., Paxton J. W., Drug
Metab. Rev., 35, 35—98 (2003).
6
)
Lundahl J., Regardh C. G., Edgar B., Johnsson G., Eur. J. Clin. Phar-
macol., 52, 139—145 (1997).
75%), sennoside A (approximately 60%) and sennidin A (ap-
proximately 25%), whereas rhein-9-anthrone showed no in-
hibitory effect. The significant inhibitory activity by rhein
was observed in the ileum and colon suggesting the inhibi-
tion activity of rhein differs in parts. An in vitro study of the
rat intestinal bacteria confirmed that sennoside A and senni-
din A does not affect the transformation of GL to GA. There-
7
)
Fukuda K., Ohta T., Oshima Y., Ohashi N., Yoshikawa M., Yamazoe
Y., Pharmacogenetics, 7, 391—396 (1997).
8) Johne A., Brockmoller J., Bauer S., Maurer A., Langheinrich M.,
Roots I., Clin. Pharmacol. Ther., 66, 338—345 (1999).
Ishihara K., Kushida H., Yuzurihara M., Wakui Y., Yanagisawa T.,
Kamei H., Ohmori S., Kitada M., J. Pharm. Pharmacol., 52, 1023—
9)
1
029 (2000).
fore, these results indicate that the inhibition of GA absorp- 10) Barone G. W., Gurley B. J., Ketel B. L., Lightfoot M. L., Abul-Ezz S.
tion occurs at the absorption phase of GA from the intestine
by interaction with sennosides and its derivatives. However,
the fact that the inhibition activity of rhein differed in parts is
difficult to account for from these results.
R., Ann. Pharmacother., 34, 1013—1016 (2000).
1
1
1) Yue Q. T., Bergquist C., Gerden B., Lancet, 355, 576—577 (2000).
2) Durr D., Stieger B., Kullak-Ublick G. A., Rentsch K. M., Steinert H.
C., Meier P. J., Fattinger K., Clin. Pharmacol. Ther., 68, 598—604
(2000).
In this study, it was clarified that sennosides and its deriva- 13) Finney R. S. H., Somers G. F., J. Pharm. Pharmacol., 10, 613—620
(
1958).
tives inhibited the absorption of GA from the rat intestine.
We conclude that a decrease of the pharmacokinetic parame-
ters of GA in human plasma observed in the clinical study of
1
4) Kiso Y., Tohkin M., Hikino H., Hattori M., Sakamoto T., Namba T.,
Planta Med., 50, 298—302 (1984).
5) Kobashi K., Akao T., Bioscience Microflora, 16, 1—7 (1997).
1
TJ-8117 is attributable to the interactive action of absorption 16) Akao T., Kida H., Kanaoka M., Hattori M., Kobashi K., J. Pharm.
from the intestinal tract by anthraquinones contained in or
derived from rhubarb. It has been reported that involvement
of MCT1, monocarboxylic acid transporter, in the intestinal
absorption of weak organic acid such as compound having
Pharmacol., 50, 1155—1160 (1998).
1
7) Takeda S., Ishihara K., Wakui Y., Amagaya S., Maruno M., Akao T.,
Kobashi K., J. Pharm. Pharmacol., 48, 902—905 (1996)
8) Wang Z., Kurosaki Y., Nakayama T., Kimura T., Biol. Pharm. Bull.,
1
1
7, 1399—1403 (1994).
29—31)
carboxylic moiety.
Based on these reports, the MCT1 19) Akao T., Kobashi K., J. Germfree Life Gnotobiol., 25, 70—72 (1995).
transporter may participate in the absorption of both rhein 20) He J. X., Akao T., Nishino T., Tani T., Biol. Pharm. Bull., 24, 1395—
1
399 (2001).
and GA from intestine. Thus, competitive inhibition of rhein
and/or the other anthraquinones on GA transportation by
MCT1 should be concerned as one of the reason to decrease
2
2
1) De Witte P., Lemli J., J. Pharm. Pharmacol., 40, 652—655 (1988).
2) Kanaoka M., Yano S., Kato H., Nakano N., Kinoshita E., Chem.
Pharm. Bull., 31, 1866—1873 (1983).
GA absorption in human subjects. In addition, the induction 23) Fairbairn J. W., Pharm. Weekbl., 100, 1493—1499 (1965).
of the metabolic enzyme or P-glycoprotein in an intestinal 24) Dreessen M., Eyssen H., Lemli J., J. Pharm. Pharmacol., 33, 679—
6
81 (1981).
tract may be cause by lowing the absorption of GA. In order
to clarify this hypothesis, further experiments related to
MCT1/the other organic anion transporters, metabolic en-
zyme, and P-glycoprotein are required.
To our knowledge, this is the first report showing a phar-
macokinetic interaction between two herbs. In the case of our
clinical study of TJ-8117, since GA was a marker component
2
5) Hattori M., Namba T., Akao T., Kobashi K., Pharmacology, 36 (Suppl.
), 172—179 (1988).
6) Lemli J., Lemmens L., Pharmacology, 20 (Suppl. 1), 50—57 (1980).
1
2
27) Lemli J., Pharmacology, 36 (Suppl. 1), 126—128 (1988).
2
8) Sasaki K., Yamauchi K., Kuwano S., Planta Med., 37, 370—378
1979).
(
2
9) Tamai I., Takanaga H., Maeda H., Sai Y., Ogihara T., Higashida H.,
Tsuji A., Biochem. Biophys. Res. Commun., 214, 482—489 (1995).
of exposure, which might not directly relate to efficacy and 30) Takanaga H., Tamai I., Inaba S., Sai Y., Higashida H., Yamamoto H.,
safety, the observed decrease in the pharmacokinetic parame-
ters was not considered a serious problem. However, atten-
tion should be paid to herbs with a low therapeutic range
when they are administered with not only new drugs, but also
with other herbs.
Tsuji A., Biochem. Biophys. Res. Commun., 217, 370—377 (1995).
1) Tamai I., Sai Y., Ono A., Kido Y., Yabuuchi H., Takanaga H., Satoh E.,
Ogihara T., Amano O., Izeki S., Tsuji A., J. Pharm. Pharmacol., 51,
3
1
113—1121 (1999).