A Deglucosylated Metabolite of Paeoniflorin of the Root of Paeonia lactiflora and its Pharmacokinetics in Rats

Article abstract of DOI:10.1055/s-2003-45192

Paeoniflorin is a bioactive monoterpene glucoside in Paeoniae Radix (PR), the roots of Paeonia lactiflora (Ranunculaceae). By oral administration to rats with the decoction of PR, the metabolism and pharmacokinetics of paeoniflorin were investigated in this study. A deglucosylated metabolite of paeoniflorin, paeoniflorgenin (PG), in serum was identified based on HPLC/MS and NMR spectral data. HPLC/UV methods were developed for determining PG in serum and feces suspension. A non-compartment model was used for the calculation of pharmacokinetic parameters. Moreover, the metabolism of paeoniflorin by various types of feces was investigated as well. The paeoniflorin levels in serum were below the detection limit throughout the study. The C_max, t _max, and AUC_0-t of PG were 8.0 μg/mL, 10 min and 487.0 μg min/mL, respectively. Paeoniflorin was found to be hydrolyzed into PG through incubation with feces of rabbit, rat, pig or human. Similar profiles of PG were shown for various types of feces, except for rabbit. In conclusion, paeoniflorin was not absorbed per se, whereas its aglycone paeoniflorgenin was absorbable and circulating in the bloodstream. Rat and pig are appropriate models for investigating the metabolism and pharmacokinetics of paeoniflorin.

Full text of DOI:10.1055/s-2003-45192

Su-Lan Hsiu1
Ya-Tze Lin2,3
Kuo-Ching Wen4
Yu-Chi Hou5
Pei-Dawn Lee Chao1
A Deglucosylated Metabolite of Paeoniflorin of the
Root of Paeonia lactiflora and its Pharmacokinetics
in Rats
Abstract
The Cmax, tmax, and AUC0-t of PG were 8.0 mg/mL, 10 min and
487.0 mg min/mL, respectively. Paeoniflorin was found to be hy-
Paeoniflorin is a bioactive monoterpene glucoside in Paeoniae drolyzed into PG through incubation with feces of rabbit, rat,
Radix ꢀPR), the roots of Paeonia lactiflora ꢀRanunculaceae). By pig or human. Similar profiles of PG were shown for various
oral administration to rats with the decoction of PR, the meta- types of feces, except for rabbit. In conclusion, paeoniflorin was
bolism and pharmacokinetics of paeoniflorin were investigated not absorbed per se, whereas its aglycone paeoniflorgenin was
in this study. A deglucosylated metabolite of paeoniflorin, paeo- absorbable and circulating in the bloodstream. Rat and pig are
niflorgenin ꢀPG), in serum was identified based on HPLC/MS and appropriate models for investigating the metabolism and phar-
NMR spectral data. HPLC/UV methods were developed for deter- macokinetics of paeoniflorin.
mining PG in serum and feces suspension. A non-compartment
model was used for the calculation of pharmacokinetic para- Key words
meters. Moreover, the metabolism of paeoniflorin by various Paeonia lactiflora ´ Ranunculaceae ´ Paeoniflorin ´ Paeoniflor-
types of feces was investigated as well. The paeoniflorin levels genin ´ pharmacokinetics
in serum were below the detection limit throughout the study.
1113
Introduction
Previous studies also reported extremely poor absorption and
low bioavailability of orally administered paeoniflorin [10], [11],
Paeoniflorin ꢀFig.1) is a bioactive monoterpene glucoside in which was readily metabolized into metabolites named paeoni-
Paeoniae Radix ꢀPR), the roots of Paeonia lactiflora Pall ꢀRanuncu- metabolins I and II ꢀFig.1) by human intestinal bacteria in vitro
laceae), which has been widely used in clinical Chinese medicine. [12], [13]. Heikal et. al. found that blood concentrations of paeo-
Paeoniflorin possesses several pharmacological activities, such nimetabolin I after oral administration of paeoniflorin were rela-
as analgesia [1], anti-inflammation, anti-allergy [2], muscle re- tively higher than that of paeoniflorin [14], [15]. Employing var-
laxation [3], cognitive enhancement [4], and inhibition of steroid ious types of feces, e.g. rabbit, rat, pig and human in this study,
protein binding [5]. In addition, anti-hyperglycemia [6], anti-hy- we successfully identified a new metabolite of paeoniflorin and
potension [7], activation of adenosine A-1 receptor [8] and pro- the pharmacokinetics of this metabolite were then investigated
longation of thrombosis time [9] activities have been reported in rats.
in recent studies.
Affiliation
1 Department of Pharmacy, China Medical University, Taichung, R.O.C.
2 Graduate Institute of Pharmaceutical Chemistry, China Medical University, Taichung, R.O.C.
3 National Laboratories of Foods and Drugs, Department of Health, Executive Yuan, R.O.C.
4 Department of Cosmeceutics, China Medical University, Taichung, Taiwan, R.O.C.
5 School of Chinese Medicine, China Medical University, Taichung, Taiwan, R.O.C.
Correspondence
Prof. Pei-Dawn Lee Chao ´ Department of Pharmacy ´ China Medical University ´ Taichung ´ Taiwan ´ R.O.C. ´
Fax: +886-4-22031028 ´ E-mail: pdlee@mail.cmu.edu.tw
Received May 19, 2003 ´ Accepted October 3, 2003
Bibliography
Planta Med 2003; 69: 1113±1118 ´ ꢁ Georg Thieme Verlag Stuttgart ´ New York ´ ISSN 0032-0943

Fig. 1 Structures of ꢀA): paeoniflorin; ꢀB)
paeoniflorgenin; ꢀC): 7S- or 7R-paeonimeta-
bolin I and ꢀD): paeonimetabolin II.
Materials and Methods
Chemicals
LC/MS conditions
A column of LiChrospherꢂ 100, RP-18e, 5 mm, 250”4.6 mm was
employed. Mobile phase consisted of acetonitrile and 0.1% phos-
Paeoniflorin ꢀpurity 98%) was purchased from Nacalai Tesque, Inc. phoric acid ꢀ15:85, v/v). Capillary voltage: 3 kV; cone voltage: 80
ꢀKyoto, Japan). Methylparaben and b-glucosidase ꢀHP-2, from V; collision energy: 25 eV; source temperature: 1008C; desolva-
Helix pomotia) were purchased from Sigma Chemical Co. ꢀSt Louis, tion temperature: 2508C. The UV detector was set at 230 nm and
MO, U.S.A.). Potassium dihydrogen phosphate was obtained from the flow rate was 0.5 mL/min.
Wako ꢀOsaka, Japan). Sodium hydroxide was purchased from
Merck ꢀDarmstadt, Germany). All other chemicals and solvents Preparation of PR decoction and quantitation
used were of analytical grade or HPLC quality. Milli-Q plus water of paeoniflorin
ꢀMillipore, Bedford, MA, U.S.A.) was used for all preparations.
Seventy-five grams of PR were soaked in 1.5 L water for 30 min,
and then boiled for longer than 2 h to a reduced volume of about
50 mL. After filtration while hot, sufficient water was added to 50
1114
Instrumentation
A Bruker Advance DPX-200 FT-NMR spectrometer was used. The mL. This solution was stored at ±308C before use.
LC/MS spectrometer used was a Micromass Quattro Ultima
equipped with a Waters 2690 Alliance LC and 996 PDA automatic Paeoniflorin was accurately weighed and dissolved in methanol
liquid sampler and injector. The HPLC/UV system consisted of a to afford a series concentrations within the range of 25.0 ~ 500.0
Hitachi Model L-6200 intelligent pump and an L-3000 photo mg/mL. An appropriate amount of internal standard solution of
diode array detector, equipped with a Shimadzu SIL-9A autosam- methylparaben in methanol was added to each solution to afford
pler. A Cosmosil 5C18-AR II column, 5 mm, 150”4.6 mm with a a final concentration of 103.0 mg/mL. A calibration curve was
guard column ꢀMetaGuard 4.6 mm Polaris 5 m C18-A, MetaChem, plotted after linear regression of the peak area ratios ꢀpaeoniflor-
Torrance) was employed.
in to internal standard) against the concentrations of paeoniflor-
in.
Preparation of paeoniflorgenin ꢀPG) from paeoniflorin
Paeoniflorin ꢀ50.0 mg) was dissolved in 5 drops of dimethyl sulf- 0.5 mL of PR decoction was added with a suitable amount of
oxide, to which 2.5 mL of b-glucosidase solution ꢀ15.0 mg/ml, 14 methylparaben and sufficient 70% methanol to make 5 mL to af-
units/mg) was added with shaking in a water bath ꢀ378C, 100 ford a final concentration of 103.0 mg/mL methylparaben. The so-
rpm). The reaction was allowed to continue for 5 h, then the mix- lutions were filtered through a 0.45 mm syringe filter before HPLC
ture was partitioned two times with ethyl acetate. The ethyl analysis. The mobile phase consisted of acetonitrile and 0.1%
acetate layers were combined, and evaporated to dryness under phosphoric acid ꢀ15:85, v/v). The UV detector was set at 230
vacuum. The residue was recrystallized from methanol to afford nm and the flow rate was 1.0 mL/min.
PG; [a]D20: ±378 ꢀc 1.0 mg/mL, MeOH); UV ꢀMeOH): lmax = 203
ꢀlog e 4.25), 227 ꢀlog e 4.33) nm. LC/MS: m/z: = 317 ꢀM ± 1, ESI±); Acid hydrolysis of PR decoction
13C-NMR ꢀCD3OD, 50 MHz): d = 100.9 ꢀC1), 52.2 ꢀC2), 111.5 ꢀC4), 1.0 mL of PR decoction was diluted with 9 mL methanol. To 1.5
78.3 ꢀC6), 35.7 ꢀC7), 60.0 ꢀC8), 101.4 ꢀC9), 20.5 ꢀC10), 130.2 ꢀC1¢), mL of the methanolic solution, 0.5 mL of 2.4 N HCl was added
1
129.1 ꢀC2¢, C6¢), 127.9 ꢀC3¢, C5¢), 132.5 ꢀC4¢), 168.4 ꢀC7¢); H-NMR and the mixture incubated in a water bath at 808C for 2 h. The
ꢀCD3OD, 200 MHz): d = 8.05 ꢀ2H, d, J = 5.9 Hz, H-2¢, H-6¢), 7.63 solution was filtered through a 0.45 mm syringe filter before
ꢀ1H, m, H-4¢), 7.51 ꢀ2H, m, H-3¢, H-5¢), 5.41 ꢀ1H, s, H-9), 1.26 ꢀ3H, HPLC analysis. The analysis conditions are identical to those for
s, H-10).
paeoniflorin.
Hsiu S-L et al. A Deglucosylated Metabolite¼ Planta Med 2003; 69: 1113±1118

Animals and drug administration
Fermentation of paeoniflorin and quantitation of PGin feces
Six Sprague-Dawley rats, weighing 250 ~ 350 g, were housed in a The fresh feces of New Zealand white rabbits and Sprague-Daw-
12-h light-dark, constant temperature environment prior to ley rats were collected from the animal center of China Medical
study. All rats were fasted 1 day before drug administration and University. Yorkshire pig feces were obtained from Animal Tech-
the fasting was continued for 4 h thereafter. Water was supplied nology Institute, Miaoli, Taiwan. Human feces was obtained from
ad libitum. The PR decoction was given orally at a dose of 110 mg a healthy child. Twenty grams of feces were accurately weighed,
kg-1 paeoniflorin. Drug administration was carried out via gastric and 60 mL artificial intestinal juice ꢀpH 7.5) were added. The
gavage. The animal study adhered to ªThe Guidebook for the Care mixture became a homogenate by using an agitator, and the
and Use of Laboratory Animals ꢀ2002)º ꢀPublished by The Chi- homogenate was filtered through gauze.
nese Society for the Laboratory Animal Science, Taiwan, R.O.C.).
Feces suspensions of rabbit, rat, pig and human ꢀ12.42 mL) were
Blood samples ꢀ0.5 mL) were withdrawn via cardiopuncture at 5, spiked with paeoniflorin standard solutions ꢀ1.38 mL, 1.0 mg/mL)
10, 30, 60, 180, 300, 420 and 540 min after drug administration. and mixed well with stirring in beakers individually. Each aliquot
All blood samples were centrifuged at 9,860 g for 15 min and the of 600 mL was placed in a dark brown glass tube, sealed with a
serum stored at ±308C until analysis.
septum and air was removed with syringe. All samples were
prepared in triplicate and kept on ice before incubation. The
tubes were then incubated in a shaking water bath ꢀ100 rpm) at
37 8C for 0, 1, 2, 4, 8, 12 and 24 h, and then samples were stored at
Calibration curve for quantitation of PGin serum and method
validation
20 mL of PG standard solution were spiked into 180 mL serum to ±308C until analysis.
afford serum standards with concentrations of 0.20, 0.39, 0.78,
1.56, 3.12, 6.25 and 12.50 mg mL±1. To 200 mL of serum standard, Feces samples were added with 600 mL ethyl acetate ꢀcontaining
800 mL of methanol containing 1.0 mg mL±1 of methylparaben as 25.0 mg/mL methylparaben as the internal standard), and vor-
internal standard were added for deproteinization. The mixture texed for 10 sec. The mixture was centrifuged at 9,860 g for 15
was vortexed for 10 sec and centrifuged at 9,860 g for 15 min, min and the supernatant was blown with N2 gas until dry. The
the supernatant was removed and evaporated to dryness by residue was reconstituted with 50 mL methanol, and 20 mL was
blowing nitrogen. The residue was reconstituted with 50 mL me- subjected to HPLC analysis. The HPLC system was identical to
thanol, of which 20 mL were subjected to HPLC analysis. The cali- that used for PR decoction as described above. Metabolic profiles
bration curve was drawn by linear regression of the peak-area ra- were drawn by plotting the peak area ratios ꢀPG to internal
tios ꢀPG to methylparaben) against concentrations of PG.
standard) against incubation time of paeoniflorin with various
types of feces.
The precision and accuracy of the analysis method was evaluated
by intra-day and inter-day analysis of triplicate serum standards
within one day and over a period of three days. By spiking PG into Results
blank serum and water in triplicates to afford concentrations of
1115
0.78, 3.12 and 12.50 mg mL±1, respectively, a recovery study was Fig. 2 shows the chromatogram of PR decoction, demonstrating
done to further assess the accuracy of this method. The concen- that paeoniflorin and the internal standard were well resolved
trations obtained in blank serum to the corresponding ones in within half an hour without any interference. A good linear rela-
water were compared. The LLOQ ꢀLower Limit of Quantitation) tionship existed for paeoniflorin over the concentration range of
represents the lowest concentration of analyte that can be deter- 25.0 ~ 500.0 mg/mL. The content of paeoniflorin in the decoction
mined with acceptable precision and accuracy, whereas the LOD was 8.1 mg from each gram of PR. Acid hydrolysis of the PR de-
ꢀLimit of Detection) represents the lowest concentration of ana- coction resulted in the transformation of paeoniflorin to PG
lyte that can be detected ꢀwith S/N > 3).
which existed only in very low concentrations before hydrolysis.
Quantitation of PGin serum
A good linear relationship existed over the concentration range
200 mL of serum sample were added with 1,800 mL of methanol of 0.20 ~ 12.50 mg/mL for PG in serum. The coefficients of varia-
containing 1.0 mg mL±1 of methylparaben as internal standard. tion were below 6.0% and the relative errors were between + 12.3
The remaining procedures followed those for calibrators. The and ±4.1% for intra-day and inter-day assays. These indicate that
mobile phase consisted of acetonitrile and 0.1% phosphoric acid the precision and accuracy were satisfactory. The LLOQ and LOD
ꢀ25:75, v/v). The UV detector was set at 230 nm and the flow rate of PG were 0.20 mg/mL and 0.10 mg/mL, respectively. The recover-
was 1.0 mL/min.
ies of PG from serum were 82.5, 81.9 and 88.7% at concentrations
of 0.78, 3.12 and 12.50 mg/mL, respectively. Throughout this
study, only PG was determined in serum, whereas paeoniflorin
Data analysis
The pharmacokinetic parameters of paeoniflorgenin were calcu- was not detected with a photo diode array detector. Fig. 3 depicts
lated using a non-compartment model with the aid of WINNON- the profile of mean serum concentrations of PG in rats after ad-
LIN ꢀversion 1.1, SCI software, Statistical Consulting, Inc., Apex, ministration of PR decoction, and the pharmacokinetic param-
NC). The peak plasma concentration ꢀCmax) and the time to peak eters of PG are listed in Table 1.
concentration ꢀTmax) were obtained from experimental ob-
servations. The area under the serum concentration-time curve A high-performance liquid chromatographic method was devel-
ꢀAUC0-t) was calculated using the trapezoidal rule to the last oped for determining paeoniflorgenin in rabbit, rat, pig and hu-
point.
man feces suspensions. The respective retention times of PG
Hsiu S-L et al. A Deglucosylated Metabolite¼ Planta Med 2003; 69: 1113±1118

[13], [14], [15]. The 13C-NMR spectra of PG did not reveal the ab-
sorptions of all carbons, probably due to the fact that C3 and C5
were hidden under the solvent peak at d = 46.1 ~ 48.7. The LC/
MS result shows the molecular weight of PG is 318, correspond-
ing to the deglucosylated derivative of paeoniflorin whose mole-
cular weight is 480.
Each rat received the PR decoction at a dose of 13.5 g/kg of crude
drug containing 110 mg/kg of paeoniflorin, that was similar to
the usual daily dose of 15 g for clinical use. After oral administra-
tion of PR decoction to rats, a metabolite with identical retention
time and UV spectrum with PG was detected in the serum of all
rats. Accordingly, this compound can be inferred as the degluco-
sylated metabolite of paeoniflorin. Moreover, acid hydrolysis of
PR decoction yielded a new component that also showed identi-
cal retention time and UV spectrum with PG, that gave more
proof to the finding. This fact further confirmed that hydrolysis
was a major metabolic fate of paeoniflorin. No conjugation me-
tabolism was found for PG, that was in agreement with the bio-
logical fate of glycyrrhetic acid, a triterpene aglycone of glycyr-
rhizin [18]. In contrast to these alcoholic aglycones, phenolic
Fig. 2 HPLC chromatogram of Paeoniae Radix decoction, 1: paeoni-
florin; 2: methylparaben ꢀinternal standard).
and methylparaben ꢀI.S.) were 21.1 and 24.6 min as shown in aglycones were generally further metabolized to conjugated me-
Fig. 4. Fig. 5 depicts the metabolic profiles of paeoniflorin to tabolites, suggesting that the activities of phase II metabolic en-
form PG during incubation with rabbit, rat, pig and human feces. zymes toward phenols were significantly higher than alcohols
At the beginning of fermentation, PG was not detected in various [19].
types of feces. One h after incubation, PG emerged in the feces
of rabbit, rat and humans, but was still missing in pig feces. At Heikal et. al. [15] reported pharmacokinetic parameters of paeo-
8 h, PG almost remained at a constant concentration in various niflorin and paeonimetabolin I after oral administration of 5 mg/
feces.
kg paeoniflorin to rats, the Cmax and AUC of paeoniflorin and
paeonimetabolin I were 20.3, 101.7 ng/mL and 1,174, 12,358 ng
min/mL, respectively. However, the parent compound paeoni-
florin was not detected in rat serum in the present study. That
the EIA method was more sensitive than the UV method [20]
Discussion
1116
The aglycone of paeoniflorin PG was prepared by hydrolysis of could be an explanation for this discrepancy. In regard to the
paeoniflorin with b-glucosidase and its structure could be deter- other metabolites reported in the above literature, e.g., paeoni-
mined by comparing the NMR spectral data with those of paeo- metabolins I and II, they were not detected in this study due to
niflorin [17]. The 1H-NMR and 13C-NMR spectra of PG revealed the absence of the benzene ring chromophore in their structure
the presence of a benzene ring moiety, indicating that PG was to absorb UV.
different from those metabolites reported previously [11], [12],
Our study demonstrated that paeoniflorin was readily metabo-
lized into PG and Tmax was the first sampling time at 10 min after
drug administration. The Cmax of PG was 8.0 mg/mL and its elimi-
nation half-life was 126 min. As compared to the pharmacoki-
Table 1 Mean ꢀ S.E.) pharmacokinetic parameters of paeoniflorgenin
after oral administration of Paeoniae Radix decoction in 6 rats
Parameters
Mean  S.E.
Tmaxa ꢀmin)
Cmaxb ꢀmg/mL)
T1/2c ꢀmin)
AUCd ꢀmg min/mL)
MRTe ꢀmin)
10.0  4.1
8.0  0.6
126.2  4.0
487.0  95.6
135.5  2.4
a Time of peak serum level.
b Concentration of peak serum level.
c Elimination half life.
d Area under serum concentration-time curve to the last point.
e Mean residence time.
Fig. 3 Mean ꢀ S.E.) serum concentration-time profile of paeoniflor-
genin after oral administration of Paeoniae Radix decoction to six rats.
Hsiu S-L et al. A Deglucosylated Metabolite¼ Planta Med 2003; 69: 1113±1118

Fig. 4 HPLC chromatograms of ꢀA): blank human
feces; ꢀB): human feces with Paeoniae Radix decoc-
tion. 1: paeoniflorgenin; 2: methylparaben ꢀinternal
standard).
netics of paeoniflorin and paeonimetabolin I reported in the pre- identical with that in serum after oral administration of PR de-
vious study [15], PG showed a much higher Cmax and AUC based coction by HPLC/UV analysis and was accordingly identified as
on dose correction. In spite of being a major metabolite of paeo- PG. The profiles indicated that PG was not further degraded by
niflorin, to our great surprise PG has never been reported before fecal flora. This phenomenon is very different from flavonoid
in the literature. The bioactivity investigations on PG by in vitro aglycones which were markedly degraded by fecal flora [21],
models could partially account for the in vivo efficacy of paeoni- suggesting that terpenes are more resistant to enterobacterial
florin or PR decoction. However, the information on PG bioactiv- degradation than heterocyclic aromatic compounds. The meta-
ity is still lacking in the literature and is worthy of more investi- bolic pattern of paeoniflorin was similar among rats, pigs and hu-
gations.
mans, whereas rabbit feces hydrolyzed paeoniflorin to a lesser
extent.
1117
The metabolic profiles of paeoniflorin in various types of feces
showed that a common metabolite was gradually generated in In summary, PG is a major metabolite of paeoniflorin in vivo. It is
the presence of fecal bacteria. This metabolite was found to be suggested that rat and pig are appropriate animal models for the
pharmacokinetic study of paeoniflorin or PR decoction.
References
1
Sugishita E, Amagaya S, Ogihara Y. Studies on the combination of Glycyr-
rhizae Radix in Shakuyakukanzo-To. J Pharmacobiodyn 1984; 7: 427±35
2 Yamahara J, Yamada T, Kimura H, Sawada T, Fujimura H. Biologically
active principles of crude drugs. II. Anti-allergic principles in ªShoseir-
yu-Toº anti-inflammatory properties of paeoniflorin and its deriva-
tives. J Pharmacobiodyn 1982; 5: 921±9
3 Kimura M, Kimura I, Nojima H. Depolarizing neuromuscular blocking
action induced by electropharmacological coupling in the combined ef-
fect of paeoniflorin and glycyrrhizin. Jpn J Pharmacol 1985; 37: 395±9
4 Watanabe H. Candidates for cognitive enhancer extracted from med-
icinal plants: paeoniflorin and tetramethylpyrazine. Behav Brain Res
1997; 83: 135±41
5 Tamaya T, Sato S, Okada HH. Possible mechanism of steroid action of
the plant herb extracts glycyrrhizin, glycyrrhetinic acid, and paeoni-
florin: inhibition by plant herb extracts of steroid protein binding in
the rabbits. Am J Obstet Gynecol 1986; 155: 1134±9
6 Hsu FL, Lai CW, Cheng JT. Antihyperglycemic effects of paeoniflorin
and 8-debenzoylpaeoniflorin, glucosides from the root of Paeonia
lactiflora. Planta Med 1997; 63: 323±5
7 Cheng JT, Wang JC, Hsu FL. Paeoniflorin reverses guanethidine-in-
duced hypotension via activation of central adenosine A1 receptors
in Wistar rats. Clin Exp Pharmacol Physiol 1999; 26: 815±6
Fig. 5 Mean ꢀ S.D.) area ratio-time profiles of paeoniflorgenin dur-
ing incubation of paeoniflorin in rabbit, rat, pig and human feces
ꢀn = 3).
Hsiu S-L et al. A Deglucosylated Metabolite¼ Planta Med 2003; 69: 1113±1118

15
16
8 Lai CW, Hsu FL, Cheng JT. Stimulatory effect of paeoniflorin on adeno-
sine A-1 receptors to increase the translocation of protein kinase C
ꢀPKC) and glucose transporter ꢀGLUT 4) in isolate rat white adipocytes.
Life Sci 1999; 62: 1591±5
Heikal OA, Akao T, Takeda S, Hattori M. Pharmacokinetic study of
paeonimetabolin I, a major metabolite of paeoniflorin from Paeony
roots. Biol Pharm Bull 1997; 20: 517±21
Shu YZ, Hattori M, Namba T, Mibu K, Akao T, Kobashi KJ. Metabolic ac-
tivation of components of oriental medicines by human intestinal
flora. on paeoniflorin and albiflorin, constituents of paeony roots.
Pharmacobio-Dyn 1987; 10: s±58
Yamasaki K, Kaneda M, Tanaka O. Carbon-13 NMR spectral assign-
ments of paeoniflorin homologues with the aid of spin-lattice relaxa-
tion time. Tetrahedron Lett. 1976; 44: 3965±8
Ching H., Hsiu SL, Hou YC, Chen CC, Chao PDL. Comparison of pharma-
cokinetics between glycyrrhizin and glycyrrhetic acid in rabbits. J
Food Drug Anal 2001; 9: 67±71
9 Ye J, Duan H, Yang X, Yan W, Zheng X. Anti-thrombosis effect of paeo-
niflorin: Evaluated in a photochemical reaction thrombosis model in
vivo. Planta Med 2001; 67: 766±7
10
17
18
19
Takeda S, Isono T, Wakui Y, Matsuzaki Y, Sasaki H, Amagaya S, Maruno
M. Absorption and excretion of paeoniflorin in rats. J Pharm Pharma-
col 1995; 47: 1036±40
Takeda S, Isono T, Wakui Y, Mizuhara Y, Amagaya S, Maruno M, Hattori
M. In-vivo assessment of extrahepatic metabolism of paeoniflorin in
rats: relevance to intestinal flora metabolism. J Pharm Pharmacol
1997; 49: 35±9
Hattori M, Shu YZ, Shimizu M, Hayashi T, Morita N, Kobashi K, Xu G,
Namba T. Metabolism of paeoniflorin and related compounds by hu-
man intestinal bacteria. Chem Pharm Bull 1985; 33: 3838±46
Shu YZ, Hattori M, Akao T, Kobashi K, Kagi K, Fukuyama K, Tsukihara T,
Namba T. Metabolism of paeoniflorin and related compounds by hu-
man intestinal bacteria. II. Structures of 7S- and 7R-paeonimetabo-
lines I and II formed by Bacteroides fragilis and Lactobacillus brevis.
Chem Pharm Bull 1987; 35: 3726±33
11
Chao PDL, Hsiu SL, Hou YC. Flavonoids in Herbs: biological fates and
potential interactions with xenobiotics. J Food Drug Anal. 2002; 11:
35±45
12
13
20 Hattori M, Yang X, Shu Y, Heikal OA, Miyashiro H, Kanaoka M, Akao T,
Kobashi K, Namba T. Enzyme immunoassay for paeonimetabolin I, a
major metabolite of paeoniflorin by intestinal bacteria. J Trad Med
1996; 13: 73±80
21
Lin YT, Hsiu SL, Hou YC, Chen HY, Chao PDL. Degradation of flavonoid
aglycones by rabbit, rat and human fecal flora. Biol Pharm Bull 2003;
26: 747±51
14
Heikal OA, Miyashiro H, Akao T, Hattori M. Quantitative determination
of paeoniflorin and its major metabolites, paeonimetabolin I, in the rat
plasma by enzyme immunoassay. J Trad Med 1997; 14: 34±40
1118
Hsiu S-L et al. A Deglucosylated Metabolite¼ Planta Med 2003; 69: 1113±1118