Stereoselective oxidation metabolism of 20(S)-protopanaxatriol in human liver microsomes and in rats

Article abstract of DOI:10.3109/00498254.2014.986562

1. In this study, the oxidative metabolites of 20(S)-protopanaxatriol (PPT) were identified in human liver microsomes (HLMs) and in rats using liquid chromatography-electrospray ionization tandem mass spectrometry. 2. Twelve oxidative metabolites were found in HLM, eight of which have not been previously reported. Twenty-four oxidative metabolites were found in rat feces after oral administration and 20 of these, including six found in HLM, were first reported. The results indicated PPT was more extensively metabolized in rats than in HLM. C20-24 epoxides, a pair of epimers (namely, M1-1 and M1-2) were the major metabolites, and other metabolites were derived from their further metabolism. 3. Enzyme kinetics experiments showed that the apparent formation Vmax of M1-1 was 10.4 folds and 2.4 folds higher than that of M1-2 in HLM and in rat liver microsomes (RLMs), respectively. The depletion rate of M1-2 was 11.0 folds faster than M1-1 in HLM, and was similar in RLM. Hence, the remarkable species differences of PPT metabolism mainly resulted from the stereoselective formation and further metabolic elimination of M1-1 and M1-2. 4. Chemical inhibition study and recombinant human P450 isoforms analysis showed that CYP3A4 was the predominant isoform involved in the oxidative metabolism of M1-1 and M1-2.

Full text of DOI:10.3109/00498254.2014.986562

http://informahealthcare.com/xen
ISSN: 0049-8254 (print), 1366-5928 (electronic)
Xenobiotica, Early Online: 1–11
2014 Informa UK Ltd. DOI: 10.3109/00498254.2014.986562
!
RESEARCH ARTICLE
Stereoselective oxidation metabolism of 20(S)-protopanaxatriol
in human liver microsomes and in rats
Wenyan Wang, Yingying Ni, Li Wang, Xin Che, Wanhui Liu, and Qingguo Meng
School of Pharmacy, Yantai University, Yantai, China
Abstract
Keywords
1
. In this study, the oxidative metabolites of 20(S)-protopanaxatriol (PPT) were identified in 20(S)-Protopanaxatriol, enzyme kinetics,
human liver microsomes (HLMs) and in rats using liquid chromatography-electrospray
ionization tandem mass spectrometry.
liver microsomes, metabolite identification,
P450 isoforms, stereoselectivity
2
. Twelve oxidative metabolites were found in HLM, eight of which have not been previously
reported. Twenty-four oxidative metabolites were found in rat feces after oral administration
and 20 of these, including six found in HLM, were first reported. The results indicated PPT
was more extensively metabolized in rats than in HLM. C20–24 epoxides, a pair of epimers
History
Received 18 September 2014
Revised 7 November 2014
Accepted 7 November 2014
Published online 28 November 2014
(namely, M1-1 and M1-2) were the major metabolites, and other metabolites were derived
from their further metabolism.
3
. Enzyme kinetics experiments showed that the apparent formation Vmax of M1-1 was 10.4
folds and 2.4 folds higher than that of M1-2 in HLM and in rat liver microsomes (RLMs),
respectively. The depletion rate of M1-2 was 11.0 folds faster than M1-1 in HLM, and was
similar in RLM. Hence, the remarkable species differences of PPT metabolism mainly resulted
from the stereoselective formation and further metabolic elimination of M1-1 and M1-2.
. Chemical inhibition study and recombinant human P450 isoforms analysis showed
that CYP3A4 was the predominant isoform involved in the oxidative metabolism of M1-1
and M1-2.
4
Introduction
However, the absolute bioavailability of the PPT was
estimated as 3.69% (Kong et al., 2013), which might result
from its extensive metabolism in the liver. The study of the
metabolic pathway of PPT has become very necessary to
explore pharmacological potency and toxicology. Utilizing
the NMR or MS analysis, many studies reported the microbial
transformation products of PPT in vitro (Chen et al., 2013;
Tian et al., 2005; Zhang et al., 2007) to investigate a possible
biotransformation pathway in vivo. Kasai et al. (2000) have
reported that PPT was metabolized to its C20–24 epoxides, by
liver microsomes of rats in vitro. He et al. (2014) identified
Gensenosides content was 2–5% in ginseng roots and are
known as the main active components of ginseng, having
bioactive and pharmacological activities (Liu et al., 2010;
Shibata, 2001). Furthermore, it has been reported that
deglycosylation ginsenosides are more readily absorbed into
the bloodstream and act as active components (Hasegawa
et al., 1996; Lai et al., 2009; Sun et al., 2005; Tawab et al.,
2
003). 20(S)-protopanaxatriol (PPT; Figure 1) is a deglyco-
sylation aglycone of the ginsenosides (such as Re, Rf, Rg1,
Rg2 and Rh1) catalyzed by gastric acid and intestinal bacteria
in vivo (Hasegawa, 2004; Liu et al., 2010). A number of
studies have shown that PPT exhibits many pharmacological
activities, including chemoprevention of cancer or inflamma-
tory diseases (Oh et al., 2004), affecting the production of
cortisol (Hasegawa et al., 2013) and anti-diabetic effect (Han
et al., 2006), especially the antitumor effects (Chen et al.,
22 metabolites of PPT in rats after oral administration,
including its C20–24 epoxides. Wang et al. (2010) identified
four metabolites of PPT after oral administration in mice,
which were completely different from the metabolites in rats.
Furthermore, detecting and characterizing reactive metab-
olites in both experimental animals and humans is a highly
important initial step to evaluate timely the hidden risks
2
2
013; Hasegawa et al., 2002; Wang et al., 2009; Usami et al.,
008).
(
Evans & Baillie, 2005), which can help to explore the species
differences and extrapolate the results to humans. Han et al.
2011) reported PPT and one of its C20–24 epoxides exerted
(
cardioprotective effects against myocardial ischemic injury,
but another isoform had no effect. However, to the best of our
knowledge, the metabolism information on PPT in humans is
scant. In this study, the metabolism of PPT in mixed human
Address for correspondence: Wenyan Wang, School of Pharmacy, Yantai
University, No 30 Qingquan Road, Laishan District, Yantai 264005,
China. Tel: +86 535 6706030-8015. Fax: +86 535 6706066. E-mail:
ytuwwy@163.com; Qingguo Meng, E-mail: qinggmeng@163.com

2
W. Wang et al.
Xenobiotica, Early Online: 1–11
2
Figure 1. Chemical structure of PPT and its full scan mass spectrum (A) and MS spectrum (B) in positive ionization modes from ESI source.
liver microsomes (HLMs) was examined. Meanwhile, the
metabolites of PPT in rats after an oral administration were
also identified to explore the species differences. The
objectives of this study were as follows: (1) to characterize
the structures of oxidative metabolites of PPT and compare
the species differences; (2) to study the stereoselective
formation and metabolic elimination of 20,24-epoxide epi-
meric metabolites of PPT; (3) to identify the cytochrome
P450 (P450) enzymes responsible for the biotransformation of
the epimeric metabolites.
samples without NADPH were included; each incubation was
performed in duplicate. The incubations of the metabolites
M1-1 (25 mM) and M1-2 (25 mM) in pooled HLM were
carried out as described above for PPT, respectively.
Synthesis and identification of M1-1 and M1-2
m-CPBA (50 mg, 320 mmol) was slowly added to the PPT
solution (50 mg, 105 mmol) in 25 ml of dichloromethane at
room temperature. After stirring for 2 h, the reaction mixture
was added to sodium carbonate-saturated ice water and
extracted thrice with ethyl acetate. The organic extract was
evaporated to dryness in a vacuum to obtain a residue. The
residue was subjected to column chromatography on a C18
reversed-phase silica gel and eluted with methanol:water
Materials and methods
Chemicals and reagents
2
0(S)-Protopanaxatriol and schisandrin (internal standard, IS)
(
80:20) to produce M1-1 (21.6 mg) and M1-2 (19.2 mg). NMR
1
with a purity of over 98% were provided by Chengdu Mansite
Pharmaceutical Co., Ltd (Chengdu, China). Pooled HLMs (11
male Mongolian aged from 21 to 38 years) and Sprague–
Dawley (SD) rat liver microsomes (RLM, 100 male SD rats
aged from 2 to 3 months) were purchased from Rild Research
Institute for Liver Diseases (Shanghai, China). b-Nicotinamide
adenine dinucleotide phosphate, reduced form (NADPH) was
purchased from Sigma-Aldrich (St. Louis, MO). Methoxsalen,
quinidine and ketoconazole were purchased from J&K
Scientific Co., Ltd (Beijing, China); and sulfaphenazole,
ticlopidene, quercetin were purchased from China Food and
Drug Inspection Institute (Beijing, China). cDNA-expressed
human P450 isoforms (CYP1A2, CYP2C8, CYP2C9,
CYP2C19, CYP2D6, CYP3A4 and CYP3A5) were purchased
from BD Biosciences (San Jose, CA). Acetonitrile was of high-
performance liquid chromatography (HPLC) grade (Merck
KgaA, Darmstadt, Germany) and all other reagents were of
analytical grade.
13
1
1
spectra of M1-1 and M1-2 ( H and C NMR, H– H COSY,
HSQC, HMBC and NOESY) were recorded on a Bruker
1
Avance 400 spectrometer (400 MHz for H and 100 MHz for
1
3
C), and chemical shifts were recorded in parts per million
using tetramethylsilane as an internal standard. Deuterium
chloroform (CDCl ) was used as a solvent.
3
Feces sample collection
SD male rats (239–264 g) were provided by the experimental
animal center of Luye Pharma (Yantai, China). The animals
were housed in standard cages in a light-controlled room at
ꢀ
19 ± 1 C and 50 ± 5% relative humidity and given a standard
pellet diet and water. All studies were conducted in accord-
ance with the principles of Laboratory Animal Care (NIH
publication no. 92–93, revised in 1985) and were approved by
the local ethics committees for animal experimentation. The
PPT was prepared as 10 mg/ml solution in 0.5% carbox-
ymethyl cellulose sodium salt, 0.5% polysorbate-20. The
solution was stirred at room temperature pending use. Five
rats were housed in metabolic cages (one rat per cage) and
given a single dose of PPT at 100 mg/kg by gavage. The feces
samples were collected within a period of 0–48 h. The blank
feces samples were collected prior to dosing. The feces
Incubations of PPT in human liver microsomes
A stock solution of PPT was prepared in acetonitrile. The
final concentration of acetonitrile in the incubation was 0.5%
(v/v). PPT (50 mM) was incubated with the pooled HLM or
RLM (1 mg/ml) in 100 mM phosphate buffer (pH 7.4) with
ꢀ
samples were frozen at ꢁ20 C immediately after collection.
1
0 mM MgCl . The reaction volume was 200 ml. After 3 min
2
ꢀ
pre-incubation at 37 C, the incubation reactions were
initiated by the addition of NADPH (2.0 mM). Incubations
Enzyme kinetics of M1-1 and M1-2 in RLM and HLM
ꢀ
were performed at 37 C in a water bath shaker for 1 h. The
reactions were terminated with an equal volume of ice-cold
Reaction mixtures containing pooled HLM or RLM (0.5 mg/
ml) in 100 mM sodium phosphate buffer (pH 7.4) with 10 mM
ꢀ
ꢀ
acetonitrile, and then stored at ꢁ20 C until analysis. Control
MgCl were incubated at 37 C in a water bath shaker. The
2

DOI: 10.3109/00498254.2014.986562
Stereoselective metabolism of 20(S)-protopanaxatriol
3
ꢀ
incubation volume was 200 ml. The apparent formation
kinetics of M1-1 and M1-2 were determined by incubating
PPT at eight different concentrations (0.5, 1, 2, 5, 10, 20, 40
and 80 mM) in RLM and HLM, respectively. After 3 min pre-
to dryness under a stream of nitrogen at 40 C. The residue
was reconstituted in 100 ml of mixed solution of acetonitrile
and water containing 0.1% formic acid (50:50, v/v). After
grind and mixing, feces samples (0.2 g) was extracted by
ultrasonic treatment with methanol for 30 min and then
centrifuged at 13 000 ꢂ g for 10 min. The following treat-
ments were same as given above. A 20 ml aliquot of the
reconstituted solution was injected into the LC-MS/MS
system for analysis.
ꢀ
incubation at 37 C, the incubation reactions were initiated by
ꢀ
the addition of NADPH (1.0 mM), and conducted at 37 C for
1
5 min. The reactions were terminated with an equal volume
ꢀ
of ice-cold acetonitrile and stored at ꢁ20 C until analysis.
The reaction velocity was linear with regard to both time and
enzyme concentrations. Meanwhile, the depletion rates of this
pair of epimers were determined after incubating M1-1 and
M1-2 (0.5ꢁ40 mM) in RLM and HLM, respectively. Control
samples without NADPH were included. Each incubation was
performed in triplicate. The reaction velocities were calcu-
lated in the unit of picomoles per minute per milligram of
microsomal protein.
LC-MS/MS conditions
LC-MS/MS was performed by an Agilent 1100 series HPLC
system (Agilent Technologies, Waldbronn, Germany) and a
TSQ Quantum Access tandem mass spectrometer (Thermo
Electron Corporation, San Jose, CA) with an electrospray
ionization (ESI) source. The analytes were separated with an
XR-ODS C18 column (50 mm ꢂ 2.1 mm i.d., 2.2 mm,
Shimadzu, Japan). For the metabolite identification in HLM
and in rat feces, the gradient elution was used to separate the
metabolites with a mixture of A (acetonitrile) and B (water
containing 0.1% formic acid) with the following procedures:
Microsomal incubations of M1-1 and M1-2 in the
presence of inhibitors
Methoxsalen, quercetin, sulfaphenazole, ticlopidine, quinidine
and ketoconazole were used as selective inhibitors of CYP1A2,
CYP2C8, CYP2C9, CYP2C19, CYP2D6 and CYP3A, respect-
ively. The inhibitors were dissolved in methanol, which was
evaporatedbefore incubation. The experiments were performed
in pooled HLM (0.5 mg/ml) at 10 mM M1-1 or M1-2 in the
presence of each selective inhibitor at two different concentra-
tions (1 and 10 mM) in 100 mM phosphate buffer (pH 7.4) with
1
0% A for 0–2 min, 50% A for 5–8 min, 90% A for 10–13 min
and 10% A for 14–20 min. A flow rate was 0.2 ml/min.
The mass spectrometer was operated in positive ionization
mode. The spray voltage was 4 kV. Sheath gas and auxiliary
gas pressures were 30 and 5 psi, respectively. The capillary
ꢀ
temperature was 350 C and argon gas pressure was 1.5 mTorr.
Data acquisition was performed in full-scan of LC-MS and
1
0 mM MgCl , respectively. The other conditions were same as
2
2
MS modes. For MS spectra, the scan was ranged from m/z 100
above-mentioned enzyme kinetics experiments. For the mech-
anism-based inhibitor, methoxsalen was pre-incubated with all
2
to 1000, and the scan time was 0.8 s. For MS , the product ion
spectra were produced via the collision-induced dissociation of
ꢀ
incubation constituents at 37 C for 15 min before adding
+
the molecular ions [M + H] of all analytes at their respective
substrates, and ticlopidine was investigated with or without pre-
incubation. The controls without the inhibitors or NADPH
were included. Each incubation was performed in duplicate.
The inhibitors were also incubated without any substrates under
the same conditions to ensure that the inhibitors would
not interfere with quantification of M1-1 and M1-2.
HPLC retention times. All data obtained were processed using
Xcalibur Workstation, version 1.4.1.
As for the quantitative analysis, the mobile phase was a
combination of acetonitrile and water (containing 0.1% formic
acid; 70:30, v/v) delivered at a flow rate of 0.2 ml/min. The
selective reaction monitoring (SRM) in the positive ionization
mode was applied to detect M1-1 and M1-2 (m/z 493.2 > 457.2)
and IS (m/z 295.0 > 248.9) using the collision-induced decom-
position voltages 20 and 28 eV, respectively. In addition, the
dehydrogenation metabolites of the M1 were monitored
semiquantitatively by SRM at m/z 491.2 > 455.2.
Metabolism of M1-1 and M1-2 in recombinant
human P450 isoforms
M1-1 and M1-2, at 10 mM concentrations, were incubated
with human recombinant P450 isoforms (CYP1A2, CYP2C8,
CYP2C9, CYP2C19, CYP2D6, CYP3A4 and CYP3A5)
at 10 pmol in 100 mM phosphate buffer (pH 7.4) with
Data analysis
1
0 mM MgCl and 1 mM NADPH, respectively. The reactions
2
were initiated by adding the substrate, and carried out under
the microsomal incubation conditions described in the
enzyme kinetics experiments. The production of M2-3 and
M2-4 were monitored by liquid chromatography tandem mass
spectrometry (LC-MS/MS) and semi-quantitated according to
the standard curves of M1-1 and M1-2, respectively.
In enzyme kinetics experiments, the apparent kinetic param-
eters (Vmax and K ) were calculated using Prism 5 (GraphPad
m
Software, Inc., San Diego, CA) based on the Michaelis–
Menten equation: V ¼ V ꢃ [S]/([S] + K ), where K is the
max
m
m
substrate concentration at half the maximum velocity (Vmax)
of the reaction and [S] is the substrate concentration.
Sample preparations for LC-MS/MS Analysis
Results
All the in vitro incubation samples were prepared using the
same methods. Methanol (400 ml) was added to a 200 ml
incubation aliquot which had been terminated by acetonitrile,
then vortex-mixed and centrifuged at 13 000 ꢂ g for 10 min.
The supernatant was transferred into a glass tube, evaporated
Identification the metabolites of PPT
Mass spectral properties of PPT
The full scan of the analytes in positive and negative
ionization modes of the ESI source showed that the signals

4
W. Wang et al.
Xenobiotica, Early Online: 1–11
obtained in positive ionization mode had sufficient abun-
2
dance, which could be subjected to MS analysis and provided
more structural information. Due to the presence of four
hydroxyl groups in the molecule, PPT is prone to generate in-
source dissociation ions. Under the experimental conditions,
the protonated molecule of PPT at m/z 477 was not detected,
instead it was in-source dissociation ions at m/z 441 ([M + H –
+
2
2
fragment ions at m/z 423 ([M + H – 3H O] ) and 405 ([M + H
H O] ) which further dehydrated in MS to produce the
2
+
2
+
– 4H O] ), as shown in Figure 1. The retention time of the
2
PPT was at 15.4 min in the chromatography.
Identification metabolites of PPT in HLM
Compared with the control sample (without NADPH), 12
oxidation metabolites were found, including dehydrogenation,
mono-oxidation, di-oxidation and tri-oxidation products. A
series of extracting ion chromatograms of incubation samples
of PPT, M1-1 and M1-2 in HLM are shown in Figure 2. By
comparing the retention time and MS spectra with the
incubation samples of M1-1 and M1-2 (the epimeric metab-
olites of PPT) in HLM, it could be inferred that other
metabolites of PPT were produced by the sequential metab-
olism of M1-1 and M1-2. The structures were elucidated by
their mass spectral fragmentation patterns (Figure 3).
Metabolite M1. As shown in Figure 2(A), M1-1 and M1-2
were detected at 12.8 and 14.6 min, respectively, with a
protonated molecular weight of 493 (m/z 477 + 16), indicating
the addition of one oxygen atom into the PPT molecule. The
2
MS spectra (Figure 3A) showed fragment ions at m/z 475
+
+
(
[M + H – H O] ), 457 ([M + H – 2H O] ), 439([M + H –
2 2
+
+
3H O] ) and 421 ([M + H – 4H O] ). The latter two was 16 Da
2 2
higher than the corresponding dehydrated ions of PPT,
respectively. In addition, the fragment ions of m/z 143 and
2
07 were beneficial to identify the oxidation position. The
+
Figure 2. Extracted ion [M + H] chromatograms of PPT metabolites
after incubation PPT (50 mM), M1-1 and M1-2 (25 mM) with HLM. (A)
M1 (m/z 493); (B) M2 (m/z 491); (C) M3 (m/z 489); (D) M4 (m/z 507);
(E) M5 (m/z 509); (F) M6 (m/z 525). (a and d) PPT; (b and e) M1-2; (c
and f) M1-1. (a, b and c), with NADPH. (d, e and f) without NADPH.
standard references of M1-1 and M1-2 were synthesized
chemically using m-CPBA. The data of NMR spectra were
listed in Table 1, indicating that they were a pair of epimers,
and that their difference lied only in the configuration at C24.
By comparing the chromatographic retention time and mass
spectral fragments with the synthesis metabolites (Figure 2A: e
and f), M1-1 and M1-2 were confirmed as the (20S,24S)-
epoxy-dammarane-3,6,12,25-tetrol and (20S,24R)-epoxy-
dammarane-3,6,12,25-tetrol, respectively.
M2-1 was difficult to judge. By comparing the chromato-
graphic retention time and mass spectral fragmentation
patterns with the incubation sample of M1-1 and M1-2
(
Figure 2B), it could be inferred M2-1 and M2-4 were the
Metabolite M2. M2-1, M2-2, M2-3 and M2-4, with a quasi-
molecular ion peak at m/z 491, were detected at 11.3, 13.2,
dehydrogenation metabolites of M1-2; M2-2 and M2-3 were
the metabolites of M1-1.
2
3.9 and 15.3 min, respectively. The MS spectra (Figure 3B–
1
D) showed identical fragment ions at m/z 473 ([M + H –
+
+
+
H O] ), 455 ([M + H – 2H O] ), 437([M + H – 3H O] ) and
Metabolite M3. M3-1 was detected at 10.1 min with a quasi-
2
molecular ion peak at m/z 489. The MS spectra (Figure 3E)
+
showed the fragment ions at m/z 471 ([M + H – H O] ), 453
2
2
2
+
419 ([M + H – 4H O] ), which were 2 Da lower than the
2
corresponding fragment ions of the M1. In addition, the
2
2
+
+
[M + H – 2H O] ) and 435 ([M + H – 3H O] ), which was
(
fragment ions of m/z 143 and 205 were found in the MS
spectra of M2-2, which indicated the dehydrogenation
occurring at C6. Only the fragment ion of m/z 143 was
2
found in the MS spectra of M2-3 and M2-4 that indicated the
2
2
2 Da lower than the corresponding fragment ions of the M2.
Due to the absence of the inferential fragment ions at the low
molecular weight region, the dehydrogenation was speculated
to be occurring at the bonds between the side chain and the
host molecular moiety. Comparison of the chromatography
and mass spectral characteristics with the incubation samples
dehydrogenation occurring at C3. However, no inferential
fragment ions were found at the low molecular weight region
in the spectra of M2-1, so the dehydrogenation position of

DOI: 10.3109/00498254.2014.986562
Stereoselective metabolism of 20(S)-protopanaxatriol
5
2
Figure 3. MS spectra of PPT metabolites after incubation in HLM and oral administration to rats. (A) M1 (m/z 493); (B) M2-1* (m/z 491); (C) M2-2
and M2-5# (m/z 491); (D) M2-3 and M2-4 (m/z 491); (E) M3-1* (m/z 489); (F) M3-2# and M3-3# (m/z 489); (G) M4-1 and M4-2# (m/z 507); (H) M5-1
(
(
m/z 509); (I) M5-2 (m/z 509); (J) M6-1 and M6-3# (m/z 525); (K) M6-2 and M 6-4# (m/z 525); (L) M7# (m/z 505); (M) M8# (m/z 521); (N and O) M9#
m/z 523). * only found in HLM incubations; # only detected in rat feces.
of M1-1 and M1-2 (Figure 2C) indicated that M3-1was the
dehydrogenation metabolite of M1-2.
143 suggested the occurrence of another oxidation modifica-
tion on the host molecular moiety. Comparison of
the chromatography and mass spectral characteristics with
the metabolites of the M1 (Figure 2D) suggested that M4-1
was the further dehydrogenation and oxidation product
of M1-1.
Metabolite M4. M4-1 was detected at 11.5 min with a quasi-
2
molecular ion peak at m/z 507. The MS spectra (Figure 3G)
+
showed the fragment ions at m/z 489 ([M + H – H O] ), 471
2
+
+
(
[M + H – 2H O] ), 453([M + H – 3H O] ) and 435 ([M + H
2 2
+
4H O] ), which was 16 Da higher than the corresponding
–
fragment ion of the M2. Presence of the fragment ion of m/z
Metabolite M5. M5-1 and M5-2 were detected at 11.2 and
11.7 min, respectively, with a protonated molecular weight of
2

6
W. Wang et al.
Xenobiotica, Early Online: 1–11
Table 1. NMR data for M1-1 and M1-2 in CDCl
3
.
M1-1
M1-2
ꢀ
H
(J in Hz)
ꢀ
C
HMBC
2, 5
1
1, 28, 29
NOESY
ꢀ
H
(J in Hz)
ꢀ
C
NOESY
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1.72, m; 1.05, m
1.68, m; 1.59, m
3.18, dd (10.0, 4.3)
38.62
26.96
78.06
39.20
61.14
68.56
46.79
40.79
49.59
39.03
31.41
70.36
48.28
51.90
32.12
28.45
48.74
16.91
17.30
87.03
28.78
31.56
25.02
87.36
70.05
24.07
27.89
30.76
15.41
17.66
1.69, m; 1.03, m
1.66, m; 1.59, m
3.17, dd (11.0, 3.8)
38.67
26.98
78.39
39.21
61.18
68.53
46.86
40.78
49.88
38.99
31.13
70.80
48.92
51.76
32.55
28.52
47.79
16.84
17.37
86.41
27.51
31.13
24.90
85.33
70.09
26.06
27.85
30.76
15.40
18.07
5, 28
3, 9
5, 28
3, 9
0.90, m
4.12, td (10.6, 3.0)
1.62, m; 1.52, m
28, 29
5, 10
5, 9, 10
0.90, m
4.11, t (10.2)
1.63, m; 1.50, m
1.50, t (12.4)
1.49, m
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
1.94, m; 1.15, m
3.52, td (9.8, 4.3)
1.68, m
8
9, 14
11, 17, 30
1.92, m; 1.07, m
3.51, td (10.2, 4.3)
1.67, m
9. 17, 30
9, 21, 30
9, 18
1.54, m; 1.12, m
1.98, m; 1.32, m
2.25, td (11.0, 3.0)
1.09, s
13, 17
13, 14
15, 20
7, 11, 14
1, 5, 9, 10
1.59, m; 1.41, m
2.01, m; 1.32, m
2.19, t (9.4)
1.06, s
18, 21
0.94, s
0.90, s
1.27, s
17, 20, 22
17, 26
27
1.28, s
24
1.90, m; 1.78, m
2.05, m; 1.87, m
3.88, dd (10.2, 4.3)
1.85, m; 1.66, m
2.03, m; 1.88, m
3.85, t (7.7)
25
25
21, 27
1.11, s
1.23, s
1.31, s
0.98, s
0.94, s
24
24
3, 4, 5
3, 4, 5
8, 13, 15
24
24
3, 5
1.09, s
1.28, s
1.31, s
0.97, s
0.93, s
24
24
3, 5
5
atoms into the PPT molecule. The MS spectra (Figure 3H
09 (m/z 477 + 32), indicating the introduction of two oxygen
2
Identification metabolites of PPT in rat feces
A series of extracting ion chromatograms of experimental
feces samples are shown in Figure 4. After oral admin-
istration of PPT to rats, a total of 24 metabolites were
detected in feces and tentatively identified. Excluding M1
and M5, whose structures were similar to M1 and M5
identified in HLM, other 20 metabolites were first
reported according to the ocotillol-type structure. Among
the 20 metabolites, 6 metabolites were consistent with
those characterized in HLM: M2-2, M2-3, M2-4, M4-1,
M6-1 and M6-2. Other 14 metabolites are described as
follows.
+
and I) showed the fragment ions at m/z 491 ([M + H – H O] ),
2
+
+
4
73 ([M + H – 2H O] ), 455([M + H – 3H O] ) and 437
2 2
+
(
[M + H – 4H O] ), which was 16 Da higher than the
2
corresponding fragment ions of the M1. The presence
fragment ions of m/z 159 and 207 in M5-1 showed another
oxidation happened in the side chain moiety of the M1, and
the fragment ion of m/z 143 in M5-2 indicated another
oxidation occurred on the host molecular moiety of the M1. In
addition, by comparing spectra characteristics with the
incubation sample of M1 (Figure 2E), it could be concluded
M5 can be produced by the metabolism of M1-1.
Metabolite M2. M2-5 was detected at 14.2 min besides M2-2,
M2-3 and M2-4, with a quasi-molecular ion peak at m/z 491.
Metabolite M6. M6-1 and M6-2 were detected at 9.7 and
1
5
0.2 min, respectively, with a protonated molecular weight of
25 (m/z 477 + 48), indicating the introduction of three
2
The MS spectrum of M2-5 was identical with that of M2-2
2
derived from M1-1. According to the retention time, M2-5
was supposed to be produced by the sequential metabolism of
M1-2.
oxygen atoms. The MS spectra (Figure 3J and K) showed
+
identical fragment ions at m/z 489 ([M + H – 2H O] ), 471
2
+
+
(
[M + H – 3H O] ) and 453 ([M + H – 4H O] ), which was
2 2
3
However, the presence fragment ion of m/z 159 and 143 in the
2 Da higher than the corresponding fragment ions of the M1.
Metabolite M3. M3-2 and M3-3 were detected at 13.0 and
14.9 min, respectively, with a quasi-molecular ion peak at m/z
2
MS spectra of M5-1 and M5-2, respectively, indicated the
2
third oxidation happening in the side chain moiety in M5-1
and the host molecular moiety in M5-2. In addition,
comparison of chromatography and mass spectra profiles
showed that M6-1 and M6-2 could be produced by the
metabolism of M1-1 (Figure 2F).
489. The fragment ion of m/z 143 was found in their MS
spectra, which showed that the dehydrogenation occurred at
the host molecular moiety (Figure 3F). According to the
retention time, M3-2 and M3-3 were inferred as the further
dehydrogenation metabolites of the M1.

DOI: 10.3109/00498254.2014.986562
Stereoselective metabolism of 20(S)-protopanaxatriol
7
+
Figure 4. Extracted ion [M + H] chromatograms of PPT metabolites in rat feces after oral administration of PPT. (A) The feces of 0–24 h after
administration and (B) the feces prior to administration.
Metabolite M4. M4-2 was detected at 13.0 min besides M4-
at 11.6 min, with a quasi-molecular ion peak at m/z 507.
10.2 min, with a quasi-molecular ion peak at m/z 525. M6-1
and M6-2 had been identified as the further oxidation
1
2
2
products of M1-1 in HLM incubations. The MS spectrum
Their MS spectra were same. It has been inferred that M4-1
was the further dehydrogenation and oxidation product of
M1-1 in HLM incubations. Therefore, it was supposed that
M4-1 and M4-2 were the further metabolites of M1-1 and
M1-2 in rats, respectively, according to the chromatography
retention time.
of M6-3 and M6-4 was same as that of M6-1 and M6-2,
respectively (Figure 3J and K). Hence, it was inferred that
M6-1 and M6-2 were the further oxidation products of M1-1
in rats, and M6-3 and M6-4 were the metabolites of M1-2.
Metabolites M7. M7-1 and M7-2 were detected at 12.4 and
14.0 min, respectively, with a quasi-molecular ion peak at m/z
2
505. The MS spectra (Figure 3L) showed the fragment ions
Metabolite M6. M6-3 and M6-4 were detected at 10.8 and
1
2.6 min, respectively, besides M6-1and M6-2 at 9.7 and

8
W. Wang et al.
Xenobiotica, Early Online: 1–11
Figure 5. (A) Formation kinetic plots and (B) elimination kinetic plots of M1-1 and M1-2 in RLM and HLM. PPT (0.5–80 mM), M1-1 and M1-2
ꢀ
(
values of three separate determinations.
0.5–40 mM) were, respectively, incubated with microsomes (0.5 mg/ml) containing 1 mM NADPH for 15 min at 37 C. Data are expressed as the mean
+
+
at m/z 487 ([M + H – H O] ), 469 ([M + H – 2H O] ), 451
and the time point after 15 min incubation. The metabolic
elimination profiles are shown in Figure 5(B), the depletion
rate of M1-2 was 11.0 folds faster than that of M1-1 in HLM,
and the depletion rate of M1-1 and M1-2 was similar in RLM
(Table 2). The results indicated that there were stereoselective
differences and species differences in the formation and
metabolic elimination between M1-1 and M1-2.
2
2
+
+
(
[M + H – 3H O] ) and 433 ([M + H – 4H O] ), which was
2 2
1
6 Da higher than the corresponding fragment ions of the M3.
In addition, the fragment ion m/z 143 was found, so it was
presumed another oxidation modification occurred on the host
molecular moiety.
Metabolite M8. M8-1 and M8-2 were detected at 11.7 and
1
5
4.4 min, respectively, with a quasi-molecular ion peak at m/z
2
21. The MS spectra (Figure 3M) showed the fragment ions
Identification of P450 enzymes responsible for M1-1
and M1-2 metabolism
+
+
at m/z 503 ([M + H – H O] ), 485 ([M + H – 2H O] ), 467
2
2
+
+
After the incubation of M1-1 and M1-2 (10 mM) with six
types of inhibitors of microsomal P450 isoform enzymes
(
[M + H – 3H O] ) and 449 ([M + H – 4H O] ), which was
2 2
1
6 Da higher than the corresponding fragment ions of the M7.
(CYP1A2, CYP2C8, CYP2C19, CYP2C9, CYP2D6 and
In addition, the fragment ion m/z 143 was present, so it was
presumed the third oxidation happened at the host molecular
moiety.
CYP3A) in HLM, the inhibition data were expressed as
relative activity compared with the control incubations
without inhibitors or NADPH. Among the inhibitors used,
only ketoconazole showed a significant inhibitory effect on
the oxidative metabolism of M1-1 and M1-2, yielding about
95% inhibition at 10 mM (Figure 6), whereas other isoform
enzyme inhibitors had no inhibitory effects on their metab-
olism essentially.
Metabolite M9. M9-1, M9-2, M9-3 and M9-4 were detected
at 10.8, 11.7, 12.4 and 13.7 min, respectively, with a quasi-
2
molecular ion peak at m/z 523. The MS spectra (Figure 3N
and O) showed identical fragment ions at m/z 505 ([M + H –
+
H O] ), 487 ([M + H – 2H O] ), 469 ([M + H – 3H O] ) and
+
+
2
2
2
+
51 ([M + H – 4H O] ), which was 16 Da higher than the
4
corresponding fragment ions of the M4. In addition, the
2
fragment ion of m/z 159 was found in MS spectra of M9-1
In seven types of recombinant human P450 isoforms
tested, as expected, only CYP3A4 and CYP3A5 at a minor
extent (less than 5% that of CYP3A4) exerted catalytic
activity for M1-1 and M1-2, meanwhile the dehydrogenation
metabolites of M1-1 and M1-2 were detectable. The catalytic
activity of other isoforms tested to M1-1 and M1-2 was
negligible. These data showed that CYP3A4 was probably the
major isoform enzyme responsible for the further oxidative
metabolism of M1-1 and M1-2.
2
and M9-3, so it was presumed the third oxidation happened at
the side chain moiety. Similarly, the ion of m/z 143 was found
in M9-2 and M9-4, so it was presumed the third oxidation
happened in the host molecular moiety.
Stereoselective formation and metabolism of M1-1
and M1-2 in HLM and RLM
Discussion
As shown in Figure 5(A), the apparent formation rate of M1-1
was higher than that of M1-2 when incubating PPT with RLM
and HLM. The Vmax of M1-1 was 2.4 folds and 10.4 folds
higher than that of M1-2 in RLM and HLM, respectively
Both negative and positive modes of ESI mass spectra were
examined in this study. Generally, in the negative mode, fewer
product ions were observed in quadrupole mass analyzer;
+
(Table 2).
while in the positive ion mode, [M + H] ions with sufficient
2
abundance were subjected to MS analysis and provided more
structural information. Proper metabolite identification using
The average depletion rates of M1-1and M1-2 were
estimated from the differences between the zero time point

DOI: 10.3109/00498254.2014.986562
Stereoselective metabolism of 20(S)-protopanaxatriol
9
Table 2. Enzyme kinetic parameters of formation or metabolic elimination for M1-1 and M1-2, the epimeric
metabolites of PPT, in RLM and HLM.
Formation
Elimination
Vmax (pmol/min/mg protein)
K
m
(mM)
V
max (pmol/min/mg protein)
m
K (mM)
RLM
M1-1
M1-2
HLM
M1-1
M1-2
a
844.5 ± 62.0
354.9 ± 34.5
21.6 ± 1.8
54.1 ± 4.6
1089 ± 80.4
1173 ± 68.6
11.2 ± 1.1
6.3 ± 0.7
b
2593 ± 125
249.3 ± 31.0
78.8 ± 5.7
88.9 ± 7.8
411.7 ± 15.1
4539 ± 308
3.6 ± 0.47
25.0 ± 2.6
c
m
The maximum velocity (Vmax) and the michaelis constant (K , the substrate concentration at half Vmax) were
estimated from the Michaelis–Menten equation by non-linear regression analysis of the mean values of
three separate determinations.
p50.01 versus M1-2 in RLM.
p50.01 versus M1-2 in HLM.
p50.01 versus M1-1 in HLM.
a
b
c
the LC-MS/MS approach requires a comprehensive under-
standing of the fragmentation behavior of the compound to be
tested. In this study, a total of 12 oxidation metabolites were
found in HLM incubations and 24 oxidation metabolites were
found in rat feces by LC-MS/MS methods. It suggested that
there were more extensive oxidation metabolism in rats than
in humans. In order to demonstrate whether there were other
metabolic pathways besides the hepatic metabolism, the
metabolism study of PPT in RLM was also carried out. Since
the results were similar to that in rat feces, the data are not
shown here.
In order to explain the extensive pharmacological activities
of PPT and explore the toxicological potency, the metabolic
pathway study of PPT has been undertaken in vitro and
in vivo. Microbial biotransformation studies showed that PPT
mainly yielded the compounds of dehydrogenation at C12 or
hydroxylation at C15, C26, C27, C28, C29 and 20S,24R-
epoxy-dammaran-3b,6a,25-triol-12-one by the fungus Mucor
spinosus or Absidia corymbifera AS (Chen et al., 2013; Tian
et al., 2005; Zhang et al., 2007). Wang et al. (2010) revealed
one phase I intestine metabolite of PPT in mice. PPT carried
out dehydrogenation at C12 and hydroxylation at C15, as well
as formation of an ester bond at C20 of the side chain. In a
recent research report (He et al., 2014), 17 phase I metabolites
were found in rats, and it was proposed that the predominant
metabolic pathways of PPT were oxidation to yield two of
C20–24 epoxides, i.e. M1-1 and M1-2 in our study; and form
corresponding carboxylic acid at C26/27. Hao et al. (2010)
proposed that the formation of C20–24 epoxides (ocotillol-
type ginsenosides) were the main oxygenation metabolism
pathway of PPT-type ginsenosides by microsomal P450-
mediated. In this study, by comparing the chromatographic
retention times and mass spectral fragments of incubation
samples of PPT with those of M1-1 and M1-2 in HLM, it was
probable to infer the further oxidative metabolites were
produced by sequential metabolism of M1-1 and M1-2. The
proposed metabolic pathways of PPT were different from the
previous report. Moreover, earlier studies (Li et al., 2011,
Figure 6. Effects of isoform-selective P450 inhibitors on the metabolism
of M1-1 (A) and M1-2 (B) in incubation with HLM. Reactions were
performed in the presence of M1-1 or M1-2 (10 mM) with different
isoform enzyme inhibitors at two concentrations (1.0 and 10 mM) in the
microsomes (0.5 mg/ml) containing 1 mM NADPH at 37 C for 15 min.
Data are reported as the mean values of two separate determinations.
ꢀ
the dehydrogenation and/or oxygenation of the C20–24
epoxides. Hence, C20–24 epoxides were important inter-
mediate metabolites of PPT in HLM and its further dehydro-
genation metabolite with higher abundance (Figure 2), which
was aid to support the further active compounds and
toxicological studies.
Compared with HLM, oxidation metabolites were found
more in rat feces samples. In addition, this study showed that
the apparent formation quantity of M1-1 was far more than
that of M1-2 after the incubation of PPT in HLM. As P450
enzymes are important in the metabolism of xenobiotics and
are highly stereoselective (Li et al., 2005; Shimshoni et al.,
2
2
012) found that the predominant metabolic pathway of
0(S)-protopanaxatdiol (PPD) and 20(S)-ginsenoside Rh2 in
HLM was the oxidation that yielded the 20,24-epoxide form.
The further sequential metabolites were also detected through
2012; Sun et al., 2013), many chiral compounds are produced

1
0
W. Wang et al.
Xenobiotica, Early Online: 1–11
by the metabolism. The further kinetic studies in the apparent
formation and metabolic elimination of M1-1 and M1-2 in
HLM and in RLM discovered the formation Vmax of M1-1 was
Declaration of interest
This work was supported by the National Natural Science
Foundation of China (No. 81102880) and the Natural Science
Foundation of Shandong Province (No. ZR2012HM036).
1
0.4 folds higher than that of M1-2, and the depletion rate of
M1-2 was 11.0 folds faster than M1-1 in HLMs. The apparent
formation differences between M1-1 and M1-2 might ascribe
to the fast depletion rate of M1-2. However, the apparent
formation Vmax of M1-1 was 2.4 folds higher than that of M1-
References
Chen G, Yang X, Zhai X, Yang M. (2013). Microbial transformation of
2
0(S)-protopanaxatriol by Absidia corymbifera and their cytotoxic
activities against two human prostate cancer cell lines. Biotechnol Lett
35:91–5.
2
, and the depletion rates of M1-1 and M1-2 were similar in
RLM. This indicated there were remarkable stereoselectivity
differences in the elimination and formation between M1-1
and M1-2. The elimination difference was mainly manifested
in HLM incubation, and the formation difference was
confirmative in RLM. Obviously, the stereoselectivity metab-
olism of both epimers had species differences between rats
and humans. It suggests that besides the differences in isoform
composition, expression and catalytic activities of drug-
metabolizing enzymes across species, the stereoselectivity
difference was also important between species. In addition,
stereochemistry is an important dimension in pharmacology
and it often results in different pharmacological activities of
chiral xenobiotics. The pharmacological activity study of PPT
and its epimeric metabolites showed that PPT and M1-2
exerted cardioprotective effects against myocardial ischemic
injury by enhancing the anti-free-radical actions of heart
tissues, and M1-1 had no such effect (Han et al., 2011). This
study suggested that M1-1, 24S-epimer was the predominant
metabolite of PPT in clinical situations.
Chiu NT, Guns ES, Adomat H, et al. (2014). Identification of human
cytochrome P450 enzymes involved in the hepatic and intestinal
biotransformation of 20(S)-protopanaxadiol. Biopharm Drug Dispos
3
5:104–18.
Evans DC, Baillie TA. (2005). Minimizing the potential for metabolic
activation as an integral part of drug design. Curr Opin Drug Discov
Devel 8:44–50.
Han B, Meng Q, Li Q, et al. (2011). Effect of 20(S)-protopanaxatriol and
its epimeric derivatives on myocardial injury induced by iso-
proterenol. Arzneimittelforschung 61:148–52.
Han KL, Jung MH, Sohn JH, Hwang JK. (2006). Ginsenoside 20S-
protopanaxatriol (PPT) activates peroxisome proliferator-activated
receptor gamma (PPARgamma) in 3T3-L1 adipocytes. Biol Pharm
Bull 29:110–13.
Hao H, Lai L, Zheng C, et al. (2010). Microsomal cytochrome P450-
mediated metabolism of protopanaxatriol ginsenosides: metabolite
profile, reaction phenotyping, and structure-metabolism relationship.
Drug Metab Dispos 38:1731–9.
Hasegawa E, Nakagawa S, Miyate Y, et al. (2013). Inhibitory effect of
protopanaxatriol ginseng metabolite M4 on the production of
corticosteroids in ACTH-stimulated bovine adrenal fasciculata cells.
Life Sci 92:687–93.
Hasegawa H. (2004). Proof of the mysterious efficacy of ginseng: basic
and clinical trials: metabolic activation of ginsenoside: deglycosyla-
tion by intestinal bacteria and esterification with fatty acid.
J Pharmacol Sci 95:153–7.
Hasegawa H, Sung JH, Matsumiya S, Uchiyama M. (1996). Main
ginseng saponin metabolites formed by intestinal bacteria. Planta Med
62:453–7.
Hasegawa H, Suzuki R, Nagaoka T, et al. (2002). Prevention of growth
and metastasis of murine melanoma through enhanced natural-killer
cytotoxicity by fatty acid-conjugate of protopanaxatriol. Biol Pharm
Bull 25:861–6.
It has been well known that CYP3A is the most abundant
human hepatic P450 isoforms and accounts for the metabol-
ism of approximately 50% of drugs in clinical use. Hao et al.
(2010) found CYP3A4 mediated oxygenation metabolism of
PPT type ginsenosides and the metabolites of C20–24
epoxides (M1 and M5) were produced. The studies (Chiu
et al., 2014; Li et al., 2012) demonstrated that CYP3A was the
sole isoform involved in the oxygenation metabolism of
gingenosides Rh2 and PPD, and compared with CYP3A5,
CYP3A4 should be the predominant isoform. Both chemical
inhibition and recombinant human P450 isoform assays in the
present study indicated that CYP3A4 was the predominant
isoform responsible for the further oxidation metabolism of
M1-1 and M1-2. Potential drug–drug interaction should be
kept in mind when drugs influencing CYP3A4 activity or
being substrates of CYP3A are co-administered with PPT.
He C, Zhou D, Li J, et al. (2014). Identification of 20(S)-protopanax-
atriol metabolites in rats by ultra-performance liquid chromatography
coupled with electrospray ionization quadrupole time-of-flight
tandem mass spectrometry and nuclear magnetic resonance spectros-
copy. J Pharm Biomed Anal 88:497–508.
Kasai R, Hara K, Dokan R, et al. (2000). Major metabolites of ginseng
sapogenins formed by rat liver microsomes. Chem Pharm Bull 48:
1
226–7.
Kong LT, Wang Q, Xiao BX, et al. (2013). Different pharmacokinetics of
the two structurally similar dammarane sapogenins, protopanaxatriol
and protopanaxadiol, in rats. Fitoterapia 86:48–53.
Lai L, Hao H, Liu Y, et al. (2009). Characterization of pharmacokinetic
profiles and metabolic pathways of 20(S)-ginsenoside Rh1 in vivo and
in vitro. Planta Med 75:797–802.
Li L, Chen X, Li D, Zhong D. (2011). Identification of 20(S)-
protopanaxadiol metabolites in human liver microsomes and human
hepatocytes. Drug Metab Dispos 39:472–83.
Li L, Chen X, Zhou J, Zhong D. (2012). In vitro studies on the oxidative
metabolism of 20(s)-ginsenoside Rh2 in human, monkey, dog, rat, and
mouse liver microsomes, and human liver S9. Drug Metab Dispos 40:
2041–53.
Li XQ, Weidolf L, Simonsson R, Andersson TB. (2005). Enantiomer/
enantiomer interactions between the S- and R- isomers of omeprazole
in human cytochrome P450 enzymes: major role of CYP2C19 and
CYP3A4. J Pharmacol Exp Ther 315:777–87.
Conclusion
In this study, 12 and 24 oxidation metabolites of PPT were
found in HLM and rats, respectively. C20–24 epoxides, M1-1
and M1-2 were the major metabolites. Moreover, other
metabolites were inferred from the sequential metabolism of
the M1 by comparing ESI-MS/MS spectra and chromatog-
raphy characteristics. PPT underwent more extensive oxida-
tive metabolism in rats than in HLM. The enzyme kinetic
studies of M1-1 and M1-2 in HLM and in RLM discovered
that the formation and further metabolic elimination of
both epimers was highly stereoselective, and the stereoselect-
ivity resulted in the metabolism species differences of PPT.
In addition, M1-1 and M1-2 were metabolized mainly by
CYP3A4 isoform.
Liu L, Gu LJ, Zhang DL, et al. (2010). Microbial conversion of rare
ginsenoside Rf to 20(S)-protopanaxatriol by Aspergillus niger. Biosci
Biotechnol Biochem 74:96–100.
Oh GS, Pae HO, Choi BM, et al. (2004). 20(S)-Protopanaxatriol, one of
ginsenoside metabolites, inhibits inducible nitric oxide synthase and

DOI: 10.3109/00498254.2014.986562
Stereoselective metabolism of 20(S)-protopanaxatriol 11
cyclooxygenase-2 expressions through inactivation of nuclear factor-
kappaB in RAW 264.7 macrophages stimulated with lipopolysacchar-
ide. Cancer Lett 205:23–9.
Shibata S. (2001). Chemistry and cancer preventing activities of ginseng
saponins and some related triterpenoid compounds. J Korean Med Sci
Tawab MA, Bahr U, Karas M, et al. (2003). Degradation of ginsenosides
in humans after oral administration. Drug Metab Dispos 31:1065–71.
Tian Y, Guo H, Han J, Guo D. (2005). Microbial transformation of
20(S)-protopanaxatriol by Mucor spinosus. J Nat Prod 68:678–80.
Usami Y, Liu YN, Lin AS, et al. (2008). Antitumor agents. 261. 20(S)-
protopanaxadiol and 20(s)-protopanaxatriol as antiangiogenic agents
and total assignment of (1)H NMR spectra. J Nat Prod 71:478–81.
Wang YZ, Chen J, Chu SF, et al. (2009). Improvement of memory in
mice and increase of hippocampal excitability in rats by ginsenoside
Rg1’s metabolites ginsenoside Rh1 and protopanaxatriol. J Pharmacol
Sci 109:504–10.
1
6:28–37.
Shimshoni JA, Roberts AG, Scian M, et al. (2012). Stereoselective
formation and metabolism of 4-hydroxy-retinoic acid enantiomers by
cytochrome p450 enzymes. J Biol Chem 287:42223–32.
Sun J, Wang G, Haitang X, et al. (2005). Simultaneous rapid
quantification of ginsenoside Rg1 and its secondary glycoside Rh1
and aglycone protopanaxatriol in rat plasma by liquid chromatog-
Wang YZ, Wang YS, Chu SF, et al. (2010). Protopanaxatriol metabolites
identified by LC-MS/MS after oral administration in mice. Int J Clin
Pharmacol Ther 48:282–90.
raphy-mass spectrometry after solid-phase extraction.
Biomed Anal 38:126–32.
J Pharm
Zhang J, Guo H, Tian Y, et al. (2007). Biotransformation of 20(S)-
protopanaxatriol by Mucor spinosus and the cytotoxic structure
activity relationships of the transformed products. Phytochemistry 68:
Sun SY, Wang YQ, Li LP, et al. (2013). Stereoselective interaction
between tetrahydropalmatine enantiomers and CYP enzymes in
human liver microsomes. Chirality 251:43–7.
2
523–30.