1521-009X/42/4/735–743$25.00
DRUG METABOLISM AND DISPOSITION
Copyright ª 2014 by The American Society for Pharmacology and Experimental Therapeutics
http://dx.doi.org/10.1124/dmd.113.055962
Drug Metab Dispos 42:735–743, April 2014
Hepatic Glucuronidation of Isoneochamaejasmin A from the
Traditional Chinese Medicine Stellera Chamaejasme L. Root
Lushan Yu, Jianbin Pu, Minjuan Zuo, Xia Zhang, Yang Cao, Shifeng Chen, Yan Lou, Quan Zhou,
Haihong Hu, Huidi Jiang, Jianzhong Chen, and Su Zeng
Department of Pharmaceutical Analysis and Drug Metabolism, Zhejiang Province Key Laboratory of Anti-Cancer Drug Research,
College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang, China (L.Y., M.Z., X.Z., Y.L., H.H., H.J., S.Z.); Institute
of Materia Medica, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang, China (J.P., Y.C., S.C., J.C.); and
Department of Pharmacy, 2nd Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, China (Q.Z.)
Received November 14, 2013; accepted January 22, 2014
ABSTRACT
Among the recombinant human UGT isoform test and correlation
analysis, UGT1A1, UGT1A3, and UGT1A9 were found to mediate M1
formation, whereas only UGT1A3 mediated M2 formation. Kinetic
studies showed obvious species differences between human,
mouse, rat, dog, and pig liver microsomes. UGT1A1, HLMs, and
human intestinal microsomes, but not human kidney microsomes,
exhibited substrate inhibition for the formation of M1. UGT1A1mediated formation of M1 showed a 6- and 11-fold higher Vmax than
did UGT1A3- and UGT1A9-mediated formation of M1, respectively.
The results of the relative activity factor assay showed that UGT1A1
contributed approximately 75% in the formation of M1. These
findings collectively indicate that UGT1A1 is the major enzyme in
the formation of M1, whereas UGT1A3 is the major enzyme in the
formation of M2.
Introduction
The glucuronidation conjugation reaction is an important elimination and detoxification mechanism for xenobiotics and endogenous
compounds. UDP glucuronosyltransferases (UGTs) glucuronidate
compounds by transferring glucuronic acid from its cofactor UDP
glucuronic acid (UDPGA) to lipophilic substrates, thereby transforming them into hydrophilic glucuronides and facilitating their
subsequent elimination via biliary or renal routes.
Isoneochamaejasmin A (INCA; Fig. 1), a biflavonoid, is one of the
main active ingredients in the dried root of S. chamaejasme L. with
a high content level of up to 1% (w/w) (Feng et al., 2004). In the present
study, we report the identity of the human liver UGTs responsible for
the O-glucuronide metabolism of INCA in an in vitro system. Metabolic
screening was performed with a battery of recombinant human UGTs to
identify the specific UGT isoforms involved in O-glucuronidation.
Inhibition studies and correlation studies were further performed to
identify the UGT isoforms mediating O-glucuronidation of INCA. In
addition, the two O-glucuronidation metabolites of INCA were obtained
by chemical synthesis. Interestingly, although INCA is a single
enantiomer, stereoselective metabolism was observed in INCA Oglucuronidation.
Stellera chamaejasme L., a traditional Chinese medicine, is
widely used for the treatment of scabies, ringworm, stubborn skin
ulcers, chronic tracheitis, and liver and lung cancers in China and in
other Asian countries (Niwa et al., 1984; Yang, 1993). Pharmacological studies have revealed that S. chamaejasme L. possesses
multiple bioactivities, including antiviral, antitumor, antimitotic,
antifungal, and immunomodulating activities (Yoshida et al., 1996;
Endo et al., 1998; Yang et al., 2005; Liu and Zhu, 2012).
Flavonoids and biflavonoids are two types of important constituents in S. chamaejasme L., which exhibit antitumor, antiviral,
antimitotic, and antifungal activities (Yang et al., 2005; Liu et al.,
2008; Asada et al., 2013).
This research was supported by the Natural Science Foundation of China
[Grants 81230080, 81102500, and 81172983]; National Major Projects of China
[Grants 2011CB710800 and 2012ZX09506001-004]; and the Zhejiang Natural
Science Foundation [LH12H31007].
L.Y. and J.P. contributed equally to this work.
dx.doi.org/10.1124/dmd.113.055962.
ABBREVIATIONS: 7-HFC, 7-hydroxy-4-trifluoromethylcoumarin; CLint, intrinsic clearance; DLM, dog liver microsome; DMSO, dimethylsulfoxide;
HIM, human intestinal microsome; HKM, human kidney microsome; HLM, human liver microsome; HPLC, high-performance liquid chromatography;
INCA, isoneochamaejasmin A; IS, internal standard; M1, 7-O-glucuronide; M2, 49-O-glucuronide; MLM, mouse liver microsome; MS, mass
spectrometry; NMR, nuclear magnetic resonance; PLM, pig liver microsome; RAF, relative activity factor; RLM, rat liver microsome; TOF, time of
flight; UDPGA, UDP glucuronic acid; UGT, UDP glucuronosyltransferase.
735
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Isoneochamaejasmin A (INCA), a biflavonoid, is one of main active
ingredients in the dried root of Stellera chamaejasme L., a widely
used traditional Chinese medicine. In the present study, we
identified the glucuronidation metabolite of INCA and characterized
the UDP glucuronosyltransferases (UGTs) responsible for INCA
glucuronidation. 7-O-glucuronide (M1) and 49-O-glucuronide (M2)
were identified by incubation of INCA with human liver microsomes
(HLMs) in the presence of UDP glucuronic acid, and their structures
were confirmed by high-resolution mass spectrometry and nuclear
magnetic resonance analyses. Although INCA is a single enantiomer molecule, its M1 metabolite showed two equal-size peaks on
a pNAP stationary phase but only one peak on a C18 stationary
phase, indicating that the 7-/799- and 49-/4999-hydroxyl groups of
INCA were in different spatial configurations relative to each other.
736
Yu et al.
Materials and Methods
Preparation of INCA. INCA was isolated and purified from the dried root
of S. chamaejasme L. The dried root of S. chamaejasme L. (10.0 kg) was
purchased from Nanjing Qingze Medical Technological Development Co. Ltd.
(Nanjing, China). A voucher specimen had been deposited at the College of
Pharmaceutical Sciences at Zhejiang University (Zhejiang, China). The dried
root was pulverized and then extracted three times in a 10-fold volume (w/v) of
95% ethanol at ambient temperature. The 95% ethanol solution was condensed
and the condensate was successively extracted three times in a 10-fold volume
(v/v) of ethyl acetate at ambient temperature. The ethyl acetate fraction was
condensed and subjected to silica gel (200–300 mesh) column chromatography
with a gradient elution by using petroleum ether (60°C–90°C) ethyl acetate.
The eluted solution (60:40) was evaporated to dryness in a rotary evaporator at
40°C under vacuum. The residue was redissolved in methanol and further
purified by a Waters RP-prep (Waters Corporation, Milford, MA) highperformance liquid chromatography (HPLC) system with a mobile phase
consisting of 55% methanol and 45% water/formic acid (100:0.1, v/v). The
purity was .98% based on the HPLC determination. The structure of prepared
INCA was identified by mass spectrometry (MS) as well as 13C nuclear
magnetic resonance (NMR) and 1H NMR analysis and was consistent with the
literature (Feng and Pei, 2002; Feng et al., 2004).
Chemicals and Enzyme Sources. 17b-estradiol, UDPGA, alamethicin,
trifluoperazine, and b-D-glucuronidases from Escherichia coli, and 4-trifluoromethyl7-hydroxycoumarin glucuronide were purchased from Sigma-Aldrich (St. Louis,
MO). 7-Hydroxy-4-trifluoromethylcoumarin (7-HFC) was obtained from Acros
Organics (Geel, Belgium). Propofol was purchased from ICN Biomedicals Inc.
(Irvine, CA). Quercetin was purchased from the National Institutes for Food and
Drug Control (Beijing, China). The internal standard (IS) [3-(2-ethyl phenyl)-5(3-methoxy phenyl)-1H-1,2,4-triazol] was kindly donated by Xianju Pharmaceutical Factory (Zhejiang, China). All other reagents and solvents used were
either of analytical or of HPLC grade.
Pooled and individual human liver microsomes (HLMs), pooled human
intestinal microsomes (HIMs), and pooled human kidney microsomes (HKMs)
were purchased from the Research Institute for Liver Diseases (Shanghai,
China). Pig liver microsomes (PLMs) from the Bama miniature pig were kindly
provided by the Department of Laboratory Animal Science, College of Basic
Medical Sciences, Third Military Medical University (Chongqing, China). The
following human recombinant UGT supersomes expressed in insect cells were
purchased from BD Biosciences (Woburn, MA): UGT1A1, UGT1A3, UGT1A4,
UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B4, UGT2B7,
UGT2B15, and UGT2B17.
Preparation of Mouse, Rat, and Dog Liver Microsomes. ICR mice and
Sprague-Dawley rats were purchased from the Laboratory Animal Center of
Zhejiang University (Hangzhou China), and Beagle dogs were purchased from
Jiaxin Jiaan Laboratory Animal Culture Co. Ltd. (Jiaxin, China). The pooled
mouse liver microsomes (MLMs; n = 8), rat liver microsomes (RLMs; n = 8),
and dog liver microsomes (DLMs; n = 4) were prepared as previously described
(Gibson and Skett, 1994) and all manipulations were carried out in an ice-cold
bath. Pellets were resuspended in sucrose-Tris buffer (pH 7.4; 95:5, w/v) and
immediately stored at 280°C. Microsomal protein concentrations were
determined by the modified Lowry method, using bovine serum albumin as
the standard.
All experiments with animals were performed according to the National
Institutes of Health Guide for the Care and Use of Laboratory Animals, and
were approved by the Institutional Animal Care and Use Committee of
Zhejiang University.
Determination of Enzyme Activity of Liver Microsomes and Recombinant UGTs. The glucuronosyltransferase activities of mouse, rat, dog, pig, and
human liver microsomes, HIMs, HKMs, and recombinant UGTs were
determined according to our previous methods (Yu et al., 2007; Chen et al.,
2008). Trifluoperazine was used as the probe substrate of UGT1A4, and 7-HFC
was used as the probe substrate of the other UGT isoforms. The results
indicated that all of the liver microsomes and recombinant UGTs had good
activities (data not shown).
In Vitro Metabolism of INCA by Liver Microsomes and Recombinant
UGTs. A glucuronidation assay was performed in a 100-ml incubation mixture
that contained 50 mM Tris-HCl buffer (pH 7.5), 10 mM MgCl2, 2 mM
UDPGA, 20 mg/ml alamethicin, 0.4 mg/ml enzyme protein (recombinant
UGTs, or pooled liver microsomes), and 50 mM INCA. INCA was dissolved in
dimethylsulfoxide (DMSO). The final concentration of DMSO in the reaction
mixture was 1% (v/v). After preincubation at 37°C for 5 minutes, the reaction
was initiated by the addition of 1 ml UDPGA (2 mM final concentration) and
was run for 30 minutes at 37°C in a shaking water bath. The reaction was
terminated by the addition of 200 ml acetonitrile on ice and then centrifuged at
15,700 g for 5 minutes at 4°C. Aliquots of the supernatant were analyzed by
HPLC. Chromatography was performed using an Agilent 1200 HPLC system
(Agilent Technologies, Santa Clara, CA), equipped with a UV detector.
Separation was performed on a GraceSmart RP18 column (4.6 mm 150 mm,
5 mm; Thermo Fisher Scientific Inc., Rockford, IL), and the UV wavelength
was set at 280 nm. The mobile phase (1.0 ml/min) consisted of water (A) and
acetonitrile (B) containing 0.1% formic acid in a gradient program. The
gradient, expressed as changes in mobile phase B, was as follows: 0–5 minutes,
a linear increase from 25% to 60% B; 5–7 minutes, hold at 60% B. The
retention times of two glucuronide metabolites of INCA were 5.7 and 6.2
minutes, respectively.
Hydrolysis with b-Glucuronidase. A b-glucuronidase assay was performed as follows. The glucuronidation incubation mixtures that contained
HLMs, in a total volume of 200 ml, were heated for 30 minutes at 95°C and
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 1. Structures of INCA and its glucuronides. glu, glucuronic acid group.
UGTs Involved in Isoneochamaejasmin A Metabolism
737
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 2. Chromatograms of INCA glucuronides after incubation with HLMs and recombinant human UGTs in the presence of UDPGA. (A) Blank control. (B) HLMs. (C)
UGT1A1. (D) UGT1A3. (E) UGT1A9. (F) M1 and M2 by chemical synthesis.
then equally divided into two parts. Ten microliters of KH2PO4 buffer (0.1 M,
pH 5.0) containing 1000 U b-glucuronidase was added into one of them and
incubated at 37°C for 2 hours, whereas the other one was treated in parallel
without b-glucuronidase. Each reaction was then stopped with the addition of
100 ml ice-cold acetonitrile. After removal of the protein by centrifugation, the
supernatant was subjected to HPLC as described above.
738
Yu et al.
ethyl acetate and evaporated to give crude product. Glucuronides were purified
by preparative HPLC. Chromatography was performed using an Agilent PrepStar
218 preparation LC system (Agilent Technologies), equipped with a UV
detector. Separation was performed on a COSMOSIL pNAP column (10 mm 250 mm, 5 mm; Nacalai Tesque, Inc., Kyoto, Japan), and the UV wavelength was
set at 280 nm. The mobile phase (3.0 ml/min) consisted of 39% methanol and
61% water containing 0.1% formic acid.
Preparation of M2 was as follows. INCA (650.2 mg), Ag2O (192.6 mg), dry
CaSO4 (235.6 mg), and N,N-Diisopropylethylamine (1.5 ml) were stirred
protected from right under a N2 atmosphere at 0°C for 10 minutes. Excess
glucuronyl bromide (3251.0 mg) was then slowly added, and stirring was
continued at 0°C for 8 hours. The other process was the same as above except
that the mobile phase consisted of 62% methanol and 38% water containing
0.1% formic acid.
1
H NMR and 13C NMR spectra were recorded on Bruker 500 MHz
spectrometers (Bruker Bioscience, Billerica, MA) using tetramethylsilane as an
IS. Samples were dissolved in deuterated DMSO.
Correlation Analysis. A correlation analysis was performed between the
activities of INCA glucuronidation versus propofol (UGT1A9), quercetin
(UGT1A3), and 17b-estradiol (UGT1A1) glucuronidation in HLMs of 10
individual donors, respectively (Yu et al., 2007; Ma et al., 2012). Glucuronidation
activity for the form of M1 of INCA was measured and the substrate and protein
concentrations were 40 mM and 0.4 mg/ml, respectively. The activity of each UGT
Fig. 3. Representative high-resolution mass spectra of INCA glucuronide formed from INCA by HLMs. The spectra were taken at the retention time of M1a (A), M1b (B),
and M2 (C) in Fig. 2B.
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Analysis of the Glucuronide Conjugates of INCA by Liquid
Chromatography–Time of Flight Mass Spectrometry. To identify the
glucuronide metabolites of INCA, an AB SCIEX (Framingham, MA) TripleTOF
4600 liquid chromatography (LC)/time of flight (TOF)/MS system, coupled with
an Eksigent ekspert ultraLC 100 ultra-HPLC system, was used. The
chromatographic conditions were the same as mentioned above. The TurboIonSpray interface was operated in the negative ion mode at 5500 V and 550°C. The
operating conditions were as follows: nebulizing gas flow, 8 psi; curtain gas flow,
15 psi; and collision energy, 25 eV. MS/MS spectra were obtained in the range
of m/z 50–750.
Chemical Synthesis of the Glucuronide Conjugates of INCA. The
intermediate a-glucuronyl bromide, methyl-2,3,4-tri-O-acetyl-a-D-glucopyranosyl
urinate bromide, can be prepared in three steps from D-glucurone. Glucuronyl
bromide was used as the starting material for O-glucuronide synthesis.
Preparation of M1 was as follows. INCA (120.4 mg), glucuronyl bromide
(138.5 mg), Ag2O (35.6 mg), dry CaSO4 (43.6 mg), and quinoline (1.5 ml)
were stirred protected from light under a N2 atmosphere at ambient temperature
for 8 hours, and were then diluted with ethyl acetate and filtered and the filtrate
was washed with H2SO4/H2O (pH = 2) solution to remove quinoline. The
washed filtrate was then evaporated and the residue was suspended in 50%
methanol/H2O (72 ml). Na2CO3 solution (0.5 M, 7 ml) was added drop-wise to
remove the acyl groups, stirring at 0°C for 7 hours. The pH of the supernatant
was adjusted to below 3.0 with H2SO4 solution. The mixture was washed with
739
UGTs Involved in Isoneochamaejasmin A Metabolism
Results
Glucuronidation of INCA in HLMs. Two metabolites (M1 and
M2) were generated after incubation of INCA with HLMs in the
presence of UDPGA (Fig. 2B). These two metabolites were confirmed
as glucuronide conjugates by LC-TOF-MS analysis (Fig. 3). The
corresponding high-resolution MS data indicated that the molecular
formulas of M1 and M2 were C36H30C16 (found, 717.1461; calculated, 717.1456, [M 2 H]2) and C36H30C16 (found, 717.1513; calculated, 717.1456, [M 2 H]2), respectively. The subsequent product
ion mass spectra of metabolites generated identical fragment ion
patterns, which gave the major fragment ion at m/z 541, indicating the
loss of a glucuronic acid moiety (176 Da). Hydrolysis studies indicated that M1 and M2 were easily hydrolyzed with b-glucuronidase
and converted to the parent compound INCA.
Structures of INCA Glucuronides. It is difficult to identify the
glucuronide position on which phenolic hydroxyl group of INCA by
LC-TOF-MS data. Therefore, to identify the metabolite structures, M1
and M2 were synthesized by chemical synthesis. The retention times
of the synthesized compounds corresponded to those of M1 and M2
(Fig. 2F). The mass spectra of the synthesized compounds also
matched those of M1 and M2. 1H NMR and 13C NMR data confirmed
M1 as the 7-O-glucuronide and M2 as the 49-O-glucuronide (Table 1).
Interestingly, M1 showed only one peak on a C18 stationary phase;
however, it showed two equal-size peaks on a pNAP stationary phase
(4.6 mm 250 mm, 5 mm; Nacalai Tesque, Inc.) with a mobile phase
consisting of methanol and 0.1% formic acid water solution (65:35, v/v)
both in synthesis samples and in HLM incubation samples (Fig. 4).
The first peak was designated M1a and the second peak was
designated M1b. M1a and M1b were then resolved by using prepared
LC on a semi-prepared COSMOSIL pNAP column (10 mm 250
mm, 5 mm; Nacalai Tesque, Inc.). The 1H NMR and 13C NMR data
and the subsequent product ion mass spectra of M1a and M1b were
too similar to distinguish their absolute configurations (Fig. 3, A and
B; Table 1). However, M1a and M1b exhibited different specific
optical rotations with values of +0.5 ° and –1.7 °, respectively,
determined using a Jasco automatic P1030 polarimeter (JASCO
International Co., Kyoto, Japan) with a 589 nm Na lamp. M2 showed
only one peak in this chromatographic condition with a pNAP column
(Fig. 4).
TABLE 1
NMR assignments of ICNA, M1a, M1b, and M2
1
13
H NMR
C NMR
Position
M2
ICNA
M1a
M1b
299
399
499
599
699
4.76, 2H, m
3.71, 2H, m
ICNA
4.79, 2H, m
3.71, 2H, m
4.78, 2H, m
3.71, 2H, m
4.78, 2H
3.71, 2H
5.76, 2H, d, 1.5 Hz
5.76, 1H, d, 1.5 Hz;
6.06, 1H, d, 1.0 Hz
5.77, 1H, d, 1.5 Hz;
6.07, 1H, d, 1.0 Hz
5.76, 2H
81.0
47.6
195.3
163.9
96.7
81.0, 81.4
49.1
195.1
163.9
96.8
81.0, 81.4
49.1
195.1
163.9
96.8
81.0
47.6
195.1
163.9
96.8
7, 799
8, 899
5.89, 2H, d, 1.5 Hz
5.89, 1H, d, 1.5 Hz;
6.15, 1H, d, 1.5 Hz
5.90, 1H, d, 1.5 Hz;
6.15, 1H
5.90, 2H
167.4
95.5
167.6, 165.6
95.6
167.6, 165.6
95.6
167.6
95.6
162.9
101.5
127.1
129.9
116.0
162.9
101.4, 99.4
126.9
129.9
116.0
162.9
101.4, 99.4
126.8
129.9
116.0
162.9
101.5
127.1
123.0
116.0, 116.7
158.8
158.9
158.9
158.8, 158.1
2,
3,
4,
5,
6,
9, 999
10, 1099
19, 1999
29, 2999, 69, 6999
39, 3999, 59, 5999
6.99, 4H, d, 8.5 Hz
6.79, 4H, d, 8.5 Hz
49, 4999
d, doublet; m, multiplet.
M1a
7.00, 4H, d, 8.0 Hz
6.79, 4H, d, 8.5 Hz
M1b
7.01, 4H, d, 8.5 Hz
6.79, 4H, d, 8.0 Hz
7.06, 7.12, 4H
6.79, 2H; 6.95,
6.99, 2H
M2
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
isoform was determined by HPLC as described previously for 17b-estradiol
glucuronidation (Ma et al., 2012) as well as quercetin and propofol glucuronidation
(Yu et al., 2007). Pearson’s product-moment correlation coefficients were
calculated by GraphPad Prism 5 software (version 5.01; GraphPad Software
Inc., San Diego, CA). P , 0.05 was considered statistically significant.
Kinetic Analysis with Liver Microsomes and Recombinant UGTs. The
kinetic studies were performed using pooled human, dog, pig, rat, and mouse
liver microsomes, as well as HIMs, HKMs, and human recombinant UGTs
(UGT1A1, UGT1A3, and UGT1A9). The optimal conditions for microsomal
incubation were determined in the linear range for the formation of glucuronide
metabolites from INCA. The final concentration of organic solvent (DMSO) was
1% (v/v) in the final incubation mixtures. Typical incubation mixtures (total
volume of 100 ml), containing either 0.4 mg/ml of liver microsomes or
recombinant UGT, INCA (2–100 mM), 50 mM Tris-HCl buffer (pH 7.5), 10 mM
MgCl2, and 20 mg/ml alamethicin, were preincubated on ice for 30 minutes. The
reaction was initiated by adding UDPGA (2 mM) and run for 30 minutes at 37°C
in a shaking water bath. The reaction was terminated by the addition of 100 ml
acetonitrile with IS (20 mM), and the mixture was centrifuged at 15,700 g for
10 minutes. Aliquots of the supernatant were used for HPLC analysis.
Relative Activity Factor Method Application. To achieve an accurate
understanding of the contribution of UGT enzymes in the formation of M1 in
liver, M1 formation activity was scaled from recombinant systems to HLMs
through the relative activity factor (RAF) method (Crespi and Miller, 1999; Zhu
et al., 2012). The relative contribution of the individual isoform (Contributioni)
is calculated as follows (Zhu et al., 2012): 1) RAFi = Vprobe reaction in HLM/
Vprobe reaction in recombinant UGT; 2) Vi = vi RAFi; and 3) Contributioni = Vi/Vtotal
100%. Vi, vi, and Vtotal are the M1 formation rate of the individual isoforms
in HLMs, the M1 formation rate of the recombinant UGTs, and the total M1
formation rate in HLMs, respectively. Glucuronidation rates of 17b-estradiol
and propofol were reported to be probe reactions for catalytic activity of human
UGT1A1 and UGT1A9, respectively (Soars et al., 2003; Zhu et al., 2012). The
final concentrations of INCA in HLMs and recombinant UGT incubation
mixtures were 10 and 40 mM.
Data Analysis. All results are expressed as the mean 6 S.D. in triplicate
experiments. Kinetic parameters were estimated using GraphPad Prism software
(version 5.01; GraphPad Software Inc.) designed for a nonlinear least-squares fit
to the standard Michaelis–Menten equation or the Hill equation (V = Vmax · Sn/
(S50n + Sn)) or the substrate inhibition equation (V = Vmax/(1 + Km/S + S/Ki). The
calculated parameters include the maximum rate of formation (Vmax), the
Michaelis–Menten constant (apparent Km), the intrinsic clearance (CLint = Vmax/
apparent Km), the Hill coefficient (n), and the constant describing the substrate
inhibition interaction (Ki). The t test was used for statistical analysis and
statistical significance was defined as P , 0.05.
740
Yu et al.
Identification of UGT Isoforms Involved in Glucuronidation of
INCA. Two concentrations of INCA (10 and 100 mM) were incubated
with a panel of recombinant UGT isoforms expressed in insect cells
(UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8,
UGT1A9, UGT1A10, UGT2B4, UGT2B7, UGT2B15, and UGT2B17).
The formation rates of M1 and M2 after incubation of INCA with UGT
isoforms are shown in Fig. 5. M1 formation was catalyzed simultaneously by UGT1A1, UGT1A3, and UGT1A9, whereas UGT1A1
showed the highest catalytic activity. M2 formation was catalyzed only
by UGT1A3 among the 12 UGT isoforms. Correlation analyses with
activities of UGT isoforms in HLMs were performed to confirm the
involvement of UGT1A3 and UGT1A9 in the formation of M1. As
shown in Fig. 6, the M1 formation rates were significantly correlated
with glucuronidation of 17b-estradiol (r2 = 0.7819, P = 0.0007),
glucuronidation (r2 = 0.6280, P = 0.0063), and quercetin (r2 = 0.5603,
P = 0.0127).
Contributions of UGT1A1 and UGT1A9 to the Formation of
M1 in HLMs. The RAF value of 17b-estradiol glucuronidation
between recombinant UGT1A1 and HLMs was 0.84. The RAF value
of propofol glucuronidation between recombinant UGT1A9 and
HLMs was 0.41. When M1 formation rates by HLMs, UGT1A1,
and UGT1A9 at a substrate concentration of 10 mM were determined,
the results showed that UGT1A1 contributed approximately 75% of
this activity in HLMs, whereas UGT1A9 contributed 8%. Repeating
these assays at a substrate concentration of 40 mM showed that
UGT1A1 and UGT1A9 contributed approximately 77% and 9%,
respectively, to the formation of M1.
Enzyme Kinetic Analysis. Kinetic analyses were performed with
pooled from human, mouse, rat, dog, and pig liver microsomes, as
well as HIMs, HKMs, and recombinant UGT1A1, UGT1A3, and
UGT1A9 using 2–100 mM INCA. Under the experimental conditions
used, the formation of M1 catalyzed by HLMs, HIMs, MLMs, DLMs,
and recombinant human UGT1A1 exhibited substrate inhibition with
Ki values of 53.0, 21.9, 25.0, 20.0, and 47.2 mM, respectively (Fig. 7;
Table 2). The formation of M1 catalyzed by HKMs, RLMs,
recombinant UGT1A3, and UGT1A9 fitted a typical Michaelis–
Menten equation. The catalytic activities of HIMs and HKMs were
Fig. 5. Formation of M1 and M2 by recombinant human UGT isoforms. (A)
Formation of M1. (B) Formation of M2. INCA at two different concentrations of 10
and 100 mM was incubated with each of 12 recombinant UGT isoforms at a protein
concentration of 0.2 mg/ml, and other conditions are the same with HLMs. The
values are averages of triplicate (6 S.D.) incubations.
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 4. Chromatograms of INCA and its metabolites on a pNAP column. (A) M1a, M1b, M2, and INCA dissolved in mobile phase solution. (B) Sample of INCA (60 mM)
in HLM incubation after 20 minutes.
UGTs Involved in Isoneochamaejasmin A Metabolism
741
low for the formation of M2. M2 was detected only at a high
concentration of INCA in the incubation solutions (.20 mM). The
catalytic behaviors of all enzymes (excepting PLMs) for the formation
of M2 exhibited typical Michaelis–Menten modes (Fig. 7; Table 2).
Specifically, the glucuronidation of INCA mediated by PLMs fitted
the Hill equation with Hill coefficients of 2.6 and 1.9 for the
formations of M1 and M2, respectively (Table 2). UGT1A1-mediated
formation of M1 displayed a 6- and 11-fold higher Vmax than did the
UGT1A3 and UGT1A9-mediated formations of M1, respectively. On
the basis of the Vmax and CLint data, UGT1A1 was considered to make
a greater contribution to M1 formation than UGT1A3 and UGT1A9.
Discussion
Biflavonoid constituents in S. chamaejasme L. have noticeable
pharmacological activities, but they also have certain toxicities such as
hepatic and renal toxicity (Song et al., 1996; Du and Liu, 1999).
Metabolism is an important pathway to promote the elimination of
some endogenous and exogenous compounds in vivo. Although we
observed no product peak and no obvious decrease in the INCA peak
area by HPLC after INCA was incubated with HLMs and RLMs in the
presence of NADPH (data not shown), two metabolites (M1 and M2)
of INCA were found in human, mouse, rat, dog, and pig liver
microsomes in the presence of UDPGA in this study. These two
metabolites were confirmed as single glucuronide metabolites by MS
and by b-glucuronidase assay, but their structures were difficult to
identify by MS spectra alone. The glucuronic acid moiety of M1 and
M2 was easily lost even in low-energy collisions in MS/MS analysis.
Therefore, similarly to other flavonoid glucuronides, it was difficult to
confirm the glucuronide position in INCA from the molecular ion
fragment information (Fig. 3).
To identify the O-glucuronide position of INCA, all possible
glucuronide metabolites of INCA were synthesized by chemical
synthesis. The synthesis processes for M1 and M2 were optimized for
high yields. Comparing their retention times with those of metabolites
in liver microsome incubations under different HPLC conditions (using
a C18 stationary phase), M1 and M2 were confirmed as 7-O-glucuronide
and 49-O-glucuronide, respectively. Initially, we assumed that site 7 and
site 799 were the same site for the biflavonoid because M1 always
showed one peak regardless of the mobile phase conditions. However,
to our surprise, M1 from chemical synthesis and biosynthesis were both
separated into two similar peaks (designated M1a and M1b) on a pNAP
stationary phase, whereas M2 showed only one peak under the HPLC
conditions (Fig. 4). A pNAP column is packed with naphthalene-bound
silica and offers improved separation of compounds such as positional
isomers that are difficult to separate with alkyl group-bonded materials.
M1a and M1b were then resolved by a semi-preparation pNAP column.
Unfortunately, it was difficult to confirm their absolute configurations
by their specific optical rotations and NMR data. We have made several
attempts to grow single crystals for X-ray diffraction analysis, but to
date we have been unsuccessful.
It is well known that racemic compounds are mirror-symmetrical.
Although INCA (as a biflavonoid) is a single enantiomer molecule, its
two single flavonoid groups are mirror-symmetrical to each other (Fig.
1). Therefore, INCA can be approximated as a racemate and its 5-/599-,
7-/799-, and 49-/4999-hydroxyl groups are different in spatial configuration to each other. Glucuronide metabolites of some racemates,
such as flurbiprofen (Wang et al., 2011), propranolol (Yu et al., 2004,
2010), propafenone (Xie and Zeng, 2010), and sarpogrelate
metabolites (Kim et al., 2013), might be separated on a nonchiral
stationary phase column. We believe that M2 also has two
constitutional isomers even if they cannot be separated under the
present HPLC conditions. This point was confirmed by the 1H NMR
spectra of M2, which showed two sets of overlapping signals
produced by a mixture of two diastereomers, although those were
not observed from 13C NMR data (Table 1).
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 6. Correlations between formation of the M1 and quercetin glucuronidation (A), propofol glucuronidation (B), and 17b-estradiol glucuronidation (C) in HLMs from 10
donors. P , 0.05 indicates statistical significance. Each data point represents the mean of duplicate determinations. The values are averages of triplicate (6 S.D.) incubations.
742
Yu et al.
Identification of UGT isoforms responsible for the glucuronide
metabolism of INCA was demonstrated by several approaches. The
results in 12 recombinant human UGT isoform incubations showed
that UGT1A1, UGT1A3, and UGT1A9 catalyzed the 49- and 499-Oglucuronide of INCA (M1) and only UGT1A3 catalyzed the 7- and/or
799-O-glucuronide of INCA (M2) (Fig. 5). Statistically significant
correlations were shown between the M1 formation and UGT1A1mediated 17b-estradiol glucuronidation rates (P = 0.0007), between
the M1 formation and UGT1A9-mediated propofol glucuronidation
rates (P = 0.0063) (Yu et al., 2007), and between the M1 formation
and UGT1A3-mediated quercetin glucuronidation rates (P = 0.0127)
(Yu et al., 2007) (Fig. 6). Our results collectively indicated that
UGT1A1, UGT1A3, and UGT1A9 catalyzed the formation of M1,
whereas only UGT1A3 catalyzed the formation of M2.
Obvious species differences were found in the kinetic parameters of
five species (Table 2). HLMs, HIMs, MLMs, and DLMs exhibited
a substrate inhibition behavior in the 7-O-glucuronidation of INCA,
whereas HKMs and RLMs fitted the typical Michaelis–Menten
equation. Except for PLMs, all other liver microsomes fitted the
typical Michaelis–Menten equation in the 49-O-glucuronidation of
INCA. The 7- and 49-O-glucuronidations of INCA in PLMs were both
characterized by Hill kinetics, suggesting the involvement of a single
enzyme or more than one enzyme with similar affinities (Seo et al.,
2010). Although M1 contains M1a and M1b, the two compound
formations were similar in liver microsomes and recombinant UGTs
incubation mixtures with 40 mM INCA after incubating for 30 minutes
in 37°C (data not shown). Therefore, the kinetic parameters were fitted
using M1. The intrinsic clearance levels of M1 formation were
significantly higher than that of M2 formation in HLMs, MLMs, and
PLMs; the opposite was found in RLMs and DLMs. The kinetic
parameters (including Km, Vmax, CLint, and Ki) of recombinant human
UGT1A1-mediated M1 formation were close to that of HLMs and
HIMs and the CLint of UGT1A1-mediated M1 formation was potently
higher than that of UGT1A3 and UGT1A9, suggesting that UGT1A1
was a major enzyme in the 7-O-glucuronidation of INCA. The results
of the RAF contribution assay also indicated that UGT1A1 played
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
Fig. 7. Michaelis–Menten plots for the formation of M1 and M2 in HLMs (A), M1 in HIMs (B), M1 in HKMs (C), M1 in UGT1A1 (D), M1 and M2 in UGT1A3 (E), and
M1 in UGT1A9 (F). Microsomes or recombinant UGTs were incubated with 2–100 mM INCA at 37°C for 30 minutes in the presence of UDPGA (2 mM). The values are
averages of triplicate (6 S.D.) incubations.
743
UGTs Involved in Isoneochamaejasmin A Metabolism
TABLE 2
Kinetic parameters of INCA glucuronidation by microsomes and recombinant human UGTs (n = 3)
Metabolites
7-O-Glucuronide (M1)
4-O-Glucuronide (M2)
HLM
HIM
HKM
MLM
RLM
DLM
PLM
UGT1A1
UGT1A3
UGT1A9
HLM
MLM
RLM
DLM
PLM
UGT1A3
42.91
38.45
18.81
34.79
125.4
59.70
22.18
51.83
21.50
9.66
56.72
39.41
258.2
74.96
19.80
50.62
Km
Vmax
CLint
Kia
mM
nmol/min per
mg protein
ml/min per
mg protein
mM
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
17.29
10.59
4.94
12.88
2.85
5.78
23.58
12.10
4.74
5.80
2.59
6.17
9.04
10.85
4.55
1.07
53.00 6 21.18
21.94 6 10.34
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
14.21
12.64
3.14
17.64
16.8
18.69
1.44
13.25
3.01
1.32
15.76
6.38
85.34
16.12
1.57
12.73
0.742
0.407
0.093
0.448
0.358
0.345
0.523
0.627
0.102
0.056
0.147
0.243
2.335
0.813
0.090
0.054
0.184
0.135
0.005
0.175
0.029
0.132
0.022
0.227
0.005
0.002
0.020
0.017
1.047
0.096
0.004
0.006
Hill Coefficient
24.97 6 13.36
19.96 6 11.39
47.23 6 14.21
2.55
1.90
Ki, constant describing the substrate inhibition interaction.
a major contribution in the formation of M1 in HLMs. The kinetic
results for HKMs (Table 2) suggest that UGT1A9 is the main
contributor to INCA glucuronidation in this tissue.
In summary, the present results demonstrate that INCA was
transformed to M1 and M2 in HLMs. The two single flavonoid groups
in the INCA molecule are mirror-symmetrical to each other, which led
to the production of M1a and M1b when the 7-/799-hydroxyl groups of
INCA were conjugated with UDPGA. UGT1A1, UGT1A3, and
UGT1A9 were responsible for the formation of M1, and only
UGT1A3 was responsible for the formation of M2. In addition, the
formation of M1 was mainly metabolized by UGT1A1.
Authorship Contributions
Participated in research design: Yu, J. Chen, Zeng.
Conducted experiments: Yu, Pu, Zuo, Zhang, Cao, S. Chen, Lou, Zhou, Hu.
Contributed new reagents or analytic tools: Pu, Zhou.
Performed data analysis: Yu, Pu, Zuo, Jiang.
Wrote or contributed to the writing of the manuscript: Yu, Pu, J. Chen,
Zeng.
References
Asada Y, Sukemori A, Watanabe T, Malla KJ, Yoshikawa T, Li W, Kuang X, Koike K, Chen
CH, and Akiyama T, et al. (2013) Isolation, structure determination, and anti-HIV evaluation of
tigliane-type diterpenes and biflavonoid from Stellera chamaejasme. J Nat Prod 76:852–857.
Chen YK, Xie SG, Chen SQ, and Zeng S (2008) Glucuronidation of flavonoids by recombinant
UGT1A3 and UGT1A9. Biochem Pharmacol 76:416–425.
Crespi CL and Miller VP (1999) The use of heterologously expressed drug metabolizing enzymes
—state of the art and prospects for the future. Pharmacol Ther 84:121–131.
Du J and Liu J (1999) The toxicity study on organisms of the water extract of Stellera chamaejsme
L. Shan Dong Yi Yao Gong Ye 18:49–50.
Endo Y, Maruno M, Miura N, Baba M, and Ikegawa T (1998) Novel diterpenes and their use as
antiviral and anti-HIV agents with toxicity. Japanese patent 10287617.
Feng BM, Pei YH, Zhang HL, Hua HM, and Wang YQ (2004) Chemical constituents from roots
of Stellera chamaejsme L. Chin Tradit Herb Drugs 35:12–14.
Feng BM and Pei YH (2002) Research on the Anti-Epilepsy Constituents of Stellera chamaejasme
L. and Citrus grandis Osbeck. Chinese Topic Doctorate Dissertation. Shenyang Pharmaceutical
University, Shenyang, China.
Gibson GG and Skett P (1994) The clinical relevance of drug metabolism, in Introduction to Drug
Metabolism, 2nd ed., pp 217–219, Blankie Academic & Professional, London.
Kim HJ, Jeong ES, Seo KA, Shin KJ, Choi YJ, Lee SJ, Ghim JL, Sohn DR, Shin JG, and Kim DH
(2013) Glucuronidation of a sarpogrelate active metabolite is mediated by UDPglucuronosyltransferases 1A4, 1A9, and 2B4. Drug Metab Dispos 41:1529–1537.
Liu WK, Cheung FW, Liu BP, Li C, Ye W, and Che CT (2008) Involvement of p21 and FasL in
induction of cell cycle arrest and apoptosis by neochamaejasmin A in human prostate LNCaP
cancer cells. J Nat Prod 71:842–846.
Liu X and Zhu X (2012) Stellera chamaejasme L. extract induces apoptosis of human lung cancer
cells via activation of the death receptor-dependent pathway. Exp Ther Med 4:605–610.
Ma L, Sun J, Peng Y, Zhang R, Shao F, Hu X, Zhu J, Wang X, Cheng X, and Zhu Y, et al. (2012)
Glucuronidation of edaravone by human liver and kidney microsomes: biphasic kinetics and
identification of UGT1A9 as the major UDP-glucuronosyltransferase isoform. Drug Metab
Dispos 40:734–741.
Niwa M, Chen XF, Liu GQ, Tatematsu H, and Hirata Y (1984) Structures of isochamaejasmin
from Stellera chamaejasme L. Chem Lett 13:1587–1590.
Seo KA, Bae SK, Choi YK, Choi CS, Liu KH, and Shin JG (2010) Metabolism of 19- and 4hydroxymidazolam by glucuronide conjugation is largely mediated by UDP-glucuronosyltransferases
1A4, 2B4, and 2B7. Drug Metab Dispos 38:2007–2013.
Soars MG, Ring BJ, and Wrighton SA (2003) The effect of incubation conditions on the enzyme
kinetics of UDP-glucuronosyltransferases. Drug Metab Dispos 31:762–767.
Song XP, Cao GR, and Li SJ (1996) Abstraction and toxicity of total-flavonoid in Stellera
chamaejsme L. Acta Univ Agric Boreali-occidentalis 24:35–38.
Wang H, Yuan L, and Zeng S (2011) Characterizing the effect of UDP-glucuronosyltransferase
(UGT) 2B7 and UGT1A9 genetic polymorphisms on enantioselective glucuronidation of
flurbiprofen. Biochem Pharmacol 82:1757–1763.
Xie S and Zeng S (2010) Stereoselective glucuronidation of propafenone and its analogues by
human recombinant UGT1A9. Chem Pharm Bull (Tokyo) 58:879–883.
Yang CL (1993) Materia Medica of Toxic Herbs, pp 1021–1023, Chinese Medicine and Materials
Publisher, Beijing.
Yang GH, Liao ZX, Xu ZY, Zhang HP, and Chen DF (2005) Antimitotic and antifungal C-3/C30-biflavanones from Stellera chamaejasme. Chem Pharm Bull (Tokyo) 53:776–779.
Yoshida M, Feng WJ, Saijo N, and Ikekawa T (1996) Antitumor activity of daphnane-type
diterpene gnidimacrin isolated from Stellera chamaejasme L. Int J Cancer 66:268–273.
Yu LS, Lu SJ, Lin YG, and Zeng S (2007) Carboxyl-glucuronidation of mitiglinide by human
UDP-glucuronosyltransferases. Biochem Pharmacol 73:1842–1851.
Yu LS, Luan LJ, Shao Q, and Zeng S (2004) Direct determination of S-(-)- and R-(+)-propranolol
glucuronide in rat hepatic microsomes by RP-HPLC. Biomed Chromatogr 18:833–837.
Yu LS, Qian MR, Liu Y, Yao TW, and Zeng S (2010) Stereoselective metabolism of propranolol
glucuronidation by human UDP-glucuronosyltransferases 2B7 and 1A9. Chirality 22:456–461.
Zhu L, Ge G, Zhang H, Liu H, He G, Liang S, Zhang Y, Fang Z, Dong P, and Finel M, et al.
(2012) Characterization of hepatic and intestinal glucuronidation of magnolol: application of
the relative activity factor approach to decipher the contributions of multiple UDPglucuronosyltransferase isoforms. Drug Metab Dispos 40:529–538.
Address correspondence to: Jianzhong Chen or Su Zeng, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou 310058, China. E-mail: chjz@
zju.edu.cn or [email protected]
Downloaded from dmd.aspetjournals.org at ASPET Journals on June 16, 2017
a
Enzymes