Synthesis of (-)-[4-3H]epigallocatechin gallate and its metabolic fate in rats after intravenous administration.

Article abstract of DOI:10.1021/jf0011236

Because a great deal of attention has been focused on the metabolism of (-)-epigallocatechin gallate (EGCg), quantitative analysis of this compound is required. For this purpose we developed a method of chemical synthesis of [4-(3)H]EGCg. Synthesized [4-(3)H]EGCg showed 99.5% radiochemical purity and a specific activity of 13 Ci/mmol. To clarify the excretion route of EGCg, the radioactivity levels of bile and urine were quantified after intravenous administration of [4-(3)H]EGCg to bile-duct-cannulated rats. Results showed that the radioactivity of the bile sample excreted within 48 h accounted for 77.0% of the dose, whereas only 2.0% of the dose was recovered in the urine. The excretion ratio of bile to urine was calculated to be about 97:3. These results clearly showed that bile was the major excretion route of EGCg. Time-course analysis of the radioactivity in blood was also performed to estimate the pharmacokinetic parameters following intravenous administration of [4-(3)H]EGCg. In addition, EGCg metabolites excreted in the bile within 4 h after the intravenous dose of [4-(3)H]EGCg were analyzed by HPLC. The results showed that 4',4"-di-O-methyl-EGCg was present in the conjugated form and made up about 14.7% of the administered radioactivity.

Full text of DOI:10.1021/jf0011236

1042
J. Agric. Food Chem. 2001, 49, 1042−1048
Syn th esis of (-)-[4-3H]Ep iga lloca tech in Ga lla te a n d Its Meta bolic
F a te in Ra ts a fter In tr a ven ou s Ad m in istr a tion
Toshiyuki Kohri,*,† Fumio Nanjo,† Masayuki Suzuki,† Ryota Seto,† Natsuki Matsumoto,†
Mana Yamakawa,† Hiroshi Hojo,† Yukihiko Hara,† Dhimant Desai,‡ Shantu Amin,‡
C. Clifford Conaway,§ and Fung-Lung Chung§
Food Research Laboratories, Mitsui Norin Company, Limited, 223-1 Miyabara, Fujieda City, Shizuoka Pref.
426-0133, J apan, Division of Organic Synthesis Facility and Carcinogenesis and Molecular Epidemiology,
American Health Foundation, 1 Dana Road, Valhalla, New York 10595
Because a great deal of attention has been focused on the metabolism of (-)-epigallocatechin gallate
(EGCg), quantitative analysis of this compound is required. For this purpose we developed a method
of chemical synthesis of [4-3H]EGCg. Synthesized [4-3H]EGCg showed 99.5% radiochemical purity
and a specific activity of 13 Ci/mmol. To clarify the excretion route of EGCg, the radioactivity levels
of bile and urine were quantified after intravenous administration of [4-3H]EGCg to bile-duct-
cannulated rats. Results showed that the radioactivity of the bile sample excreted within 48 h
accounted for 77.0% of the dose, whereas only 2.0% of the dose was recovered in the urine. The
excretion ratio of bile to urine was calculated to be about 97:3. These results clearly showed that
bile was the major excretion route of EGCg. Time-course analysis of the radioactivity in blood was
also performed to estimate the pharmacokinetic parameters following intravenous administration
of [4-3H]EGCg. In addition, EGCg metabolites excreted in the bile within 4 h after the intravenous
dose of [4-3H]EGCg were analyzed by HPLC. The results showed that 4′,4′′-di-O-methyl-EGCg was
present in the conjugated form and made up about 14.7% of the administered radioactivity.
Keyw or d s: (-)-Epigallocatechin gallate; (-)-[4-3H]epigallocatechin gallate; catechin; tea; biliary
excretion
INTRODUCTION
16), although ingested EGCg was detected mostly in
conjugated forms in the human plasma. These findings
suggested that bile could be the main excretion route
of EGCg. Recently, Suganuma et al. (17) reported on
the distribution of radioactivity in various organs and
blood, and on the excretion of radioactivity in the urine
and feces after oral administration of [3H]EGCg to mice.
They showed that maximum levels of radioactivity in
most of the organs and blood were observed at 24 h
postdose, and about 6% of radioactivity of dose was
excreted in the urine within 24 h. However, no informa-
tion was available with respect to biliary excretion of
[3H]EGCg. In addition, although Suganuma et al.
showed the exchange rate of [3H]EGCg with water in
vitro was less than 5% for 24-h incubation, it still
remains that the labeled EGCg may be susceptible to
hydrogen-tritium exchange in vivo due to the presence
of tritium atoms in the aromatic B ring and/or the
galloyl moiety of EGCg. In this study, we established
chemical synthesis of [4-3H]EGCg, which is resistant to
hydrogen-tritium exchange. With this labeled EGCg,
the excretion route of EGCg was traced after intrave-
nous administration of [4-3H]EGCg to rats. Further,
apparent pharmacokinetic parameters of EGCg were
calculated on the basis of radioactivity in the blood after
intravenous dose, and some biliary metabolites were
also identified by HPLC analysis.
Green tea derived from the leaves of Camellia sinensis
contains a large amount of catechins. Green tea cate-
chins have been reported to have a wide range of
physiological functions, including antibacterial (1), an-
tiviral (2), antioxidative (3), hypocholesterolemic (4, 5),
hypoglycemic (6), and anticarcinogenic activities (7, 8).
In particular, because (-)-epigallocatechin gallate (EGCg)
is the most abundant component of green tea catechins
and has stronger physiological activities than the other
catechin compounds, there are expectations that it will
exert beneficial effects on human health. Accordingly,
there has also been a growing interest in determining
the metabolic profiles of EGCg in order to clarify the
mechanisms of the physiological activities in vivo. It has
been demonstrated that ingested EGCg was absorbed,
at least in part, into the portal vein in rats (9, 10), and
that a part of ingested EGCg entered into the systemic
circulation in rats (11-13) and humans (12, 14-16). In
addition, Chen et al. (13) have reported that the
bioavailability of ingested EGCg was less than 2% of
the dose in rats. With regard to excretion, EGCg and
its conjugates were not detected in human urine (14,
* To whom correspondence should be addressed. Phone:
+81-54-639-0080. Fax: +81-54-648-2001. E-mail: J DD04347@
nifty.ne.jp.
† Mitsui Norin Company.
MATERIALS AND METHODS
‡ Division of Organic Synthesis Facility, American Health
Foundation.
Gen er a l. EGCg was prepared from green tea by the method
previously described (18). Sodium borodeuteride (NaB2H4) was
obtained from Aldrich Chemical Co. (15681-89-7) and sodium
§ Carcinogenesis and Molecular Epidemiology, American
Health Foundation.
10.1021/jf0011236 CCC: $20.00 © 2001 American Chemical Society
Published on Web 01/13/2001

[4-3H]EGCg Synthesis and Its Metabolic Fate in Rats
J. Agric. Food Chem., Vol. 49, No. 2, 2001 1043
boro[3H]hydride (NaB3H4) was obtained from Amersham-
Pharmacia (TRK838), and all other reagents were of analytical
grade. For structural analyses of EGCg peracetate, 4-bromo-
EGCg peracetate, and [4-2H]EGCg, mass spectra were re-
corded on a J EOL DX-303 mass spectrometer and NMR
spectra were obtained on a J EOL Lambda-500 system. The
NMR chemical shifts were given in ppm (δ) with tetrameth-
ylsilane (TMS) as an internal standard. High-resolution fast
atom bombardment mass spectrometry (HR-FABMS) was
measured on a Finnigan Mat 95 instrument at the Mass
Spectrometry Service, University of Minnesota. Elemental
analysis was performed at Microanalytica Narita (J apan).
Mass and 3H NMR spectra of [4-3H]EGCg were measured at
Amersham Pharmacia Biotec (UK).
concentrated to remove acetonitrile. The product deposited in
the concentrated solution was collected with a membrane filter
(Advantec) and then washed with 500 mL of distilled water.
The resulting product was dried under reduced pressure to
give the compound 3 (600 mg, 46% yield). The purity of the
product was found to be 99% by analytical HPLC under the
same conditions as described in the preparation of 2. Full
assignments of 1H and 13C NMR spectra were done by HMBC
and HMQC experiments.
HR-FABMS (positive) m/z: 873.0935 [M+H]+, 875.0915
[M+H]+ (Calcd. for C38H33O19BrH, 873.0878; 875.0858). 1H
NMR (500 MHz, CDCl3): δ 2.24 (s, 6H, CH3-Ac × 2), 2.26 (s,
3H, CH3-Ac), 2.27 (s, 3H, CH3-Ac), 2.28 (s, 6H, CH3-Ac ×
2), 2.29 (s, 3H, CH3-Ac), 2.37 (s, 3H, CH3-Ac), 5.38 (d, 1H,
H-4, J ) 2.0 Hz), 5.64 (dd, 1H, H-3, J ) 0.9, 2.0 Hz), 5.87 (br
s, 1H, H-2), 6.74 (d, 1H, H-8, J ) 2.5 Hz), 6.77 (d, 1H, H-6, J
) 2.5 Hz), 7.30 (s, 2H, H-2′, 6′), 7.59 (s, 2H, H-2′′, 6′′). 13C NMR
(125 MHz, CDCl3): δ 20.2 (CH3-Ac × 2), 20.6 (CH3-Ac × 4),
21.1 (CH3-Ac × 2), 37.2 (C-4), 72.4 (C-3), 72.7 (C-2), 108.1
(C-8), 109.9 (C-6), 110.4 (C-4a), 118.8 (C-2′′, 6′′), 122.5 (C-2′,
6′), 126.6 (C-1′′), 134.1 (C-4′), 134.7 (C-1′), 139.3 (C-4′′), 143.4
(C-3′′, 5′′), 143.6 (C-3′, 5′), 150.1 (C-7), 151.9 (C-5), 154.3 (C-
8a), 163.2 (CdO), 166.1 (CdO-Ac), 166.7 (CdO-Ac), 167.4
(CdO-Ac × 2), 167.6 (CdO-Ac × 2), 167.9 (CdO-Ac), 168.5
(CdO-Ac).
P r ep a r a tion of EGCg P er a ceta te (2). EGCg (1 g, 2.18
mmol) (1) was added to a mixture of pyridine (14 mL) and
acetic anhydride (16 mL, 157 mmol) with stirring, and the
mixture was kept for 20 h at 45 °C in an oil bath. The reaction
mixture was then poured into ice-cold water (500 mL) with
vigorous stirring and was left to stand for 1 h to form a white
precipitate. The resulting precipitate was collected by filtration
with a sintered glass funnel (4 G), washed with 500 mL of
chilled water, and dried under reduced pressure. The resulting
product was chromatographed on a column of silica gel 60
(100-210 µm, 25 mm i.d. × 500 mm), with n-hexane/ethyl
acetate (1:3) (solvent A) as the eluent. According to TLC
analysis with solvent A, the fractions containing the compound
showing Rf value of 0.57 were pooled and evaporated to give
the compound 2 (1.7 g, 98.3% yield). The purity of the product
was found to be 99% by analytical HPLC. The analytical HPLC
was performed in a J ASCO liquid chromatograph apparatus
with a J ASCO 870 UV detector. Following are the conditions:
column, CAPCELL PAK C18 UG120 (Shiseido Co., Ltd., 4.6
mm i.d. × 250 mm); mobile phase, 50% aqueous acetonitrile
containing 0.01% trifluoroacetic acid (TFA); flow rate, 1 mL/
min; temperature, 35 °C; detection, UV 280 nm. Full assign-
ments of 1H and 13C NMR spectra were done by heteronuclear
multiple bond correlation (HMBC) and heteronuclear multiple
quantum coherence (HMQC) experiments.
P r ep a r a tion of [4-2H]EGCg (4). The compound 3 (200 mg,
229 µmol) was dissolved in 15 mL of anhydrous methanol, and
then 360 mg of NaB2H4 was added to the solution. After the
solution was gently stirred for 6 h at room temperature, 5%
phosphoric acid (80 mL) was added to the reaction mixture to
terminate the reaction. The resulting mixture was concen-
trated to remove methanol and then was applied to a DIAION
HP-20 (Mitsubishi Chemica Co.) column (20 mm i.d. × 150
mm) that had been equilibrated with distilled water. After
washing with 200 mL of distilled water, the column was eluted
with 100 mL of acetonitrile and the eluent was evaporated to
dryness. The resulting residue was further purified by pre-
parative HPLC. The preparative HPLC conditions were the
same as those for the purification of 3 except for using
acetonitrile/ethyl acetate/0.05% phosphoric acid (12:0.6:90, by
volume) as a mobile phase. The peak with a retention time of
31 min was collected and concentrated to remove organic
solvents. Desalting of the concentrated solution was performed
with the DIAION HP-20 column under the same conditions
as described above. The eluent was evaporated to dryness
under reduced pressure to obtain the compound 4 (55 mg, 52%
yield). The purity of the product was found to be more than
99% by HPLC under the same analytical conditions used for
3 except for using acetonitrile/ethyl acetate/0.05% phosphoric
acid (12:0.6:90, by volume) as a mobile phase. Full assignments
1
EI-MS m/z: 794 (M+). H NMR (500 MHz, CDCl3): δ 2.21
(CH3-Ac × 2), 2.23 (CH3-Ac), 2.25 (CH3-Ac), 2.26 (CH3-Ac
× 3), 2.27 (CH3-Ac), 2.98 (dd, 1H, H-4b (provisionally deter-
mined), J ) 2.3, 17.9), 3.04 (dd, 1H, H-4a (provisionally
determined), J ) 4.6, 17.9), 5.16 (br s, 1H, H-2), 5.62 (m, 1H,
H-3), 6.59 (d, 1H, H-8, J ) 2.2 Hz), 6.71 (d, 1H, H-6, J ) 2.2
Hz), 7.24 (s, 2H, H-2′, 6′), 7.60 (s, 2H, H-2′′, 6′′). 13C NMR (125
MHz, CDCl3): δ 20.1-21.1 (CH3-Ac × 8), 25.9 (C-4), 68.0 (C-
3), 76.5 (C-2), 108.1 (C-8), 109.0 (C-6), 109.4 (C-4a), 118.8 (C-
2′′, 6′′), 122.4 (C-2′, 6′), 127.5 (C-1′′), 134.4 (C-4′), 135.1 (C-1′),
139.0 (C-4′′), 143.3 (C-3′′, 5′′), 143.5 (C-3′, 5′), 149.7 (C-7), 149.8
(C-5), 154.8 (C-8a), 163.5 (CdO), 166.2 (CdO-Ac), 166.7 (Cd
O-Ac), 167.4 (CdO-Ac × 2), 167.5 (CdO-Ac × 2), 168.4 (Cd
O-Ac), 168.8 (CdO-Ac).
1
of H and 13C NMR were done by HMBC and HMQC experi-
ments.
Elemental analysis calcd. for C22H17D1O11‚2H2O: C, 53.32%;
H(+D), 4.68%. Found: C, 53.42%; H(+D), 4.63%. FABMS
(positive) m/z: 460 [M+H]+. FABMS (negative) m/z: 458
P r ep a r a tion of 4-Br om o-EGCg P er a ceta te (3). The
compound 2 (1.2 g, 1.51 mmol) was dissolved in carbon
tetrachloride (150 mL) by heating at 60 °C. N-Bromosuccin-
imide (290 mg, 1.63 mmol) and R,R′-azobisisobutyronitrile (3
mg) were added to the solution. The mixture was refluxed for
1 h. The reaction mixture was then cooled to room temperature
and evaporated to dryness under reduced pressure. The
resulting product was chromatographed on a column of silica
gel 60 (100-210 µm, 25 mm i.d. × 500 mm) with solvent A as
the eluent. The aliquot of each fraction was checked by the
same TLC analysis described in the method for the preparation
of 2. The fractions (Rf value of 0.63) were pooled and evapo-
rated to dryness under reduced pressure. For further purifica-
tion, the residue (1.23 g) was dissolved in acetonitrile (3 mL)
and the aliquots (3 × 1 mL) of the solution were subjected to
preparative HPLC. The preparative HPLC was performed with
a CAPCELL PAK C18 UG120 (Shiseido Co., Ltd., 20 mm i.d.
× 250 mm) column in the same HPLC systems described
above. Following are the HPLC conditions: mobile phase, 50%
aqueous acetonitrile containing 0.01% TFA; flow rate, 9.9 mL/
min; temperature, room temperature; detection, UV 280 nm.
The peak with a retention time of 61 min was collected and
1
[M-H]-. H NMR (500 MHz, CD3OD): δ 2.82 (d, 1H, H-4, J
) 2.5 Hz), 4.96 (br s, 1H, H-2), 5.52 (dd, 1H, H-3, J ) 1.5, 2.5
Hz), 5.95 (s, 2H, H-6, 8), 6.49 (s, 2H, H-2′, 6′), 6.94 (s, 2H, H-2′′,
6′′). 13C NMR (125 MHz, CD3OD): δ 28.1 (C-4), 69.9 (C-3), 78.7
(C-2), 95.9 (C-8), 96.6 (C-6), 99.4 (C-4a), 106.9 (C-2′′, 6′′), 110.3
(C-2′, 6′), 121.5 (C-1′′), 130.9 (C-4′), 133.8 (C-1′), 139.9 (C-4′′),
146.4 (C-3′′, 5′′), 146.7 (C-3′, 5′), 157.3 (C-8a), 157.9 (C-7), 158.0
(C-5), 167.7 (CdO).
P r ep a r a tion of [4-3H]EGCg (5). The compound 3 (10 mg,
11.5 µmol) was dissolved in 0.7 mL of anhydrous methanol,
and then 17 Ci of NaB3H4 (11 mg, 283 µmol) was added. The
reaction mixture was left to stand overnight at room temper-
ature. Subsequently, the reaction mixture was left to stand
for 3 h at room temperature after the addition of NaBH4 (8
mg, 211 µmol). Then the reaction was quenched by the addition
of 5% phosphoric acid (4 mL). The resulting mixture was
concentrated to about half of its original volume in vacuo. The
concentrated solution was extracted four times with ethyl
acetate (3 mL). The combined extracts were washed three
times with distilled water (3 mL) and evaporated to dryness.

1044 J. Agric. Food Chem., Vol. 49, No. 2, 2001
Kohri et al.
The residue was purified by preparative HPLC. The HPLC
was carried out on a Prodigy ODS-2 in a liquid chromatograph
apparatus equipped with a UV detector (280 nm) and a
â-radioactivity detector. The elution was done with methanol/
water/acetic acid (20:80:0.1, by volume) at a flow rate of 3 mL/
min at room temperature. The main radioactive peak was
collected, and the solvent was evaporated to dryness (4.2 mg,
19% yield). The residue was dissolved in ethanol solution
containing L-ascorbic acid at 8 mg/mL (30 mL). The ethanol
solution showed 30 mCi of radioactivity. The radiochemical
purity of the product was found to be 99.5% by HPLC. The
HPLC was carried out on a CAPCELL PAK C18 UG120
(Shiseido Co., Ltd., 4.6 mm i.d. × 250 mm) column in a liquid
chromatograph apparatus equipped with a UV detector (280
nm) and a â-radioactivity detector. The elution was done with
acetonitrile/ethyl acetate/0.05% phosphoric acid (12:0.6:90, by
volume) at a flow rate of 1 mL/min at 40°. The specific activity
was 13 Ci/mmol calculated by MS analysis. This procedure
was carried out at Amersham Pharmacia Biotec (UK). Fol-
lowing are the analytical data: FABMS (positive) m/z: 459
[M+H]+; m/z: 461 [M+H]+. 3H NMR (500 MHz, CD3OD): δ
2.98.
An im a ls a n d Diets. Male Wistar rats (6 weeks of age, 180-
210 g) purchased from Charles River J apan, Inc., were given
a polyphenol-free diet (10) for a week. Water was provided ad
libitum throughout this experiment (for bile-duct-cannulated
rats, 0.9% NaCl solution was supplied as drinking water). The
rats were fasted overnight before dosing labeled EGCg and
the diet was resumed 4 h after dosing.
Su r gica l P r oced u r e for Bile-Du ct Ca n n u la tion . After
the rats were anesthetized with pentobarbital sodium, the skin
of their head was incised and the abdomen was opened. A
micropolyethylene tube (outside diameter 0.61 mm) was fed
from the incision of the head skin to the abdomen through the
back subcutaneously and then the end of the tube was inserted
into the bile duct. The other end of the tube was put into a
vial containing 0.1 M sodium acetate buffer (pH 5.0) to collect
the bile. The skin and abdomen wall were sutured and
bandaged. The rats were left overnight to recover from
anesthetization prior to dosing.
buffer (pH 5.0) containing â-glucuronidase type H-1 (748 units
of â-glucuronidase and 39.2 units of sulfatase). The mixture
was incubated at 37 °C for 2 h with gentle shaking, then it
was extracted three times with the same volume of ethyl
acetate. The organic phase was concentrated to dryness and
the residue was dissolved in 1 mL of acetonitrile/ethyl acetate/
0.05% phosphoric acid (12:0.6:90, by volume) then subjected
to analytical HPLC. The HPLC was carried out on a CAPCELL
PAK C18 UG120 (Shiseido Co., Ltd., 4.6 mm i.d. × 250 mm)
in a J ASCO liquid chromatograph apparatus equipped with a
J ASCO UV-870 detector (280 nm) and a â-radioactivity detec-
tor (Beckman). The radioactivity detector was operated in
homogeneous liquid scintillation counting mode with addition
of 0.8 mL/min of Ultima-Flou M (Packard Bio Science B. V.)
to the eluent passed through the UV detector. The elution was
done with acetonitrile/ethyl acetate/0.05% phosphoric acid (12:
0.6:90, by volume) at a flow rate of 0.8 mL/min at 40 °C.
RESULTS
Syn t h et ic Con sid er a t ion s. We have developed
chemical synthesis of [4-2H]EGCg and [4-3H]EGCg. The
4 position of EGCg was selected for the labeling position
of deuterium or tritium to avoid hydrogen exchange. The
synthetic routes to [4-2H]EGCg and [4-3H]EGCg are
outlined in Figure 1. Thus, EGCg (compound 1) was first
peracetylated with acetic anhydride in pyridine and
EGCg peracetate (compound 2) was obtained in a good
yield (98%). Compound 2 was brominated with N-
bromosuccinimide and R, R′-azobisisobutyronitrile in
carbon tetrachloride for 1 h at reflux. The molecular
formula of the product was established as C38H33O19Br
1
by HR-FABMS analysis. The H NMR chemical shifts
of the brominated product were found to be similar to
those of 2. However, in the brominated product one of
two double-doublet signals (δ 3.04, 2.98) derived from
H-4a and H-4b protons of 2 disappeared and another
one (δ 5.38) emerged as a doublet with the downfield
shift. The 13C NMR chemical shifts of the brominated
product were analogous to those of 2, except for the
downfield shifts of C-4 (about 11 ppm) and C-3 (about
4 ppm), and the upper-field shift of C-2 (about 4 ppm).
These observations indicated that a hydrogen atom at
the 4 position of the compound 2 was replaced by a
bromine atom in the product. As a result, the product
was identified as 4-bromo-EGCg peracetate (the com-
pound 3). Compound 3 was treated with NaB2H4 in
anhydrous methanol at room temperature. The positive
and negative FABMS data of the product showed
[M+H]+ ion peak at m/z 460 and [M-H]- ion peak at
m/z 458, respectively. The MS analysis suggested that
the product is monodeuterated EGCg. The 1H NMR
chemical shifts of the deuterated product were found
to be superimposable on those of 1 (data not shown)
except for the disappearance of the signal (δ 2.98)
derived from the H-4b proton of 1. In addition, the 13C
NMR data of the product were almost the same as those
of 1. These results demonstrated that substitution of a
bromine at C-4 for a deuterium and complete deacetyl-
ation of 3 occurred simultaneously. Consequently, the
product was identified as [4-2H]EGCg (the compound
4).
Dose. The [4-3H]EGCg preparation (13 Ci/mmol) was
diluted with unlabeled EGCg to a specific activity of 22.9 mCi/
mmol. Thus, 200 µCi of the [4-3H]EGCg preparation (13 Ci/
mmol) was added to 2 mL of physiological saline containing
cold EGCg (4 mg). The solution was intravenously adminis-
tered to the rats (n ) 6) at a dose of 2 mL per kg.
Sa m p le Collection . After dosing, the bile-duct-cannulated
rats were placed in stainless steel metabolism cages. Urine
was collected in chilled vessels with dry ice. Bile was collected
as described in surgical procedure. The time intervals of the
collection were 0-4, 4-8, 8-12, 12-24, and 24-48 h after
dosing. Serial blood samples (50 µL) were collected from the
caudal vein at 1, 2, 4, 8, and 24 h. All rats were sacrificed at
48 h and blood samples (200 µL) were collected from inferior
vena cava. These samples were stored at -20 °C until use.
Det er m in a t ion of Tot a l R a d ioa ct ivit y in Sa m p les.
Each urine and bile sample was weighed, and aliquots (200
mg) of both samples were placed in scintillation vials. Each
blood sample (50 or 200 µL) was also put into a scintillation
vial. To the above vials containing urine, bile, and blood
samples was added 1 mL of Solvable (Packard Bio Science B.
V.). The scintillation vials were incubated for several hours
at 45 °C, then 0.5 mL of propanol and 200-300 µL of 30%
H2O2 were added, and the vials were incubated for 1 h at 50
°C to decolorize the samples. After the vials were cooled, 10
mL of Hionic Flour (Packard Bio Science B. V.) was added and
the resulting mixtures were left to stand in a darkroom for 1
h. The radioactivities of the mixtures were determined by
liquid scintillation spectroscopy (Aloka, LSC-3100).
Synthesis of [4-3H]EGCg (the compound 5) from 3 was
conducted by modification of the procedure for prepara-
tion of 4, as described in Materials and Methods. Thus,
compound 3 was first treated overnight with NaB3H4,
followed by additional NaBH4 for 3 h. This process
contributed to reducing the usage of NaB3H4 markedly.
The HPLC analysis revealed that the retention time of
the product coincided with that of compound 1. The
HP LC An a lysis of Meta bolites in Bile. An aliquot (1 mL)
of bile sample collected from six rats during the first 4 h was
evaporated to dryness, and the residue was dissolved in 1 mL
of 0.1M sodium acetate buffer (pH 5.0) containing 1% ascorbic
acid and 0.15 mM ethylenediaminetetraacetic acid (EDTA).
The solution was mixed with 20 µL of 0.1M sodium acetate

[4-3H]EGCg Synthesis and Its Metabolic Fate in Rats
J. Agric. Food Chem., Vol. 49, No. 2, 2001 1045
F igu r e 1. Synthetic route to [4-2H]EGCg and [4-3H]EGCg.
positive FABMS data of the product showed [M+H]+
ion peak at m/z 459 and [M+H]+ ion peak at m/z 461,
suggesting that the product is a mixture of compound
1 and monotritiated EGCg. The 3H NMR analysis of the
product showed a signal only at δ 2.98. The chemical
shift was superimposable on that of the signal (δ 2.98)
derived from H-4b proton of compound 1, which disap-
peared in the case of 4. From these data, [4-3H]EGCg
(the compound 5) was confirmed to be produced from
3. The radiochemical purity of 5 was found to be 99.5%
by HPLC analysis, and the specific activity was calcu-
lated to be 13 Ci/mmol by MS analysis. This is the first
report on labeled EGCg in which deuterium or tritium
is located at the 4 position.
E xcr et ion of R a d ioa ct ivit y in Bile a n d Ur in e.
[4-3H]EGCg was administered intravenously to bile-
duct-cannulated rats to clarify the excretion route of
EGCg. The cumulative excretion of radioactivity in bile
and urine, expressed as percentage of the dose, is
illustrated in Figure 2. Only 2.0% of the dose was
recovered in the urine over a period of 48 h. On the other
hand, cumulative biliary excretion of radioactivity ac-
counted for 77.0% of the dose in 48 h and rapid biliary
excretion (56.9% of the dose) was observed within the
first 4 h after intravenous dosing. These results showed
that bile was the major excretion route for EGCg and
the excretion ratio of bile to urine was about 97:3.
P h a r m a cok in etic P a r a m eter s of Ra d ioa ctivity
Der ived fr om [4-3H]EGCg in Blood . Figure 3 shows
the time-course of the radioactivity in the blood after
intravenous administration of [4-3H]EGCg to bile-duct-
cannulated rats. These time-course data were analyzed
by the MULTI program (19). The best fit was achieved
with a two-compartment mode, and apparent pharma-
cokinetic parameters for EGCg were then calculated.
Distribution half-life (T1/2R), elimination half-life (T1/2â),
and distribution volume (Vd) were 1.04 h ( 0.6, 9.34 h
( 2.8. and 14.45 ( 9.2 dl/kg, respectively. The area
F igu r e 2. Cumulative excretion of radioactivity in bile (b)
and urine (O) after intravenous administration of [4-3H]EGCg
to bile-duct-cannulated rats. The amount of excreted radio-
activity was expressed as cumulative percentage of adminis-
tered dose. Each point represents the mean ( SD of 6 rats.
under the curve (AUC0-∞) was estimated to be 2242.8
( 1222 µg-equivalent‚min/mL.
HP LC An a lysis of Meta bolites in Bile. In this
study, to deconjugate the biliary metabolites, bile
samples collected during the first 4 h after intravenous
administration of [4-3H]EGCg to rats were treated with
glucuronidase/sulfatase mixture as described in Materi-
als and Methods. Then, the metabolites formed were
extracted with ethyl acetate and analyzed by HPLC
equipped with radioisotope and UV detectors. Identifi-
cation of the metabolites was performed by comparison
with the retention time of six deconjugated biliary
metabolites (EGCg, 4′′-O-methyl-EGCg, 3′′-O-methyl-
EGCg, 3′-O-methyl-EGCg, 4′-O-methyl-EGCg, and 4′,4′′-
di-O-methyl-EGCg) identified in the rat bile collected
after oral dose of EGCg in our previous study (20).

1046 J. Agric. Food Chem., Vol. 49, No. 2, 2001
Kohri et al.
positions labeled were estimated to be in the B ring and/
or galloyl moiety. In general, it is likely that tritium
labeled in aromatic rings will be susceptible to undergo-
ing hydrogen exchange. With this concern in mind, in
this study we developed chemical synthesis of [4-3H]-
EGCg with high radiochemical purity (99.5%) and
specific activity (13 Ci/mmol). The radiochemical purity
of [4-3H]EGCg stored at -20 °C in ethanol solution was
still 98.6% after three months. Further, after an oral
dose of [4-3H]EGCg to rats the urine sample excreted
within 48 h was analyzed by HPLC, but no tritiated H2O
peak was observed (data not shown). J udging from these
results, it is thought that the tritium at the 4 position
of EGCg is not replaced by hydrogen from water in the
body and there is no concern over the stability of the
labeled EGCg.
In some previous studies on the metabolism of EGCg,
it has been reported that part of the EGCg orally
administered to rats or humans was detected in the
plasma as EGCg or the glucuronide/sulfate conjugated
forms thereof (11-16). On the other hand, neither EGCg
nor its conjugates were detected in the urine (13, 14).
In light of these results it is supposed that EGCg
absorbed into the body is excreted mainly in the bile.
However, studies up until now have still been lacking
in a clear explanation of the excretion route of EGCg.
We attempted to define this point using [4-3H]EGCg
intravenously administered to bile-duct-cannulated rats.
Results showed that most of the radioactivity derived
from [4-3H]EGCg was excreted in the bile, and the
excretion ratio of bile to urine was 97:3. Millburn et al.
(21) proposed that an important factor in facilitating the
excretion of aromatic compounds into the bile was a
molecular weight of more than 325 ( 50. Millburn et
al. (21) and Bu¨lles et al. (22) reported that facilitation
of excretion of a compound into the bile was influenced
not only by the original molecular weight of a compound,
but also by an increase in the molecular weight and the
polarity due to conjugation of a compound with gluc-
uronic acid and sulfate. In the case of EGCg, not only
does it have a molecular weight of 458, but it also can
be found mostly in the forms of glucuronide and/or
sulfate conjugates in the plasma (14) and bile (20). The
present study confirmed that EGCg is mostly excreted
in the bile, and hence EGCg is a representative com-
pound which follows the biliary excretion rules proposed
by Millburn et al. (21) and Bu¨lles et al. (22).
In the present study we calculated the apparent
pharmacokinetic parameters of EGCg on the basis of
radioactivity (total amount of EGCg and its metabolites)
in the blood after intravenous administration of [4-3H]-
EGCg. Previously, Chen et al. (13) had calculated the
pharmacokinetic parameters of EGCg based on the
amount of EGCg and its conjugates in the plasma.
Results of the present study gave values of T1/2R (1.04
h) and T1/2â (9.34 h) which were much larger than those
(T1/2R (11.8 min) and T1/2â (135.1 min)) reported by Chen
et al. Moreover, although the dosage in this study was
lower than that used by Chen et al., the value of AUC
(2242.8 µg-equivalent‚min/mL) was much larger than
that (AUC (143.2 µg-equivalent‚min/mL)) reported by
Chen et al. A possible explanation for these differences
may be that the parameters in the present study were
calculated on the basis of the total amount of all
metabolites, whereas Chen et al. calculated the param-
eters on the basis of only EGCg and its conjugates.
Accordingly if metabolites with a slower elimination rate
F igu r e 3. Blood concentration-time profiles of radioactivity
after intravenous administration of [4-3H]EGCg to bile-duct-
cannulated rats. Nanogram equivalent of EGCg was calculated
from the amount of radioactivity in blood. Each point repre-
sents the mean ( SD of 6 rats.
F igu r e 4. Representative HPLC profile of deconjugated forms
of radioactive metabolites in rat bile sample. Prior to HPLC
analysis bile samples were treated with â-glucuronidase/
sulfatase mixture and the metabolites formed were extracted
with ethyl acetate as described in Materials and Methods.
Figure 4 shows a representative HPLC profile of the
deconjugated forms of radioactive metabolites in the
bile. Four radioactive peaks of M-1, M-2, M-3, and M-4
were detected with a retention time of 13.2 , 22.5, 27.7,
and 48.5 min, respectively. The peaks of M-1, M-2, M-3,
and M-4 were identified as EGCg, 4′′-O-methyl-EGCg,
4′-O-methyl-EGCg, and 4′,4′′-di-O-methyl-EGCg, re-
spectively; whereas 3′-O-methyl-EGCg and 3′′-O-methyl-
EGCg were not detected. The radioactivity in EGCg, 4′′-
O-methyl-EGCg, 4′-O-methyl-EGCg, and 4′,4′′-di-O-
methyl-EGCg accounted for 0.8, 1.5, 0.6, and 14.7% of
administered radioactivity, respectively. Furthermore,
a major unknown peak with the retention time of 100
min was detected, and the radioactivity accounted for
8% of the dose.
DISCUSSION
Radiolabeling technique is an invaluable tool for
investigating metabolic fates of compounds in in vivo
studies. Suganuma et al. (17) have reported that [3H]-
EGCg labeled with tritium gas was synthesized, and the

[4-3H]EGCg Synthesis and Its Metabolic Fate in Rats
J. Agric. Food Chem., Vol. 49, No. 2, 2001 1047
ABBREVIATIONS USED
are produced from EGCg and its conjugates in the blood,
it could be expected that the values of half-life and AUC
would increase. Piskula and Terao (23) reported that
methylated (-)-epicatechin (EC) conjugates were me-
tabolites with a slower elimination rate than EC and
its conjugates. Thus, it seems likely that methylated
EGCg conjugates were metabolites with a slower elimi-
nation rate and this may possibly explain the reason
for the increase in values of half-life and AUC.
EGCg, (-)-epigallocatechin gallate; EDTA, ethylene-
diaminetetraacetic acid; TMS, tetramethylsilane; HR-
FABMS, high-resolution fast atom bombardment mass
spectrometry; T1/2R, distribution half-life; T1/2â, elimina-
tion half-life; Vd, distribution volume; AUC, area under
the blood concentration vs time curve.
ACKNOWLEDGMENT
We thank Professor Naoto Oku of the School of
Pharmaceutical Sciences, University of Shizuoka, for
allowing us the use of the radioisotope laboratory.
In our previous study (20) we found that after oral
administration of EGCg, most EGCg metabolites in the
bile were in glucuronide and/or sulfate forms. In addi-
tion, following treatment of the bile with â-glucuroni-
dase/sulfatase, the deconjugated forms of the biliary
metabolites were found to be EGCg, 3′-O-methyl-EGCg,
3′′-O-methyl-EGCg, 4′-O-methyl-EGCg, 4′′-O-methyl-
EGCg, and 4′,4′′-di-O-methyl-EGCg. In this study, using
the above six EGCg metabolites as standards, we
identified the biliary metabolites after intravenous
administration of [4-3H]EGCg to rats. Following are the
biliary metabolites that we detected: 4′,4′′-di-O-methyl-
EGCg (14.7% of dose), 4′′-O-methyl-EGCg (1.5% of dose),
EGCg (0.8% of dose), and 4′-O-methyl-EGCg (0.6% of
dose). In addition, a major unknown peak with a longer
retention time (100 min) than that of 4′,4′′-di-O-methyl-
EGCg (48.5 min) was detected. The amount of the peak
was estimated to be 8% of the dose. No information is
available on this peak, but if it is a methyl-derivative
of EGCg, judging from the retention time obtained from
HPLC analysis it could be assumed to be trimethylated
or further methylated EGCg. Furthermore, 3′-O-methyl-
EGCg and 3′′-O-methyl-EGCg were not observed in this
study. The main biliary metabolite (in the deconjugated
form) was EGCg in the previous study but was 4′,4′′-
di-O-methyl-EGCg in the present study. Thus, there are
some differences in the metabolites between oral and
intravenous administration. After intravenous admin-
istration, EGCg enters the liver directly. On the other
hand, in the case of oral administration, since there is
absorption process from the intestine, a large part of
EGCg absorbed may first undergo glucuronidation (23),
after which the EGCg conjugate enters the liver.
Subsequently, intact EGCg and its conjugate may
undergo further modification, in particular methylation
in the liver. Although intact EGCg seems to be suscep-
tible to methylation, the conjugated EGCg may be more
resistant to methylation than intact EGCg, and this
could be surmised to be the reason for the differences
in the biliary metabolites according to the administra-
tion route.
LITERATURE CITED
(1) Fukai, K.; Ishigami, T.; Hara, Y. Antibacterial activity
of tea polyphenols against phytopathogenic bacteria.
Agric. Biol. Chem. 1991, 55, 1895-1897.
(2) Nakayama, M.; Suzuki, K.; Toda, M.; Okubo, S.; Hara,
Y.; Shimamura, T. Inhibition of the infectivity of influ-
enza virus by tea polyphenols. Antiviral Res. 1993, 21,
289-299.
(3) Nanjo, F.; Honda, M.; Okushio, K.; Matsumoto, N.;
Ishigaki, F.; Ishigami, T.; Hara, Y. Effects of dietary tea
catechins on R-tocopherol levels, lipid peroxidation, and
erythrocyte deformability in rats fed on high palm oil
and perilla oil diets. Biol. Pharm. Bull. 1993, 16, 1156-
1159.
(4) Muramatsu, K.; Fukuyo, M.; Hara, Y. Effect of green
tea catechins on plasma cholesterol level in cholesterol-
fed rats. J . Nutr. Sci. Vitaminol. 1986, 32, 613-622.
(5) Chisaka, T.; Matsuda, H.; Kubomura, Y.; Mochizuki, M.;
Yamahara, J .; Fujimura, H. The effect of crude drugs
on experimental hypercholesteremia: Mode of action of
(-)-epigallocatechin gallate in tea leaves. Chem. Pharm.
Bull. 1988, 36, 227-233.
(6) Matsumoto, N.; Ishigaki, F.; Ishigaki, A.; Iwashina, H.;
Hara, Y. Reduction of blood glucose levels by tea
catechin. Biosci. Biotechnol. Biochem. 1993, 57, 525-
527.
(7) Xu, Y.; Ho, C.-T.; Amin, S. G.; Han, C.; Chung, F.-L.
Inhibition of tobacco-specific nitrosamine-induced lung
tumorigenesis in A/J mice by green tea and its major
polyphenol as antioxidants. Cancer Res. 1992, 52, 3875-
3879.
(8) Matsumoto, N.; Kohri, T.; Okushio, K.; Hara, Y. Inhibi-
tory effects of tea catechins, black tea extract and oolong
tea extract on hepatocarcinogenesis in rats. J pn. J .
Cancer Res. 1996, 56, 1034-1038.
(9) Okushio, K.; Matsumoto, N.; Suzuki, M.; Nanjo, F.;
Hara, Y. Absorption of (-)-epigallocatechin gallate into
rat portal vein. Biol. Pharm. Bull. 1995, 18, 190-191.
(10) Okushio, K.; Matsumoto, N.; Kohri, T.; Suzuki, M.;
Nanjo, F.; Hara, Y. Absorption of tea catechins into rat
portal vein. Biol. Pharm. Bull. 1996, 19, 326-329.
(11) Unno, T.; Takeo, T. Absorption of (-)-epigallocatechin
gallate into the circulation system of rats. Biosci. Bio-
technol. Biochem. 1995, 59, 1558-1559.
In this study we developed the method of chemical
synthesis of [4-3H]EGCg which is resistant to hydrogen-
tritium exchange. With this labeled EGCg we confirmed
that the main excretion route of EGCg is by way of the
bile. By comparison of apparent pharmacokinetic pa-
rameters of EGCg calculated in the present study with
those calculated by Chen et al., it can be postulated that
metabolites with a slower elimination rate than that of
EGCg and its conjugates, were present in blood after
intravenous administration of EGCg to rats. In addition,
HPLC analysis of the bile showed that the main biliary
metabolite of EGCg dosed intravenously was the con-
jugated form of 4′,4′′-di-O-methyl-EGCg and it is sug-
gested that the metabolites produced varied according
to different administration routes (oral and intrave-
nous).
(12) Nakagawa, K.; Miyazawa, T. Chemiluminescence-high-
performance liquid chromatographic determination of
tea catechin, (-)-epigallocatechin 3-gallate, at picomole
levels in rat and human plasma. Anal. Biochem. 1997,
248, 41-49.
(13) Chen, L.; Lee, M.-J .; Li, H.; Yang, C. S. Absorption,
distribution, and elimination of tea polyphenols in rats.
Drug Metabol. Dispos. 1997, 25, 1045-1050.
(14) Lee, M.-J .; Wang, Z.-Y.; Li, H.; Chen, L.; Sun, Y.; Gobbo,
S.; Balentine, D. A.; Yang, C. S. Analysis of plasma and
urinary tea polyphenols in human subjects. Cancer
Epidemiol. Biomark. Prev. 1995, 4, 393-399.
(15) Unno, T.; Kondo, K.; Itakura, H.; Takeo, T. Analysis of
(-)-epigallocatechin gallate in human serum obtained
after ingesting green tea. Biosci. Biotechnol. Biochem.
1996, 60, 2066-2068.

1048 J. Agric. Food Chem., Vol. 49, No. 2, 2001
Kohri et al.
(16) Yang, C. S.; Chen, L.; Lee, M.-J .; Balentine, D.; Kuo,
M. C.; Schantz, S. P. Blood and urine levels of tea
catechins after ingestion of different amounts of green
tea by human volunteers. Cancer Epidemiol. Biomark.
Prev. 1998, 7, 351-354.
(17) Suganuma, M.; Okabe, S.; Oniyama, M.; Tada, Y.; Ito,
H.; Fujiki, H. Wide distribution of [3H](-)-epigallocat-
echin gallate, a cancer preventive tea polyphenol, in
mouse tissue. Carcinogenesis 1998, 19, 1771-1776.
(18) Matsuzaki, T.; Hara, Y. Antioxidative activity of tea leaf
catechins. Nippon Nogeikagaku Kaishi (in J apanese).
1985, 59, 129-134.
(21) Millburn, P.; Smith, R. L.; Williams, R. T. Biliary
excretion of foreign compounds. Biochem. J . 1967, 105,
1275-1281.
(22) Bu¨lles, V. H.; Bu¨lles, J .; Krumbiegel, G.; Mennicke, W.
H.; Nitz, D. Untersuchungen zur verstoffwechselung
und zur ausscheidung von silybin bei der ratte. Arz-
neim.-Forsch. 1975, 25, 902-905.
(23) Piskula, M. K.; Terao, J . Accumulation of (-)-epicatechin
metabolites in rat plasma after oral administration and
distribution of conjugation enzymes in rat tissues. J .
Nutr. 1998, 128, 1172-1178.
(19) Yamaoka, K.; Tanigawara, Y.; Nakagawa, T.; Uno, T.
A pharmacokinetic analysis program (MULTI) for micro-
computer. J . Pharm. Dyn. 1981, 4, 879-885.
(20) Kida, K.; Suzuki, M.; Matsumoto, N.; Nanjo, F.; Hara,
Y. Identification of biliary metabolites of (-)-epigallo-
catechin gallate in rats. J . Agric. Food Chem. 2000, 48,
4151-4155.
Received for review September 8, 2000. Revised manuscript
received November 21, 2000. Accepted November 24, 2000.
J F0011236