Metabolism of (-)-epigallocatechin gallate by rat intestinal flora

Article abstract of DOI:10.1021/jf903375s

Anaerobic metabolism of ( - )-epigallocatechin gallate (EGCg) by rat intestinal bacteria was investigated in vitro. First, intestinal bacteria which are capable of hydrolyzing EGCg to ( - )epigallocatechin (EGC) and gallic acid (2) were screened with 169 strains of enteric bacteria. As a result, Enterobacter aerogenes, Raoultella planticola, Klebsiella pneumoniae susp. pneumoniae, and Bifidobacterium longum subsp. infantis were found to hydrolyze EGCg. Subsequent steps of EGCg metabolism are degradation of EGC (1) by intestinal bacteria. Then, EGC was incubated with rat intestinal bacteria in 0.1 M phosphate buffer (pH 7.1) and the degradation products were analyzed with time by HPLC or LC-MS. Further, the products formed from EGC were isolated and identified by LC-MS and NMR analyses. The results revealed that EGC was converted first to 1-(3', 4', 5'-trihydroxyphenyl)-3-(2 , 4 , 6 -trihydroxyphenyl)propan-2-ol (3) by reductive cleavage between 1 and 2 positions of EGC, and subsequently metabolite 3 was converted to 1-(3', 5'-dihydroxyphenyl)-3-(2 , 4 , 6 -trihydroxyphenyl)propan-2-ol (4) followed by the conversion to 5-(3, 5-dihydroxyphenyl)-4-hydroxyvaleric acid (5) by decomposition of the phloroglucinol ring in metabolite 4. This degradation pathway was considered to be the major route of EGCg metabolism in the in vitro study, but two minor routes were also found. In addition to the in vitro experiments, metabolites 3, 4, 5, and 6 were detected as the metabolites after direct injection of EGC into rat cecum. When EGCg was administered orally to the rats, metabolites 4, 5, 6, 11, and 12 were found in the feces. Among the metabolites detected, metabolite 5 was dominant both in the cecal contents and feces. These findings suggested that the metabolic pathway of EGCg found in the in vitro study may be regarded as reflecting its metabolism in vivo.

Full text of DOI:10.1021/jf903375s

J. Agric. Food Chem. 2010, 58, 1313–1321 1313
DOI:10.1021/jf903375s
Metabolism of (-)-Epigallocatechin Gallate by Rat
Intestinal Flora
A
KIKO
T
AKAGAKI
*
AND
F
UMIO
NANJO
Laboratory of Tea Function R&D Food Research Laboratories, Mitsui Norin Co., Ltd. 223-1 Miyabara,
Fujieda-shi, Shizuoka, 426-0133, Japan
Anaerobic metabolism of (-)-epigallocatechin gallate (EGCg) by rat intestinal bacteria was
investigated in vitro. First, intestinal bacteria which are capable of hydrolyzing EGCg to (-)-
epigallocatechin (EGC) and gallic acid (2) were screened with 169 strains of enteric bacteria. As
a result, Enterobacter aerogenes, Raoultella planticola, Klebsiella pneumoniae susp. pneumoniae,
and Bifidobacterium longum subsp. infantis were found to hydrolyze EGCg. Subsequent steps of
EGCg metabolism are degradation of EGC (1) by intestinal bacteria. Then, EGC was incubated with
rat intestinal bacteria in 0.1 M phosphate buffer (pH 7.1) and the degradation products were
analyzed with time by HPLC or LC-MS. Further, the products formed from EGC were isolated and
identified by LC-MS and NMR analyses. The results revealed that EGC was converted first to
1-(30,40,50-trihydroxyphenyl)-3-(200,400,600-trihydroxyphenyl)propan-2-ol (3) by reductive cleavage be-
tween 1 and 2 positions of EGC, and subsequently metabolite 3 was converted to 1-(30,50-
dihydroxyphenyl)-3-(200,400,600-trihydroxyphenyl)propan-2-ol (4) followed by the conversion to
5-(3,5-dihydroxyphenyl)-4-hydroxyvaleric acid (5) by decomposition of the phloroglucinol ring in
metabolite 4. This degradation pathway was considered to be the major route of EGCg metabolism
in the in vitro study, but two minor routes were also found. In addition to the in vitro experiments,
metabolites 3, 4, 5, and 6 were detected as the metabolites after direct injection of EGC into rat
cecum. When EGCg was administered orally to the rats, metabolites 4, 5, 6, 11, and 12 were found
in the feces. Among the metabolites detected, metabolite 5 was dominant both in the cecal contents
and feces. These findings suggested that the metabolic pathway of EGCg found in the in vitro study
may be regarded as reflecting its metabolism in vivo.
KEYWORDS: Metabolism; epigallocatechin gallate; intestinal flora
INTRODUCTION
These compounds are thought to be the ring-fission metabolites
of EGCg and/or EGC by intestinal microflora. On the basis of the
findings, the metabolic pathway of EGCg have been proposed in
gut tract (12). Hattori et al. (19, 20) has identified EGCg
metabolites formed by human intestinal bacteria and then initial
stage of EGCg metabolism was proposed. However, up until now
there has been no detailed study on the whole metabolic pathway
of EGCg by intestinal bacteria.
In this study, we investigated in detail the metabolic pathways
of EGCg by rat intestinal bacteria in vitro. Further, we examine
whether or not the in vitro metabolic pathway mirrors its
metabolism in vivo.
It is well recognized that tea catechins have various physiolo-
gical functions including antioxidative, blood cholesterol low-
ering, blood sugar level lowering, and cancer preventive
activities (1-4). Many studies have shown that (-)-epigalloca-
techin gallate (EGCg), the most abundant catechin, has strong
physiological activity. Accordingly, there has also been great
interest in the metabolism of EGCg because of the importance of
elucidating the mechanism responsible for its biological activity.
Many studies to date have addressed the metabolism of EGCg in
mammalia such as mouse, rat, and human. After oral adminis-
tration of EGCg, it has been found in rat and human
plasma (5-9). Distribution and excretion of EGCg (10-12) have
been also reported in addition to its pharmacokinetics study (13,
14). The metabolites of EGCg excreted in bile and urine were
further identified (10, 15-18). Among the metabolites identified,
5-(30,40-dihydroxyphenyl)-γ-valerolactone, 5-(30,40,50-trihydro-
xyphenyl)-γ-valerolactone, and 5-(30,50-dihydroxyphenyl)-γ-
valerolactone were found in human (17, 18) and rat urine (12).
MATERIALS AND METHODS
Chemicals. (-)-Epigallocatechin-3-O-gallate (EGCg) and (-)-epigallo-
catechin (EGC) were obtained from Sigma-Aldrich Japan (Tokyo). General
anaerobic medium (GAM) was obtained from Nissui Pharmaceutical
(Tokyo, Japan). All other chemicals were available products of analytical
grade or HPLC grade.
Animals. Male Wistar rats were purchased from Charles Liver
Laboratories Inc. (Yokohama, Japan) and housed in a room at 23 (
3 °C and 50 ( 5% relative humidity, with a 12 h light-dark cycle. The rats
were given normal pelleted food (Oriental Yeast Ltd., Tokyo, Japan)
*Corresponding author: tel: þ81-54-648-2600; fax: þ81-54-648-
2001; e-mail: tgaki@mitsui-norin.co.jp.
© 2009 American Chemical Society
Published on Web 12/31/2009
pubs.acs.org/JAFC

1314 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Takagaki and Nanjo
during experimental period. In this study, 15-21 weeks of age of rats were
used. All experimental procedures were in accordance with the guidelines
for animal experiments of Food Research Laboratories, Mitsui Norin Co.,
Ltd.
Screening of Enteric Bacteria Capable of Hydrolyzing EGCg.
After sterilization of 9 mL of 0.1 M phosphate buffer (pH 7.1) by
autoclaving at 121 °C for 15 min, EGCg solution (1 mL, 10 mg/mL)
filtered through sterilized membrane filter DISMIC-25cs, 0.2 μm (Toyo
Roshi Kaisha Ltd., Tokyo, Japan) was added to the buffer. Each bacterial
strain (22 genera, 169 strains) was inoculated to the solution and then 2 mL
of sterile mineral oil was layered over. The mixture was incubated at 37 °C
for 24 h and then was centrifuged at 3500g for 10 min at room temperature.
The supernatant was analyzed by HPLC to select the enterobacteria with
the ability to hydrolyze EGCg. Analytical HPLC was carried out with a
250 ꢀ 4.6 mm i.d., 5 μm, Capcell Pak C18 UG120 (Shiseido Co. Ltd.,
Tokyo, Japan) in an Waters 2695 Separations Module (Nihon Waters,
Tokyo, Japan) apparatus equipped with an Waters 2489 UV/VS Detector
(270 nm). Elution was done with distilled water/methanol/acetonitrile/
phosphoric acid (85/10/5/0.01, v/v/v/v) at a flow rate of 1 mL/min at 40 °C.
Time-Course Analysis of the Metabolite Formation from EGC
(1) by Rat Intestinal Flora. Incubation method of EGC with rat
enterobacteria was done according to that of Meselhy et al. (22). Fresh
cecal contents (3.3 g) obtained from three male Wistar rats (16 weeks of
age) were put into 300 mL of GAM broth and then incubated at 37 °C
under anaerobic condition with an Anaeropack system (Mitsubishi Gas
Chemical, Tokyo, Japan). After 3 days, the culture was centrifuged by
Avanti J-25 (Beckman Coulter, Tokyo, Japan) at 15000g for 20 min at
4 °C, and the harvested cells were washed once with sterile water (200 mL).
The cells were suspended in 150 mL of prefiltered 0.1 M phosphate buffer
(pH 7.1), and 200 mg of EGC in 20 mL of 5% aqueous methanol
prefiltrated with the sterilized membrane filter was added to the cell
suspension. The mixture was incubated under anaerobic condition with
the Anaeropack system. After incubation for 24, 48, 72, 96, and 168 h,
aliquots (1 mL) of the incubation mixture were withdrawn in an anaerobic
glovebox under CO2 atmosphere. These samples were centrifuged by MX-
301 centrifuge (TOMY Seiko, Tokyo, Japan) at 15000g for 10 min at 4 °C.
The resulting supernatants were filtered with the membrane filter and then
analyzed by Surveyor HPLC and LCQ Deca XPplus system (Thermo
Fisher Scientific, USA). The above experiment was repeated more than 15
times with different cecal contents of different rats (15-21 weeks).
Purification of EGC Metabolites. The metabolite formation from
EGC by rat intestinal flora was performed according to the procedure as
described above. After incubation for 24, 48, and 96 h, 50 mL each samples
of the incubation mixture were taken out in the anaerobic glovebox. Each
sample was centrifuged at 15000g for 20 min at 4 °C to remove bacterial
cells. The resulting supernatant was adjusted pH to about 3.5 with 5 M
HCl and then extracted three times with equal volume of ethyl acetate. The
ethyl acetate fraction was evaporated to dryness under reduced pressure.
The residue obtained from the 24-h incubation mixture was used for the
isolation of 3 and 8, and the residue from the 48-h mixture for 4. Similarly,
the isolation of 5 to 7 was performed with the 96-h incubation mixture.
Each residue obtained from the 24- and 48-h incubation mixtures were
dissolved in 3 mL of 5% aqueous methanol and the solution was subjected
to preparative HPLC forpurification of 3, 4, and 8. Preparative HPLC was
performed with a 150 mm ꢀ 20 mm i.d., 5 μm, Capcellpak C18 MG
column (Shiseido Co. Ltd., Tokyo, Japan) in a JASCO HPLC 800 series
system (JASCO, Tokyo, Japan). The column was eluted with a linear
gradient of solvent, starting with 10% (v/v) methanol aqueous solution
containing 0.5% (v/v) acetic acid and ending with 60% (v/v) aqueous
methanol containing 0.5% (v/v) acetic acid at a flow rate of 9.7 mL/min at
40 °C. The elution pattern was monitored by measuring the absorbance at
270 nm. Each fraction containing metabolite was collected and evaporated
to dryness. The residue was dissolved in 5 mL of distilled water and then
concentrated to dryness. This procedure was repeated two more times to
remove acetic acid completely. The resultant residues were individually
dissolved in a small amount of distilled water and freeze-dried to obtain
metabolite 3 (17 mg), metabolite 4 (15 mg), and metabolite 8 (5 mg). For
the isolation of 5 to 7, the ethyl acetate fraction obtained from the 96-h
incubation mixture was extracted with the one-fifth volume of 20 mM
aqueous Na2CO3 solution. In this process, 6 remained in the organic layer,
and 5 and 7 were distributed in the aqueous layer. Metabolite 6 in the ethyl
Figure 1. Time profile of EGC (metabolite 1) metabolism by rat intestinal
flora via the pathway of route I.
acetate fraction was purified in the same manner as was metabolite 3 to
obtain 5 mg of purified 6. The aqueous solution containing 5 and 7 was
evaporated to about 5 mL, and the concentrate was adjusted to around pH
5.0 with 5 M acetic acid. After removal of the precipitate (probably sodium
acetate) by centrifugation, the supernatant was subjected to preparative
HPLC. The HPLC were performedunder the same conditions as described
above except for using mobile phases without acetic acid. Metabolite 7
fraction was evaporated under reduced pressure and then freeze-dried to
obtain purified metabolite 7 (3 mg). On the other hand, 5 fraction was
concentrated to remove the organic solvents and the resulting concentrate
was applied to an ion exchange column 10 ꢀ 65 mm, 5.1 mL of Diaion
SK1B Naþ form (Mitsubishi Chemical, Co., Tokyo, Japan) which had
been equilibrated with distilled water. The column was eluted with 5 vol of
distilled water. The eluate was concentrated to about 3 mL and then
lyophilized to obtain 21 mg of metabolite 5 as sodium salt.
When different intestinal flora was used, metabolites 9-12 were formed
from EGC in addition to 3, 5, and 6 (Figure 2A). Then, the 24-h incubation
mixture for the isolation of 9 and 10 was adjusted pH to 1.5 with 5 M
phosphoric acid and the metabolites were extracted three times with same
volume of a mixture of ethyl acetate/n-butanol (1/1, v/v). The organic layer
was then treated two times with half volume of 100 mM aqueous Na2CO3
containing 0.1% sodium ascorbate. In this procedure, 9 passed into the
aqueous layer, while 10 remained in the organic layer. The aqueous layer
was adjusted to around pH 7.0 with 5 M HCl and was concentrated to
about 15 mL. To the concentrate 75 mL of ethanol was added and then it
was centrifuged to remove insoluble material. The supernatant was
concentrated to 1 mL and then the concentrate was adjusted to pH 2-3
with 2 M HCl and subjected to preparative HPLC. The HPLC conditions
were the same as for metabolite 5 as mentioned above. Metabolite 9
fraction was collected and concentrated to remove organic solvent. The
resultant solution was treated with the Diaion SK1B Naþ form column in
the same manner as for 5. The eluate was concentrated and lyophilized to
obtain 5 mg of purified 9. The ethyl acetate-n-butanol layer containing 10
was concentrated to dryness. The resultant residue was treated in the same
manner as was 3 to obtain 3 mg of purified 10. Metabolites 11 and 12 were
purified from 72-h and 168-h incubation mixtures, respectively, using the
same procedure as for 3. Finally, 10 mg of 11 and 19 mg of 12 were
obtained. Metabolite 13 was also transformed from EGC by other rat
intestinal flora (Figure 2B). This compound was also purified from 24-h
incubation mixture using the same procedure as for 11 to obtain 8 mg of
purified 13.
Interconversion between Metabolite 5 and Metabolite 6. Inter-
conversion between metabolites 5 and 6 was examined in the pH range of
1-10. Buffers used were 0.2 M potassium-hydroxy chroride buffer (pH 1.0
and 2.0), 0.2 M citrate-phosphate buffer (pH 3.0-8.0), 0.2 M borate-
sodium hydroxide buffer (pH 9.0-10.0). Each buffer (0.9 mL) was placed
in an individual tube, and then 0.1 mL each of 1 M 5 or 6 aqueous solu-
tion was added. The mixture was incubated under CO2 gas at 37 °C. After
24 h, the incubation mixture was analyzed by HPLC.
Analysis of Metabolites Formed from EGC (1) in Rat Cecum.
Male Wistar rats (21 weeks of age, n = 6) underwent laparotomy after
intraperitoneal injection of 25 mg of thiopental sodium (Ravonal; Mitsuibishi

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 1315
Figure 2. HPLC chromatograph profile of route II and route III.
Takeda Pharma Co., Ltd.). The segment of junction between cecum and
large intestine was ligated. Then, EGC solution (3 mg/ 0.5 mL of saline)
was directly injected into the cecum and the rats were maintained under the
anesthesia gas. After 5 h, cecal contents were harvested. To the cecal
contents (1 g) was added 0.8 mL of 10% aqueous methanol and
0.2 mL of 1 M sodium acetate buffer (pH 3.5) containing 1% ascorbic
acidand 0.15 mM EDTA. After sonication for 30 minat 40 °C, the mixture
was centrifuged at 25000g for 20 min at 10 °C. The supernatant was
analyzed by the LC-MS system as described below.
Analysis of Fecal Metabolites after Oral Administration of
EGCg. Fecal metabolites after oral administration of EGCg were
determined. Male Wistar rats (15 weeks of age, n = 4) were orally
administered EGCg (25 mg in 1 mL of saline). The rats were then placed
in stainless steel metabolic cages. Fecal samples were separately collected
at time intervals of 0-8, 8-24, 24-32, 32-48 h after dosing. The samples
were frozen at -30 °C, and then lyophilized. The resulting fecal powder
(1 g) was added to 0.8 mL of 10% aqueous methanol and 0.2 mL of 1 M
sodium acetate buffer (pH 3.5) containing 1% ascorbic acid and 0.15 mM
EDTA. The mixture was homogenized well with a vortex mixer (Scientific
Industries, NY, USA) and ultrasonic bath for 1 h. The mixture was
centrifuged at 25000g for 20 min at 10 °C. The resulting supernatant was
subjected to the LC-MS system.
scanning from m/z 100 to 1000. LC-MS analysis for quantification of
metabolites in cecum and feces was carried out with a 150 ꢀ 2.0 mm i.d.,
4 μm, Synergi MAX RP-18 column (Phenomenex, Macclesfield, U.K.)
(21). The column was eluted with buffer A (distilled water/acetonitrile/
acetic acid; 98/2/0.3, v/v/v) and buffer B (distilled water/acetonitrile/acetic
acid; 60/40/0.3, v/v/v) at a flow rate of 0.2 mL/min at 40 °C. The column
was eluted initially with 100% buffer A for 2 min, followed by linear
increases in buffer B to 100% from 2 to 20 min, and with 100% buffer B up
to 24 min. The column was finally equilibrated with buffer A for 6 min.
Quantification of the metabolites in cecum and feces was performed with
the selected ion monitoring (SIM) mode. The separated [M - H]- ion
chromatograms were selected at m/z 305, 307, 291, 225, and 207 for the
specific parent ions of metabolites 1, 3, 4, 5, and 6, respectively. Metabo-
lites in cecum and feces were determined based on the standard calibration
curves obtained with the metabolites purified in this study.
NMR and Optical Rotation Analyses. NMR analysis was per-
formed by a Bruker Ultrashield 400 plus system. (1H, 400 MHz; 13C,
100 MHz: Bruker BioSpin, Ibaragi, Japan). All samples were dissolved in
methanol-d4 (Kanto Chemical, Tokyo, Japan). Chemical shifts were
referenced with tetramethylsilane (TMS) as an internal standard. Specific
20
rotation [R]D of two metabolites (5 and 6 ) was measured by a P-1020
polarimeter (JASCO, Tokyo, Japan). Metabolite 5 was dissolved in
LC-MS Analysis. LC-MS analysis of EGC metabolites by rat
intestinal flora in vitro was performed by a Finnigan Spectra System
which consisted of a Surveyor HPLC system and an LCQ Deca XPplus
system (Thermo Fisher Scientific, USA). HPLC was carried out with a
100 ꢀ 2.0 mm i.d., 5 μm, Capcell pack MG column (Shiseido Co.Ltd.,
Tokyo, Japan) at a flow rate of 0.2 mL/min at 40 °C. The column was
eluted at a flow rate of 0.2 mL/min at 40 °C with 100% buffer A (distilled
water/acetonitrile/acetic acid; 100/2.5/0.2, v/v/v) for 2 min, followed by
linear increases in buffer B (distilled water/acetonitrile/methanol/acetic
acid; 33/2.5/66/0.2, v/v/v) from 0 to 80% for 2 to 25 min. The mobile phase
was then re-equilibrated with 100% buffer A for 5 min. The column elute
was monitored by the diode array detector and subsequently was directed
to the LCQ Deca XPplus ion trap mass spectrometer incorporated with
ESI interface. The negative ion polarity mode was set for the ESI source.
Analysis was initially carried out using full scan, data-dependent MS/MS
distilled water (c 0.225) and metabolite 6 in methanol (c 0.355).
RESULTS AND DISCUSSION
Screening of Enteric Bacteria Capable of Hydrolyzing EGCg. It
has been previously suggested that the metabolism of galloylated
catechins such as ECg and EGCg by intestinal bacteria was
initiated by releasing gallic acid from the catechins (14,22,25). In
relation to this, Hackett and Griffiths (23, 24) have also reported
that3-O-methyl-(þ)-catechin, unlike (þ)-catechin, did not under-
go ring fission by intestinal flora. Thus, the free hydroxy group at
the 3 position of catechins is important for their metabolism by
intestinal flora. Accordingly, we first screened enteric bacteria
capable of hydrolyzing EGCg. Among 169 stains of the bacteria

1316 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Takagaki and Nanjo
commercially available, four bacterial strains, Enterobacter aerogenes,
Raoultella planticola (Klebsiella planticola), K. pneumoniae
subsp. pneumoniae, and Bifidobacterium longum subsp. infantis
(B. infantis) were found to be capable of hydrolyzing EGCg. Since
these bacteria are well recognized to exist commonly in the
intestinal tract, EGCg is considered to be easily hydrolyzed to
yield EGC (1) and gallic acid (2) in the gut.
Time-Course Analysis of Metabolites by Rat Intestinal Bacteria.
After hydrolyzing EGCg to EGC (1) and gallic acid (2), subse-
quent metabolic processes of EGCg are degradation of EGC by
intestinal bacteria. In this study, EGC was incubated with
intestinal bacteria from rat cecal contents in 0.1 M sodium
phosphate buffer (pH 7.1) and transformation of 1 was examined
over time by LC-MS analysis. We performed the in vitro experi-
ments more than 15 times, and finally found that EGC underwent
degradation through at least three different metabolic pathways
by rat intestinal flora. These three metabolic routes are referred to
as route I, route II, and route III.
Figure 1 shows time-course analysis of EGC (1) metabolites
produced by rat intestinal flora through the pathway of route I. In
this route, EGC was first converted to 3 within 24 h incubation,
together with a small amount of 8. Second metabolite (4) peaked
at 48 h incubation accompanied by a decrease in 3. Subsequently,
5 began to increase gradually after incubation for 48 h with
decreasing 4, reaching a plateau at about 96 h incubation and
remained as the main product throughout the incubation periods
(168 h). Two other small peaks (6 and 7) were also detected at the
same incubation period when 5 was produced.
In route II (Figure 2A), the first metabolic step of EGC (1) was
conversion into 3 just as in route I. However, 3 was converted to 9
at the second step in route II, and then 9 to 5. Metabolite 5 was
further transformed to 11 and finally 12 was produced from 11.
Thus, 5 was produced from 3 through 9 in route II, while this
compound was transformed from 3 via 4 in the case of route I.
Metabolite 5 was accumulated in route I, but was converted to 12
via 11 and hence 12is regarded asthe major product in route II. In
addition, 6 and 10 were found to be formed.
In route III (Figure 2B), EGC (1) was first converted to 3 as
well as in routes I and II. During incubation for 24 to 48 h,
metabolite 3 was observed to be converted to 9, and thereafter 9
was converted to 13 and 13 was converted to 11, successively.
Metabolite 10 was also produced after 48 h incubation. Metabo-
lite 11 was hardly converted to any other metabolites even after
extension of incubation time up to 168 h, and therefore the
compound was regarded as the major product in route III. This
metabolic route is very analogous to route II.
We have revealed three routes of EGC (1) metabolic pathway
by rat intestinal flora in the in vitro experiments. Among the three
routes, route I was shown to be more dominant pathway than
routes II and III because the conversion of EGC into the
compounds shown in route I occurred with high frequencies
(probability with more than 90%) in the repeated experiments in
vitro.
Structural Analysis of Metabolites Produced in Routes I, II, and
III. EGC metabolites detected in each metabolic route as de-
scribed above were then purified by preparative HPLC and their
structures were determined by LC-MS and NMR analyses. Thus,
six metabolites (3-8) in route I, four metabolites (9-12) in route
II, and a metabolite (13) in route III were purified and identified.
In the case of 5 and 9, extraction and purification processes were
somewhat cumbersome. As indicated in Figure 3, metabolite 5
could not be extracted from the incubation mixture with
ethyl acetate until the pH of the mixture was adjusted to below
pH 4.0. Furthermore, 5 was spontaneously converted into 6
during the concentration process, especially under acidic
condition. Accordingly, purification of the compound was per-
formed cautiously. Similar phenomena were observed in the
isolation process of 9, and hence this compound was also purified
carefully.
Metabolite 3 showed a pseudomolecular ion peak at m/z 307
[M - H]- in its electoronspray ionization (ESI)-MS analysis,
2 mass units larger than that of EGC. 1H- and 13C NMR spectra
of 3 were very similar to those of 1-(30,40-dihydroxyphenyl)-
3-(200,400,600-trihydroxyphenyl)propan-2-ol (22) and 1-(30,50-
dihydroxyphenyl)-3-(200,400,600-trihydroxyphenyl)propan-2-
ol (19) except for their aromatic signal patterns. The substitution
patterns of two aromatic rings in 3 were found to resemble those
1
of EGC. Together with H-1H shift correlation spectroscopy
(COSY), heteronuclear multiple quantum coherence (HMQC)
and heteronuclear multiple bond correlation (HMBC) experi-
ments, metabolite 3 was identified to be 1-(30,40,50-trihydroxy-
phenyl)-3-(200,400,600-trihydroxyphenyl)propan-2-ol.
Metabolite 4 showed an [M - H]- ion peak at m/z 291 in
negative ESI-MS, 16 mass units less than that of 3. The 1H- and
13C NMR data of 4 could be superimposed with those of 1-(30,50-
dihydroxyphenyl)-3-(200,400,600-trihydroxyphenyl)propan-2-ol re-
ported by Wang et al. (19). Consequently, metabolite 4 was
determined to be 1-(30,50-dihydroxyphenyl)-3-(200,400,600-tri-
hydroxyphenyl)propan-2-ol.
The 1H- and 13C NMR data of 5 were closely similar to those of
5-(3,4-dihydroxyphenyl)-4-hydroxyvaleric acid reported in our
previous paper (25) except for the aromatic signal patterns. The
aromatic pattern of 5 was in harmony with that of the 30,50-
dihydroxyphenyl moiety in 4. The negative ESI-MS data of 5 also
showed [M - H]- ion peak at m/z 225, which is consistent with
that of 5-(3,4-dihydroxyphenyl)-4-hydroxyvaleric acid. From
these data together with COSY, HMQC, and HMBC experi-
ments, the chemical structure of metabolite 5 was concluded to
be 5-(3,5-dihydroxyphenyl)-4-hydroxyvaleric acid. The optical
rotation value, [R]D20 (c 0.225, H2O) is -18.1°.
The MS, 1H NMR, and 13C NMR data of 6 agreed with those
of 5-(30,50-dihydroxyphenyl)-γ-valerolactone reported pre-
viously (14). Accordingly, metabolite 6 was concluded to be
5-(30,50-dihydroxyphenyl)-γ-valerolactone. The optical rotation
20
value, [R]D -8.9o (c 0.355, CH3OH), of 6 was very similar to
20
that ([R]D -12.9o (c 0.4, CH3OH)) of the same compound
identified previously (14). Thus, it is reasonable that 6 has 4R
configuration.
Metabolite 7 showed a pseudomolecular ion peak at m/z 181
[M - H]- in negative ESI-MS. Molecular mass of this compound
was 16 mass units larger than that of 3-(3-hydroxyphenyl)
1
propionic acid (25). Aromatic signal patterns of 7 in H NMR
and chemical shifts of the aromatic ring in 13C NMR were
analogous to those of 6. From these observations together with
HMQC and HMBC experiments, metabolite 7 was determined to
be 3-(3,5-dihydroxyphenyl) propionic acid.
Negative ESI-MS data of 8 exhibited an [M - H]- ion peak at
m/z 289, 16 mass units less than that of EGC, suggesting the
compound lacked one hydroxyl group relative to EGC. The 1H-
and 13C NMR data were very similar to those of EGC, besides
chemical shifts in the B ring. The chemical shifts of 8 were
comparable to those in the 30,50-dihydroxyphenyl moiety of 4.
From further experiments with HMQC and HMBC, metabolite 8
was identified to be 40-dehydroxy EGC.
A pseudomolecular ion peak at m/z 241 [M - H]- in negative
ESI-MS was observed in 9. The 1H- and 13C NMR data indicated
good resemblance to those of 5, except for its aromatic signal
patterns. The aromatic chemical shifts of 9 were very similar to those
in the 30,40,50-trihydroxyphenyl group of 3. Therefore, it is concluded
that metabolite 9 is 5-(3,4,5-trihydroxyphenyl)-4-hydroxyvaleric

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 1317
Figure 3. Influence of pH range on extraction process of metabolite 5.
acid. This conclusion was further substantiated by COSY, HMQC,
and HMBC experiments.
of 3 and 9. Accordingly, metabolite 13 was identified to be
5-(3,4,5-trihydroxyphenyl)valeric acid. This result was further
supported by COSY, HMQC, and HMBC experiments.
The 1H NMR and 13C NMR spectroscopic data of the purified
metabolites were summarized in Tables 1 and 2, respectively. The
chemical structures of the metabolites are also shown in Figure 6.
Interconversion between Metabolite 5 and Metabolite 6. Figure 4
shows the interconversion ratio of metabolites 5 and 6 in the pH
range of 1.0-10.0. It was found that 5 is stable under alkaline
conditions, but the conversion into 6 occurred under acidic
conditions below pH 4. In particular, 90% or more of 5 was
spontaneously converted to 6 at pH below 2. On the contrary,
metabolite 6 was stable under acidic conditions and began to
change to 5 at above pH 7. Similar phenomenon also occurred
between 9 and 10. Accordingly, much attention should be paid to
handling these metabolites 5, 6, 9, and 10, especially in the
purification process of these compounds.
Metabolism of EGCg in Vivo. In order to obtain further
understanding of EGCg metabolism in the intestinal tract, we
examined metabolites of EGC (1) after direct injection into rat
cecum. The metabolites produced in the cecum 5 h after dosing of
EGC (3 mg, 9.8 μmol) were determined by LC-MS system. As
indicated in Figure 5, the amounts of each metabolite detected in
the cecum were 0.026, 0.028, 0.857, 1.701, and 0.618 μmol in the
cecal contents for metabolite 1, 3, 4, 5, and 6 respectively. These
results demonstrated that EGC was transformed in vivo through
the metabolic pathway of route I found in the in vitro experiments
as described above. In addition, metabolite 5 was also the most
dominant metabolite in rat cecum as well as in our in vitro
experiments.
Negative ESI-MS data of metabolite 10 exhibited an [M - H]-
ion peak at m/z 223, 18 mass units less than that of 9, suggesting
1
the compound was dehydrated from 9. The H- and 13C NMR
spectra indicated that 10 has γ-valerolactone structure as well as
6. The signal patterns and chemical shifts of the aromatic ring in
10 were found to be nearly identical with those in 9. Together with
COSY, HMQC, and HMBC experiments, metabolite 10 was
determined to be 5-(30,40,50-trihydroxyphenyl)-γ-valerolactone.
In the case of 11, a pseudomolecular ion peak at m/z 209 [M -
H]- was observed, 16 mass units less than that of 5, suggesting the
compound lacked one hydroxyl group relative to metabolite 5.
The 1H- and 13C NMR data of 11 were compatible with those of
5-(30-hydroxyphenyl)valeric acid previously reported by Meselhy
et al. (22) except for the aromatic signal patterns of 11. The
aromatic patterns of 11 were observed to be closely similar to
those of 7. These observations suggested that this compound is
5-(3,5-dihydroxyphenyl)valeric acid. This suggestion was con-
firmed by further experiments with COSY, HMQC, and HMBC.
The 1H- and 13C NMR data of 12 and 13 were analogous to
those of 11, except that the signal patterns and chemical shifts of
aromatic rings in three compounds were different from each
other. The whole signal patterns and chemical shifts in 12 were
superimposable on those of 5-(30-hydroxyphenyl)valeric acid
reported by Meselhy et al. (22). The ESI-MS data of 12 showed
a pseudomolecular ion peak at m/z 193 [M - H]-, which was 16
mass units less than that of 11. These observations clearly
demonstrated that metabolite 12 is 5-(3-hydroxyphenyl)valeric
acid. In metabolite 13, the ESI-MS data exhibited a pseudo-
molecular ion peak at m/z 225 [M - H]-, 16 mass units larger
Further, fecal metabolites after oral administration of EGCg
(25 mg, 54.5 μmol) were determined. Nine metabolites were
detected in feces, but no EGCg was found. The amounts of the
1
than that of 11. The H NMR and 13C NMR spectra of the
aromatic signal patterns of 13 were very analogous to those

1318 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Table 1. 1H-NMR Spectra Data of Catechin Metabolitesa
Takagaki and Nanjo
3
4
5
6
7
8
9
10
11
12
13
H-1a
H-1b
2.601 dd
2.942 dd
(13.7, 7.6) (13.8, 6.5)
2.432 dd 2.719 dd
(13.7, 8.3) (13.8, 6.9)
H-2 (or 2a) 3.938 dddd 4.898 dddd 2.329 ddd
(8.3, 7.6, (8.3, 7.5, (7.5, 7.4, 7.3) (17.9, 9.4, 9.0)
4.3, 4.0) 6.9, 6.5)
2.532 ddd
2.510 t 4.620 d
(6.6)
2.441 ddd 2.485 ddd
(15.7, 10.6, 7.4) (17.9, 10.3, 8.2)
2.286 m 2.301t
(6.8)
2.286 t
(7.0)
(6.8)
H-2b
2.256 ddd 2.408 ddd
(7.4, 7.4, 7.2) (17.9, 8.9,4.7)
2.326 ddd 2.341 ddd
(15.7, 8.8, 6.8) (17.9, 11.3, 4.9)
H-3 (0r 3a) 2.849 dd
3.010 dd
1.657 dddd
2.262 dddd
(12.2, 9.4,
6.9, 4.7)
2.743t 4.000 m
(7.6)
1.805 dddd
(13.5, 10.6,
6.8, 3.3)
1.945 dddd
(15.0, 11.1, 8.2, 7.0)
1.609 m 1.620 m 1.586 m
(13.8, 4.3) (13.1, 8.3) (14.3, 7.5,
7.4, 7.4)
H-3b
2.646 dd
2.737 dd
(13.8, 4.0) (13.1, 7.5) (14.3, 7.3,
1.810 dddd
1.964 dddd
(12.2, 9.0, 8.9, 7.2)
1.588 dddd
(13.5, 8.8,
7.4,4.4)
2.243 dddd
(15.0, 10.3, 7.3, 4.9)
7.2, 3.7)
H-4 (or 4a)
H-4b
H-5a
H-5b
H-6
3.784 m
4.744 dddd
(6.9,7.2,6.5,6.4)
2.522 dd
(16.2,7.4)
2.761 dd
(16.2,5.0)
3.708 dddd 4.709 dddd
(9.4,6.3,4.4,3.3) (7.3,7.0,6.4,6.3)
1.609 m 1.620 m 1.586 m
2.623 dd
2.870 dd
(13.9, 6.5)
2.771 dd
(13.9,6.4)
2.715 dd
(14.0, 6.0)
2.816 dd
(14.0,6.0)
2.464 m 2.545 t
(6.8)
2.408 t
(6.8)
(13.5, 6.8)
2.562 dd
(13.5, 6.2)
5.883 or 5.926d
(2.1)
H-8
5.883 or 5.926 d
(2.1)
H-20
6.220 s
6.240 d
(2.1)
6.195 d
(2.4)
6.193 d
(2.1)
6.165 d 6.333 d
(1.9) (2.1)
6.085 d 6.202 t
(1.9) (2.1)
6.217 s
6.217 s
25
6.244 s
6.244 s
25
6.133 d 6.637 t
(2.0) (2.4)
6.078 t 6.597 dd
6.180 s
6.180 s
25
H-40
6.151 t
(2.1)
6.109 t
(2.2)
6.141 t
(2.1)
(2.0)
(7.9, 2.0)
7.064 dd
(7.9, 8.1)
H-50
H-60
6.220 s
5.863 s
5.863 s
6.240 d
(2.1)
6.195 d
(2.4)
6.193 d
(2.1)
6.165 d 6.333 d
(1.9)
6.133 d 6.650 d
(8.1)
(2.1)
(2.0)
H-300
H-500
5.806 d
(2.0)
5.773 d
(2.0)
temp (°C) 25
25
40
25
25
25
25
25
a Chemical shifts are expressed in ppm downfield from the signal of TMS in methanol-d4. Coupling constants in Hz are in parentheses.
metabolites determined are summarized in Table 3. The metabo-
lites were not detected at 0-8 h post dose and emerged mainly at
8-24 h and 24-31 h. In the feces, 5 (6.16 μmol/feces) was
abundant, followed by 4 (3.45 μmol) and 6 (3.27 μmol). In this
experiment, EGCg was considered to be converted via the mixed
metabolic pathways of routes I and II found in the in vitro
experiments.
The above in vivo experiments suggested that the metabolic
pathway of EGCg found in the in vitro experiments probably
reflected the pathway in vivo.
On the basis of the time-course analysis of the metabolites
derived from EGCg and their structural identification, we pro-
posed three metabolic pathways of EGCg by rat intestinal
bacteria (route I to III) as illustrated in Figure 6. EGCg is
hydrolyzed to EGC (1) and gallic acid (2) at the first step of
metabolism. This step was common in the three pathways. In the
most dominant metabolic pathway (route I) in this study, EGC
formed from EGCg is converted to 3 by reductive opening
between the 1 and 2 positions of EGC. The resulting 3 is
transformed to 4 by dehydroxylation at 40 position of 3. Further,
metabolite 4 undergoes ring-fission of the phloroglucinol moiety
to yield 5 as the major metabolite in route I. At the same time,
metabolite 6 may be formed by lactonization of 5 just after the
ring-fission. Metabolite 5 is converted into 7, but its conversion is
onlyveryslight. In addition, while the dominant pathway is 1 to 3,
a very small amount of 1 is converted to 8 by dehydroxylation at
the 40 position. This metabolite appeared not to be metabolized
any more by the intestinal bacteria in this study. Wang et al. (19)
have discussed the possibility that the presence of 40-hydroxyl
group of EGC may be necessary for the reductive opening of its
heterocyclic ring, which is essential for further degradation of
EGC. This consideration is seemedto bethe casefor 8. Inroute II,
one of the two minor pathways, metabolite 3 formed from EGC
undergoes degradation of the phloroglucinol ring to produce 9.
This compound is converted into 5 by dehydroxylation at the 40
position. Furthermore, metabolite 5 undergoes dehydroxylation
at the 4 position to form 11, followed by transformation to 12
by elimination of the 50-hydroxyl group in 11 in route II. Meta-
bolite 3 is transformed to 13 via 9 in route III. Finally, meta-
bolite 13 is converted to 11 by undergoing dehydroxylation at the
40 position of 13. Metabolite 10 is also produced just after ring-
fission of the phloroglucinol moiety of 3 in addition to 9 in routes
II and III, as well as the formation of 5 and 6 in route I.
Thus, we determined the three metabolic pathways of EGCg
by rat intestinal bacteria. The differences in the metabolic pattern
of EGCg are considered to be a reflection of differences in the
intestinal microflora of the different rats used. We further
examined whether or not the metabolic pathways proposed in
the in vitro experiments reflect the EGCg metabolism in gut tract
in vivo. The metabolites derived from EGC were determined 5 h

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 1319
Table 2. 13C-NMR Spectra Data of Catechin Metabolitesa
3
4
5
6
7
8
9
10
11
12
13
C-1
C-2
44.19
75.38
31.64
43.17
85.73
32.76
182.97
34.24
35.91
74.09
178.03
28.11
29.53
83.00
144.84
37.838
32.63
178.4
32.03
32.87
73.26
180.42
29.56
27.93
83.31
178.19
35.25
25.89
31.91
177.56
34.92
25.70
31.95
177.83
34.91
25.68
32.17
82.73
68.7
C-3
C-4
27.82
100.65
157.68
96.37
157.91
95.57
156.71
C-4a
C-5
45.13
42.25
44.75
41.78
36.66
36.47
36.16
C-6
C-7
C-8
C-80a0
C-100
C-200
C-300
C-4
C-500
C-600
C-10
C-20
C-30
C-40
C-50
C-60
temp (°C)
105.75
158.29
95.92
104.52
159.55
96.06
157.76
95.92
162.87
90.56
158.29
132.35
109.47
146.74
132.01
146.74
109.47
25
159.55
141.29
108.92
159.68
101.78
155.38
108.92
25
142.79
109.16
159.40
101.65
159.40
109.16
40
139.88
108.93
159.49
102.10
159.49
108.93
25
143.17
106.49
159.55
103.13
159.55
106.49
25
131.11
109.42
146.9
132.55
146.9
109.42
25
128.46
109.55
147.11
133.11
147.11
109.55
25
145.88
107.97
159.38
101.13
159.38
107.97
25
145.07
116.23
158.23
113.69
130.29
120.81
25
134.67
108.36
146.87
132.03
146.87
108.36
25
107.78
159.52
101.44
159.52
107.78
25
a Chemical shifts are expressed in ppm downfield from the signal of TMS in methanol-d4.
Figure 6. Putative metabolic pathway of EGCg by rat intestinal flora.

1320 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Takagaki and Nanjo
after direct injection of EGC into the cecum of the operating rats.
As a result, metabolites 1, 3, 4, 5, and 6 were detected in the cecal
contents and 5 was found to be the most dominant metabolite
among them. We also determined the metabolites in rat feces after
oral administration of EGCg. Metabolites 4, 5, and 6 were found
as the major metabolites and metabolite 5 was dominant in the
feces. Therefore, it is likely that the in vitro metabolic pathway
proposed in this study reflects the in vivo metabolism of EGCg in
the rats.
In this study, we have found that metabolite 5 (5-(3,5-
dihydroxyphenyl)-4-hydroxyvaleric acid) is the major metabolite
produced by rat intestinal bacteria. However, there have been
no reports that 5-hydroxyphenyl-4-hydroxyvaleric acids were
identified as the metabolites produced from catechins such as
(þ)-catechin (23) and (-)-epicatechin gallate (ECg) (22) by
mammalian intestinal bacteria. Thus, there is an apparent dis-
crepancy between our results and those obtained in previous
reports. Although we have no sufficient explanation to this
discrepancy at present, we have noticed that metabolite 5
(5-(3,5-dihydroxyphenyl)-4-hydroxyvaleric acid) was converted
into metabolite 6 (5-(30,50-dihydroxyphenyl)-γ-valerolactone)
during the isolation process of 5 in this study. The lactonization
of 5 to 6 was found to take place easily during the concentration
process, particularly under acidic conditions (data not shown).
Such a lactonization reaction is well-known as γ-hydroxy-
carboxylic acid being converted rapidly to the corresponding
γ-lactone under acidic conditions. Therefore, a possible explana-
tion for no detection of 5-hydroxyphenyl-4-hydroxyvaleric acids
in the previous papers is that the valeric acids were converted
artificially, at least in part, to the corresponding γ-valerolactone
during the isolation process. In addition, it was observed that
5-(3,5-dihydroxyphenyl)-4-hydroxyvaleric acid in aqueous solu-
tion could not be extracted with ethyl acetate when pH of the
solution was at higher than 5, while the compound was shifted to
the organic layer at the pH below 4 (Figure 3). This may be
another reason that 5-hydroxyphenyl-4-hydroxyvaleric acids
were not isolated as the metabolites of catechins.
Similarly, 5-hydroxyphenyl-4-hydroxyvaleric acids were still
not found in urinary metabolites after EGCg dosing (14, 18). We
have reported that a kind of valeric acid, 5-(3,4-dihydroxy-
phenyl)-4-hydroxyvaleric acid, was found in rat urine after oral
administration of ECg, but was a minor metabolite (25). To the
contrary, lactonization products of 5-hydroxyphenyl-4-hydroxy-
valeric acids, 5-hydroxyphenyl-γ-valerolactones such as 5-(30, 40-
dihydroxyphenyl)-γ-valerolactone, 5-(30,40,50-triihydroxyphenyl)-
γ-valerolactone, and 5-(30,50-dihydroxyphenyl)-γ-valerolactone
have been reported to be the major metabolites of (þ)-cate-
chin (23), ECg (22, 25), and EGCg (12, 17). Negligible detection
of 5-hydroxyphenyl-4-hydroxyvaleric acids as the urinary meta-
bolites of catechins in previous papers may be explained by
lactonization of the compounds to the γ-valerolactones during
their isolation process as discussed above. However, further work
is needed to elucidate whether or not the conversion of valeric
acids into the corresponding γ-valerolactones takes place in the
body.
Figure 4. Interconversion of metabolite 5 to metabolite 6 (A) and
metabolite 6 to metabolite 5 (B) in the pH range 1.2-10.2. These values
are evaluated by peak area resulting HPLC analysis after 24 h incubation
at 37 °C. Both metabolites have the maximum absorption of wavelength at
275 nm. (gray bars: metabolite 5; black bars: metabolite 6).
We also examined the effects of pH on interconversion of 5 into
6 to ascertain their stability in our in vitro experiments with
intestinal bacteria. As shown in Figure 4, more than 90% of 5 was
converted into 6 at pH values below 2.0 within 24 h at 37 °C, but
the conversion rate decreased with increasing pH and no change
was observed at pH 5.0 and above. On the other hand, although 6
was stable at about pH 6.9 and below, it changed to 5 in alkaline
pH. Thus, the interconversion between 5 and 6 was caused by the
change in the pH of the solution, but this conversion scarcely
occurred in the pH range of about 5.0 to 6.9. Indeed, when EGC
was incubated with intestinal bacteria in 0.1 M phosphate buffer
(pH 7.0), pH of the incubation mixture was maintained between
pH 6.0 and 7.0 throughout the experimental periods. Hence, the
two compounds were considered unconverted in the in vitro
Figure 5. Cecal metabolites after direct injection of EGC (metabolite 1)
into rat cecum. Data expressed as mean ( SD values in μmol of six rats,
five hours after injection.
Table 3. Concentration of Metabolites Excreted in Feaces After Oral Administration of EGCga
EGCg
1
3
4
5
6
9
10
11
12
0-8 h
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8-24 h
24-32 h
32-48 h
total
0.38 ( 0.31
0.07 ( 0.05
0.05 ( 0.02
0.50 ( 0.38
0.12 ( 0.17
0.02 ( 0.04
0.01 ( 0.01
0.15 ( 0.22
3.44 ( 3.25
0.01 ( 0.01
0
3.00 ( 1.47
2.52 ( 1.66
0.64 ( 0.29
6.16 ( 3.41
1.60 ( 1.57
1.37 ( 0.87
0.30 ( 0.13
3.27 ( 2.57
0.02 ( 0.03
0.01 ( 0.01
0
0.01 ( 0.01
0.01 ( 0.01
0
0.06 ( 0.09
0.40 ( 0.22
0.45 ( 0.26
0.09 ( 0.57
0
0.02 ( 0.03
0.07 ( 0.03
0.09 ( 0.06
3.45 ( 3.26
0.03 ( 0.05
0.01 ( 0.02
a Data expressed as mean ( SD values in μmol of four rats.

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 1321
study. The pH of the cecal contents and feces in our in vivo
experiments were measured under the extraction conditions
described. The pH was found to be in the range of 5.0-7.0
(data not shown), and therefore the interconversion between 5
and 6 was thought not to take place. These findings implied that
both 5 and 6 detected in this study would be actually transformed
from EGC and EGCg by rat intestinal bacteria. Similarly, 9 and
10 were also considered to be truly produced from EGC.
In this study, we revealed the whole metabolic pathway of
EGCg by rat intestinal bacteria through both in vitro and in vivo
experiments. It was also presumed that the metabolic pathway
changed somewhat according to differences in the composition of
rat intestinal bacteria. Substantial amounts of the metabolites
from EGCg would be produced in the gut tract since intact EGCg
is reported to be poorly absorbed in the body (12,13). Therefore,
it will be important to examine the biological activities of the
metabolites and their metabolic fate in the body.
(11) Suganuma, M.; Okabe, S.; Oniyama, M.; Tada, Y.; Ito, H.; Fujiki,
H. Wide distribution of [3H](-)-epigallocatechin gallate, a cancer
preventive tea polyphenol, in mouse tissue. Carcinogenesis 1998, 19,
1771–1776.
(12) Kohri, T.; Matsumoto, N.; Yamakawa, M.; Suzuki, M.; Nanjo, F.;
Hara, Y.; Oku, N. Metabolic fate of (-)-[4-3H]epigallocatechin
gallate in rats after oral administration. J. Agric. Food Chem.
2001, 49, 4102–4112.
(13) Chen, L.; Lee, M.-J.; Li, H.; Yang, C. S. Absorption, distribution,
and elimination of tea polyphenols in rats. Drug Metab. Dispos.
1997, 25, 1045–1050.
(14) Kohri, T.; Nanjo, F.; Suzuki, M.; Seto, R.; Matsumoto, N.;
Yamakawa, M.; Hojo, H.; Hara, Y.; Desai, D.; Amin, S.; Conaway,
C. C.; Chung, F.-L. Synthesis of (-)-[4-3H]epigallocatechin gallate
and its metabolic fate in rats after intravenous administration.
J. Agric. Food Chem. 2001, 49, 1042–1048.
(15) Kida, K.; Suzuki, M.; Matsumoto, N.; Nanjo, F.; Hara, Y. Identi-
fication of biliary metabolites of (-)-epigallocatechin gallate in rats.
J. Agric. Food Chem. 2000, 48, 4151–4155.
(16) Sang, S.; Lambert, J. D.; Hong, J.; Tian, S.; Lee, M.-J.; Stark, R. E.;
Ho, C.-T.; Yang, C. S. Synthesis and structure identification of thiol
conjugates of (-)-epigallocatechin gallate and their urinary levels in
mice. Chem. Res. Toxicol. 2005, 18, 1762–1769.
(17) Li, C.; Lee, M.-J.; Sheng, S.; Meng, X.; Prabhu, S.; Winnik, B.;
Huang, B.; Chung, J. Y.; Yan, S.; Ho, C.-T.; Yang, C. S. Structural
identification of two metabolites of catechins and their kinetics in
human urine and blood after tea ingestion. Chem. Res. Toxicol. 2000,
13, 177–184.
(18) Sang, S.; Lee, M.-J.; Yang, I.; Buckley, B.; Yang, C. S. Human
urinary metabolite profile of tea polyphenols analyzed by liquid
chromatography/electrospray ionization tandem mass spectrometry
with data-dependent acquisition. Rapid Commun. Mass Spectrom.
2008, 22, 1567–1578.
ABBREVIATIONS USED
EGCg, (-)-epigallocatechin gallate; EGC, (-)-epigallocate-
chin.
ACKNOWLEDGMENT
We acknowledge the following persons of Mitsui Norin Co.,
Ltd. Daisuke Arai conducted the bacterial screening for us, and
Sayaka Ishikawa and Shuichi Otani assisted with the animal
experiments.
LITERATURE CITED
(1) Hara, Y. Prophylactic functions of tea. ASC Symp. Ser. 1994, 547,
34–50.
(2) Gupta, J.; Siddique, Y. H.; Beg, T.; Ara, G.; Afzal, M. A review on
the benefical effects of tea polyphenols on human health. Int. J.
Pharmacol. 2008, 4, 314–338.
(19) Wang, L.-Q.; Meselhy, M. R.; Li, Y.; Nakamura, N.; Min, B.-S.;
Qin, G.-W.; Hattori, M. The heterocyclic ring fission and dehydrox-
ylation of catechins and related compounds by Eubacterium sp.
Strain SDG-2, a human intestinal bacterium. Chem. Pharm. Bull.
2001, 49, 1640–1643.
(3) Naghma, K.; Hasan, M. Tea polyphenols for health promotion. Life
Sci. 2007, 81, 519–533.
(4) Carmen, C.; Reyes, A.; Rafael, G. Benefical effects of green tea - A
review. J Am. Coll. Nutr. 2006, 25, 79–99.
(5) 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. Biomerk. Prev.
1995, 4, 393–399.
(6) 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.
(7) Unno, T.; Takeo, T. Absorption of (-)-epigallocatechin gallate into
circulation system of rats. Biosci. Biotech. Biochem. 1995, 59, 1558–1559.
(8) 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.
(9) Unno, T.; Kondo, K.; Itakura, H.; Takei, T. Analysis of (-)-epi-
gallocatechin gallate in human serum obtained after ingesting green
tea. Biosci. Biotech. Biochem. 1996, 60, 2066–2068.
(20) Hattori, M. Metabolism of epicatechin 3-O-gallate and epicatechin
3-O-gallate by human intestinal flora. Proceedings of 2001 Inter-
national Conference on O-CHA (Tea) Culture and Science, Session III
Health and Benefits, October 5 -8, 2001; pp 38-41.
(21) Auger, C.; Mullen, W.; Hara, Y.; Crozier, A. Bioavailability of
polyphenon E flavan-3-ols in humans with an lleostomy1-4. J. Nutr.
2008, 138, 1535s–1542s.
(22) Meselhy, M. R.; Nakamura, N.; Hattori, M. Biotransformation of
(-)-epicatechin 3-O-gallate by human intestinal bacteria. Chem.
Pharm. Bull. 1997, 45, 888–893.
(23) Griffiths, L. A. Mammalian metabolism of flavonoids. In The
Flavonoids: Advances in Research; Harbone, J. B., Mabry, T. J., Eds;
Chapman and Hall: London, 1982; pp 681-718.
(24) Hackett, A. M.; Griffiths, L. A. The metabolism and excretion of
3-O-methyl-(þ)-catechin in the rat, mouse, and marmoset. Drug
Metab. Dispos. 1981, 9, 54–59.
(25) Kohri, T.; Suzuki, M.; Nanjo, F. Identificationofmetabolitesof (-)-epi-
catechin gallate and their metabolic fate in the rat. J. Agric. Food Chem.
2003, 51, 5561–5566.
(10) Nakagawa, K.; Miyazawa, T. Absorption and distribution of tea
catechin, (-)-epigallocatechin-3-gallate, in the rat. J. Nutr. Sci.
Vitaminol. 1997, 43, 679–684.
Received for review September 25, 2009. Revised manuscript received
December 7, 2009. Accepted December 9, 2009.