Polymer-like polyphenols of black tea and their lipase and amylase inhibitory activities

Article abstract of DOI:10.1248/cpb.56.266

Lipase and amylase inhibitory activities of black tea were examined. After solvent partitioning of a black tea extract with the ethyl acetate and n-butanol, the two soluble fractions showed comparable inhibitory activities. Activity in the ethyl acetate f

Full text of DOI:10.1248/cpb.56.266

266
Chem. Pharm. Bull. 56(3) 266—272 (2008)
Vol. 56, No. 3
Polymer-Like Polyphenols of Black Tea and Their Lipase and Amylase
Inhibitory Activities
Rie KUSANO, Hisashi ANDOU, Miho FUJIEDA, Takashi TANAKA,* Yosuke MATSUO, and Isao KOUNO*
Graduate School of Biomedical Sciences, Nagasaki University; 1–14 Bunkyo Machi, Nagasaki 852–8521, Japan.
Received September 25, 2007; accepted December 11, 2007; published online December 13, 2007
Lipase and amylase inhibitory activities of black tea were examined. After solvent partitioning of a black tea
extract with the ethyl acetate and n-butanol, the two soluble fractions showed comparable inhibitory activities.
Activity in the ethyl acetate fraction was mainly attributable to polyphenols with low-molecular weights, such as
theaflavin gallates. On the other hand, the active substance in the n-butanol layer was ascertained to be a poly-
1
mer-like substance. H- and 13C-NMR spectra showed signals arising from the flavan A-ring and galloyl groups,
although signals due to flavan B-rings were not detected, suggesting that the polymer-like substances were gener-
ated by oxidative condensation of flavan B-rings, a result which was previously deduced from our results of
in vitro catechin oxidation experiments. Enzymatic oxidation of epicatechin 3-O-gallate produced a similar poly-
mer-like substance and suggested that condensation between a B-ring and galloyl groups was involved in the
polymerization reaction.
Key words black tea; lipase; amylase; polyphenol; oxidation; thearubigin
Black tea, which is one of the most widely drunk bever- Result and Discussion
ages in the world, is produced by fermentation of the fresh
Lipase and a-Amylase Inhibitory Activity of Black Tea
leaves of Camellia sinensis. During the fermentation process, Commercial black tea was first extracted with cold water and
the constituents of the leaves are enzymatically converted to then with aqueous acetone. The aqueous acetone extract was
numerous secondary products that contribute to the charac- further fractionated by solvent partitioning to produce ether
teristic color and flavor of black tea.1,2) The tea catechins, in (yields of 3.2% of the black tea dry weight), EtOAc (4.7%),
particular, are major constituents of fresh tea leaves and are n-BuOH (10.6%) and aqueous (9.0%) layers, along with an
mainly composed of (ꢀ)-epicatechin, (ꢀ)-epigallocatechin insoluble paste-like precipitate (0.5%). The first cold water
and their galloyl esters. These constituents are oxidized dur- extract contained mainly sugars and other polar substances,
ing fermentation to yield a complex mixture of secondary and the ether fraction contained chlorophylls and other non-
polyphenols including theaflavins,3) theasinensins4,5) and oo- polar compounds; these fractions showed only weak lipase
longtheanins.5) However, most of the secondary polyphenols inhibitory activity (data not shown). On the other hand, the
in black tea have not yet been chemically characterized be- EtOAc, n-BuOH, aqueous layers and insoluble precipitate
cause of their complexity and the difficulties associated with showed inhibitory activities of 47.7%, 46.8%, 28.3% and
their separation and purification.6) Despite this lack of knowl- 61.1%, respectively (Table 1).
edge concerning their chemical structures, the health benefits
HPLC analysis of the EtOAct layer (Fig. 2), which showed
of black tea polyphenols are well known, and include antioxi- strong inhibition, revealed sharp peaks due to caffeine,
dant,7) anticancer8,9) and anti-inflammatory10,11) effects. The monomeric catechins [(ꢀ)-epigallocatechin 3-O-gallate (1),
black tea polyphenols, represented by the theaflavins and (ꢀ)-epicatechin 3-O-gallate (2), and epicatechin], theaflavins
theasinensins, have larger molecular weights than those of the (3—5), and theasinensin A (6) (Fig. 3). The activities of
monomeric tea catechins; therefore, the extent of direct ab-
sorption of the polyphenols by the digestive tract is expected
to be lower than for tea catechins.12) When we consider the
contribution of black tea to human health, the inhibition of
digestive enzymes may have considerable importance, since
polyphenols with high molecular weights have an inherent
ability to interact with proteins by forming hydrophobic and
hydrogen bonds,13,14) which usually results in enzyme inhibi-
tion.15) Amylase and lipase are digestive enzymes that hy-
drolyze starch and triglyceride, respectively, and inhibition of
these enzymes has been linked to the decreased incidence of
common diseases caused by diets rich in carbohydrates and
fats. It has been reported that theaflavins exhibit inhibitory
activities against these enzymes16—18); however, the concen-
trations of theaflavins in black tea infusions were much lower
than other water-soluble polyphenols (Fig. 1).6,13) This paper
Fig. 1. HPLC Analysis of Polyphenol Fraction Obtained from Commer-
cial Bottled Black Tea
describes the importance of the polymer-like polyphenols in
black tea by showing their enzyme inhibitory activities, as
well as presenting their chemical characterization and possi-
ble biogenesis.
Isomers of tea catechins produced during heat sterilization were observed. GC: gallo-
catechin, EGC: (ꢀ)-epigallocatechin, Cat: catechin, EC: (ꢀ)-epicatechin, ECg: (ꢀ)-
epicatechin 3-O-gallate, EGCg: (ꢀ)-epigallocatechin 3-O-gallate, GCg: (ꢀ)-gallocate-
chin 3-O-gallate, Cg: (ꢀ)-catechin 3-O-gallate, Fla: flavonol glycosides.
∗ To whom correspondence should be addressed. e-mail: t-tanaka@nagasaki-u.ac.jp
© 2008 Pharmaceutical Society of Japan

March 2008
267
Table 1. Lipase Inhibition Activities of Fractions and Polyphenolsa)
Inhibition (%)
AcOEt L.
BuOH L.
Aqeous L.
Insoluble precipitates
Epigallocatechin 3-O-gallate (1)
Theaflavin (3)
47.7ꢁ2.6
46.8ꢁ5.4
28.3ꢁ8.5
61.1ꢁ9.8
14.5ꢁ13.2
0.0ꢁ0
Theaflavin 3-O-gallate (4)
Theaflavin 3,3ꢂ-di-O-gallate (5)
Procyanidin-B-4 (8)
Procyanidin-B-3 (9)
Theasinensin A (6)
Galloyl oolongtheanin (7)
BLS1 (caffeine and small amount of POP)
BLS2 (POPꢃtheaflavin gallates)
BLS3 (POP)
BLP1 (caffeine and small amount of POP)
BLP2 (POPꢃtheaflavin gallates)
BLP3 (POP)
29.6ꢁ7.5
48.9ꢁ2.8
31.8ꢁ1.6
13.4ꢁ4.7
24.6ꢁ1.6
55.9ꢁ13.6
5.3ꢁ9.1
33.8ꢁ16.4
65.3ꢁ9.7
14.6ꢁ20.3
65.2ꢁ8.8
80.3ꢁ4.4
a) 0.04 mg/ml.
Fig. 2. HPLC Analysis of the Fractions of Black Tea Obtained from an
Aqueous Acetone Extract
Fig. 3. Structure of Catechin Oxidation Products and Procyanidins
some structure-defined polyphenols are listed in Table 1, and similar to that of the n-BuOH fraction. A large part of the
the activity of the EtOAc layer was probably attributable to POPs in the n-BuOH layer was precipitated from aqueous
these catechin oxidation products. Galloyl oolongtheanin (7) methanol to produce a brown powder designated as BLPs
showed a stronger inhibition than the theaflavin digallate (5), (precipitates obtained from n-BuOH layer), and the super-
and this was in agreement with the results of Nakai et al.18) natant (BLS) also contained POPs. The BLPs were separated
In contrast, the HPLC analysis of the n-BuOH layer showed by Sephadex LH-20 column chromatography into three parts
the presence of uncharacterized polyphenols that were de- that contained mainly caffeine (BLP-1), POPs with a small
tected as a broad hump on the HPLC baseline (polymer-like amount of theaflavin gallates (BLP-2), and pure POPs (BLP-
oxidation products; POPs) (Fig. 2). In a TLC analysis, the 3), respectively. Similarly, BLS was also separated into three
POPs were detected at the origin and positive to the FeCl3 fractions (BLS-1, 2, 3). The lipase inhibitory activities of
reagent (dark greenish-blue) and vanillin-HCl reagent (red) these fractions are listed in Table 1, and BLP-3 showed the
characteristic to catechin related compounds. The activity of strongest inhibition. HPLC analysis revealed that the peak
the BuOH layer was comparable to that of the EtOAc layer, summit retention times of BLP-3 and BLS-3 differed (Fig.
and the yield (10.6% of black tea dry weight) was much 4A), and BLP-3 was retained in the HPLC column more
higher than that of the EtOAc layer (4.7%); thus, the BuOH strongly. Direct size-exclusion chromatography of these frac-
soluble fraction was important in the total inhibition activity tions using an acetone-containing acidic aqueous urea solu-
of black tea. Although the yield (0.5%) was very low, the tion as the elution solvent19) suggested that the molecular size
insoluble precipitate showed the strongest activity (61.1%) of of BLP-3 was larger than that of BLS-3, indicating that POPs
the fractions. HPLC analysis of the precipitate indicated that with larger molecular weights showed stronger inhibitory ac-
it was mainly composed of POPs, and the HPLC profile was tivities (Fig. 4B). The POPs also showed strong a-amylase

268
Vol. 56, No. 3
1
inhibitory activity (Table 2). The result suggested that the the same resonance peaks in the H- and 13C-NMR spectra.
POPs of black tea also affect starch digestion, because the The results supported the proposal that POPs obtained from
concentration of POPs in the black tea infusion was higher black tea were oxidation products of tea catechins. In ad-
than that of theaflavin-3,3ꢂ-di-O-gallate (5), a black tea pig- dition, the absence of catechol- and pyrogallol-type B-ring
ment used as a positive control (Fig. 1).16)
signals in the NMR spectra suggested that the POPs were
Chemical Study on the Polymer-Like Oxidation Prod- produced by oxidative condensation of B rings, which was
ucts Since production of POPs during tea fermentation was probably related to the formation of theasinensins (6) and
accompanied by a decrease in concentrations of catechins, it theaflavins (3—5) (Fig. 3). Gel Permeation Chromatography
was deduced that POPs were oxidation products of catechins. (GPC) of BLP-3 using TSK-gel a-3000 suggested that the
1
The H-NMR spectrum of BLP-3 showed broad peaks at d peak top molecular weight was 9700.20) Measurement of the
7.0, 6.0 and 2.8 that were attributable to galloyl H-2, 6, cate- total polyphenol content in BLP-3 by the Folin-Ciocalteau
chin A-ring H-6, 8 and C-ring H-4, respectively (Fig. 5). The method21) using 1 as the standard showed that the polyphenol
inverse-gated decoupling 13C-NMR spectrum also showed content of BLP-3 was only 50%. This suggested that pheno-
broad signals attributable to the galloyl group [d 121 (C-1), lic benzene rings may be oxidatively degraded to non-pheno-
109 (C-2, 6), 145 (C-3, 5), 138 (C-4), 165 (C-7)] catechin A- lic forms, such as carboxylic acid.22) HPLC analysis of the
ring [d 95 (C-6, 8), 100 (C-4a), 155-157 (C-5, 7, 8a)], indi- thiol degradation products using 2-mercaptoethanol as a
cating that POPs are a mixture of oxidation products of tea nucleophile23) showed the production of trace amounts of epi-
catechins.
catechin, epicatechin 3-O-gallate, catechin-4-hydroxyethylthi-
The enzymatic oxidation of a mixture of four pure tea cat- oether, and epicatechin-4-hydroxyethylthioether. This indi-
echins, and subsequent purification followed by size-exclu- cated the presence of procyanidin-like C-4—C-8(6) interfla-
sion chromatography resulted in POPs which showed almost van linkages; however, the apparent decrease in size of the
broad hump on the base-line was not observed in this degra-
dation reaction. Therefore, the procyanidin-type linkage does
not appear to be an important component of POPs. Ethanoly-
sis in ethanolic hydrogen chloride yielded gallic acid ethyl
ester as the major product. The amounts of the galloyl group
in BLP-3 and BLS-3 were 5.5ꢁ0.1% and 7.0ꢁ0.1%, respec-
tively. Since the percentage of the galloyl group in 1 is 33%,
the proportion of the galloyl group in POPs was not high.
The lipase inhibitory activity of BLP-3 was higher than BLS-
3, where the molecular size of the former is larger than that
Table 2. a-Amylase Inhibition Activities of Fractions and Polyphenolsa)
Inhibition (%)
AcOEt L.
BuOH L.
Aqeous L.
65.4ꢁ4.1
57.0ꢁ1.8
5.3ꢁ1.1
BLS1 (caffeine and small amount of POP)
BLS2 (POPꢃtheaflavin gallates)
BLS3 (POP)
10.4ꢁ1.1
64.8ꢁ3.2
62.3ꢁ2.9
36.8ꢁ0.2
60.4ꢁ5.1
53.7ꢁ3.3
81.6ꢁ2.8
BLP1 (caffeine and small amount of POP)
BLP2 (POPꢃtheaflavin gallates)
BLP3 (POP)
Theaflavin 3,3ꢂ-di-O-gallate (5)b)
Fig. 4. Reversed-Phase and Size-Exclusion HPLC of POPs
(A) Reversed-phase HPLC. (B) Size-exclusion HPLC.
a) 0.04 mg/ml. b) Positive control.
Fig. 5. 1H- and 13C-NMR Spectra of BLP-3

March 2008
269
of the latter, as mentioned previously (Fig. 4). Therefore, the from 2 (Fig. 6) were related to those of BLP-3, and the peak-
results indicated that the molecular size of POPs is more im- top molecular weight was estimated to be 8100 by GPC. The
portant for enzyme inhibitory activity than the amounts of 13C-NMR spectrum of the polymeric substance showed sig-
galloyl esters. Furthermore, the lower content of galloyl nals attributable to the galloyl groups, as well as those due to
groups in BLP-3 suggested that galloyl groups were oxidized the A- and C-rings; the galloyl C-2, 6 signals were smaller
when the molecule was extended.
than the A-ring signals. This observation suggested that sig-
Previous studies suggested that intermolecular oxidative nificant amounts of the galloyl groups were oxidized during
couplings of catechins mainly occur between two B-rings, production of the polymeric substance. The major product of
and generally give dimeric products such as 3—7. Further the reaction was theaflavate A (10),25) which is a dimer pro-
extension of the molecule at the B-rings of the products were duced by intermolecular oxidative condensation between the
not observed,1,4,5,22) except for the oxidative coupling at catechol B-ring and the galloyl group (Chart 1). In addition,
the benzotropolone moieties of theaflavin, which yielded theaflavate C (11), an extended trimer of 2, was isolated from
bistheaflavins A and B.24) However, when epicatechin 3-O- the same reaction mixture. Furthermore, bistheaflavate A
gallate (2) was enzymatically oxidized with the polyphe- (12), a tetramer of 2 produced by oxidative coupling between
noloxidase of the Japanese pear homogenate,22,24) a poly- two benzotropolone moieties of 10 were also produced.26)
meric substance similar to POPs was produced. The HPLC Since intermolecular coupling between the B-ring and gal-
and 13C-NMR spectra of the polymeric substance derived loyl group extends the molecular size as shown in 10 and 11,
Fig. 6. HPLC (A) and 13C-NMR (B) Spectra of a POP-Like Substance Produced from Epicatechin 3-O-Gallate
Chart 1. Possible Mechanism for Production of POPs

270
Vol. 56, No. 3
Materials Glyceryl trioleate, pancreatic lipase (type II) and butylated
hydroxyanisole were purchased from Sigma (St. Louis, U.S.A.). Lecithin
(from egg), bathocuproine, sodium cholate, 3,5-dinitrosalycilic acid (DNS),
and Cu (NO3)2·3H2O were purchased from Wako Pure Chemical Industries,
Ltd. (Osaka, Japan). N-Tris(hydroxymethyl)methyl-2-aminoethane sulfonic
the polymeric substance may be produced by a related poly-
merization reaction. This was supported by the MALDI-
TOF-MS analysis of a fraction containing the oxidation prod-
ucts of 2 with lower molecular weights obtained from the
same reaction mixture, which exhibited peaks corresponding acid (TES) was purchased from Kishida Chemical Co., Ltd. (Osaka, Japan).
to the tetramer and pentamer [m/z: 875 (10, [MꢃNa]ꢃ), 1285
Epicatechin 3-O-gallate, epigallocatechin and epigallocatechin 3-O-gallate
were isolated from green tea.4) Theaflavin, theaflavin 3-O-gallate, theaflavin-
(corresponding to a trimer 11, [MꢃNa]ꢃ), 1740 (tetramer)
3,3ꢂ-di-O-gallate,32) oolongtheanin gallate,5,22) and theasinensins A22,33) were
obtained by enzymatic oxidation of tea catechins in our previous research.
and 2175 (pentamer)]. Similar oxidative couplings of galloyl
groups attached to theasinensins and theaflavins occur during
tea fermentation. In particular, the production of trimer 13
from 1 by enzymatic oxidation was previously demonstrated
(Chart 1).27,28) Moreover, the extension of the molecule may
also occur at the benzotropolone ring in a manner similar to
the formation of 12 and bistheaflavins A and B.24)
HPLC Analysis of Commercial Black Tea Bottled commercial black
tea (heat sterilized) (3.0 l) was directly applied to Sephadex LH-20 (25ꢄ
3.0 cm i.d.) and washed with 25% MeOH to remove sugars and caffeine.
The polyphenols adsorbed on the gel was eluted out with MeOH (200 ml)
and then 50% aqueous acetone (400 ml). After removal of the organic sol-
vents by rotary evaporator, the aqueous solution of the polyphenols was
lyophilized to give brown powder (1.69 g), which was analyzed by reversed
phase HPLC (Fig. 1).
Measurement of Amylase Inhibitory Activity Amylase inhibition was
measured according to a published method.16,34) Briefly, a sample in 0.1 ml
of 20 mM phosphate buffer (pH 6.9) containing 6.7 mM NaCl was mixed with
0.5 ml of amylase solution (0.03 mg/ml in the above buffer) and incubated at
37 °C for 10 min. To the mixture, 0.4 ml of a starch solution (3 mg/ml in the
above buffer) was added and incubated further for 1 h. A portion (0.4 ml) of
the reaction mixture was taken and mixed with 0.4 ml of DNS reagent34) and
heated at 100 °C for 5 min in a screw-capped vial. After cooling on ice, the
solution was diluted with the addition of 4 ml of water and the absorption at
540 nm was measured. Theaflavin 3,3ꢂ-di-O-gallate was used as the positive
Conclusion
Black tea contains a variety of polyphenols including tea
catechins and chemically uncharacterized oxidation products,
where the complexity of black tea polyphenols has hampered
the evaluation of their biological functions. The complexity
is caused by cascade-type complex oxidation between four
tea catechins during black tea production.22,28,29) The impor-
tance of some characteristic polyphenols, such as theaflavins,
is already known; however, the present study has shown the control.16)
Measurement of Lipase Inhibitory Activity The investigation of li-
importance of POPs (obtained by extraction in n-BuOH) as
the active components inhibiting pancreatic lipase and a-
amylase. BLP-3 represented the fraction of POPs with the
strongest enzyme inhibitory activities. Although this fraction
was not soluble in water when purified, it remained in the
water layer of the black tea extract on partitioning with
EtOAc. It is likely that these POPs were solubilized by inter-
action with coexisting water-soluble substances such as
flavonol glycosides, as observed for polymeric proantocyani-
dins.14,30) POPs are a mixture of catechin oxidation products
that could not be detected as sharp peaks by HPLC analysis.
These POPs may be partly identical with thearubigins, which
are uncharacterized black tea polyphenols as designated by
Roberts,6,31) and are widely considered to account for a major
part of polyphenols in black tea infusions.6) The results of the
in vitro oxidation of 2 suggested that oxidative coupling of
galloyl groups with other B-rings was important for the pro-
duction of POPs such as BLP-3, in addition to B–B-ring cou-
plings (Chart 1). Further studies concerning the structure and
mechanism of the production of POPs are now in progress.
pase inhibitory activity was based on the method of Han et al.35) using glyc-
eryl trioleate as the substrate. The substrate solution was prepared by sonica-
tion of a mixture of glyceryl trioleate (80 mg), lecithin (10 mg) and sodium
cholate (5 mg) in 9 ml of 0.1 M TES buffer (pH 7). The substrate solution
(0.1 ml) was mixed with a sample solution (0.1 ml) and preincubated at
37 °C for 5 min. After addition of 0.05 ml of a lipase solution (0.05 mg/ml)
the mixture was further incubated for 30 min. Free fatty acids generated by
this procedure were measured according to the method of Tsujita and
Zapf.36,37) The reaction mixture (125 ml) was added to 3 ml of CHCl3–hep-
tane (1 : 1, v/v) containing 2% (v/v) MeOH and shaken horizontally for
10 min. The mixture was centrifuged at 2000 g for 10 min and 1 ml of the
lower organic phase was taken and mixed with 1 ml of copper reagent.37) The
test tube was shaken vigorously for 10 min and centrifuged at 2000 g for
10 min. The upper organic phase (0.5 ml) containing fatty acid copper salts
was reacted with 0.5 ml of 0.1% (w/v) bathocuproine in CHCl3 containing
0.05% (w/v) butylated hydroxyanisole to produce a pale orange-colored
complex, and absorption of the solution at 480 nm was measured. Prepara-
tion of the copper reagent was as follows: Cu(NO3)2 (1.21 g) was dissolved
in a 1 M triethanolamine solution (10 ml) and mixed with a 1 M NaOH solu-
tion (6 ml). NaCl (33 g) was added to the solution and the total volume was
then adjusted to 100 ml with water.
Extraction and Isolation for Comparison of Lipase Inhibitory Activ-
ity Commercial black tea (919 g, tea bag produced in India) was extracted
with water, three times at room temperature. The extract was applied directly
to a Sephadex LH-20 column (10ꢄ31 cm) and washed with H2O to remove
Experimental
General Experimental Procedures UV spectra were obtained with a sugars and caffeine. The adsorbed polyphenols were fractionated into frac-
1
JASCO V-560 UV/VIS spectrophotometer. H- and 13C-NMR spectra were tions 1—10 by a 20% stepwise gradient elution with water–MeOH (0—
recorded in a mixture of acetone-d6 and D2O (19 : 1, v/v) at 27 °C with a
100%) and then MeOH–H2O–acetone (8 : 1 : 1→3 : 1 : 1). The residue re-
Varian Unity Plus 500 spectrometer operating at 500 MHz for 1H and maining after extraction with water was further extracted with 50% aqueous
125 MHz for 13C. Mass spectra (MS) were recorded on a Voyager-DE Pro acetone three times and the acetone was evaporated in vacuo. The resulting
spectrometer, and 2,5-dihydroxybenzoic acid (10 mg/ml in 50% acetone aqueous solution (2 l) was successively partitioned with ether (1 lꢄ3),
containing 0.05% trifluoroacetic acid) was used as the matrix for MADLI- EtOAc (1 lꢄ4), and n-BuOH (1 lꢄ4) to produce an ether soluble fraction
TOF-MS measurements. Column chromatography was performed using (29.4 g, 3.2% of black tea dry weight), an EtOAc fraction (43.6 g, 4.7%), an
Sephadex LH-20 (25—100 mm, Pharmacia Fine Chemical Co. Ltd.). Thin
n-BuOH fraction (97.0 g, 10.6%), and a water fraction (82.4 g, 9.0%). The
layer chromatography was performed on precoated Kieselgel 60 F254 plates paste-like precipitate (4.3 g, 0.5%) remaining in the final aqueous layer was
(0.2-mm thick, Merck) with benzene–ethyl formate–formic acid (1 : 7 : 1,
v/v), CHCl3–MeOH–water (14 : 6 : 1, v/v) and Cellulose F254 (0.2-mm thick,
collected. The lipase inhibitory activity of the ten fractions obtained from
the initial water extract, and that of the ether and CHCl3 fractions, was weak
Merck) with 2% AcOH. Spots were detected using ultraviolet (UV) illumi- (less than 30% inhibition) and these fractions were not examined further.
nation and by spraying with 2% ethanolic FeCl3 or 10% sulfuric acid reagent
Separation of the n-BuOH Layer The n-BuOH fraction (65 g) was dis-
followed by heating. Analytical HPLC was performed on a Cosmosil 5C18- solved in 50% MeOH (500 ml) by heating and left to stand at room tempera-
AR II (Nacalai Tesque, Inc.) column (250ꢄ4.6 mm i.d.) with gradient elu-
ture for 10 h. The resulting brown precipitate was collected by centrifuga-
tion from 4—30% (39 min) and 30—75% (15 min) of CH3CN in 50 mM
H3PO4 (flow rate, 0.8 ml/min; detection, JASCO photodiode array detector duced a further precipitate. The precipitates were combined to form an
tion. The addition of H2O to the supernatant (→ about 30% MeOH) pro-
MD-910).
n-BuOH layer precipitate (BLP) fraction (41.7 g). The final supernatant was

March 2008
271
concentrated to produce an n-BuOH layer supernatant (BLS) fraction pletely removed. The resulting aqueous solution was applied to a Sephadex
(18.7 g). These fractions were fractionated separately into three fractions LH-20 column (26ꢄ4.5 cm i.d.) with 0—100% MeOH (10% stepwise elu-
using Sephadex LH-20 column chromatography (50—100% MeOH and tion, each 300 ml) and 50% aqueous acetone to yield 7 fractions. Fr. 1 (15.9
then 50% aqueous acetone) to produce fractions BLP-1 (13.5 g, yield:
32.5%), BLP-2 (10.5 g, 25.3%) and BLP-3 (7.7 g, 18.4%) from BLP, and
BLS-1 (9.6 g, 51%), BLS-2 (7.8 g, 41.6%) and BLS-3 (1.4 g, 7.4%) from
BLS. HPLC analysis demonstrated that BLP-1 and BLS-1 contained caf-
mg) and Fr. 4 (172 mg) were identified as epicatechin and 2, respectively,
using TLC and HPLC comparisons. MALDI-TOF-MS of Fr. 5 showed peaks
at m/z 875 (10, [MꢃNa]ꢃ), 1285 (corresponding to a trimer 11, [MꢃNa]ꢃ),
and Fr. 7 exhibited the peaks at m/z 1740 (tetramer) and 2175 (pentamer). Fr.
feine and small amounts of polymer-like oxidation products (POPs), BLP-2 6 (266 mg) contained polymeric substances and was further separated using
and BLS-2 contained POPs and small amounts of mono- and di-galloyl
esters of theaflavin, and BLP-3 and BLS-3 contained only POPs. BLP-3 was
a Sephadex LH-20 column (20ꢄ1.5 cm i.d.) with 8 M urea in 60% acetone
(adjusted to pH 2 by conc. HCl). Tubes containing polymeric substance and
obtained as a dark-brown amorphous powder with the following elemental 10 were collected separately and concentrated by evaporation until the ace-
constituents: C, 58.38%; H, 4.44%; N, 0.30%. tone was removed. The resulting aqueous solutions were applied separately
Size-Exclusion HPLC of BLP-3 and BLS-3 Size-exclusion HPLC of to a MCI-gel CHP 20P column (20ꢄ3.0 cm i.d.). After washing the column
BLP-3 and BLS-3 (Fig. 4B) was performed according to Yanagida et al.19) with water to remove urea, the polyphenols were eluted with H2O–MeOH
using a TSK gel a-2500 column (300ꢄ7.8 mm i.d.). The mobile phase was (10% stepwise elution from 0 to 60%, each 100 ml) to produce the poly-
acetone–6 M urea (pH 2) (3 : 2), the flow rate was 0.1 ml/min, and the eluate
was monitored at 220 nm.
meric substance (45.1 mg) and 10 (28.5 mg).
Thiolysis of Polymer-Like Substance A solution of 0.1% (w/v) POPs
Acknowledgements The authors are grateful to Mr. K. Inada and Mr. N.
in 70% EtOH (0.2 ml) was mixed with 5% mercaptoethanol in 60% EtOH Yamaguchi for NMR and MS measurements. This work was supported by a
containing 0.1% HCl (0.8 ml) and heated at 70 °C for 7 h. HPLC analysis Grant-in-Aid for Scientific Research No. 18510189 from the Japan Society
showed four small peaks. The retention time and UV absorption of the peaks for the Promotion of Science.
coincided with those of epicatechin, catechin-4-hydroxyethylthioether, epi-
catechin 4-hydroxyethylthioether and epicatechin-3-O-gallate, and repre- References
sents the order of their elution.23) However, the broad hump on the base-line
did not decrease.
1) Hashimoto F., Nonaka, G., Nishioka, I., Chem. Pharm. Bull., 40,
1383—1389 (1992).
GPC Analysis of POPs POPs were dissolved in dimethylformamide
and applied to a TSK-gel a3000 column (300ꢄ7.8 mm i.d.) at 40 °C with
dimethylformamide containing 10 mM LiCl as the elution solvent at a flow
rate of 0.5 ml/min. The peaks were detected by monitoring UV absorption at
254 nm. Polystyrene standards (molecular weights of 4000, 25000, 50000
and 170000), toluene (molecular weight 92), and three ellagitannins [1-O-
galloylpedunculagin (MW 936),38,39) sanguiin H-6 (MW 1870),40) and
rubusuaviin C (MW 2804)39)] were used as standards. Peak-top molecular
weights were recorded by an Integrator 807-IT (JASCO Corporation).
Ethanolysis of BLP-3 and BLS-3 BLP-3 and BLS-3 were dissolved
separately in 1 M HCl in EtOH (2 mg/ml) and heated at 100 °C. The concen-
tration of ethyl gallate in the reaction mixture was analyzed by HPLC at 12,
24, 36, 48 and 60 h. The peak area did not increase after 48 h. The yields of
ethyl gallate from BLP-3 and BLS-3 were 7.1ꢁ0.1% (5.5ꢁ0.1% as galloyl
group) and 9.1ꢁ0.1% (7.0ꢁ0.1%), respectively. The calibration curve was
2) Wang D., Kurasawa E., Yamaguchi Y., Kubota K., Kobayashi A., J.
Agric. Food Chem., 49, 1900—1903 (2001).
3) Takino Y., Imagawa H., Horikawa H., Tanaka A., Agric. Biol. Chem.,
28, 64—71 (1964).
4) Nonaka G., Kawahara O., Nishioka I., Chem. Pharm. Bull., 31, 3906—
3914 (1983).
5) Hashimoto F., Nonaka G., Nishioka I., Chem. Pharm. Bull., 36,
1676—1684 (1988).
6) Haslam E., Phytochemistry, 64, 61—73 (2003).
7) Luczaj W., Skrzydlewska E., Preventive Medicine 40, 910—918
(2005).
8) Way T.-Z., Lee H.-H., Kao M.-C., Lin J.-K., Eur. J. Cancer, 40,
2165—2174 (2004).
9) Krishnan R., Maru G. B., J. Agric. Food Chem., 52, 4261—4269
(2004).
constructed using ethyl gallate prepared by a similar ethanolysis of tannic 10) Maity S., Ukil A., Karmakar A., Datta N., Chaudhuri T., Vedasiromoni
acid. Preparation of the standard ethyl gallate was as follows: tannic acid
was heated in 1 M HCl in EtOH under reflux for 12 h. After solvent partition-
J. R., Ganguly D. K., Das P. K., Eur. J. Pharmacol., 470, 103—112
(2003).
ing between EtOAc and water, the EtOAc layer was applied to a Sephadex 11) Sang S., Lambert J. D., Tian S., Hong J., Hou Z., Ryu J.-H., Stark R.
LH-20 column with EtOH and further purified using silica gel 60 (CHCl3–
MeOH–H2O, 90 : 10 : 1, v/v/v) chromatography and crystallized from CHCl3–
MeOH.
E., Rosen R. T., Huang M.-T., Yang C. S., Ho C.-T., Bioorg. Med.
Chem., 12, 459—467 (2004).
12) Mulder T. P. J., van Platerink C. J., Wijnand Schuyl P. J., van
Amelsvoort J. M. M., J. Chromatogr. B, 760, 271—279 (2001).
Production of POPs Using
a Model Fermentation Experiment
Japanese pear fruit (600 g) was homogenized with water (100 ml) in a me- 13) Haslam E., “Practical Polyphenolics, from Structure to Molecular
chanical blender and filtered through four layers of gauze. The homogenate
(300 ml) was mixed with an aqueous solution (200 ml) containing (ꢀ)-epi-
catechin (0.5 g), (ꢀ)-epigallocatechin (1.0 g), (ꢀ)-epicatechin-3-O-gallate (0.5
g) and (ꢀ)-epigallocatechin-3-O-gallate (2.5 g) and stirred vigorously for
Recognition and Physiological Action,” Cambridge University Press,
Cambridge, 1988, pp. 138—177.
14) Zhang Y.-J., Tanaka T., Betsumiya Y., Kusano R., Matsuo A., Ueda T.,
Kouno I., Chem. Pharm. Bull., 50, 258—262 (2002).
3 h. The mixture was heated at 90 °C for 2 min to inactivate enzymes and 15) Inokuchi J., Okabe H., Yamaichi T., Nagamatsu A., Nonaka G., Nishi-
decompose unstable quinone intermediates.22,41) After cooling, the mixture
was poured into acetone (2 l) and filtered through No. 2 filter paper. The fil-
oka I., Chem. Pharm. Bull., 33, 264—269 (1985).
16) Hara Y., Honda M., Agric. Biol. Chem., 54, 1939—1945 (1990).
trate was concentrated until the organic solvent was removed, and the result- 17) Matsui T., Tanaka T., Tamura S., Toshima A., Tamaya K., Miyata Y.,
ing aqueous solution (400 ml) was successively partitioned with EtOAc Tanaka K., Matsumoto K., J. Agric. Food Chem., 55, 99—105 (2007).
(400 mlꢄ4) and n-BuOH (300 mlꢄ4). The n-BuOH layer was concentrated 18) Nakai M., Fukui Y., Asami A., Toyoda-Ono Y., Iwashita T., Shibata
in vacuo and the residue was fractionated by size-exclusion chromatography
using Sephadex LH-20 (26ꢄ1.5 cm i.d.) with acetone: 6 M urea (3 : 2, v/v,
H., Mitsunaga T., Hashimoto F., Kiso Y., J. Agric. Food Chem., 53,
4593—4598 (2005).
adjusted to pH 2 with diluted HCl).19) The POPs with larger molecular 19) Yanagida A., Shoji T., Shibusawa Y., J. Biochem. Biophys. Methods,
weights were eluted faster than catechins and theaflavins, and the fractions
containing only POPs were combined. After evaporation of acetone, the 20) Fujiwara T., Nakagawa-Izumi A., Osawa T., Numata O., Biosci.
aqueous solution was passed through Sephadex LH-20 (12ꢄ1.5 cm i.d.) and Biotechnol. Biochem., 71, 711—719 (2007).
washed with water to remove urea and HCl. The POPs adsorbed on the gel 21) Piccinelli A. L., Simone F., Passi S., Rastrelli L., J. Agric. Food Chem.,
were eluted using 50% aqueous acetone to give POPs (70 mg). 52, 5863—5868 (2004).
Enzymatic Oxidation of 2 Japanese pear (150 g) was homogenized in 22) Tanaka T., Watarumi S., Matsuo Y., Kamei M., Kouno I., Tetrahedron,
56, 311—322 (2003).
H2O (150 ml) and filtered through four layers of gauze. The homogenate
(300 ml) was mixed with an aqueous solution of 2 (1.0 g in 200 ml) and vig-
orously stirred for 4 h at room temperature. The reaction was monitored by
59, 7939—7947 (2003).
23) Tanaka T., Takahashi R., Kouno I., Nonaka G., J. Chem. Soc. Perkin
Trans. 1, 1944, 3013—3022 (1994).
reversed-phase HPLC. The mixture was poured into acetone (2 l), gently 24) Tanaka T., Inoue K., Betsumiya Y., Mine C., Kouno I., J. Agric. Food
stirred for 30 min, and the insoluble material was then removed by filtration. Chem., 49, 5785—5789 (2001).
The filtrate was concentrated by evaporation until the acetone was com- 25) Wan X., Nursten H. E., Cai Y., Davis A. L., Wilkins J. P. G., Davies A.

272
Vol. 56, No. 3
P., J. Sci. Food Agric., 74, 401—408 (1997).
26) Kusano R., Tanaka T., Matsuo Y., Kouno I., Chem. Pharm. Bull., 55,
1768—1772 (2007).
27) Tanaka T., Matsuo Y., Kouno I., J. Agric. Food Chem., 53, 7571—7578
(2005).
28) Li Y., Tanaka T., Kouno I., Phytochemistry, 68, 1081—1088 (2007).
29) Tanaka T., Matsuo Y., Li Y., Kouno I., “Polyphenols Communications
34) Bernfeld P., “Methods in Enzymology,” Vol. 1, ed. by Colowick S. P.,
Kaplan N. O., Elsevier, Amsterdam, 1955, pp. 149—158.
35) Han L.-K., Gong X.-J., Kawano S., Saito M., Kimura Y., Okuda H.,
Yakugaku Zasshi, 125, 213—217 (2005).
36) Tsujita T., Okuda H., Eur. J. Biochem., 133, 215—220 (1983).
37) Zapf J., Schoenle E., Waldvogel M., Sand I., Froesch E. R., Eur. J.
Biochem., 113, 605—609 (1981).
2006,” ed. by Daayf F., El Hadrami A., Adam L., Ballance G. M., Uni- 38) Okuda T., Yoshida T., Ashida M., Yazaki K., J. Chem. Soc., Perkin.
versity of Manitoba, Manitoba, 2006, pp. 115—116.
30) Tanaka T., Kataoka M., Tsuboi N., Kouno I., Chem. Pharm. Bull., 48,
201—207 (2000).
Trans 1, 1983, 1765—1772 (1983).
39) Li H., Tanaka T., Zhang Y.-J., Yang C.-R., Kouno I., Chem. Pharm.
Bull., 55, 1325—1331 (2007).
31) Roberts E. A. H., “The Chemistry of Flavonoid Compounds,” ed. by
Geissman T. A., Pergamon Press, Oxford, 1962, pp. 468—512.
40) Tanaka T., Nonaka G., Nishioka I., J. Chem. Res. (M), 1985, 2001—
2029.
32) Tanaka T., Mine C., Inoue K., Matsuda M., Kouno I., J. Agric. Food 41) Tanaka T., Mine C., Watarumi S., Fujioka T., Mihashi K., Zhang Y.-J.,
Chem., 50, 2142—2148 (2002).
Kouno I., J. Nat. Prod., 65, 1582—1587 (2002).
33) Matsuo Y., Tanaka T., Kouno I., Tetrahedron, 62, 4774—4783 (2006).