Carcinogenesis vol.25 no.9 pp.1567--1574, 2004
doi:10.1093/carcin/bgh168
Generation of hydrogen peroxide primarily contributes to the induction of
Fe(II)-dependent apoptosis in Jurkat cells by (ÿ)-epigallocatechin gallate
Hiroshi Nakagawa1,2, Keiji Hasumi2, Je-Tae Woo3,
Kazuo Nagai3 and Masaaki Wachi1,4
1
Department of Bioengineering, Tokyo Institute of Technology, 4259
Nagatsuta-cho, Midori-ku, Yokohama 226-8501, 2Department of Applied
Biological Science, Tokyo Noko University, 3-5-8 Saiwai-cho, Fuchu,
Tokyo 183-8509 and 3Department of Biological Chemistry, College of
Bioscience and Biotechnology, Chubu University, 1200 Matsumoto-cho,
Kasugai, Aichi 487-8501, Japan
4
To whom correspondence should be addressed
Email: [email protected]
Although (ÿ)-epigallocatechin gallate (EGCG) has been
reported to induce apoptosis in a variety of tumor cells,
detailed mechanisms remain to be explored. In the present
study, we investigated the antitumor mechanism of EGCG
by using human T-cell acute lymphoblastic leukemia Jurkat cells. We focused on the involvement of reactive oxygen
species, as we found previously that EGCG caused apoptotic cell death in osteoclastic cells due mainly to promotion
of the reduction of Fe(III) to Fe(II) to trigger Fenton reaction, which affords hydroxyl radical from hydrogen peroxide [H2O2 ‡ Fe(II) ! OH ‡ OHÿ ‡ Fe(III)]. EGCG
(12.5--50 mM) decreased the viability of Jurkat cells and
caused concomitant increase in cellular caspase-3 activity.
Catalase and the Fe(II)-chelating reagent o-phenanthroline
suppressed the EGCG effects, indicating involvements of
both H2O2 and Fe(II) in the mechanism. Unexpectedly,
epicatechin gallate (ECG), which has Fe(III)-reducing
potency comparable with EGCG, failed to decrease the
viability of Jurkat cells, while epigallocatechin (EGC),
which has low capacity to reduce Fe(III), showed cytotoxic
effects similar to EGCG. These results suggest that, unlike
in osteoclastic cells, a mechanism other than Fe(III) reduction plays a role in catechin-mediated Jurkat cell death. We
found that EGCG causes an elevation of H2O2 levels in
Jurkat cell culture, in cell-free culture medium and sodium
phosphate buffer. Catechins with a higher ability to produce H2O2 were more cytotoxic to Jurkat cells. Hydrogen
peroxide itself exerted Fe(II)-dependent cytotoxicity.
Amongst tumor and normal cell lines tested, cells exhibiting lower H2O2-eliminating activity were more sensitive to
EGCG. From these findings, we propose the mechanism
that make catechins cytotoxic in certain tumor cells is due
to their ability to produce H2O2 and that the resulting
increase in H2O2 levels triggers Fe(II)-dependent formation of highly toxic hydroxyl radical, which in turn induces
apoptotic cell death.
Abbreviations: EC, (ÿ)-epicatechin; ECG, (ÿ)-epicatechin gallate; EGC,
(ÿ)-epigallocatechin; EGCG, (ÿ)-epigallocatechin gallate; MTT reagent, 3[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; NHDF, normal
human dermal fibroblast; PBS(ÿ), phosphate-buffered saline without
Ca2‡ and Mg2‡.
Carcinogenesis vol.25 no.9 # Oxford University Press 2004; all rights reserved.
Introduction
Catechins, contained in some types of foods and plants
(especially in green tea), are a subspecies of polyphenols
and have been demonstrated in vitro to have a wide range of
pharmacological properties such as antioxidative (1,2), antibacterial (3), antimutagenic (4), antiresorptive (5) and antitumor effects (6--28). Among these effects, the most
extensively studied one has been the antitumor effect in
which several characteristic phenomena of apoptosis were
observed: cell cycle arrest (6--11), injury of DNA (8--21) and
activation of caspases (17--23). The cytotoxicity of catechins
has been recognized to be relatively specific to tumor cells
as compared with normal cells (8,10,14,23). Phases I and II
clinical trials and in vivo study were performed with green
tea extract as an anticancer drug, and catechins have been
demonstrated to be beneficial substances with little side
effects (29--31).
The in vitro antitumor mechanism of (ÿ)-epigallocatechin
gallate (EGCG), a major catechin in green tea, has been suggested to be modulation of the expression of key molecules in
cell cycle progression (cyclin kinase inhibitor, cyclin, cyclindependent kinase and p27Kip1) (6,7,10,11) and in transcription
(inhibitor of nuclear factor-kB) (7), activation of mitogenactivated protein (MAP) kinase cascade (especially apoptosisregulating kinase-1, MAP kinase kinase, c-Jun N-terminal
kinase and p38 MAP kinase) (21,22), inhibition of telomerase
(16) and interaction with Fas (19). However, comprehensive
mechanisms to explain the diverse effects of EGCG in causing
apoptotic cell death remain to be explored.
In the previous study, we found that o-phenanthroline, a
Fe(II)-chelating reagent, and catalase, a H2O2-scavenging
enzyme, suppressed EGCG-induced apoptotic cell death in
cultured osteoclastic cells and demonstrated in a cell-free
system that reduction of Fe(III) by EGCG triggered a
Fenton reaction to form a highly reactive hydroxyl radical
from H2O2 [H2O2 ‡ Fe(II) ! OH ‡ OHÿ ‡ Fe(III)] (5).
The present study aims to test whether similar mechanisms
are involved in the antitumor effect of EGCG in vitro using
human T-cell acute lymphoblastic leukemia Jurkat cells.
The results demonstrated that EGCG caused Fe(II)- and
H2O2-dependent Jurkat cell apoptosis. However, experiments using EGCG analogs with a different potency to
reduce Fe(III) demonstrated that, unlike in osteoclastic
cells, elevation of H2O2 levels in culture rather than reduction of Fe(III) plays an important role in the cytotoxic
mechanism. Exogenously added H2O2 showed Fe(II)dependent cytotoxicity, and the catechin cytotoxicity changed among cells with a different ability to eliminate H2O2.
We propose a novel mechanism that the EGCG-mediated
generation of H2O2 primarily triggers Fe(II)-dependent
formation of highly toxic molecules (possibly hydroxyl
radicals), which in turn induce apoptotic cell death in Jurkat
cells.
1567
H.Nakagawa et al.
Materials and methods
Chemicals
EGCG, o-phenanthroline hydrochloride, human recombinant superoxide dismutase (SOD), hydrogen peroxide (H2O2) and 0.1% (w/v) xylenol orange
solution were purchased from Wako Pure Chemical Industries (Osaka,
Japan). (ÿ)-Epicatechin (EC), (ÿ)-epicatechin gallate (ECG) and (ÿ)-epigallocatechin (EGC) were from Nagara Science (Gifu, Japan). RPMI-1640 medium (Cat. No. R8758) and Dulbecco's modified Eagle's medium (DMEM, Cat.
No. D6046), 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide
(MTT reagent), catalase from bovine liver and D-sorbitol was obtained from
Sigma Chemical (St Louis, MO). Protease inhibitor cocktail (CompleteTM ),
0.25 M ammonium iron(II) sulfate solution and acetyl-Asp-Glu-Val-Asp-a-(4methyl-coumaryl-7-amide) (Ac-DEVD-MCA) were obtained from Roche
(Mannheim, Germany), Kanto Chemical (Tokyo, Japan) and Peptide Institute
(Osaka, Japan), respectively.
Cell culture
HeLa, HL-60, Jurkat and U937 cell lines were maintained in medium A [RPMI
1640 medium supplemented with 10% (v/v) heat-inactivated fetal calf serum
(Valley Biomedical, Winchester, VA), 100 U/ml penicillin and 100 mg/ml
streptomycin] at 37 C under 5% CO2 and 95% air. Normal human dermal
fibroblast (NHDF) was maintained in DMEM supplemented with 10% (v/v)
heat-inactivated fetal calf serum, 100 U/ml penicillin and 100 U/ml streptomycin at 37 C under 5% CO2 and 95% air. HeLa cells and NHDF were
transferred to 96-well culture plates (Corning, Corning, NY) at 1 105 cells/
well or 6-well culture plates (Corning, Corning, NY) at 4 105 cells/well 12 h
prior to use in experiments. All the cells were cultured in medium A (not
containing sodium pyruvate) in experimental use.
MTT assay
HL-60, Jurkat and U937 cells (4 105 cells/well) placed on 6-well culture
plates and HeLa cells and NHDF pre-cultured for 12 h on 96-well culture
plates were treated with catechins (EC, 50 mM; ECG, 50 mM; EGC, 50 mM;
EGCG, 0--50 mM) (dissolved in methanol, the final concentrations of which in
culture medium were 50.5%) or H2O2 (50 mM) for 6 h in medium A. After
incubation, HL-60, Jurkat and U937 cells were washed with fresh medium A
and placed on 96-well culture plates at 1 105 cells/well in 100 ml of medium
A containing 50 mg of MTT reagent. HeLa cells and NHDF were also
incubated with 100 ml of medium A containing 50 mg of MTT reagent after
washing with fresh medium A. After incubation for 2 h, 100 ml of 10% (w/v)
SDS in phosphate-buffered saline without Ca2‡ and Mg2‡ [PBS(ÿ)] was
added to solubilize MTT-formazan. After overnight incubation, absorbance
at 595 nm was measured. MTT reagent-reducing activity of the cells (MTTreducing activity) was expressed as percentage of control or percent inhibition
by catechins. Data are expressed as the mean SD of triplicate cultures in one
of two to three experiments. For statistical analysis of the results, groups were
compared with Student's t-test.
Assay for caspase-3 activity
Jurkat cells (1 106 cells/well of 6-well culture plates) were treated with
EGCG (0--50 mM) for 6 h in medium A. After incubation, cells were collected
by centrifugation and washed with ice-cold PBS(ÿ). The cells were lysed in
lysis buffer [100 mM Tris--HCl (pH 7.5), 1 mM DTT, 1% (v/v) Triton X-100
and protease inhibitor cocktail (CompleteTM )], and then passed 10 times
through a 27G needle. After centrifugation at 3000 r.p.m. for 10 min, the
supernatant [50 mg protein, determined using Bio-Rad Protein Assay kit
(Bio-Rad) with bovine serum albumin as a standard] was incubated with
20 mM Ac-DEVD-MCA at 37 C for 1 h. Fluorescence of the liberated
aminomethylcoumarin was determined by excitation at 365 nm and emission
at 450 nm using a MTP-32 microplate reader (Corona Electric).
Assay for Fe(III)-reduction
To assess Fe(III)-reducing activity of catechins, catechins (1 mM) were added
to 1 ml of working solution (1.2 mg/ml o-phenanthroline and 1 mM FeCl3), and
the mixture was incubated at room temperature for 20 s. After the incubation,
absorbance of Fe(II)--o-phenanthroline complex at 510 nm was immediately
measured. Fe(III)-reducing activity of catechins was assessed from the amount
of Fe(II)--o-phenanthroline complex in solutions and expressed as relative to
EGCG, of which activity is taken as 100%. Data are expressed as the mean SD of triplicate determinations in one of two to three experiments.
Measurement of the concentration of H2O2 in serum-containing medium A in
the presence or absence of Jurkat cells
Catechins (50 mM) were added to 6-well culture plates filled with 2 ml of
medium A containing Jurkat cells (1 106 cells/well) or no cells, and the plates
were placed at 37 C under 5% CO2. After 15, 30, 60, 120, 240 and 360 min, part
of the medium was collected to measure the concentration of H2O2 by the
1568
ferrous ion oxidation--xylenol orange method with slight modification (32,33).
Briefly, medium (30 ml) was mixed with 0.3 ml of working solution [250 mM
ammonium iron(II) sulfate, 25 mM H2SO4, 100 mM sorbitol and 125 mM
xylenol orange], followed by vortexing and incubation at room temperature
for 20 min. After incubation, absorbance of Fe(III)--xylenol orange complex at
595 nm was measured. The concentration of H2O2 was calculated from a
standard curve, which was obtained by determining H2O2 concentrations
immediately after addition of H2O2 into the culture medium. This assay method
was applicable to determine H2O2 concentrations as low as 0.2 mM. Data are
expressed as the mean SD of three independent experiments. For statistical
analysis of the results, groups were compared with Student's t-test.
Measurement of the concentration of H2O2 in various solutions
EGCG (50 mM) was added to 6-well culture plates containing serum-free
medium A or 100 mM phosphate buffer (pH 5.8, 6.8 and 7.8). Where indicated,
100 U/ml catalase or 500 U/ml SOD was included in the solution. Serum-free
medium A and phosphate buffers were incubated at 37 C under 5% CO2 and at
37 C under air, respectively. After incubation for 1 h, measurement of H2O2
concentration was carried out by the ferrous ion oxidation--xylenol orange
method as described above. The concentration of H2O2 was calculated from
a standard curve that had been obtained by determining H2O2 concentrations
immediately after addition of H2O2 into each solution. Data are expressed as
the mean SD of three independent experiments. For statistical analysis of the
results, groups were compared with the Student's t-test.
Assay for H2O2-eliminating activity in cells
HL-60, Jurkat and U937 cells (4 105 cells/well) were placed on 6-well
culture plates and HeLa cells and NHDF were pre-cultured for 12 h on 6well culture plates in medium A. After addition of 50 mM H2O2, each culture
was incubated at 37 C under 5% CO2 for 15, 30, 45 and 60 min. The
concentration of H2O2 remaining in the medium was determined by the ferrous
ion oxidation--xylenol orange method. Data are expressed as the mean SD of
triplicate determinations in one of two to three experiments.
Absorption spectrum
EGCG (50 mM) was added to 100 mM phosphate buffer at pH 5.8, 6.8 or 7.8.
After incubation for 1 h at 25 C, the solutions were subjected to measurement
of absorption spectrum on a model 320 spectrophotometer (Hitachi Instruments Service, Tokyo, Japan) at a wavelength range of 200--400 nm.
Results
Hydrogen peroxide- and Fe(II)-dependent cytotoxicity of
EGCG in Jurkat cells
To investigate antitumor mechanism of EGCG, human T-cell
acute lymphoblastic leukemia Jurkat cells, widely used in the
study of apoptosis, were used in the present study. The cytotoxic effect of EGCG in Jurkat cells was examined by MTT
assay (Figure 1A). MTT-reducing activity of Jurkat cells was
dose-dependently decreased in the culture with 12.5--50 mM
EGCG for 6 h. The decrease was accompanied by an increase
of caspase-3 activity in the cells (Figure 1B). In both experiments, vehicle (methanol) alone did not affect both activities at
0.5% (v/v) (Figure 1A and B). These results indicate that
EGCG cause apoptotic cell death in Jurkat cells and that the
MTT-reducing activity of Jurkat cells represents the viability
of the cells. As shown in Figure 1C and D, catalase (100 U/ml)
completely suppressed the cytotoxic inhibitory effect of
EGCG. o-Phenanthroline, a Fe(II)-specific chelator, also
restored the Jurkat cell viability, whereas its effect was partial
(75% restoration, Figure 1C). These results that not all but
most of the cytotoxic effects of EGCG in Jurkat cells was
mediated both by H2O2 and Fe(II), suggests an involvement
of H2O2- and Fe(II)-dependent hydroxyl radical formation in
the EGCG-mediated cell death.
EGCG-mediated H2O2 generation causes cell death in
Jurkat cells
To elucidate the H2O2- and Fe(II)-dependent cytotoxic
mechanism of EGCG in Jurkat cells, three EGCG analogs
Fe(II)-dependent apoptosis in Jurkat cells by EGCG
Fig. 1. Effects of EGCG on the viability of Jurkat cells. Jurkat cells were
treated in medium A with EGCG (A and B, indicated concentrations; C and
D, 50 mM) or vehicle [0.5% (v/v) methanol, open circle and open column]
for 6 h in the presence (C and D) or absence (A and B) of 100 U/ml catalase
(CL) or 50 mM o-phenanthroline (PNT). (A and C) After incubation, cells
were further incubated in fresh medium A containing MTT reagent for 2 h to
determine MTT-reducing activity. Data are expressed as percentage of
control (mean SD of triplicate cultures). P 5 0.005 as compared with
control (A) or between the indicated groups (C). (B and D) After incubation,
cell lysates were prepared, and then caspase-3 activity in the lysates was
determined using fluorogenic substrate. Similar results were obtained in two
other experiments.
(EC, ECG and EGC) (Figure 2A) with different activities to
reduce Fe(III) were tested for cytotoxicity. Among the three
analogs, only EGC decreased the viability of Jurkat cells at
50 mM, and the others showed no effect at 50 mM (Figure 2B).
EGC showed a three times lower Fe(III)-reducing activity than
the non-cytotoxic analog ECG, which is comparable with
EGCG in the activity to reduce Fe(III) (Figure 2B and C).
These results partially conflict with the previous observations
that osteoclastic cell death is caused by Fenton reaction due
mainly to a EGCG-mediated reduction of Fe(III) to Fe(II).
Thus, these results suggest that a mechanism other than
Fe(III) reduction is involved in the EGCG effects in Jurkat cell
death. Although hydrogen peroxide, that ubiquitously exists in
cells, can be eliminated by antioxidative systems (34--36),
an increase of H2O2 levels should cause Fe(II)-dependent
radical formation. So we next examined the levels of H2O2 in
Jurkat cell culture. Cultures in the presence of cytotoxic catechins (EGC and EGCG) showed higher levels of H2O2 than
those in the presence of non-cytotoxic catechins (EC and ECG)
(Figure 2B and D). Similar but higher H2O2 accumulations
were observed when 50 mM catechins were added to cell- and
serum-free medium A, followed by incubation for 1 h at 37 C
(EC, 2.9 0.8 mM; ECG, 7.9 1.6 mM; EGC, 43 2.0 mM;
EGCG, 52 2.2 mM). These results indicate that cytotoxic
catechins produce H2O2. Since the activity of catechins to
generate H2O2 was related to their cytotoxicity (Figure 2B
and D), cytotoxicity of exogenously added H2O2 was examined. As shown in Figure 3A, exogenous H2O2 decreased
viability of Jurkat cells at H2O2 concentrations comparable
with those generated in Jurkat cell culture in the presence of
EGCG (3.125--50 mM). The cytotoxic effect of H2O2 was
suppressed by 100 U/ml catalase and 50 mM o-phenanthroline
(Figure 3B). Once again, the effect of o-phenanthroline was
partial, and this result suggests that there are Fe(II)-dependent
Fig. 2. Involvement of H2O2 in cytotoxic effect of catechins. (A) Structures of catechins. (B) Jurkat cells were treated with 50 mM catechins for 6 h in medium A.
After incubation, cells were further incubated in fresh medium A containing MTT reagent for 2 h to determine MTT-reducing activity. Data are expressed as
inhibition of MTT-reducing activity (%) compared with control (mean SD of triplicate cultures). (C) Each catechin (1 mM) was added to a mixture of 1.2 mg/ml
o-phenanthroline and 1 mM FeCl3. Twenty seconds after the addition, the absorbance at 510 nm was measured to determine the amount of Fe(II). The amounts
of Fe(II) formed after incubation with catechins are expressed as relative to that after incubation with EGCG, which is taken as 100%. Data represent the
mean SD of triplicate determinations. (D) Each catechin (50 mM) was added to Jurkat cell culture in medium A. Thirty minutes after the addition of
catechins, the concentration of H2O2 in the medium was determined by the ferrous ion oxidation--xylenol method. Data are expressed as the mean SD
of three independent experiments.
1569
H.Nakagawa et al.
Fig. 3. Effects of H2O2 on the viability of Jurkat cells. Hydrogen peroxide (A,
indicated concentrations; B, 50 mM) was added to Jurkat cell culture with (B)
or without (A) 100 U/ml catalase (CL) or 50 mM o-phenanthroline (PNT).
After incubation for 6 h, cells were further incubated in fresh medium A
containing MTT reagent for 2 h to determine MTT-reducing activity. Data are
expressed as percentage of control (mean SD of triplicate cultures).
P 5 0.005 as compared with control (A) or between the indicated groups (B).
and -independent mechanisms in H2O2 cytotoxicity as well as
in EGCG cytotoxicity. Nevertheless, all the results shown here
suggest that the production of H2O2 by EGCG primarily contributes to the Fe(II)-dependent cytotoxicity in Jurkat cells.
Cells may be able to escape from Fenton reaction-mediated
cell death by eliminating H2O2 in spite of the ubiquitous existence of H2O2 and Fe(II) in cells. As shown in Figure 4A, H2O2
concentration increased rapidly after addition of EGCG to
Jurkat cell culture, and the levels decreased time-dependently,
suggesting an existence of H2O2-eliminating ability in Jurkat
cells. In the presence of catalase, EGCG-mediated H2O2 generation was not observed (Figure 4A). If cytotoxicity of EGCG
is primarily due to its activity to generate H2O2, cells with a
higher capacity to eliminate H2O2 might be more resistant to
EGCG cytotoxicity. To test this possibility, five different cell
lines were examined for their ability to eliminate H2O2 as well
as for their sensitivity to EGCG and H2O2. As shown in Figure
4B, all cell lines tested had an evident but distinct capability to
eliminate H2O2. The half-life of the added H2O2, as determined
from the pseudo-first-order decay curve, was ~10 min in HeLa
and NHDF, ~14 min in U937 and 18--20 min in Jurkat and HL60 cell cultures. As expected, HL-60 cells, as well as Jurkat
cells, were sensitive to both EGCG and H2O2 (Figure 4C). On
the other hand, HeLa, NHDF and U937 cells, which have a
higher ability to eliminate H2O2, were resistant to EGCG and
H2O2 at concentrations up to 50 mM.
Mechanism of EGCG-mediated generation of H2O2
As mentioned above, catechins promote H2O2 generation in
Jurkat cell culture. We next examined dose dependency of
EGCG and the effect of catalase on H2O2 generation. As
shown in Figure 5A, the generation of H2O2 gradually increased as the concentration of EGCG elevated (12.5--50 mM).
The increase in H2O2 levels was completely inhibited by
100 U/ml catalase. The generation of H2O2 was also observed
in the absence of Jurkat cells and serum (Figure 5B). These
results indicate that the EGCG-mediated H2O2 generation is a
cell- and serum-independent process. On the other hand,
EGCG failed to generate H2O2 in pure water, ethanol or
methanol (data not shown). So we explored the essential constituents of the medium for EGCG-induced H2O2 generation.
As shown in Figure 6A, EGCG-mediated H2O2 generation was
detected in 100 mM sodium phosphate buffer at pH 7.8, but
H2O2 generation was very small at pH 6.8 and not detected
at pH 5.8. This result suggests that EGCG-mediated H2O2
1570
Fig. 4. Hydrogen peroxide-eliminating ability and EGCG susceptibility in
various cell lines. (A) EGCG (50 mM) was added to Jurkat cell culture in the
presence (closed circles in A) or absence (open circles in A) of 100 U/ml
catalase. After indicated time, the concentration of H2O2 in the medium was
determined by the ferrous ion oxidation--xylenol method. Data are expressed
as the mean SD of three independent experiments. P 5 0.05 as compared
with both zero time and catalase-treated cultures at each time point.
(B) Hydrogen peroxide (50 mM) was added to the cultures of HeLa
(open triangles), HL-60 (open squares), Jurkat (open diamonds), NHDF
(cross) and U937 (plus) cells as well as to cell-free medium A (open circles).
After indicated times, the concentration of H2O2 in the medium was
determined by the ferrous ion oxidation--xylenol method. Data are expressed
as the mean SD of three determinations. (C) Hydrogen peroxideeliminating ability and sensitivities to EGCG and H2O2 in a variety of cell.
a
The concentrations of EGCG and H2O2 to reduce the cell viability by 50%
(IC50) as determined by the MTT assay. bHalf life of H2O2 in cell cultures
calculated from the pseudo-first-order decay curves in (B).
generation is dependent on pH levels. The absorption spectrum
of EGCG in 100 mM sodium phosphate buffer at pH 5.8 was
distinct from that at pH 7.8: the absorption at 300--370 nm
increased with elevation of the pH levels (Figure 6B). This
Fe(II)-dependent apoptosis in Jurkat cells by EGCG
Fig. 5. Effects of EGCG on H2O2 generation under various conditions.
EGCG (indicated concentrations) was added to Jurkat cell culture (A), cellfree medium A (B) or cell- and serum-free medium A (B) in the presence
(closed symbols) or absence (open symbols) of 100 U/ml catalase. After 1 h,
the concentration of H2O2 in the medium was determined by the ferrous ion
oxidation--xylenol method. Data are expressed as the mean SD of three
independent experiments. P 5 0.05; P 5 0.005, as compared with
control.
indicates a pH-dependent change in aromatic resonance structure in the EGCG molecule. The pKa1 value for EGCG is
reported to be at 7.59--7.75 (37,38), and the change in absorption spectrum may represent the deprotonation of phenolic
hydroxyl group(s).
As H2O2 can be generated from superoxide anion (Oÿ
2 )
‡
under physiological conditions (2Oÿ
2 ‡ 2H ! O2 ‡ H2O2),
we next examined the involvement of Oÿ
in the EGCG2
mediated H2O2 generation by using SOD, which catalyses
H2O2 formation from Oÿ
2 . SOD inhibited EGCG-mediated
H2O2 generation in 100 mM sodium phosphate buffer at pH
7.8 (Figure 6C). SOD also inhibited EGCG-mediated
H2O2 generation in Jurkat cell culture and suppressed cytotoxicity of EGCG to Jurkat cells in a dose-dependent manner
(20--500 U/ml) (Figure 7A--C), while bovine serum albumin,
tested as a control, did not exert such activities at the same
concentration (2 mg protein/ml) as those of SOD (data not
shown). On the other hand, the cytotoxic effect of H2O2 was
not suppressed by SOD (Figure 7D), excluding the possibility
that Oÿ
plays a role after H2O2 is formed. These results
2
Fig. 6. Generation of H2O2 in sodium phosphate buffer by EGCG. EGCG
(50 mM) was added to 100 mM sodium phosphate buffer at pH 5.8 (A and
solid line a in B), 6.8 (A and broken line b in B) or 7.8 (A, bold solid line c in
B and C). (A and C) One hour after the addition of EGCG, the concentration
of H2O2 in the buffers was determined by the ferrous ion oxidation--xylenol
method. Data are expressed as the mean SD of three independent
experiments. P 5 0.005, as compared between the indicated groups. (B)
One hour after the addition of EGCG to 100 mM phosphate buffers,
absorption spectra were measured at a wavelength range of 200--400 nm.
Similar results were obtained in two other experiments.
suggest that Oÿ
2 participates in the process of H2O2 generation
by EGCG.
Discussion
In the present study, we investigated the antitumor mechanism
of EGCG by using human T-cell acute lymphoblastic leukemia
Jurkat cells. EGCG was found to reduce viability of the cells
1571
H.Nakagawa et al.
Fig. 7. Suppression by SOD of EGCG-mediated H2O2 generation and Jurkat
cell death. Jurkat cells were cultured in the presence or absence of either
SOD (A and B, indicated concentrations; C and D, 500 U/ml) or catalase (C,
100 U/ml) in medium A. (A, B and C) After incubation with 50 mM EGCG or
50 mM H2O2 (D) for 6 h, cells were further incubated in fresh medium A
containing MTT reagent for 2 h to determine MTT-reducing activity (A and
D) or directly used for lysate preparation to determine caspase-3 activity (C).
(A and D) Data are expressed as percentage of control (mean SD of
triplicate cultures). (B) Similar results were obtained in two other
experiments. (C) Thirty minutes after the addition of EGCG (50 mM), the
concentration of H2O2 in the medium was determined by the ferrous ion
oxidation--xylenol method. Data are expressed as the mean SD of triplicate
determinations. P 5 0.005, as compared between the indicated groups.
and activate cellular caspase-3. These results were consistent
with the previous findings that the antitumor effect of EGCG
resulted from apoptosis (17--23). Yang et al. reported that
catalase suppressed the antitumor effect of EGCG (12,24),
indicating a major role of H2O2 in the cytotoxic effect of
EGCG. The present study demonstrated that the Fe(II)-specific
chelator o-phenanthroline, as well as catalase, suppressed the
cytotoxic effect of EGCG in Jurkat cells. Although the suppression by catalase was complete, the effect of o-phenanthroline
was partial (~75% restoration of viability). In addition,
o-phenanthroline partially blocked the H2O2 cytotoxicity itself
(~60% restoration of viability). These results suggest that
cytotoxicity of EGCG is due to its ability to generate H2O2
and that resulting H2O2 exerts cytotoxicity through two distinct mechanisms, Fe(II)-dependent and -independent ones. In
Jurkat cells, the Fe(II)-dependent mechanism may play a major
role because o-phenanthroline restores ~75% of EGCG cytotoxicity. Our results also suggest that EGCG itself is not
directly involved in Jurkat cell death, but a substance formed
in the presence of H2O2 and Fe(II) may act as a direct effector
in the cytotoxic effect. One possible candidate is hydroxyl
radical, a highly reactive oxygen species that can be formed
from H2O2 in the presence of Fe(II) through a Fenton reaction.
Our previous findings have suggested that EGCG acted
mainly in reducing Fe(III) to Fe(II), which in turn promoted
radical formation via Fenton chemistry, leading to apoptotic
cell death in osteoclastic cells (5). However, experiments using
EGCG analogs with a different activity to reduce Fe(III)
1572
suggested that the Fe(III)-reducing activity of catechins was
not related to their cytotoxic effects in the Jurkat cell system.
On the other hand, we have found that catechins with a pyrogallol moiety have H2O2-producing activity and this activity is
responsible for the cytotoxic effect in Jurkat cells. The amount
of H2O2 generated by EGCG was demonstrated to be sufficient
to exert Fe(II)-dependent cytotoxicity in Jurkat cells. Antunes
and Cadenas have shown that, following the exposure to H2O2,
the cytosolic concentration of H2O2 in Jurkat cells reaches a
steady-state level at the few seconds, with a ratio of intracellular to extracellular concentration of 1:7 (39). They also
have shown that H2O2 induces apoptosis in Jurkat cells
at intracellular concentrations of 1--3 mM (40). Thus, the
intracellular concentration of H2O2 in EGCG-treated Jurkat
cells could be assumed to be ~1 mM (1 h after treatment with
50 mM EGCG).
The results that some catechins produce H2O2 in culture
medium are consistent with observations by other investigators
(33,41), while its relationship with cell death remains to be
elucidated in those studies. Hydrogen peroxide has been
reported to act as a second messenger that is involved in
chemokine expression in macrophages (42), activation of lymphocytes (43) and activation of transcription factors (44).
Hydrogen peroxide is potentially harmful when it oxidizes
proteins, lipids and nucleic acids (45,46). Furthermore, it
becomes more harmful when it reacts with transition metal
ions [such as Fe(II) and Cu(II)] to afford hydroxyl radicals via
a Fenton reaction (47,48). As mentioned above, we have
shown that cytotoxicity of EGCG was due to its activity to
produce H2O2 and that the effect of H2O2, as well as the effect
of EGCG, was dependent mainly on Fe(II). These results
suggest an important role of the Fe(II)-dependent mechanism,
rather than the role as a second messenger or direct modifier of
biological molecules in the H2O2 cytotoxicity in Jurkat cells.
Taken together, we propose a mechanism that EGCG primarily
acts to generate H2O2 and that the resulting H2O2 triggers an
Fe(II)-dependent reaction to form highly toxic radicals, which
in turn induce apoptotic cell death in Jurkat cells. The idea that
cytotoxic potency of EGCG is due to its activity to generate
H2O2 is further supported by the observation that cells with
lower activity to eliminate H2O2 are more sensitive to EGCG.
Higher activity to eliminate H2O2 in normal cells, as demonstrated with NHDF, may explain relative specificity of EGCG
to certain tumor cells as compared with normal cells.
Although detailed mechanism of EGCG to generate H2O2 in
cell culture medium has not been investigated fully, the observation that EGCG caused H2O2 generation even in sodium phosphate buffer indicates that most of the medium components,
including amino acids, vitamins and inorganic salts, were not
essential to EGCG-mediated H2O2 generation. Rather, pH
levels may be an important factor, as EGCG-mediated H2O2
generation prefers alkaline pH and is not observed at pH 5.8.
The experiments using some EGCG analogs indicated that a
pyrogallol moiety in the catechin molecule played an essential
role in H2O2 generation and cytotoxic effects. As the pKa1
value of EGCG is reported to be at 7.59--7.75 (37,38),
the process for deprotonation in the pyrogallol moiety and/or
deprotonated form of EGCG may play a crucial role in H2O2
generation. It was reported previously that varying amount
of H2O2 was generated when EGCG was added to several
cell culture media (33). Our preliminary experiments
suggested that pyruvate, which is included in some types
of medium, suppressed EGCG-mediated H2O2 generation.
Fe(II)-dependent apoptosis in Jurkat cells by EGCG
Pyruvate non-enzymatically reacts with H2O2 to afford acetate, carbon dioxide and water (CH3COCOOÿ ‡ H2O2 !
CH3COOÿ ‡ CO2 ‡ H2O) (49). It is probable that such variation is due to the difference in the amount of antioxidative
components such as pyruvate in the medium. We have found
that SOD inhibited EGCG-mediated H2O2 generation as well
as the cytotoxic effect of EGCG, but not of H2O2. These
observations raise the possibility that Oÿ
2 plays a key role in
EGCG-mediated H2O2 generation, and further investigation on
the role of Oÿ
2 may provide us with important information to
understand the entire mechanism.
In conclusion, a novel antitumor mechanism of EGCG was
proposed here, and this mechanism may be a potential candidate that explains the diverse effects of EGCG.
Acknowledgements
This work was supported partly by the Grant of the 21st Century COE Program
from the Ministry of Education, Culture, Sports, Science and Technology of
Japan.
References
1. Anderson,R.F., Fisher,L.J., Hara,Y., Harris,T., Mak,W.B., Melton,L.D.
and Packer,J.E. (2001) Green tea catechins partially protect DNA from
OH radical-induced strand breaks and base damage through fast chemical
repair of DNA radicals. Carcinogenesis, 22, 1189--1193.
2. Higdon,J.V. and Frei,B. (2003) Tea catechins and polyphenols: health
effects, metabolism and antioxidant functions. Crit. Rev. Food Sci. Nutr.,
43, 89--143.
3. Ikigai,H., Nakae,T., Hara,Y. and Shimamura,T. (1993) Bactericidal
catechins damage the lipid bilayer. Biochim. Biophys. Acta, 1147,
132--136.
4. Toering,S.J., Gentile,G.J. and Gentile,J.M. (1996) Mechanism of antimutagenic action of (‡)-catechin against the plant-activated aromatic
amine 4-nitro-o-phenylenediamine. Mutat. Res., 361, 81--87.
5. Nakagawa,H., Wachi,M., Woo,J.T., Kato,M., Kasai,S., Takahashi,F.,
Lee,I.S. and Nagai,K. (2002) Fenton reaction is primarily involved in a
mechanism of (ÿ)-epigallocatechin-3-gallate to induce osteoclastic cell
death. Biochem. Biophys. Res. Commun., 292, 94--101.
6. Ahmad,N., Cheng,P. and Mukhtar,H. (2000) Cell cycle dysregulation by
green tea polyphenol epigallocatechin-3-gallate. Biochem. Biophys. Res.
Commun., 275, 328--334.
7. Nam,S., Smith,D.M. and Dou,Q.P. (2001) Ester bond-containing tea
polyphenols potently inhibit proteasome activity in vitro and in vivo.
J. Biol. Chem., 276, 13322--13330.
8. Ahmad,N., Feyes,D.K., Nieminen,A.L., Agarwal,R. and Mukhtar,H.
(1997) Green tea constituent epigallocatechin-3-gallate and induction of
apoptosis and cell cycle arrest in human carcinoma cells. J. Natl Cancer
Inst., 89, 1881--1886.
9. Tan,X., Hu,D., Li,S., Han,Y., Zhang,Y. and Zhou,D. (2000) Differences of
four catechins in cell cycle arrest and induction of apoptosis in LoVo cells.
Cancer Lett., 158, 1--6.
10. Ahmad,N., Gupta,S. and Mukhtar,H. (2000) Green tea polyphenol
epigallocatechin-3-gallate differentially modulates nuclear factor kappaB
in cancer cells versus normal cells. Arch. Biochem. Biophys., 376,
338--346.
11. Gupta,S., Hussain,T. and Mukhtar,H. (2003) Molecular pathway for (ÿ)epigallocatechin-3-gallate-induced cell cycle arrest and apoptosis of
human prostate carcinoma cells. Arch. Biochem. Biophys., 410, 177--185.
12. Yang,G.Y., Liao,J., Kim,K., Yurkow,E.J. and Yang,C.S. (1998) Inhibition
of growth and induction of apoptosis in human cancer cell lines by tea
polyphenols. Carcinogenesis, 19, 611--616.
13. Otsuka,T., Ogo,T., Eto,T., Asano,Y., Suganuma,M. and Niho,Y. (1998)
Growth inhibition of leukemic cells by (ÿ)-epigallocatechin gallate, the
main constituent of green tea. Life Sci., 63, 1397--1403.
14. Chen,Z.P., Schell,J.B., Ho,C.T. and Chen,K.Y. (1998) Green tea
epigallocatechin gallate shows a pronounced growth inhibitory effect on
cancerous cells but not on their normal counterparts. Cancer Lett., 129,
173--179.
15. Saeki,K., Sano,M., Miyase,T., Nakamura,Y., Hara,Y., Aoyagi,Y. and
Isemura,M. (1999) Apoptosis-inducing activity of polyphenol compounds
derived from tea catechins in human histiolytic lymphoma U937 cells.
Biosci. Biotechnol. Biochem., 63, 585--587.
16. Naasani,I., Oh-Hashi,F., Oh-Hara,T., Feng,W.Y., Johnston,J., Chan,K. and
Tsuruo,T. (2003) Blocking telomerase by dietary polyphenols is a major
mechanism for limiting the growth of human cancer cells in vitro and
in vivo. Cancer Res., 63, 824--830.
17. Islam,S.,
Islam,N.,
Kermode,T.,
Johnstone,B.,
Mukhtar,H.,
Moskowitz,R.W., Goldberg,V.M., Malemud,C.J. and Haqqi,T.M. (2000)
Involvement of caspase-3 in epigallocatechin-3-gallate-mediated apoptosis
of human chondrosarcoma cells. Biochem. Biophys. Res. Commun., 270,
793--797.
18. Pan,M.H., Liang,Y.C., Lin-Shiau,S.Y., Zhu,N.Q., Ho,C.T. and Lin,J.K.
(2000) Induction of apoptosis by the oolong tea polyphenol theasinensin A
through cytochrome c release and activation of caspase-9 and caspase-3 in
human U937 cells. J. Agric. Food Chem., 48, 6337--6346.
19. Hayakawa,S., Saeki,K., Sazuka,M., Suzuki,Y., Shoji,Y., Ohta,T., Kaji,K.,
Yuo,A. and Isemura,M. (2001) Apoptosis induction by epigallocatechin
gallate involves its binding to Fas. Biochem. Biophys. Res. Commun., 285,
1102--1106.
20. Sakagami,H., Arakawa,H., Maeda,M., Satoh,K., Kadofuku,T., Fukuchi,K.
and Gomi,K. (2001) Production of hydrogen peroxide and methionine
sulfoxide by epigallocatechin gallate and antioxidants. Anticancer Res.,
21, 2633--2641.
21. Chen,C., Shen,G., Hebbar,V., Hu,R., Owuor,E.D. and Kong,A.N. (2003)
Epigallocatechin-3-gallate-induced stress signals in HT-29 human colon
adenocarcinoma cells. Carcinogenesis, 24, 1369--1378.
22. Saeki,K., Kobayashi,N., Inazawa,Y., Zhang,H., Nishitoh,H., Ichijo,H.,
Isemura,M. and Yuo,A. (2002) Oxidation-triggered c-Jun N-terminal
kinase (JNK) and p38 mitogen-activated protein (MAP) kinase pathways
for apoptosis in human leukaemic cells stimulated by epigallocatechin-3gallate (EGCG): a distinct pathway from those of chemically induced and
receptor-mediated apoptosis. Biochem. J., 368, 705--720.
23. Vergote,D., Cren-Olive,C., Chopin,V., Toillon,R.A., Rolando,C.,
Hondermarck,H. and Le Bourhis,X. (2002) (ÿ)-Epigallocatechin (EGC)
of green tea induces apoptosis of human breast cancer cells but not of their
normal counterparts. Breast Cancer Res. Treat., 76, 195--201.
24. Yang,G.Y., Liao,J., Li,C., Chung,J., Yurkow,E.J., Ho,C.T. and Yang,C.S.
(2000) Effect of black and green tea polyphenols on c-jun phosphorylation
and H2O2 production in transformed and non-transformed human
bronchial cell lines: possible mechanisms of cell growth inhibition and
apoptosis induction. Carcinogenesis, 21, 2035--2039.
25. Morre,D.J., Bridge,A., Wu,L.Y. and Morre,D.M. (2000) Preferential
inhibition by (ÿ)-epigallocatechin-3-gallate of the cell surface NADH
oxidase and growth of transformed cells in culture. Biochem. Pharmacol.,
60, 937--946.
26. Wang,Y.C. and Bachrach,U. (2002) The specific anti-cancer activity of
green tea (ÿ)-epigallocatechin-3-gallate (EGCG). Amino Acids, 22,
131--143.
27. Morre,D.J., Morre,D.M., Sun,H., Cooper,R., Chang,J. and Janle,E.M.
(2003) Tea catechin synergies in inhibition of cancer cell proliferation and
of a cancer specific cell surface oxidase (ECTO-NOX). Pharmacol.
Toxicol., 92, 234--241.
28. Naasani,I., Seimiya,H. and Tsuruo,T. (1998) Telomerase inhibition,
telomere shortening and senescence of cancer cells by tea catechins.
Biochem. Biophys. Res. Commun., 249, 391--396.
29. Pisters,K.M., Newman,R.A., Coldman,B., Shin,D.M., Khuri,F.R.,
Hong,W.K., Glisson,B.S. and Lee,J.S. (2001) Phase I trial of oral green
tea extract in adult patients with solid tumors. J. Clin. Oncol., 19,
1830--1838.
30. Jatoi,A., Ellison,N., Burch,P.A. et al. (2003) A phase II trial of green tea in
the treatment of patients with androgen independent metastatic prostate
carcinoma. Cancer, 97, 1442--1446.
31. Gupta,S., Hastak,K., Ahmad,N., Lewin,J.S. and Mukhtar,H. (2001)
Inhibition of prostate carcinogenesis in TRAMP mice by oral infusion
of green tea polyphenols. Proc. Natl Acad. Sci. USA, 98, 10350--10355.
32. Nourooz-Zadeh,J., Tajaddini-Sarmadi,J. and Wolff,S.P. (1994)
Measurement of plasma hydroperoxide concentrations by the ferrous
oxidation-xylenol orange assay in conjunction with triphenylphosphine.
Anal. Biochem., 220, 403--409.
33. Long,L.H., Clement,M.V. and Halliwell,B. (2000) Artifacts in cell culture:
rapid generation of hydrogen peroxide on addition of (ÿ)-epigallocatechin, (ÿ)-epigallocatechin gallate, (‡)-catechin and quercetin to
commonly used cell culture media. Biochem. Biophys. Res. Commun.,
273, 50--53.
1573
H.Nakagawa et al.
34. Chance,B., Sies,H. and Boveris,A. (1979) Hydroperoxide metabolism in
mammalian organs. Physiol. Rev., 59, 527--605.
35. Takagi,Y., Mitsui,A., Nishiyama,A., Nozaki,K., Sono,H., Gon,Y.,
Hashimoto,N. and Yodoi,J. (1999) Overexpression of thioredoxin in
transgenic mice attenuates focal ischemic brain damage. Proc. Natl Acad.
Sci. USA, 96, 4131--4136.
36. Bai,J., Rodriguez,A.M., Melendez,J.A. and Cederbaum,A.I. (1999)
Overexpression of catalase in cytosolic or mitochondrial compartment
protects HepG2 cells against oxidative injury. J. Biol. Chem., 274,
26217--26224.
37. Jovanovic,S.V., Hara,Y., Steenken,S. and Simic,M.G. (1995) Antioxidant
potential of gallocatechins. A pulse radiolysis and laser photolysis study.
J. Am. Chem. Soc., 117, 9881--9888.
38. Kumamoto,M., Sonda,T., Nagayama,K. and Tabata,M. (2001) Effects of
pH and metal ions on antioxidative activities of catechins. Biosci.
Biotechnol. Biochem., 65, 126--132.
39. Antunes,F. and Cadenas,E. (2000) Estimation of H2O2 gradients across
biomembranes. FEBS Lett., 475, 121--126.
40. Antunes,F. and Cadenas,E. (2001) Cellular titration of apoptosis with
steady state concentrations of H2O2: submicromolar levels of H2O2 induce
apoptosis through Fenton chemistry independent of the cellular thiol state.
Free Radic. Biol. Med., 30, 1008--1018.
41. Dashwood,W.M., Orner,G.A. and Dashwood,R.H. (2002) Inhibition of
beta-catenin/Tcf activity by white tea, green tea and epigallocatechin-3gallate (EGCG): minor contribution of H2O2 at physiologically relevant
EGCG concentrations. Biochem. Biophys. Res. Commun., 296, 584--588.
1574
42. Jaramillo,M. and Olivier,M. (2002) Hydrogen peroxide induces murine
macrophage chemokine gene transcription via extracellular signalregulated kinase- and cyclic adenosine 50 -monophosphate (cAMP)dependent pathways: involvement of NF-kappa B, activator protein 1
and cAMP response element binding protein. J. Immunol., 169,
7026--7038.
43. Reth,M. (2002) Hydrogen peroxide as second messenger in lymphocyte
activation. Nat. Immunol., 3, 1129--1134.
44. Rhee,S.G. (1999) Redox signaling: hydrogen peroxide as intracellular
messenger. Exp. Mol. Med., 31, 53--59.
45. Stadtman,E.R. (1992) Protein oxidation and aging. Science, 257,
1220--1224.
46. Stadtman,E.R. and Berlett,B.S. (1998) Reactive oxygen-mediated protein
oxidation in aging and disease. Drug Metab. Rev., 30, 225--243.
47. Halliwell,B. and Gutteridge,J.M. (1990) Role of free radicals and
catalytic metal ions in human disease: an overview. Methods Enzymol.,
186, 1--85.
48. Stohs,S.J. and Bagchi,D. (1995) Oxidative mechanisms in the toxicity of
metal ions. Free Radic. Biol. Med., 18, 321--336.
49. Hinoi,E., Fujimori,S., Takemori,A. and Yoneda,Y. (2002) Cell death by
pyruvate deficiency in proliferative cultured calvarial osteoblasts.
Biochem. Biophys. Res. Commun., 294, 1177--1183.
Received October 29, 2003; revised March 14, 2004;
accepted April 9, 2004