Antiviral effect of catechins in green tea on influenza

Abstract

Polyphenolic compound catechins ((-)-epigallocatechin gallate (EGCG), (-)-epicatechin gallate (ECG) and (-)-epigallocatechin (EGC)) from green tea were evaluated for their ability to inhibit influenza virus replication in cell culture and for potentially direct virucidal effect. Among the test compounds, the EGCG and ECG were found to be potent inhibitors of influenza virus replication in MDCK cell culture and this effect was observed in all influenza virus subtypes tested, including A/H1N1, A/H3N2 and B virus. The 50% effective inhibition concentration (EC50) of EGCG, ECG, and EGC for influenza A virus were 22-28, 22-40 and 309-318 microM, respectively. EGCG and ECG exhibited hemagglutination inhibition activity, EGCG being more effective. However, the sensitivity in hemagglutination inhibition was widely different among three different subtypes of influenza viruses tested. Quantitative RT-PCR analysis revealed that, at high concentration, EGCG and ECG also suppressed viral RNA synthesis in MDCK cells whereas EGC failed to show similar effect. Similarly, EGCG and ECG inhibited the neuraminidase activity more effectively than the EGC. The results show that the 3-galloyl group of catechin skeleton plays an important role on the observed antiviral activity, whereas the 5'-OH at the trihydroxy benzyl moiety at 2-position plays a minor role. The results, along with the HA type-specific effect, suggest that the antiviral effect of catechins on influenza virus is mediated not only by specific interaction with HA, but altering the physical properties of viral membrane.

Full text

Antiviral Research 68 (2005) 66–74
Antiviral effect of catechins in green tea on influenza virus
Jae-Min Song, Kwang-Hee Lee, Baik-Lin Seong∗
Department of Biotechnology, College of Engineering, Yonsei University, 134, Shinchon-dong, Seodaemun-gu, Seoul 120-749, South Korea
Received 22 February 2005; accepted 28 June 2005
Abstract
Polyphenolic compound catechins ((−)-epigallocatechin gallate (EGCG), (−)-epicatechin gallate (ECG) and (−)-epigallocatechin (EGC))
from green tea were evaluated for their ability to inhibit influenza virus replication in cell culture and for potentially direct virucidal effect.
Among the test compounds, the EGCG and ECG were found to be potent inhibitors of influenza virus replication in MDCK cell culture
and this effect was observed in all influenza virus subtypes tested, including A/H1N1, A/H3N2 and B virus. The 50% effective inhibition
concentration (EC50) of EGCG, ECG, and EGC for influenza A virus were 22–28, 22–40 and 309–318 ␮M, respectively. EGCG and ECG
exhibitedhemagglutinationinhibitionactivity,EGCGbeingmoreeffective.However,thesensitivityinhemagglutinationinhibitionwaswidely
different among three different subtypes of influenza viruses tested. Quantitative RT-PCR analysis revealed that, at high concentration, EGCG
and ECG also suppressed viral RNA synthesis in MDCK cells whereas EGC failed to show similar effect. Similarly, EGCG and ECG inhibited
the neuraminidase activity more effectively than the EGC. The results show that the 3-galloyl group of catechin skeleton plays an important
role on the observed antiviral activity, whereas the 5-OH at the trihydroxy benzyl moiety at 2-position plays a minor role. The results, along
with the HA type-specific effect, suggest that the antiviral effect of catechins on influenza virus is mediated not only by specific interaction
with HA, but altering the physical properties of viral membrane.
© 2005 Elsevier B.V. All rights reserved.
Keywords: Influenza virus; Catechins; Green tea; EGCG; ECG; Hemagglutination
1. Introduction
Green tea is produced from the leaves of the evergreen
plantCamellia sinensis.Themajoractiveingredientsofgreen
tea are polyphenolic compounds, known as catechins. Cat-
echins of green tea leaves are account for about 10% of
dry weight, including (−)-epigallocatechin gallate (EGCG),
(−)-epigallocatechin (EGC), (−)-epicatechin gallate (ECG)
(Fig.1) and (−)-epicatechin(EC), where EGCGaccounts for
approximately50% of thetotal amounts of catechins in green
tea (Balentine et al., 1997).
Various biological and pharmacological activities have
been reported for EGCG, including antioxidative (Valcic
et al., 1999), antibacterial (Toda et al., 1992), antitumor
and antiviral activity. The antitumor effects of EGCG have
been investigated in detail and the compound demonstrated
inhibitory action against carcinogenesis on several differ-
∗Corresponding author. Tel.: +82 2 2123 2885; fax: +82 2 362 7265.
E-mail address: blseong@yonsei.ac.kr (B.-L. Seong).
ent organs in animal models (Gensler et al., 1996; Katiyar
and Mukhtar, 1996; Mimoto et al., 2000; Yamane et al.,
1995; Yang et al., 2002). Studies have shown that antitumor
effect correlates with inactivation of tumor-related proteases
(Jankun et al., 1997), nitric oxide synthase (NOS) (Lin and
Lin, 1997), and is mediated by PI3-to Akt kinase in the NF-
kB pathway (Pianetti et al., 2002). Recently, laminin receptor
has been identified as a receptor for EGCG mediating the anti
cancer activity (Tachibana et al., 2004).
With regard to the antiviral activities, EGCG acts as a
strong inhibitor of HIV replication in cultured peripheral
blood cells (Fassina et al., 2002), and EGCG and ECG were
found more effective than EGC or EC in the inhibition of the
HIV-1 reverse transcriptase in vitro (Nakane and Ono, 1990).
EGCG also binds directly to CD4 molecule with consequent
inhibition of gp120 binding (Kawai et al., 2003). EGCG
also induced inactivation of virus in vitro by deformation
of phospholipids (Yamaguchi et al., 2002). Distinct antiviral
activities of EGCG were reported for Epstein–Barr virus
inhibition of expression of viral proteins (Chang et al., 2003)
0166-3542/$ – see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.antiviral.2005.06.010

J.-M. Song et al. / Antiviral Research 68 (2005) 66–74 67
Fig. 1. Structures of (−)-epigallocatechin (EGC) (a), (−)-epicatechin gallate (ECG) (b), (−)-epigallocatechin gallate (EGCG) (C), (+)-catechin (C) and
(−)-epicatechin (EC).
and inhibition of host factors (Weber et al., 2003). Antiviral
effects of EGCG have also been reported for influenza virus.
EGCG affected the infectivity of influenza virus in cell cul-
ture, and it was shown to agglutinate the viruses, preventing
the viruses from absorbing to MDCK cells (Nakayama et
al., 1993). It was also shown that green tea extract exerted an
inhibitory effect on the acidification of intracellular compart-
ments such as endosomes and lysosomes, resulting in inhibi-
tion of growth of influenza virus in cell culture (Imanishi et
al., 2002). So far, however, biochemical studies on the anti-
influenza effects of green tea polyphenols have focused on
EGCG. With a view to investigate the structure-activity rela-
tionshipsofthe green tea polyphenolic compounds, we inves-
tigated in this report the antiviral effects of various catechin
compounds, EGCG, ECG and EGC, on influenza virus. The
studies were further extended to all three currently circulat-
ing virus subtypes, including two different H3N2 and H1N1
influenza A types and one influenza B type. Here, we found
that ECG and EGCG were much more effective than EGC,
and besides the known inhibition of hemagglutination, the
compounds also exerted inhibitory effect on neuraminidase
and affects viral RNA synthesis at high concentration.
2. Materials and methods
2.1. Preparation of catechins
Green tea extract was prepared by infusing the leaves
(Sulloc Cha, Amore-Pacific Co. Ltd., Korea) with 75◦C
distilled water in the ratio of 1:7 (w/w). After 20min of
infusion, the tea extract was quickly separated from the
tea leaves by filtration and the tea extract was freeze-
dried for further tests. To isolate catechins, green tea
leaves were extracted with five times of 90◦C distilled
water for 5h, and the tea extract was washed with chlo-
roform and extracted with ethyl acetate three times. The
ethyl acetate layers were combined and vacuum evaporated.
The powder remained was re-dissolved in 95% ethanol,
and loaded to Sephadex LH-20 HPLC column for further
purification. The catechin fractions collected were freeze-
dried and stored in the freezer (−20◦C) before use. Both
the green tea extract and isolated catechins were ana-
lyzed for their composition by C18 reverse phase column
chromatography (elution with 22% THF at the flow of
1 ml/min). Each purified catechins were dissolved in DMSO
(dimethylsulfoxide) and stored in the freezer (−20◦C).
(−)-Epicatechin and (+)-catechin (C) were purchased from
Sigma-Aldrich.
2.2. Viruses and cells
Influenza A/Chile/1/83 (H1N1), A/Sydney/5/97 (H3N2)
and B/Yamagata/16/88 viruses were propagated in 11-day
chick embryos. Allantoic fluids were harvested and stored at
the freezer (−80◦C). Madin–Darby canine kidney (MDCK)
cells were cultured as monolayers in the MEM medium (JBI,
Korea) supplemented with 10% heat-inactivated fetal bovine
serum (BioWhittaker, USA).

68 J.-M. Song et al. / Antiviral Research 68 (2005) 66–74
2.3. Plaque inhibition assay
For plaque inhibition assays, confluent monolayer
MDCK cells cultured in a 6-well tissue culture plate
(1×105cells/cm2) were infected with a mixture of approx-
imately 500PFU/ml of virus. After 60min for virus adsorp-
tion, the solution was removed and the cells were washed
twice with pre-warmed MEM medium, and replaced with
overlay medium (DMEM containing 10␮g/ml trypsin, 1%
low melting agarose, without serum), containing catechins
at different concentration. After incubating cultures for 2–3
days at 37◦C with 5% CO2, monolayers were fixed with
4% formaldehyde solution for 30min and the agarose was
then removed by flowing water and stained with 1% (w/v)
crystal violet solution. The plaques were counted by visual
examination and percentage of plaque inhibition was calcu-
lated as relative to the control without catechins. A required
concentration to reduce the 50% plaque number (EC50), was
calculatedbyregressionanalysisof the dose–response curves
generated from these data.
2.4. Virus growth inhibition assay
Confluent monolayers of MDCK cells in 12-well plates
were washed once with phosphate buffered saline (PBS) and
then infected with influenza A/Chile/1/83 (H1N1) virus at
0.1 multiplicity of infection (MOI). The plates were regu-
larly inverted on a shaker for 45min at room temperature in
compounds-free conditions for virus adsorption. The solu-
tion was removed and replaced with MEM medium con-
taining catechins of various concentrations. Viruses were
harvested at 8, 24, 36h post-infection, and the viral yield
was estimated by plaque assay on MDCK cells. As a control,
the infected cells incubated in catechin-free medium were
included throughout the experiment.
2.5. Hemagglutination inhibition assay
Hemagglutination inhibition assay was employed to test
the effect of catechins in virus adsorption to target cells.
Catechin solutions (25␮l) with two-fold serial dilution with
PBS were mixed with equal volume of influenza virus solu-
tion (200HAU/25␮l). After incubation for 30 min at room
temperature, 50 ␮l of the solution was mixed with equal vol-
ume of 1% chicken erythrocyte suspension and incubated for
30min at room temperature.
2.6. Quantitative RT-PCR analysis
MDCK cells were grown at about 90% confluence and
infected with influenza A/Chile/1/83 virus at 0.1 MOI and
cultured in the presence of catechins at various concentra-
tions. At 16h post-infection, cells were scraped off and col-
lected by centrifugation (500g for 5min). Cell pellets were
washed with PBS twice. Total cellular and viral RNAs were
isolated from pellets using the RNeasy mini kit (QIAGEN)
followingthe manufacture’sprotocol. First-strandcDNA was
synthesized from 1␮g of total RNA with Omniscript RT kit
(QIAGEN) using specific primers. PCR reactions were per-
formed with 50 ␮l of reaction buffer [5␮l of cDNA template,
50pmols of primers, 0.1mM dNTPs, and 0.5 U of EX-Taq
polymerase (Takara)]. The amplification conditions were as
follows: 94 ◦C for 5 min (1 cycle), 94 ◦C for 1 min, 55 ◦C40s
and 72 ◦C 1min 40 s (18 cycles, respectively). NP RNA was
chosen for detection and the primer sequences used for the
detection of viral RNA were 5-TGC TGG ATT CTC GTT
CGG TC (sense) and 5-CCT TTA TGA CAA AGA AGA
AATAAG GCG (antisense). The ␤-actinwas used as internal
control of cellular RNAs, with primer sequences of 5-TCA
CCC GAG TCC ATC ACG AT (sense) and 5-GAA GTA
CCC CAT TGA GCA CGG (antisense). The reverse tran-
scription and PCR products were resolved on 1.0% agarose
gels and stained with ethidium bromide.
2.7. Neuraminidase inhibition assay
Neuraminidase inhibition assay was employed to test the
effect of catechins in neuraminidase activity in virus. Cate-
chinsolutions (25 ␮l) with two-fold serial dilutions with PBS
were mixed with equal volume of influenza A/Chile/1/83
virus solution. Equal volume (50␮l) of the substrate
solution (4-MU-NANA; 2-(4-methylumbelliferyl)-A-d-N-
acetylneuraminic acid sodium, Sigma) was added and the
mixturewasfurther incubated at 37 ◦C for2h, protected from
light. Optical density is then measured by fluorescence of
4-methylumbelliferone with fluorescence spectrophotometer
(excitation 365nm, emission 460 nm).
Relative activities were calculated by the following for-
mula:
Relative activities(%) =NA activites with catechins
NA activites without catechins
×100
2.8. Cytotoxicity test by MTT assay
MDCKcells were grown (8800 cells/well) in 96-well plate
for 24 h. The medium was replaced with that containing seri-
ally diluted catechins and the cells were further incubated for
48h. The culture medium was removed and 25␮l MTT, 3-
(4,5-dimethylthiozol-2-yl)-3,5-dipheryltetrazolium bromide
(Sigma) solution was added to each well and incubated at
37◦C for 4 h. After removal of supernatant, 50␮l of DMSO
was added for solubilization of formazan crystals and incu-
batedfor30 min. The optical density wasmeasuredat540 nm
in ELISA reader.
2.9. Statistical analysis
The results were expressed as mean±S.E.M. for three
independent experiments. Student’s unpaired t-test was used
to evaluate the difference between the test sample and

J.-M. Song et al. / Antiviral Research 68 (2005) 66–74 69
Fig. 2. Inhibitory effects of catechins on plaque formation. Six-well plates containing monolayers of MDCK cells were washed and infected with equal volume
of virus suspension (500PFU/ml). The overlay medium, mixed with 2% low melting agarose and DMEM, containing each catechin on 50␮Mor50␮g/ml
with total extracts and polyphenols mixture were added to each plate. The plates were incubated at 37 ◦C for 48h. Plaques were stained with crystal violet and
counted. The percentage of plaque formation ratio relative to the control (no catechin) was determined for each catechin.
untreated control. A Pvalue of <0.05 was considered sta-
tistically significant. And also one-way ANOVA was used to
evaluate the difference between the test samples. We used
GraphPad PRISM software for statistical analysis.
3. Results
3.1. Inhibitory effects of catechins on plaque formation
by influenza A and B viruses in MDCK cells
For initial screening of the antiviral activity, each com-
pound was tested by plaque inhibition assay in MDCK cells
at a fixed concentration. As shown in Fig. 2, EGCG and
ECG at 50 ␮M concentration inhibited more than 50% of the
plaque forming activity of influenza A and B viruses whereas
EGC exhibited little inhibition. Notably, polyphenols mix-
ture is more efficient than any other single compounds in
plaque inhibition. For a given compound, similar inhibitory
effect was observed for all influenza viruses tested, including
A/H1N1, A/H3N2 and B virus.
Eachcatechin compounds werethen tested atvariouscon-
centrations. The mean 50% inhibitory concentration (EC50)
of catechins, represented as the mean of at least three
independent experiments, are summarized in Table 1.In
summary, EGCG and ECG was about 10-fold more effec-
tive than EGC. This finding was confirmed for all differ-
ent influenza viruses, including two different subtype of A
viruses (A/Chile/1/83 (H1N1) and A/Sydney/5/97 (H3N2))
and B virus (B/Yamagata/16/88).
Onthecontrary (−)-epicatechin, the back bonecompound
ofgreen tea polyphenols, hadno antiviral effectevenat1 mM
concentration and (+)-catechin, the geometric isomer of EC,
exerted little inhibition on plaque formation. (+)-Catechin
showed plaque inhibition of only about 20% at 600␮M, and
at higher concentration (summarized in Table 1).
3.2. Inhibitory effects of catechins on influenza virus
replication in MDCK cells
Next, we analyzed the effect of catechins on virus yield
at various times post-infection of MDCK cells. The cells
were infected with 0.1 MOI influenza A virus and various
concentrations of catechins were added in culture medium.
At 8, 24 and 36h post-infection, virus yields were deter-
mined by plaque assay. As shown in Fig. 3, the virus yields
in MDCK cell were reduced by 2–6log units depending
on the concentration of the compounds. Virtually complete
reduction in virus yields was observed at highest concen-
tration tested (120, 1200 and 120␮M for EGCG, EGC and
Table 1
Inhibitory effects of catechin compounds on plaque formation on MDCK cells
Virus EC50a
EC and C (␮M) EGCG (␮M) EGC (␮M) ECG (␮M) Total extracts (␮g/ml) Polyphenols mixture (␮g/ml)
A/Chile/1/83 >600 28.4±1.4 318 ±3.0 26.4±2.0 70.5 ±1.2 13.1 ±1.4
A/Sydney/5/97 >600 22.8±1.1 309 ±1.2 22.2±1.1 63.8 ±1.2 15.2 ±1.2
B/Yamagata/16/88 >600 26.1±1.4 311.1 ±2.4 40.4±1.7 68.2 ±2.1 14.1 ±1.6
Values represent the mean ±S.E.M. for three independent experiments.
aEC50 represents the concentration of catechin necessary to reduce the plaque number by 50% relative to control wells without test compound, calculated
from dose–response data of plaque inhibition.

70 J.-M. Song et al. / Antiviral Research 68 (2005) 66–74
Fig. 3. Inhibitory effects of catechins on influenza virus yield in MDCK cells. Confluent monolayers of MDCK cells in 12-well plates were infected with
A/Chile/1/83 (H1N1) virus at 0.1 MOI. After 45min it was replaced with MEM medium containing each catechin in various concentration without serum.
Culturemedium was harvested at 8, 24and 36 h post-infection. The viral yieldwas estimated by plaque assay on MDCK cells. The asterisk indicates a significant
difference between test sample and control, *P<0.05, **p<0.001.
ECG, respectively, and 200 and 24␮g/ml for total extracts
and polyphenols mixture, respectively). For example, EGCG
reduced about 4log units of virus yields at 30␮M, and
this significant reduction (P<0.001) was observed at all
phases of infection (8, 24 and 36h post-infection). Com-
plete inhibition was observed at 120␮M, where we failed
to observe any release of viruses. In this assay, EGCG
was by far the most effective, and EGC was least effec-
tive among three isolated compounds tested. Again, we
observed that polyphenols mixture was very effective (fail-
ure to detect any viral yield at 24␮g/ml concentration). The
results further extend previous data on plaque inhibition
(Fig. 2).
Overall, the inhibitory effect is observed in concentration
dependent manner, and throughout the virus infection cycle
after initial infection. The results suggest that the antiviral
effect is exerted not only on the initially infecting viruses but
newly propagated viruses as well.

J.-M. Song et al. / Antiviral Research 68 (2005) 66–74 71
Fig. 4. Inhibitory effects of catechins on virus adsorption to chicken red blood cells. Twenty-five micro moles of influenza virus suspension (200HAU/25␮l)
were mixed with an equal volume of different concentrations of catechins and incubated for 30min at room temperature. The mixture was further incubated
with 50␮l of 1% chicken red blood cells suspension for 30 min at room temperature for hamagglutination. We determined the minimum concentration of
catechins that completely inhibited the adsorption of virus. The asterisk indicates a significant difference between test samples, *P<0.05.
3.3. Inhibitory effects of catechins on the virus
adsorption with chicken red blood cells
Influenza A virus has an ability to adsorb to chicken red
blood cells resulting in hemagglutination. We investigated if
cathechins could interfere with the viral adsorption to RBC
resulting in hemagglutiniation inhibition, and analyzed the
minimum concentration of catechins that completely inhib-
ited the adsorption ability of virus. As shown in Fig. 4,
EGCG, ECG and polyphenols mixture exhibited, to some
degree, inhibition of viral adsorption of RBC, but the effect
was surprisingly different among type of viruses and cate-
chins (P<0.05). In case of EGCG, for examples, influenza
A/Sydney/5/97 virus was three- to four-fold more sensitive
and the B/Yamagata virus was seven-fold more sensitive than
A/Chile/1/83, respectively. Albeit less pronounced, clearly
different sensitivity is also observed with ECG. In another set
of experiment, the sensitivity towards EGCG was up to 15-
fold among the viruses tested (data not shown). The result is
inclear contrast to the results of plaque inhibition assay where
difference in EC50 between the two influenza A strains was
within40%. Moreover,contrary to previous results onplaque
inhibition assay and viral growth on cell culture, total extract
was more effective than the polyphenols mixture. However,
in case of EGC, we failed to observe inhibitory effect even at
the highest concentration tested in this experiment (3.5 mM).
3.4. Measurement of the viral mRNA expression in
infected cell with quantitative RT-PCR
We then tested the effect of catechins on transcription
of viral genes in infected cells by quantitative RT-PCR of
influenza virus specific mRNA. MDCK cells were infected
with A/Chile/1/83 virus and were incubated 16 h in the pres-
enceofvariousconcentrations of catechins. TotalRNAs were
isolatedfrominfected MDCK cells and RT-PCRanalysiswas
performed using primers specific for viral NP RNA.
The results showed significant differences among various
catechin compounds (Fig. 5). Down regulation of viral RNA
synthesis was evident especially at high concentration (over
600␮MEGCG,1200 ␮M ECG, respectively).Theinhibitory
effectwas mostpronounced with polyphenols mixture (about
80% inhibition at 500␮M). However, we failed to observe
inhibitory effect with EGC even at the highest concentration
testedin this experiment (3.2 mM). Asan internal control, the
transcription of cellular ␤-actin mRNA was not affected in
all catechin concentrations tested. Therefore, the inhibitory
effect on viral replication observed at high concentration
was not due to a general cytotoxic effect on the cell. The
results suggest that, in addition to direct virucidal effect by
binding to virus particles, the catechins may have additional
inhibitory effect on viral RNA synthesis, especially at high
concentration.
3.5. Inhibitory effects of catechins on the viral
neuraminidase activity
Neuraminidase is thought to play a key role in the release
of newly made virus particle from infected cells by cleavage
of target cell receptor sialic acid moieties. Additionally,
the enzyme activity is responsible for preventing self-
aggregation of virus particles by cleavage of sialic acids still
bound to the virus surface. We therefore tested the potential
effect of catechins on the viral neuraminidase activity
(see Section 2). The activity of neuraminidase decreased
significantly by EGCG and ECG, but not by EGC (Fig. 6).
Reduction of half enzymatic activity was shown at relatively
high concentration (about 350 and 550␮M for EGCG,
ECG, respectively). The results suggest that the catechins
may have additional inhibitory effect on virus release step
through inhibition of neuraminidase, especially at high
concentration.
3.6. Cellular toxicity of catechins
We evaluated the cytotoxicity of catechins by the MTT
assay. Half-confluent MDCK monolayers were incubated
with media in the absence or presence of two-fold diluted

72 J.-M. Song et al. / Antiviral Research 68 (2005) 66–74
Fig. 5. Effect of catechins on influenza viral RNA synthesis in infected cell as analyzed by quantitative RT-PCR. Ninety percent confluent cells were infected
with virus at 0.1 MOI and cultured with MEM medium containing different concentrations of each catechin and 1% serum. Cells were lysed 16 h after infection,
and total cellular and viral RNA isolated from cell pellets. Quantitative RT-PCR was performed using specific primers for viral RNA (NP) and cellular RNA
(actin). Similar results were obtained in three other experiments.
Fig. 6. Inhibitory effects of catechins on the viral neuraminidase activity. Catechin solutions (25␮l) with two-fold serial dilution with PBS were mixed with
equal volume of influenza virus solution. Fifty microliters of the substrate solution (4-MU-NANA) was mixed with equal volume catechins/virus mixture and
incubated for 2 h at 37◦C. Neuraminidase activity was monitored by fluorescence of 4-methylumbelliferone with fluorescence spectrophotometer. Each point
represents the mean ±S.E.M. for three independent experiments.

J.-M. Song et al. / Antiviral Research 68 (2005) 66–74 73
Table 2
Cellular toxicity and selectivity index of catechins on MDCK cells
EGCG (␮M) EGC (␮M) ECG (␮M) Total extracts (␮g/ml) Polyphenols mixture (␮g/ml)
CC50a275.4 ±22.8 1233.1 ±44.9 525.9±30.7 353.3 ±11.2 330.0 ±57.9
Selectivity indexb(SI) 12.1 4.0 23.7 5.5 21.7
aCC50, 50% cell toxicity concentration determined by MTT assay. The values of CC50 are mean±S.E.M.
bSI is the ratio of CC50 to EC50.
catechins and total extracts (0–30 mM or 20mg/ml) for 48 h.
MTT reagents were then added to the monolayer of MDCK
cells. After incubation at 37 ◦C for 4h, absorbance (540 nm)
was measured by ELISA reader. The estimated doses that
reduced cell viability about 50% are described in Table 2.
Theresults showed that EGC has lowest toxicity among three
isolated compounds and ECG was about two-fold less toxic
than EGCG. The viabilities of the all test sets were at least
20% at the highest dose tested (data not shown).
4. Discussion
Green tea contains various useful chemical compounds
such as catechins, caffeine and vitamins, most notable com-
ponents being catechins including EGCG, ECG, ECG and
EC. As the most abundant component, EGCG has been most
extensively studied for various biological activities. Previ-
ous reports have demonstrated that EGCG inhibits influenza
virusinfection when they contact withinfluenzavirusdirectly
(Nakayama et al., 1993), but indirect effect on host cell that
might interfere with the virus-cell membrane fusion was also
suggested (Imanishi et al., 2002).
As an approach to structure-function relationship of anti-
influenza activities, we have tested three isolated catechin
compounds in in vitro culture of influenza viruses in MDCK
cells. Among the test compounds, the EGCG and ECG were
found to be potent inhibitors of influenza virus growth, and
thiseffectwasobservedin all virus subtypes tested,including
A/H1N1, A/H3N2 and B virus that currently afflicts human
population(Fig. 2,Table1). The EC50 values of EGCG, EGC
and ECG for influenza A virus were 22–28, 309–318 and
22–40␮M, respectively, EGCG being most effective. In all
experiments two different controls were included; polyphe-
nols mixture (major constituents EGCG, ECG, EGC) and
total green tea extracts. Considering the molecular weights
of each catechins and the composition of polyphenols mix-
ture, 1␮g/ml concentration of polyphenols mixture would
be equivalent to 3␮M of EGC or 2 ␮M of EGCG and ECG.
Considering this, the inhibitory activity in both plaque for-
mation and the viral growth of polyphenols mixture would
roughly correlate to that of EGCG and ECG.
Strong inhibitory effect (P<0.001) was observed regard-
less of the contact time of the compounds throughout the
growth. This raised possibility that the antiviral effect is
exerted not only on the initially infecting viruses but other
steps of infectious cycle. Although it is difficult to dissect the
effect at each steps involved, we are analyzed the antiviral
effects at various steps in influenza virus infection. Inhibition
of influenza virus adsorption to MDCK cells and chicken red
blood cells by EGCG was previously reported (Nakayama et
al.,1993).Further extendingthe report, here we observed that
EGCG is most effective among three isolated compounds in
hemagglutination inhibition activity. The sensitivity towards
catechin compounds was widely different among three dif-
ferent subtypes of influenza viruses tested. It should also
be mentioned that total tea extract was much more effec-
tive than any other isolated or mixture of polyphenols. This
strongly suggests that, although dietary uptake of tea would
be beneficial for direct intervention of influenza virus infec-
tion, components other than catechins are more responsible
especially for the hemagglutination inhibition. It should also
be noted that, despite differential sensitivity in HA inhibi-
tion, the growth of all subtypes of influenza viruses were
effectively inhibited by EGCG, suggesting that catechins
may affect others steps of infectious cycle as well. This may
involve interference of viral membrane fusion by inhibition
of acidification of endosome (Imanishi et al., 2002). Here, by
quantitative RT-PCR analysis of influenza-specific RNA in
infected cells showed that, at high concentration, ECG and
EGCG also suppressed viral RNA synthesis (Fig. 5). Simi-
larly, EGCG and ECG inhibited the neuraminidase activity at
relatively high concentration more effectively than the EGC
(Fig. 6). The ability to raise a resistant mutant against these
catechinsmayprovidean avenueforidentification of primary
target for catechin compounds.
However, the physiological relevance of the present in
vitro data should be interpreted with caution especially
considering the pharmacokinetic properties of catechins,
such as cellular uptake, chemical modification and distri-
bution (Kroon et al., 2004; Vaidyanathan and Walle, 2003;
Williamson, 2002). For example, the cellular uptake and
efflux of tea polyphenols in human intestinal cell line Caco2
is dependent on the transport mechanism (Vaidyanathan and
Walle, 2003). Moreover, many polyphenolic compounds are
metabolized into various conjugated form—sulfates, sul-
foglucuronide, etc., and the degree of conjugation is widely
different among catechin compounds (Kroon et al., 2004)
and the accumulation of metabolized compounds is widely
different among various tissues (Williamson, 2002).
With potential exception of different subtypes of
influenza virus, the observed inhibitory effects of vari-
ous catechin derivatives were consistent, in the order of
EGCG>ECG > EGC. The resultsshowthat thegalloylgroup
atthe 3-hydroxyl ofcatechin skeleton plays an important role
on the observed antiviral activity, whereas the 5-OH of the

74 J.-M. Song et al. / Antiviral Research 68 (2005) 66–74
trihydroxybenzylmoiety at 2-position plays a minor role (see
Fig. 1). The result is consistent and further extends previous
findingthat galloyl group is important for the anti HIV reverse
transcriptase activity (Chang et al., 1994; Yamaguchi et al.,
2002).
The results also suggest that, besides the known inhibitory
activity on viral attachment of host cells, the antiviral activ-
ities of polyphenols are associated with various steps in the
influenza virus life cycle. The differential antiviral effect
among different catechins, which was consistently observed
at various steps of influenza virus infection cycle albeit at
different concentrations, strongly suggests that primary tar-
get for catechins is likely to be membrane. The effect would
be altering physical integrity of virus particles or host mem-
brane. The interpretation is in good agreement with the
inhibitory effect of catechins on acidification of intracellu-
lar endosome compartments required for fusion of viral and
cellular membranes (Imanishi et al., 2002). In addition, as
suggestedby the subtype-specificeffecton hemagglutination
inhibition (Fig. 4), catechin is expected to affect the confor-
mation of HA or its interaction with influenza virus. The HA
type specific effect of catechins, hitherto unknown before,
remains to be explored. The structure-activity relationships
of anti-influenza properties merit further investigation and
could be further extended by chemical modification of cate-
chin derivatives.
Acknowledgements
This work was supported, in part, by the Strategic
Research Fund for Emerging/Re-emerging Viruses from the
Korean National Institute of Health (KNIH), the Chemical
and Biological Terrorism Research Fund from the Ministry
of Commerce, Industry and Energy (MOCIE) and Nano-
biotechnology Research Initiatives from the Ministry of Sci-
ence and Technology (MOST) of the Korean Government.
References
Balentine, D.A., Wiseman, S.A., Bouwens, L.C., 1997. The chemistry of
tea flavonoids. Crit. Rev. Food Sci. Nutr. 37, 693–704.
Chang, C.W., Hsu, F.L., Lin, J.Y., 1994. Inhibitory effects of polyphenolic
catechins from chinese green tea on HIV reverse transcriptase activity.
J. Biomed. Sci. 1, 163–166.
Chang, L.K., Wei, T.T., Chiu, Y.F., Tung, C.P., Chuang, J.Y., Hung, S.K.,
Li, C., Liu, S.T., 2003. Inhibition of Epstein–Barr virus lytic cycle by
(−)-epigallocatechin gallate. Biochem. Biophys. Res. Commun. 301,
1062–1068.
Fassina, G., Buffa, A., Benelli, R., Varnier, O.E., Noonan, D.M.,
Albini, A., 2002. Polyphenolic antioxidant (−)-epigallocatechin-3-
gallate from green tea as a candidate anti-HIV agent. AIDS 16,
939–941.
Gensler, H.L., Timmermann, B.N., Valcic, S., Wachter, G.A., Dorr, R.,
Dvorakova, K., Alberts, D.S., 1996. Prevention of photocarcinogenesis
by topical administration of pure epigallocatechin gallate isolated from
green tea. Nutr. Cancer 26, 325–335.
Imanishi, N., Tuji, Y., Katada, Y., Maruhashi, M., Konosu, S., Mantani,
N., Terasawa, K., Ochiai, H., 2002. Additional inhibitory effect of tea
extract on the growth of influenza A and B viruses in MDCK cells.
Microbiol. Immunol. 46, 491–494.
Jankun, J., Selman, S.H., Swiercz, R., Skrzypczak-Jankun, E., 1997. Why
drinking green tea could prevent cancer. Nature 387, 561.
Katiyar, S.K., Mukhtar, H., 1996. Tea in chemoprevention of cancer:
epidemiologic and experimental stuides. Int. J. Oncol. 8, 221–238.
Kawai, K., Tsuno, N.H., Kitayama, J., Okaji, Y., Yazawa, K., Asakage,
M., Hori, N., Watanabe, T., Takahashi, K., Nagawa, H., 2003. Epi-
gallocatechin gallate, the main component of tea polyphenol, binds
to CD4 and interferes with gp120 binding. J. Allergy Clin. Immunol.
112, 951–957.
Kroon, P.A., Clifford, M.N., Crozier, A., Day, A.J., Donovan, J.L., Man-
ach, C., Williamson, G., 2004. How should we assess the effects of
exposure to dietary polyphenols in vitro? Am. J. Clin. Nutr. 80, 15–21.
Lin, Y.L., Lin, J.K., 1997. (−)-Epigallocatechin-3-gallate blocks
the induction of nitric oxide synthase by down-regulating
lipopolysaccharide-induced activity of transcription factor nuclear
factor-kB. Mol. Pharmacol. 52, 465–472.
Mimoto, J., Kiura, K., Matsuo, K., Yoshino, T., Takata, I., Ueoka, H.,
Kataoka, M., Harada, M., 2000. (−)-Epigallocatechin gallate can pre-
vent cisplatin induced lung tumorigenesis in A/J mice. Carcinogenesis
21, 915–919.
Nakane, H., Ono, K., 1990. Differential inhibitory effects of some cate-
chin derivatives on the activities of human immunodeficiency virus
reverse transcriptase and cellular deoxyribonucleic and ribonucleic
acid polymerases and various DNA and RNA polymerases by some
catechin derivatives. Biochemistry 29, 2841–2845.
Nakayama, M., Suzuki, K., Toda, M., Okubo, S., Hara, Y., Shimamura, T.,
1993. Inhibition of the infectivity of influenza virus by tea polyphe-
nols. Antivir. Res. 21, 289–299.
Pianetti, S., Guo, S., Kavanagh, K.T., Sonenshein, G.E., 2002. Green
tea polyphenol epigallocatechin-3 gallate inhibits Her-2/neu signaling,
proliferation, and transformed phenotype of breast cancer cells. Cancer
Res. 62, 652–655.
Tachibana, H., Koga, K., Fujimura, Y., Yamada, K., 2004. A receptor for
green tea polyphenol EGCG. Nat. Struct. Mol. Biol. 11, 380–381.
Toda, M., Okubo, S., Ikigai, H., Suzuki, T., Suzuki, Y., Hara, Y., Shi-
mamura, T., 1992. The protective activity of tea catechins against
experimental infection by Vibrio cholerae O1. Microbiol. Immunol.
36, 999–1001.
Vaidyanathan, J.B., Walle, T., 2003. Cellular uptake and efflux of the tea
flavonoid (−)epicatechin-3-gallate in the human intestinal cell line
Caco-2. J. Pharmacol. Exp. Ther. 307, 745–752.
Valcic, S., Muders, A., Jacobsen, N.E., Liebler, D.C., Timmermann, B.N.,
1999. Antioxidant chemistry of green tea catechins. Identification of
products of the reaction of (−)-epigallocatechin gallate with peroxyl
radicals. Chem. Res. Toxicol. 12, 382–386.
Weber, J.M., Ruzindana-Umunyana, A., Imbeault, L., Sircar, S., 2003.
Inhibition of adenovirus infection and adenain by green tea catechins.
Antivir. Res. 58, 167–173.
Williamson, G., 2002. The use of flavonoid aglycones in in vitro systems
to test biological activities: based on bioavailability data, is this a
valid approach? Phytochem. Rev. 1, 215–222.
Yamaguchi, K., Honda, M., Ikigai, H., Hara, Y., Shimamura, T., 2002.
Inhibitory effects of (−)-epigallocatechin gallate on the life cycle
of human immunodeficiency virus type 1 (HIV-1). Antivir. Res. 53,
19–34.
Yamane, T., Takahashi, T., Kuwata, K., Oya, K., Inagake, M., Kitao,
Y., Suganuma, M., Fujiki, H., 1995. Inhibition of N-methyl-N-nitro-
N-nitrosoguanidine-induced carcinogenesis by (−)-epigallocatechin
gallate in the rat glandular stomach. Cancer Res. 55, 2081–
2084.
Yang, C.S., Maliakal, P., Meng, X., 2002. Inhibition of carcinogenesis by
tea. Annu. Rev. Pharmacol. Toxicol. 42, 25–54.