ARTHRITIS & RHEUMATISM

Vol. 54, No. 8, August 2006, pp 2393–2401

DOI 10.1002/art.22023
© 2006, American College of Rheumatology

Regulation of Interleukin-1␤–Induced Chemokine Production

and Matrix Metalloproteinase 2 Activation by
Epigallocatechin-3-Gallate in
Rheumatoid Arthritis Synovial Fibroblasts

Salahuddin Ahmed,1 Angela Pakozdi,1 and Alisa E. Koch2

Objective. To evaluate the efficacy of concentration-dependent manner. EGCG at 50 ␮M

epigallocatechin-3-gallate (EGCG), a potent antiinflam- caused a complete block of IL-1␤–induced production of
matory molecule, in regulating interleukin-1␤ (IL-1␤)– RANTES, ENA-78, and GRO␣, and reduced production
induced production of the chemokines RANTES of MCP-1 by 48% (P < 0.05). Zymography showed that
(CCL5), monocyte chemoattractant protein 1 (MCP-1/ EGCG blocked constitutive, IL-1 ␤ –induced, and
CCL2), epithelial neutrophil–activating peptide 78 chemokine-mediated MMP-2 activity. Evaluation of sig-
(ENA-78/CXCL5), growth-regulated oncogene ␣ naling events revealed that EGCG preferentially
(GRO␣/CXCL1), and matrix metalloproteinase 2 blocked the phosphorylation of PKC␦ and inhibited the
(MMP-2) activity in rheumatoid arthritis (RA) synovial activation and nuclear translocation of NF-␬B in IL-
fibroblasts. 1␤–treated RA synovial fibroblasts.
Methods. Fibroblasts obtained from RA synovium Conclusion. These results suggest that EGCG
were grown, and conditioned medium was obtained. Cell may be of potential therapeutic value in inhibiting joint
viability was determined by MTT assay. RANTES, destruction in RA.
MCP-1, ENA-78, and GRO␣ produced in culture super-
natants were measured by enzyme-linked immunosor- Rheumatoid arthritis (RA) is a chronic inflam-
bent assay. MMP-2 activity was analyzed by gelatin matory disease characterized by robust infiltration of
zymography. Western blotting was used to study the leukocytes into the synovium, resulting in hyperplasia of
phosphorylation of protein kinase C (PKC) isoforms the synovial lining, progressive cartilage destruction, and
and nuclear translocation of NF-␬B. finally erosion of the underlying bone (1). Synovial
Results. EGCG was nontoxic to RA synovial fibro- fibroblasts mediate joint destruction in RA by producing
blasts. Treatment with EGCG at 10 ␮M or 20 ␮M chemokines that facilitate the expansion and invasion of
significantly inhibited IL-1 ␤ –induced ENA-78, synovial fibroblasts into the adjacent tissue, and the
RANTES, and GRO␣, but not MCP-1 production in a regulation of these events has been a primary target of
therapeutic intervention in RA (2). In response to the
Dr. Koch’s work was supported by the NIH (grants AI-40987 proinflammatory cytokines produced by macrophages,
and AR-48267), the Frederick G. L. Huetwell and William D. Robin-
son, MD, Professorship in Rheumatology, and the Office of Research such as interleukin-1␤ (IL-1␤) and tumor necrosis factor
and Development, Medical Research Service, Department of Veterans ␣ (TNF␣), RA synovial fibroblasts produce chemokines
Affairs. that promote inflammation, neovascularization, and car-
1
Salahuddin Ahmed, PhD, Angela Pakozdi, MD: University
of Michigan Medical School, Ann Arbor; 2Alisa E. Koch, MD: VA tilage degradation via activation of matrix-degrading en-
Medical Center and University of Michigan Medical School, Ann zymes, such as matrix metalloproteinases (MMPs) (3).
Arbor, Michigan. Chemokines are a specialized family of small
Address correspondence and reprint requests to Salahuddin
Ahmed, PhD, Department of Internal Medicine/Division of Rheuma- (8–10-kd), structurally related proteins classified into 4
tology, University of Michigan Medical School, BSRB Room 4388, 109 supergene families, C, CC, CXC, and CX3C, based on
Zina Pitcher Drive, Ann Arbor, MI 48109-2200. E-mail: salahmed@ the location of cysteine residues (4). The CC and CXC
umich.edu.
Submitted for publication December 23, 2005; accepted in chemokines are well-established regulators of gene tran-
revised form May 8, 2006. scription, cell proliferation, and leukocyte trafficking to
2393
2394 AHMED ET AL

normal and inflamed tissues (5). Chemokines such as polyclonal anti–phospho-PKC␦, anti–phospho-PKC␣/␤II,
epithelial neutrophil–activating peptide 78 (ENA-78/ anti–phospho-PKC␧, and anti-rabbit horseradish peroxidase–
linked secondary antibodies were purchased from Cell Signal-
CXCL5), RANTES (CCL5), monocyte chemoattractant
ing Technology (Beverly, MA). Rabbit anti–␤-actin and anti–
protein 1 (MCP-1/CCL2), and growth-regulated onco- NF-␬B p65 were purchased from Sigma-Aldrich and Santa
gene ␣ (GRO␣/CXCL1) are potent chemotactic agents Cruz Biotechnology (Santa Cruz, CA), respectively. All signal-
that have been shown to be constitutively produced by ing inhibitors used in this study were purchased from Calbio-
RA synovial fibroblasts and up-regulated upon stimula- chem (La Jolla, CA). Enzyme-linked immunosorbent assay kits
for MCP-1, ENA-78, GRO␣, and RANTES were purchased
tion with IL-1␤ (6). These chemokines play a major role from R&D Systems.
in inducing MMP activity and expression in RA synovial Culture of human RA synovial fibroblasts. Fibroblasts
fibroblasts (7). IL-1␤– or TNF␣-mediated up-regulation were isolated from synovium obtained from RA patients who
of chemokines was found to be suppressed by the had undergone total joint replacement surgery or synovec-
tomy, according to a protocol approved by the institutional
neutralization of IL-1␤, but not TNF␣, suggesting a
review board of the University of Michigan Medical School.
predominant role of IL-1␤ (7). Fresh synovial tissues were minced and digested in a solution
Green tea (Camellia sinensis) is one of the most of Dispase, collagenase, and DNase. Cells were used at passage
commonly consumed beverages in the world, with no 5 or older, at which time they were a homogeneous population.
significant side effects (8). Extensive studies in several RA synovial fibroblasts were grown in RPMI 1640 medium
supplemented with 10% fetal bovine serum (FBS) at 37°C in a
animal models in the past 2 decades have verified the humidified atmosphere with 5% CO2. Upon confluence, cells
antioxidant, antiinflammatory, and antioncogenic prop- were passaged by brief trypsinization, as previously described
erties of a polyphenolic mixture derived from green tea (15). All treatments were performed in serum-free medium.
(9). A majority of pharmacologic properties of green tea Preparation of EGCG solution. A stock solution of
EGCG (10 mM) was prepared in water, sterile filtered with
are mimicked by its active constituent, epigallocatechin- 0.2-␮m syringe filters, and stored at ⫺20°C in aliquots. Fresh
3-gallate (EGCG) (10). In our earlier studies using EGCG solution was used in each experiment and directly
human chondrocytes derived from osteoarthritic carti- added to the culture medium.
lage, we showed EGCG to be an effective inhibitor of Treatment of RA synovial fibroblasts with IL-1␤ and
the production of catabolic mediators, such as nitric EGCG. To study the effect of EGCG on cell viability, RA
synovial fibroblasts (2 ⫻ 104/well) were plated in 96-well,
oxide (NO) and prostaglandin E2 (PGE2), by transcrip- flat-bottomed tissue culture plates (Corning, Corning, NY)
tional and translational regulation (11–13). Recently, we and cultured in RPMI 1640 plus 10% FBS for 6 hours. This was
showed that EGCG significantly inhibited the expression then replaced with fresh medium and culture continued for 24
and activities of MMP-1 and MMP-13 in human chon- hours. EGCG (10–60 ␮M) was added to RA synovial fibro-
drocytes at a physiologically achievable dose (14). None- blasts in serum-free medium and the culture was incubated at
37°C for another 24 hours. Two hours prior to termination, 20
theless, significant gaps remain in our understanding of ␮l of MTT dye (5 mg/ml in sterile phosphate buffered saline
the mechanism of action of EGCG in RA. [PBS]; Invitrogen, Carlsbad, CA) was added to each well and
Hence, the present study was undertaken to study further incubated at 37°C. At the end of incubation, cells were
the effect of EGCG on IL-1␤–induced production of washed twice with PBS, solubilized in 100 ␮l of DMSO at 37°C
chemokines and MMP-2 activation in RA synovial for 5 minutes, and read at an optical density of 570 nm.
To study the effect of EGCG on chemokine produc-
fibroblasts. EGCG down-regulated IL-1␤–induced tion, RA synovial fibroblasts were incubated with or without
RANTES, MCP-1, ENA-78, and GRO␣ production and EGCG (10–50 ␮M) in serum-free medium for 12 hours,
MMP-2 activation in human RA synovial fibroblasts, followed by stimulation with IL-1␤ (10 ng/ml) for 24 hours.
and it was also effective in blocking chemokine-induced After 24 hours, culture supernatant was collected and centri-
MMP-2 activity in RA synovial fibroblasts. We provide fuged at 10,000g for 5 minutes at 4°C to remove particulate
matter, and stored at ⫺80°C in fresh tubes. Using commer-
evidence that EGCG specifically inhibits IL-1␤–induced cially available kits (R&D Systems), culture supernatants were
protein kinase C␦ (PKC␦) phosphorylation and sup- used to determine the quantities of RANTES, MCP-1, ENA-
presses chemokine production and MMP-2 activation. 78, and GRO␣. To study the signaling mechanism of chemo-
Additionally, we show that the protective effects of kine production by IL-1␤, RA synovial fibroblasts were incu-
EGCG are also mediated via NF-␬B inhibition. bated with PKC inhibitors (Rottlerin and Gö 6976; 10 ␮M) and
NF-␬B inhibitor (pyrrolidine dithiocarbamate [PDTC]; 200
␮M) for 2 hours, followed by stimulation with IL-1␤ for 24
MATERIALS AND METHODS hours, and processed as above.
Western immunoblotting and analysis. To study the
Antibodies and reagents. EGCG was purchased from effect of EGCG on signaling events, RA synovial fibroblasts
Sigma-Aldrich (St. Louis, MO). Recombinant human IL-1␤ were incubated with or without EGCG (1–20 ␮M) in serum-
was purchased from R&D Systems (Minneapolis, MN). Rabbit free RPMI 1640 for 12 hours, followed by stimulation with
ROLE OF EGCG IN RA CHEMOKINE PRODUCTION AND MMP-2 ACTIVATION 2395

IL-1␤ for 20 minutes. Cells were lysed in cell lysis buffer

containing 100 mM Tris (pH 7.4), 100 mM NaCl, 1 mM EDTA,
1 mM EGTA, 1 mM NaF, 20 mM NaP2O4, 2 mM Na3VO4, 1%
Triton X-100, 10% glycerol, 0.1% sodium dodecyl sulfate
(SDS), 0.5% deoxycholate, 1 mM phenylmethylsulfonyl fluo-
ride [PMSF], and protease inhibitors (1 tablet/10 ml; Roche,
Indianapolis, IN). Protein was measured using a BCA protein
assay kit (Pierce, Rockford, IL). Equal amounts of protein (20
␮g) were loaded and separated by SDS–polyacrylamide gel
electrophoresis and transferred onto nitrocellulose mem-
branes (Bio-Rad, Richmond, CA). Blots were probed using
rabbit polyclonal antibodies specific for phosphorylated
PKC␣/␤II, PKC␦, and PKC␧ isoforms, and rabbit polyclonal
anti–␤-actin. The immunoreactive protein bands were visual-
ized by enhanced chemiluminescence. Densitometric analysis
of the bands was performed using UN-SCAN-IT software,
version 5.1 (Silk Scientific, Orem, UT), and the data were Figure 1. Inhibition of interleukin-1␤ (IL-1␤)–induced production of
A, RANTES (CCL5), B, monocyte chemoattractant protein 1 (MCP-
analyzed using Prism software (GraphPad Software, San Di-
1/CCL2), C, epithelial neutrophil–activating peptide 78 (ENA-78/
ego, CA).
CXCL5), and D, growth-regulated oncogene ␣ (GRO␣/CXCL1) by
Gelatin zymography. MMP-2 activity in conditioned
epigallocatechin-3-gallate (EGCG) in rheumatoid arthritis (RA) syno-
medium was measured as previously described (16). Briefly, 15 vial fibroblasts. RA synovial fibroblasts (2 ⫻ 104/well) were incubated
␮l of conditioned medium was added to 15 ␮l of 2⫻ nonre- with EGCG (10–50 ␮M) for 12 hours, followed by stimulation with
ducing sample buffer, resolved under nonreducing conditions IL-1␤ (10 ng/ml) for 24 hours. Production of RANTES, MCP-1,
on 7.5% SDS–polyacrylamide gels polymerized with 1 mg/ml ENA-78, and GRO␣ in culture supernatants was measured using a
gelatin (type B from bovine skin; Sigma) as a substrate, and commercially available enzyme-linked immunosorbent assay kit. Val-
electrophoresed at 125V. Following electrophoresis, the gels ues are the mean and SEM. ⴱ ⫽ P ⬍ 0.05 versus treatment with IL-1␤
were washed in 2.5% Triton X-100 for 30 minutes with gentle alone.
shaking, followed by a 30-minute wash in developing buffer (50
mM Tris HCl [pH 8.0], 5 mM CaCl2, and 0.02% NaN3). The
gels were incubated overnight in fresh developing buffer at Statistical analysis. Student’s t-tests were performed
37°C, stained in Coomassie brilliant blue R250, and then to calculate statistical differences between the different vari-
destained using a solution of 7% acetic acid and 5% methanol. ables. P values less than 0.05 were considered significant.
Images of the digested regions representing MMP activity were
captured using the Quantity One 1-D image analyzer (Bio-Rad) RESULTS
and analyzed using UN-SCAN-IT software.
Preparation of nuclear extracts. To study the effect of Effect of EGCG on RA synovial fibroblast viabil-
EGCG on IL-1␤–induced NF-␬B activation, RA synovial ity. The results of an MTT-based viability assay using
fibroblasts (2 ⫻ 106/well) were grown to confluence in 10-cm samples obtained from 3 different donors showed that
dishes and treated with EGCG (10 ␮M and 20 ␮M) for 12 EGCG (10–60 ␮M) had no significant effect on the
hours with and without IL-1␤ stimulation for 30 minutes. viability of cultured RA synovial fibroblasts (data not
Cytoplasmic and nuclear fractions were prepared as previously shown). The highest concentration used in the viability
described (17). Upon termination of the reaction, cells were
washed twice with ice-cold PBS (pH 7.4), scraped, collected in assay (60 ␮M) showed a mean ⫾ SEM 10 ⫾ 4.0% loss in
Eppendorf tubes, and centrifuged at 1,500g for 5 minutes at viability, which was not statistically significant (P ⬎ 0.05).
4°C. The pellet obtained was suspended in 400 ␮l buffer A (10 Effect of EGCG on IL-1␤–induced chemokine
mM HEPES buffered saline [pH 7.9], 10 mM KCl, 0.1 mM production. IL-1␤ is a potent inducer of chemokine
EDTA, 0.1 mM EGTA, 1 mM dithiothreitol [DTT], and 0.5 production in RA synovial fibroblasts (18). Excessive
mM phenylmethylsulfonyl fluoride PMSF), gently mixed, and
production of chemokines by RA synovial fibroblasts has
placed on ice for 15 minutes. Twenty-five microliters of cold
10% Nonidet P40 was added to individual samples, and the been shown to induce proliferation of these cells and
samples were vortexed and centrifuged at 14,000g for 30 facilitate invasion into the adjacent tissues (7). In the
seconds. Supernatant (cytoplasmic fraction) was collected, and present study, stimulation of RA synovial fibroblasts
the nuclear pellet obtained was suspended in 50 ␮l buffer C (20 with IL-1␤ (10 ng/ml) for 24 hours resulted in a 50-,
mM HEPES buffered saline [pH 7.9], 0.4M NaCl, 1 mM 3.6-, 5.2-, and 120-fold induction of RANTES, MCP-1,
EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF) and
ENA-78, and GRO␣ production, respectively, by RA
rocked for 45 minutes at 4°C. Samples were centrifuged for 15
minutes at 14,000g at 4°C. The supernatant (nuclear fraction) synovial fibroblasts as compared with untreated controls
was collected and stored at ⫺80°C. Nuclear cell lysate (15 ␮g) (P ⬍ 0.05) (Figures 1A–D). Treatment with EGCG
was used to detect NF-␬B p65 by Western blotting. caused a significant, but differential, inhibition of IL-1␤–
2396 AHMED ET AL

Treatment of RA synovial fibroblasts with EGCG at

concentrations of 10 ␮M and 20 ␮M resulted in a
mean ⫾ SEM 58 ⫾ 14% and 75 ⫾ 16% inhibition of
IL-1␤–induced MMP-2 activity, respectively (P ⬍ 0.05).
Interestingly, the fibroblasts treated with EGCG (10 ␮M
and 20 ␮M) alone showed a marked inhibition of
MMP-2 activity (76 ⫾ 22% and 70 ⫾ 20%, respectively),
when compared with the activity observed in untreated
samples (P ⬍ 0.05).

Figure 2. Inhibition of constitutive and IL-1␤–induced matrix metal-

loproteinase 2 (MMP-2) activity by EGCG in RA synovial fibroblasts.
RA synovial fibroblasts (2 ⫻ 105/well) were incubated with EGCG (10
␮M or 20 ␮M) for 12 hours, followed by no stimulation or stimulation
with IL-1␤ (10 ng/ml) for 24 hours. MMP-2 activity in the cell-free
supernatants from different treatment combinations was estimated by
gelatin zymography. Values are the mean and SEM. ⴱ ⫽ P ⬍ 0.05
versus treatment with IL-1␤ alone; # ⫽ P ⬍ 0.05 versus untreated
control. OD ⫽ optical density (see Figure 1 for other definitions).

induced ENA-78, RANTES, and GRO␣ production, in

a concentration-dependent manner. EGCG at 10 ␮M
blocked IL-1␤–induced ENA-78, RANTES, and GRO␣
production by 64%, 39%, and 80%, respectively. Inter-
estingly, EGCG at 50 ␮M resulted in almost 88%
inhibition of the production of ENA-78, and almost
completely blocked RANTES and GRO␣ production as
compared with samples treated with IL-1␤ alone (P ⬍
0.05). However, EGCG 10 ␮M and 20 ␮M had no
significant inhibitory effect on IL-1␤–induced MCP-1
production, whereas 50 ␮M EGCG caused almost 48%
inhibition of MCP-1 production (P ⬍ 0.05). Figure 3. Inhibition of chemokine (ENA-78, GRO␣, and RANTES)–
Effect of EGCG on IL-1␤–induced MMP-2 activ- induced matrix metalloproteinase 2 (MMP-2) activity by EGCG in RA
ity in RA synovial fibroblasts. MMPs execute a rate- synovial fibroblasts. A, RA synovial fibroblasts (2 ⫻ 105/well) were left
limiting step in synovial fibroblast invasion, inflam- untreated or treated with EGCG (10 ␮M or 20 ␮M) for 12 hours,
mation, and cartilage breakdown under abnormal followed by stimulation with IL-1␤ (10 ng/ml), ENA-78 (100 ng/ml),
GRO␣ (100 ng/ml), or RANTES (100 ng/ml) for 24 hours. MMP-2
conditions such as RA (19). The effect of EGCG on activity in the cell-free supernatants was determined by zymography
IL-1␤–induced MMP-2 activity was determined using using gelatin (1 mg/ml) as a substrate, after activation with APMA (10
zymography, and the densitometric values of the bands ␮M) for 2 hours to separate proMMP-2 and active MMP-2 bands.
were obtained for statistical analysis. RA synovial fibro- Values are the mean and SEM. B, RA synovial fibroblasts (2 ⫻
blasts were found to have a detectable basal level of 105/well) were left untreated or treated with pertussis toxin (0.75
␮g/ml) for 2 hours, followed by stimulation with ENA-78, GRO␣, or
MMP-2 activity with no stimulation. Stimulation of RA RANTES (100 ng/ml each) for 24 hours. Blots in A and B are
synovial fibroblasts with IL-1␤ resulted in a 2.5-fold representative of 3 independent samples. NS ⫽ not stimulated; C ⫽
increase in MMP-2 activity (P ⬍ 0.05) (Figure 2). chemokines (see Figure 1 for other definitions).
ROLE OF EGCG IN RA CHEMOKINE PRODUCTION AND MMP-2 ACTIVATION 2397

Effect of EGCG on chemokine-induced MMP-2

activity in RA synovial fibroblasts. Chemokines facili-
tate a unique mechanism of tissue invasion in the
diseased joint by enhancing RA synovial fibroblast pro-
liferation and MMP up-regulation (7). Regulation of
MMP-2 activity is of therapeutic value since it plays a
predominant role in processes such as angiogenesis,
inflammation, and tissue invasion (19). In the present
study, we found that ENA-78, GRO␣, and RANTES
increased RA synovial fibroblast MMP-2 activity by 3.9-,
5.4-, and 4.9-fold, respectively, as compared with un-
treated controls (P ⬍ 0.05) (Figure 3A). MMP-2 activity
induced by ENA-78, GRO␣, and RANTES was 1.1-,
1.5-, and 1.3-fold higher, respectively, than that observed
in IL-1␤–stimulated samples. Each sample was activated
with 10 ␮M APMA for 2 hours to separate proMMP-2
and active MMP-2 by gelatin zymography. Treatment of
RA synovial fibroblasts with EGCG markedly inhibited
the ability of chemokines to stimulate MMP-2 activity, in
a concentration-dependent manner (P ⬍ 0.05). EGCG
at 10 ␮M and 20 ␮M blocked MMP-2 activity induced by
ENA-78 by 33% and 75%, by GRO␣ by 52% and 80%,
and by RANTES by 55% and 84%, respectively (P ⬍
0.05). In another set of experiments, RA synovial fibro-
blasts treated with pertussis toxin (0.75 ␮g/ml) showed a
significant decrease in MMP-2 activity (Figure 3B),
which suggests that the chemokines have a direct effect
on MMP-2 activation.
Role of PKC␦ and NF-␬B in IL-1␤–induced
chemokine production and MMP-2 activation. To study
Figure 4. Involvement of protein kinase C (PKC) and NF-␬B in the
the signaling pathways involved in IL-1␤–induced pro- regulation of IL-1␤–induced chemokine production and matrix met-
duction of chemokines and MMP-2 activation, RA syno- alloproteinase 2 (MMP-2) activation. RA synovial fibroblasts (2 ⫻
vial fibroblasts were pretreated with 10 ␮M Rottlerin 105/well) were pretreated with inhibitors of PKC␦ (Rottlerin; 10 ␮M),
(PKC␦ inhibitor), 10 ␮M Gö 6976 (PKC␣ inhibitor), or PKC␣ (Gö 6976; 10 ␮M), or NF-␬B (pyrrolidine dithiocarbamate
200 ␮M PDTC (NF-␬B inhibitor) for 2 hours, followed [PDTC]; 200 ␮M) for 2 hours, followed by stimulation with IL-1␤ (10
ng/ml) for 24 hours. A, Production of chemokines ENA-78, GRO␣,
by IL-1␤ stimulation for 24 hours. The culture super- and RANTES in culture supernatants. B, MMP-2 activity. Values are
natants were used to estimate ENA-78, GRO␣, and the mean and SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.001, versus treatment
RANTES production and MMP-2 activation. The PKC␦ with IL-1␤ alone. OD ⫽ optical density (see Figure 1 for other
inhibitor almost completely blocked IL-1␤–induced definitions).
ENA-78, GRO␣, and RANTES production and MMP-2
activation (P ⬍ 0.001) (Figures 4A and B). In contrast,
PDTC treatment caused almost complete inhibition of that IL-1␤ preferentially increased PKC␦ phosphoryla-
MMP-2 activity, but induced only partial inhibition of tion to almost 3-fold that in untreated controls, without
the chemokines studied. The PKC␣ inhibitor moder- causing any marked activation of other PKC isoforms
ately decreased ENA-78 and RANTES production and (Figure 5A). Treatment of RA synovial fibroblasts with
MMP-2 activity, with little effect on GRO␣ production. EGCG (1–20 ␮M) resulted in concentration-dependent
Effects of EGCG on chemokine production and inhibition of the phosphorylation of PKC␦. In the
MMP-2 activation. In light of these observations, we present study, even the lowest concentration of EGCG
evaluated the effect of EGCG treatment on IL-1␤– (1 ␮M) suppressed phosphorylated PKC␦ levels by 35%,
induced PKC isoform phosphorylation and NF-␬B acti- whereas 20 ␮M of EGCG completely blocked the phos-
vation. Densitometric analysis of protein bands showed phorylation of PKC␦ (Figure 5A). Interestingly, EGCG
2398 AHMED ET AL

was able to reduce the levels of phosphorylated PKC␧ by

up to 30% at a 20 ␮M concentration, but did not inhibit
the phosphorylation of PKC␣/␤II in IL-1␤–treated RA
synovial fibroblasts.

Figure 6. Schematic representation of the mechanism of EGCG in

inhibiting IL-1␤–induced chemokine production and matrix metallo-
proteinase 2 (MMP-2) activation in RA synovial fibroblasts. Findings
of the present study suggest the efficacy of EGCG in regulating the
stimulating effects of IL-1␤ on RA synovial fibroblast chemokine
production and MMP-2 activation via specific inhibition of protein
kinase C␦ (PKC␦) phosphorylation and NF-␬B activation and nuclear
translocation. See Figure 1 for other definitions.

The results of Western blotting in the nuclear

lysates showed that IL-1␤ increased activation and nu-
clear translocation of NF-␬B in RA synovial fibroblasts
by 3.5-fold, when compared with unstimulated controls
(P ⬍ 0.05) (Figure 5B). Treatment with 10 ␮M and 20
␮M EGCG resulted in the inhibition of IL-1␤–induced
nuclear translocation of NF-␬B, by 51% and 58%,
respectively (P ⬍ 0.05). These results provide direct
evidence of the signaling mechanism by which EGCG
produces these regulating effects in RA synovial fibro-
Figure 5. Preferential inhibition of IL-1␤–induced protein kinase C␦ blasts (Figure 6).
(PKC␦) phosphorylation and NF-␬B activation by EGCG in RA synovial
fibroblasts. A, RA synovial fibroblasts (2 ⫻ 105/well) were treated with DISCUSSION
EGCG (1–20 ␮M) for 12 hours, followed by stimulation with IL-1␤ (10 This is the first study, to our knowledge, to
ng/ml) for 20 minutes. Cells were lysed in extraction buffer containing provide evidence of the ability of EGCG, a potent
protease inhibitors, and the phosphorylation (p) of PKC isoforms in 20
␮g of each sample was determined by Western blotting. A represen- antiinflammatory molecule derived from green tea, to
tative blot is shown. Equal loading of protein was verified by reprobing regulate IL-1␤–induced chemokine production and
blots for ␤-actin. B, RA synovial fibroblasts (2 ⫻ 106/well) were MMP-2 activation in cultured RA synovial fibroblasts. In
pretreated with EGCG (10 ␮M or 20 ␮M) for 12 hours, followed by particular, this study assigns EGCG a novel property of
stimulation with IL-1␤ (10 ng/ml) for 30 minutes. The nuclear fraction chemokine suppressor, and identifies these inhibitory
was prepared as described in Materials and Methods. Nuclear protein
(Nuc; 15 ␮g) was used to analyze nuclear NF-␬B p65. Values are the effects as being mediated via specific inhibition of the
mean and SEM. ⴱ ⫽ P ⬍ 0.05 versus treatment with IL-1␤. OD ⫽ PKC␦ signaling pathway. In addition, EGCG inhibited
optical density (see Figure 1 for other definitions). the activation and nuclear translocation of NF-␬B to
ROLE OF EGCG IN RA CHEMOKINE PRODUCTION AND MMP-2 ACTIVATION 2399

elicit its response. These findings indicate that EGCG, MMP-2 is involved in RA pathogenesis by assist-
when present in concentrations that are physiologically ing RA synovial fibroblasts in the invasion of microvas-
achievable, is able to suppress the effects of IL-1␤ by cular basement membrane and the interstitium (37,38).
blocking chemokine production, and hence may play a Chemokine-activated RA synovial fibroblasts may me-
role in regulating tissue invasion by RA synovial fibro- diate this process by their attachment to the cartilage
blasts. Other regulatory mechanisms, e.g., transcrip- surface and the release of MMPs (7). Interestingly,
tional and posttranscriptional control of messenger recent findings suggest an active involvement of selective
RNA levels of chemokines and chemokine receptors by chemokines in the destructive phase of RA, in which
EGCG, may also be important, and are currently under migration, proliferation, and MMP production by RA
study. synovial fibroblasts are characteristic features (7).
Despite the identification of potent chemokine Therefore, possible therapeutic strategies include im-
and chemokine receptor antagonists in vitro, their struc- peding the production of MMPs, blocking the active
tural diversity, species-based potency differences, and sites of MMPs, increasing production of endogenous
suboptimal pharmacokinetic and toxicity profiles sug- tissue inhibitors of MMPs, and preventing the activation
gest that novel RA therapies targeting multiple chemo- of MMPs (39).
kines and their receptors may prove to be more benefi- Recently, we showed that treatment of cultured
cial treatment options (20). In the present study, EGCG human OA chondrocytes with EGCG significantly inhib-
blocked IL-1␤–induced ENA-78, GRO␣, and RANTES ited the expression and activities of collagenases
production by RA synovial fibroblasts in a (MMP-1 and MMP-13) (14). Also, others reported that
concentration-dependent manner, with no effect on the catechins from green tea inhibited the degradation of
viability of these cells. ENA-78 and GRO␣ are members human cartilage proteoglycan and type II collagen, and
of the CXC chemokine subgroup that possess potent selectively inhibited the aggrecanases ADAMTS-1, -4,
chemoattractant and neutrophil activator functions (6, and -5 (40,41). The evidence from the present study that
21,22). RANTES is a member of the CC subfamily and EGCG blocks constitutive and cytokine-induced MMP-2
has been shown to chemoattract lymphocytes and mono- activity is important. Previous studies have shown that
cytes (23). These chemokines are constitutively pro- the lack of efficacy of direct MMP inhibitors (trocade
duced by RA synovial fibroblasts and are up-regulated and BAY12-9566) was partially attributed to their im-
upon stimulation with IL-1␤ (7,18). Interestingly, in paired bioavailability and inability to control angiogen-
animal models of arthritis, the expression of these esis (42,43), which is also involved in the early and
chemokines generally preceded the onset of clinical destructive phase of RA. The pharmacokinetics of
symptoms (24–28). EGCG in human volunteers taking a single dosage of
Recently, EGCG was shown to inhibit TNF␣- 1,600 mg/day showed a rapid absorption, with a maxi-
induced IL-8 and macrophage inflammatory protein 3␣ mum plasma concentration value of 3,392 ng/ml; the
(MIP-3␣) production in human colon epithelial cells time to reach maximum plasma concentration was 2.2
(29). In another study, EGCG was found to block hours, and the terminal elimination half-life ranged
lipopolysaccharide-induced neutrophil chemotaxis of rat between 1.9 and 4.6 hours (44). Interestingly, 10-day
macrophages (30). In human keratinocytes, preincuba- repeated administration of oral doses of EGCG of up to
tion with EGCG has been shown to act in a synergistic 800 mg per day was found to be safe and very well
manner with genistein, a dietary polyphenol, in inhibit- tolerated (45).
ing TNF␣-induced vascular growth endothelial factor Traditionally, strategies for the development of
and IL-8 production (31). EGCG-mediated inhibition of antiinflammatory treatment have been broadly divided
chemokines in the present study is consistent with the into 2 approaches: those acting outside the cell and
findings of previous studies with animal models, in which those acting inside the cell. The first approach targets
treatment with neutralization-specific antibodies, poly- receptors and cytokines, using agents such as antibodies
clonal antibodies, or receptor antagonists against ENA- or cytokine traps (e.g., anti-TNF␣ or soluble IL-1 recep-
78, GRO␣, and RANTES has led to significant improve- tor antagonist therapies) to block ligand–cell surface
ment in clinical symptoms of arthritis (27,32,33). receptor interactions. The second approach uses inhibi-
Furthermore, the regulation of multiple chemokines by tors such as cyclosporine, cyclooxygenase 2 (COX-2)
EGCG in RA synovial fibroblasts is consistent with the inhibitors, and steroids that easily enter the cell. Such
finding of synergistic effects with the use of antibodies intracellular interference with signaling pathways has
directed against multiple chemokines (34–36). generated very effective and wide-ranging antiinflamma-
2400 AHMED ET AL

tory therapies, with the caveat that they also work as duction of MMP-1 and MMP-13 in chondrocytes iso-
general immunosuppressants (20). The chemokine net- lated from arthritic joints (13,14).
work lends itself to intervention using both approaches. In conclusion, the data presented here demon-
In autoimmune diseases such as RA, it is partly strate that, in RA synovial fibroblasts, EGCG inhibits
the inappropriate leukocyte recruitment accompanied IL-1␤–stimulated increases in chemokine production
by cellular activation that results in disease symptoms and MMP-2 activation by inhibition of the PKC␦ path-
and progression (5). This is due largely to abnormal way and NF-␬B translocation to the nucleus (Figure 6).
expression of chemokines and cytokines (2). In the Activation of PKC␦ plays a critical role in the regulation
present study, EGCG specifically blocked IL-1␤– of these downstream effectors, and the regulation of
induced phosphorylation of the PKC␦ isoform, with PKC␦ by EGCG provides evidence of the potential of
little or no effect on other PKC isoforms. Some of the EGCG to block chemokine production and invasion in
effects of EGCG have been shown to be mediated via RA synovial fibroblasts.
PKC␣ isoform inhibition. EGCG was shown to inhibit
amyloid ␤ toxicity in neuronal cells by blocking PKC␣ REFERENCES
(46). It also modulated the gene expression profile of
pro-oncogenes in prostate carcinoma LNCaP cells by 1. Pope RM. Apoptosis as a therapeutic tool in rheumatoid arthritis
[review]. Nat Rev Immunol 2002;2:527–35.
completely blocking PKC␣, without affecting PKC␤, 2. Koch AE. Chemokines and their receptors in rheumatoid arthritis:
PKC␣, PKC␨, PKC␧, and PKC␦ isoforms (47). A recent future targets? [review]. Arthritis Rheum 2005;52:710–21.
study by Lee et al showed that EGCG inhibited NO- 3. Mor A, Abramson SB, Pillinger MH. The fibroblast-like synovial
cell in rheumatoid arthritis: a key player in inflammation and joint
induced apoptosis by regulation of p53 gene via PKC␦ in destruction [review]. Clin Immunol 2005;115:118–28.
dopaminergic neurons (48). In the present study, using 4. Bacon K, Baggiolini M, Broxmeyer H, Horuk R, Lindley I,
PKC␣ and PKC␦ inhibitors, we demonstrated that acti- Mantovani A, et al. Chemokine/chemokine receptor nomencla-
ture. J Interferon Cytokine Res 2002;22:1067–8.
vation of PKC␦ plays a pivotal role in the regulation of 5. Szekanecz Z, Koch AE. Therapeutic inhibition of leukocyte
IL-1␤–stimulated chemokine production and MMP-2 recruitment in inflammatory diseases [review]. Curr Opin Phar-
activity in RA synovial fibroblasts. The inhibition of macol 2004;4:423–8.
6. Unemori EN, Amento EP, Bauer EA, Horuk R. Melanoma
IL-1␤–stimulated PKC activation by EGCG provides growth-stimulatory activity/GRO decreases collagen expression by
indirect evidence that EGCG may act through PKC to human fibroblasts: regulation by C-X-C but not C-C cytokines.
inhibit the IL-1␤–stimulated increase in chemokine pro- J Biol Chem 1993;268:1338–42.
7. Garcia-Vicuna R, Gomez-Gaviro MV, Dominguez-Luis MJ, Pec
duction and MMP-2 activity. MK, Gonzalez-Alvaro I, Alvaro-Gracia JM, et al. CC and CXC
Recent studies suggest that PKC␦ regulates inter- chemokine receptors mediate migration, proliferation, and matrix
cellular adhesion molecule 1 expression via NF-␬B metalloproteinase production by fibroblast-like synoviocytes from
rheumatoid arthritis patients. Arthritis Rheum 2004;50:3866–77.
activation in human umbilical vein endothelial cells as 8. Cooper R, Morre DJ, Morre DM. Medicinal benefits of green tea.
well as NF-␬B–dependent gene expression in a human Part I. Review of noncancer health benefits [review]. J Altern
epithelial cell line (49). In another study, PKC␦ has been Complement Med 2005;11:521–8.
9. Frei B, Higdon JV. Antioxidant activity of tea polyphenols in vivo:
shown to mediate lysophosphatidic acid–induced NF-␬B evidence from animal studies [review]. J Nutr 2003;133:
activation and IL-8 secretion in human bronchial epithe- 3275S–84S.
lial cells (50). In the present study using RA synovial 10. Shimizu M, Weinstein IB. Modulation of signal transduction by
tea catechins and related phytochemicals. Mutat Res 2005;591:
fibroblasts, inhibition of MMP-2 activity and moderate 147–60.
suppression of chemokine production by the NF-␬B 11. Ahmed S, Rahman A, Hasnain A, Lalonde M, Goldberg VM,
inhibitor PDTC suggest that these effects are, at least in Haqqi TM. Green tea polyphenol epigallocatechin-3-gallate inhib-
its the IL-1␤-induced activity and expression of cyclooxygenase-2
part, controlled by PKC␦. However, this relationship was and nitric oxide synthase-2 in human chondrocytes. Free Radic
not directly examined in the current study. Our earlier Biol Med 2002;33:1097–105.
12. Singh R, Ahmed S, Islam N, Goldberg VM, Haqqi TM. Epigallo-
studies provided direct evidence of the role of EGCG in catechin-3-gallate inhibits interleukin-1␤–induced expression of
inhibiting IL-1␤–induced degradation of endogenous nitric oxide synthase and production of nitric oxide in human
I␬B␣ and NF-␬B activation to suppress inducible NO chondrocytes: suppression of nuclear factor ␬B activation by
degradation of the inhibitor of nuclear factor ␬B. Arthritis Rheum
synthase and COX-2 expression, resulting in the reduc- 2002;46:2079–86.
tion of NO and PGE2 synthesis in human chondrocytes 13. Singh R, Ahmed S, Malemud CJ, Goldberg VM, Haqqi TM.
(11,13). We also showed that EGCG differentially reg- Epigallocatechin-3-gallate selectively inhibits interleukin-1␤-
induced activation of mitogen activated protein kinase subgroup
ulated several distinct pathways and transcription factors c-Jun N-terminal kinase in human osteoarthritis chondrocytes.
NF-␬B and activation protein 1, to attenuate the pro- J Orthop Res 2003;21:102–9.
ROLE OF EGCG IN RA CHEMOKINE PRODUCTION AND MMP-2 ACTIVATION 2401

14. Ahmed S, Wang N, Lalonde M, Goldberg VM, Haqqi TM. Green 31. Trompezinski S, Denis A, Schmitt D, Viac J. Comparative effects
tea polyphenol epigallocatechin-3-gallate (EGCG) differentially of polyphenols from green tea (EGCG) and soybean (genistein)
inhibits interleukin-1␤-induced expression of matrix metallopro- on VEGF and IL-8 release from normal human keratinocytes
teinase-1 and -13 in human chondrocytes. J Pharmacol Exp Ther stimulated with the proinflammatory cytokine TNF␣. Arch Der-
2004;308:767–73. matol Res 2003;295:112–6.
15. Woods JM, Mogollon A, Amin MA, Martinez RJ, Koch AE. The 32. Plater-Zyberk C, Hoogewerf AJ, Proudfoot AE, Power CA, Wells
role of COX-2 in angiogenesis and rheumatoid arthritis. Exp Mol TN. Effect of a CC chemokine receptor antagonist on collagen
Pathol 2003;74:282–90. induced arthritis in DBA/1 mice. Immunol Lett 1997;57:117–20.
16. Xie Z, Singh M, Singh K. Differential regulation of matrix 33. Shahrara S, Proudfoot AE, Woods JM, Ruth JH, Amin MA, Park
metalloproteinase-2 and -9 expression and activity in adult rat CC, et al. Amelioration of rat adjuvant-induced arthritis by
cardiac fibroblasts in response to interleukin-1␤. J Biol Chem Met-RANTES. Arthritis Rheum 2005;52:1907–19.
2004;279:39513–9. 34. Guglielmotti A, D’Onofrio E, Coletta I, Aquilini L, Milanese C,
17. Ahmed S, Wang N, Hafeez BB, Cheruvu VK, Haqqi TM. Punica Pinza M. Amelioration of rat adjuvant arthritis by therapeutic
granatum L. extract inhibits IL-1␤–induced expression of matrix treatment with bindarit, an inhibitor of MCP-1 and TNF-␣ pro-
metalloproteinases by inhibiting the activation of MAP kinases duction. Inflamm Res 2002;51:252–8.
and NF-␬B in human chondrocytes in vitro. J Nutr 2005;135: 35. Gong JH, Yan R, Waterfield JD, Clark-Lewis I. Post-onset
2096–102.
inhibition of murine arthritis using combined chemokine antago-
18. Jeong JG, Kim JM, Cho H, Hahn W, Yu SS, Kim S. Effects of
nist therapy. Rheumatology (Oxford) 2004;43:39–42.
IL-1␤ on gene expression in human rheumatoid synovial fibro-
36. Youssef S, Maor G, Wildbaum G, Grabie N, Gour-Lavie A, Karin
blasts. Biochem Biophys Res Commun 2004;324:3–7.
N. C-C chemokine–encoding DNA vaccines enhance breakdown
19. Jackson C, Nguyen M, Arkell J, Sambrook P. Selective matrix
metalloproteinase (MMP) inhibition in rheumatoid arthritis— of tolerance to their gene products and treat ongoing adjuvant
targetting gelatinase A activation [review]. Inflamm Res 2001;50: arthritis. J Clin Invest 2000;106:361–71.
183–6. 37. Walsh DA. Angiogenesis and arthritis [review]. Rheumatology
20. Johnson Z, Power CA, Weiss C, Rintelen F, Ji H, Ruckle T, et al. (Oxford) 1999;38:103–12.
Chemokine inhibition—why, when, where, which and how? [re- 38. Murphy G, Crabbe T. Gelatinases A and B. Methods Enzymol
view]. Biochem Soc Trans 2004;32:366–77. 1995;248:470–84.
21. Koch AE, Kunkel SL, Harlow LA, Mazarakis DD, Haines GK, 39. Michaelides MR, Curtin ML. Recent advances in matrix metallo-
Burdick MD, et al. Epithelial neutrophil activating peptide-78: a proteinase inhibitors research [review]. Curr Pharm Des 1999;5:
novel chemotactic cytokine for neutrophils in arthritis. J Clin 787–819.
Invest 1994;94:1012–8. 40. Adcocks C, Collin P, Buttle DJ. Catechins from green tea (Ca-
22. Taub DD. C-C chemokines: an overview. In: Koch AE, Strieter mellia sinensis) inhibit bovine and human cartilage proteoglycan
RM, editors. Chemokines in disease. Austin (TX): RG Landes; and type II collagen degradation in vitro. J Nutr 2002;132:341–6.
1996. p. 27–54. 41. Vankemmelbeke MN, Jones GC, Fowles C, Ilic MZ, Handley CJ,
23. Schall TJ, Bacon K, Toy KJ, Goeddel DV. Selective attraction of Day AJ, et al. Selective inhibition of ADAMTS-1, -4 and -5 by
monocytes and T lymphocytes of the memory phenotype by catechin gallate esters. Eur J Biochem 2003;270:2394–403.
cytokine RANTES. Nature 1990;347:669–71. 42. Close DR. Matrix metalloproteinase inhibitors in rheumatic dis-
24. Szekanecz Z, Halloran MM, Volin MV, Woods JM, Strieter RM, eases [review]. Ann Rheum Dis 2001;60 Suppl 3:iii62–7.
Haines GK III, et al. Temporal expression of inflammatory 43. Sridhar SS, Shepherd FA. Targeting angiogenesis: a review of
cytokines and chemokines in rat adjuvant-induced arthritis. Ar- angiogenesis inhibitors in the treatment of lung cancer [review].
thritis Rheum 2000;43:1266–77. Lung Cancer 2003;42 Suppl 1:S81–91.
25. Thornton S, Duwel LE, Boivin GP, Ma Y, Hirsch R. Association 44. Ullmann U, Haller J, Decourt JP, Girault N, Girault J, Richard-
of the course of collagen-induced arthritis with distinct patterns of Caudron AS, et al. A single ascending dose study of epigallocat-
cytokine and chemokine messenger RNA expression. Arthritis echin gallate in healthy volunteers. J Int Med Res 2003;31:88–101.
Rheum 1999;42:1109–18. 45. Ullmann U, Haller J, Decourt JD, Girault J, Spitzer V, Weber P.
26. Matsukawa A, Yoshimura T, Fujiwara K, Maeda T, Ohkawara S, Plasma-kinetic characteristics of purified and isolated green tea
Yoshinaga M. Involvement of growth-related protein in lipo- catechin epigallocatechin gallate (EGCG) after 10 days repeated
polysaccharide-induced rabbit arthritis: cooperation between dosing in healthy volunteers. Int J Vitam Nutr Res 2004;74:
growth-related protein and IL-8, and interrelated regulation 269–78.
among TNF␣, IL-1, IL-1 receptor antagonist, IL-8, and growth- 46. Levites Y, Amit T, Mandel S, Youdim MB. Neuroprotection and
related protein. Lab Invest 1999;79:591–600. neurorescue against A␤ toxicity and PKC-dependent release of
27. Halloran MM, Woods JM, Strieter RM, Szekanecz Z, Volin MV, nonamyloidogenic soluble precursor protein by green tea polyphe-
Hosaka S, et al. The role of an epithelial neutrophil-activating nol (-)-epigallocatechin-3-gallate. FASEB J 2003;17:952–4.
peptide-78-like protein in rat adjuvant-induced arthritis. J Immu- 47. Wang SI, Mukhtar H. Gene expression profile in human prostate
nol 1999;162:7492–500. LNCaP cancer cells by (--) epigallocatechin-3-gallate. Cancer Lett
28. Inoue T, Yamashita M, Higaki M. The new antirheumatic drug 2002;182:43–51.
KE-298 suppresses monocyte chemoattractant protein (MCP)-1 48. Lee SJ, Kim DC, Choi BH, Ha H, Kim KT. Regulation of p53 by
and RANTES production in rats with adjuvant-induced arthritis activated protein kinase C-␦ during nitric oxide-induced dopami-
and in IL-1␤-stimulated synoviocytes of patients with rheumatoid nergic cell death. J Biol Chem 2006;281:2215–24.
arthritis. Rheumatol Int 2001;20:149–53. 49. Rahman A, Anwar KN, Uddin S, Xu N, Ye RD, Platanias LC, et
29. Porath D, Riegger C, Drewe J, Schwager J. Epigallocatechin-3- al. Protein kinase C-␦ regulates thrombin-induced ICAM-1 gene
gallate impairs chemokine production in human colon epithelial expression in endothelial cells via activation of p38 mitogen-
cell lines. J Pharmacol Exp Ther 2005;315:1172–80. activated protein kinase. Mol Cell Biol 2001;21:5554–65.
30. Takano K, Nakaima K, Nitta M, Shibata F, Nakagawa H. Inhibi- 50. Cummings R, Zhao Y, Jacoby D, Spannhake EW, Ohba M, Garcia
tory effect of (-)-epigallocatechin 3-gallate, a polyphenol of green JG, et al. Protein kinase C␦ mediates lysophosphatidic acid-
tea, on neutrophil chemotaxis in vitro and in vivo. J Agric Food induced NF-␬B activation and interleukin-8 secretion in human
Chem 2004;52:4571–6. bronchial epithelial cells. J Biol Chem 2004;279:41085–94.