The Intermediate Filament Protein Vimentin Is a New Target for Epigallocatechin Gallate*

Epigallocatechin gallate (EGCG) is the major active polyphenol in green tea. Protein interaction with EGCG is a critical step in the effects of EGCG on the regulation of various key proteins involved in signal transduction. We have identified a novel molecular target of EGCG using affinity chromatography, two-dimensional electrophoresis, and mass spectrometry for protein identification. Spots of interest were identified as the intermediate filament, vimentin. The identification was confirmed by Western blot analysis using an anti-vimentin antibody. Experiments using a pull-down assay with [3H]EGCG demonstrate binding of EGCG to vimentin with a Kd of 3.3 nm. EGCG inhibited phosphorylation of vimentin at serines 50 and 55 and phosphorylation of vimentin by cyclin-dependent kinase 2 and cAMP-dependent protein kinase. EGCG specifically inhibits cell proliferation by binding to vimentin. Because vimentin is important for maintaining cellular functions and is essential in maintaining the structure and mechanical integration of the cellular space, the inhibitory effect of EGCG on vimentin may further explain its anti-tumor-promoting effect.


From the Hormel Institute, University of Minnesota, Austin, Minnesota 55912
Epigallocatechin gallate (EGCG) is the major active polyphenol in green tea. Protein interaction with EGCG is a critical step in the effects of EGCG on the regulation of various key proteins involved in signal transduction. We have identified a novel molecular target of EGCG using affinity chromatography, two-dimensional electrophoresis, and mass spectrometry for protein identification.

Spots of interest were identified as the intermediate filament, vimentin. The identification was confirmed by Western blot analysis using an anti-vimentin antibody. Experiments using a pull-down assay with [ 3 H]EGCG demonstrate binding of EGCG to vimentin with a K d of 3.3 nM. EGCG inhibited phosphorylation of vimentin at serines 50 and 55 and phosphorylation of vimentin by cyclin-dependent kinase 2 and cAMP-dependent protein kinase. EGCG specifically inhibits cell proliferation by binding to vimentin. Because vimentin is important for maintaining cellular functions and is essential in maintaining the structure and mechanical integration of the cellular space, the inhibitory effect of EGCG on vimentin may further explain its anti-tumorpromoting effect.
A number of epidemiological studies have shown that the consumption of green tea may protect against many cancer types, including lung, prostate, and breast (1,2). The inhibition of tumorigenesis by green or black tea preparations was demonstrated in animal models at various organ sites (3)(4)(5). The structures of the four major catechins, (Ϫ)-epigallocatechin gallate (EGCG), 1 (Ϫ)-epigallocatechin (EGC), (Ϫ)-epicatechin gallate (ECG), and (Ϫ)-epicatechin (EC), are shown in Fig. 1. EGCG is the major polyphenol in green tea and may account for 50 -80% of the total catechins in tea (4,6,7). The inhibitory activity of EGCG against tumorigenesis has been demonstrated. The mechanisms responsible for these cancer-preven-tive effects of tea are not very well understood but are being intensively investigated.
Searching for the EGCG "receptor" or high affinity proteins that bind to EGCG is the first step to understanding the molecular and biochemical mechanisms of the anticancer effects of tea polyphenols. A few proteins that can directly bind with EGCG have been identified, including plasma proteins: fibronectin, fibrinogen, and histidine-rich glycoprotein (8); also fatty acid synthase (Fas) (9), laminin, and the 67-kDa laminin receptor (10,11). Plasma proteins may act as carrier proteins for EGCG. Fas might trigger the cascade of Fas-mediated apoptosis, and the fact that EGCG can bind and regulate biological functions of the 67 laminin receptor has possible implications for prion-related diseases. However, the biologic and physiologic significance for the anticancer effects of tea polyphenols is not clear. Identification of new proteins binding with EGCG should help in the design of new strategies to prevent cancer.
Mass spectrometry-based proteomic analysis is a powerful tool to identify proteins binding with EGCG. We used the JB6 mouse epidermal cell line, a system that has been used extensively as an in vitro model for tumor promotion studies (12), to identify novel proteins that bind with EGCG. The results indicated that EGCG binds with the intermediate filament (IF) protein, vimentin with high affinity (K d ϭ 3.3 nM). Vimentin, one of the type III IF proteins, is a major component of IFs and is expressed during development in a wide range of cells, including mesenchymal cells and in a variety of cultured cell lines and tumors (13,14). IFs are essential for structure and mechanical integration of the cellular space and a variety of cellular functions such as mitosis, locomotion, and organizational cell architecture, and vimentin is readily phosphorylated by numerous protein kinases, thereby regulating their function (15)(16)(17). The proteolytic derivatives indicate that the aminoterminal domain, but not the carboxyl-terminal domain, has a direct effect on filament stability and polymerization (18). We used the Swiss-Prot program and a program from the National Genomic Information center to search the MALDI-TOF data base for peptides important in binding with EGCG. The peptide (SLYSSSPGGAYVTR) contains two phosphorylation sites of vimentin, serine 50 and serine 55. Our results in vivo demonstrated that EGCG inhibited the phosphorylation of vimentin at serines 50 and 55 in a dose-dependent manner. In vitro studies revealed that the phosphorylation of IF proteins by various types of serine/threonine protein kinases induces disassembly of the filament structure (19). Vimentin is an excellent substrate for Cdc2 and PKA. Cdc2 phosphorylates vimentin at serine 55, and serine 50 is a phosphorylation site of PKA. In this study, we provide evidence that EGCG inhibited phosphorylation of vimentin by Cdc2 and PKA. EGCG is known to inhibit cell proliferation of various cancer cell lines (20 -25). Our data show that inhibition of vimentin phosphorylation by * This work was supported in part by the Hormel Foundation and National Institutes of Health Grants CA81064 and CA88961. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  EGCG was directly associated with a decrease in cell proliferation. We hypothesize that the binding of EGCG to vimentin could be very important in a cascade of vimentin-mediated cell proliferation.
Cell Culture-The JB6 Cl41 mouse epidermal cell line was grown in monolayers with minimal essential medium supplemented with 5% (v/v) heat-inactivated FBS, 2 mM L-glutamine, and 25 g/ml gentamicin. The cell lines were cultured at 37°C in a humidified atmosphere of 5% CO 2 .
Affinity Chromatography-EGCG was first coupled to the CNBractivated Sepharose 4B matrix, and the binding between vimentin and EGCG was examined by affinity chromatography according to the manufacturer's instructions. The cell lysates from JB6 Cl41 cells cultured in 10-cm dishes were disrupted in lysis buffer (50 mM Tris, pH 7.5, 5 mM EDTA, 150 mM NaCl, 1 mM dithiothreitol, 0.01% Nonidet P-40, 0.02 mM PMSF, 1ϫ protease inhibitor mixture). The lysate was sonicated and centrifuged at 15000 ϫ g for 30 min, and the supernatant fraction (10 mg/8 ml) was applied to the EGCG-Sepharose 4B column (12 ϫ 8 mm) at 4°C. The mobile phase was T buffer (50 mM Tris-HCl pH 7.5, 100 mM NaCl, 10 mM 6-aminohexanoic acid, 1 mM PMSF, 1 mM benzamidine hydrochloride, and 1 mM EDTA) running at a flow rate of 0.5 ml/min. The bound proteins were then eluted with T1 buffer (4 M urea, 1 M NaCl in T buffer). Protein elution was monitored in fractions using the Bio-Rad protein assay.
Two-dimensional Gel Electrophoresis-The samples were precipitated with trichloroacetic acid, and the ReadyPrep two-dimensional Starter kit was used as outlined in the instruction manual. The samples were then applied onto the IPG nonlinear strips, pH 5-8, for the first dimension separation. The second-dimensional separation of eluted proteins was performed by 10 -20% SDS-PAGE. The gels and proteins were stained with SYPRO Ruby protein gel stain or transferred onto nitrocellulose membranes.
MALDI-TOF Analysis-The protein spot was excised from a polyacrylamide gel and digested overnight with trypsin. The extracted peptides were mixed with the matrix ␣-cyano-4-hydroxycinammic acid (Aldrich), spotted onto a MALDI plate, and analyzed with a Voyager DePro mass spectrometer (Applied Biosystems, Birmingham, AL). The peptide masses were entered into Mascot (www.matrixscience.com/), and the NCBI data base was searched to identify the protein.

SDS-PAGE and Western
Blotting-The samples were analyzed by SDS-PAGE (27) and Western blotting as described (28) with some modification as described previously (26).
Construction of Vimentin Plasmids-A cDNA fragment encoding vimentin was inserted in-frame into the BamHI/EcoRI sites of the pcDNA3.1/V5-His vector (Invitrogen) or pGEX5X1 (Amersham Biosciences) to produce the V5 epitope-tagged construct, pcDNA3.1/V5vimentin, or pGEX5X1-vimentin. All of the positive clones containing cDNA inserts were identified by restriction enzyme mapping and sequenced at Genewiz, Inc. (North Brunswick, NJ).
Bacterial Expression and Purification of the GST-Vimentin Fusion Protein-The GST-vimentin fusion protein was expressed in Escherichia coli BL21 and purified as described (12). The GST-vimentin was used for affinity binding and in vitro kinase assays.
GST-Vimentin Affinity Binding Assay Estimation for K d -For the GST-vimentin affinity binding assay, expressed GST fusion proteins were incubated for immobilization with glutathione-Sepharose 4B beads for 1 h at room temperature. The affinity binding assay was carried out overnight at 4°C in a 500-l reaction mixture containing reaction buffer (see "In Vitro Pull-down Assay") with 1 g of GSTvimentin-Sepharose 4B or GST-Sepharose 4B and 0.5 Ci of [ 3 H]EGCG. For analyzing concentration-dependent uptake, 39 pM to 50 M concentrations of EGCG were applied. The K d value was determined through nonlinear regression analysis using the Prizm 4.0 software program (Graphpad Inc., San Diego, CA).
Assay of Phosphorylation of Vimentin at Serine 50 and Serine 55-JB6 Cl41 cells were seeded in 10-cm dishes (4 ϫ 10 5 ). After culturing at 37°C (5% FBS) for 12 h, the cells were treated with different concentrations of EGCG (1, 5, 10, or 20 M). 72 h following treatment, the cells were lysed, and equal amounts of protein were separated by SDS-PAGE. The blots were probed with phospho-specific antibodies to vimentin.
Kinase Assay-The Cdc2 kinase assay was performed at 30°C for 1 h in a 50-l reaction mixture containing kinase buffer (50 mM Tris, pH 7. The PKA kinase assay was performed at 30°C for 1 h in a 50-l reaction mixture containing kinase buffer, 0.5 l (1 ng) of enzyme, 250 M of the kinase substrates (kemptide), or 4 g of GST-vimentin and EGCG (1-20 M). The reactions were stopped by adding an equal volume of 5ϫ SDS sample buffer. After boiling for 5 min, one-half of the sample was separated by 12% SDS-PAGE, and the phosphorylated proteins were visualized by x-ray film.
siRNA Preparation and Vector Construction-Two pairs of hairpin siRNA oligonucleotides containing BamHI and HindIII sites were designed as described before (29). To design target-specific siRNA templates, we selected sequences of the type AA(N19)NN or AA(N19)UU from the open reading frame of the targeted mRNA to obtain 19nucleotide sense and antisense strands. This strategy for choosing siRNA target sites is based on the previous observation by Elbashir et al. (30) that siRNA with 3Ј overhanging UU dinucleotides are the most effective. Hairpin siRNA template oligonucleotides were chemically synthesized using a DNA synthesizer and deprotected and gel-purified as purchased from Sigma Genosys. Selected siRNA target sequences were aligned to the genome data base in a BLAST search to ensure sequences without significant homology to other genes.
Sense and antisense oligonucleotides were annealed and cloned into the pSilencer 3.1-H1 neo vector (Ambion) at the BamHI and HindIII sites as described in the manufacturer's instructions.
The siRNA sequence targeting vimentin (NM_011701) was from positions 1147-1165 relative to the start codon. As a negative siRNA control, we used two kinds of circular plasmids encoding a hairpin siRNA sequence targeting the green fluorescent protein (siRNA GFP) and a scrambled siRNA, the sequence of which lacks significant sequence homology to the mouse, human, or rat genome data base. The resulting siRNA-expressing plasmids were used for both transient and stable transfections.
Transfection and Stable Cell Lines-JB6 Cl41 cells were maintained in minimal essential medium supplemented with 5% (v/v) heat-inactivated FBS, 2 mM L-glutamine, and 25 g/ml gentamicin at 37°C in a humidified atmosphere of 5% CO 2 . Stable cell lines were obtained by transfecting 2 ϫ 10 5 cells in 6-cm dishes with 4 g of the expression plasmids. The cell lines were selected and subsequently maintained in medium containing 200 g/ml zeocin (Invitrogen).
Immunofluorescence-Transfected cells grown on glass chambers were fixed in 4% formaldehyde in PBS for 15 min and extracted in 0.5% Triton X-100 in PBS for 10 min at room temperature to allow subsequent antibody penetration. After extensive washing in PBS, the fixed and extracted cells were incubated with primary antibodies at 37°C in a humidified chamber for 1 h, washed three times in PBS, and then incubated with the fluorescein isothiocyanate-conjugated secondary antibody for an additional hour at 37°C. The cells on coverslips were viewed using an LEI-750D CE microscope (Leica, Bannockburn, IL) equipped with epifluorescence optics.
Cell Proliferation Assay-To assess proliferation, control (GFP siRNA) or vimentin siRNA cells (3 ϫ 10 3 ) were seeded in 96-well tissue culture plates and cultured at 37°C in a 5% CO 2 incubator. Cellular proliferation was then estimated at 24, 48, and 72 h using the CellTiter 96 AQueous One Solution cell proliferation assay kit (Promega) according to the manufacturer's instructions. The assay solution is added to each well, and absorbance (492 nm and 690 nm background) is read with a 96-well plate reader. Absorbance at 492 nm is directly proportional to the number of living cells. Then to test whether EGCG had an effect on proliferation of control (GFP siRNA) or vimentin siRNA cells (3 ϫ 10 3 ), the cells were seeded in 96-well tissue culture plates for 12 h and then treated with different concentrations of EGCG (1, 5, 10, 15, or 20 M). The cells were cultured for an additional 72 h, and then 20 l of the CellTiter 96 AQueous One Solution were added to each well. The cells were incubated for an additional hour, and then absorbance (492 nm and 690 nm background) was measured using a 96-well plate reader as above.
Data Analysis-The figures shown in this paper are representative of at least three independent experiments with similar results. The statistical differences were evaluated using Student's t test and considered significant at p Յ 0.05.

Identification of Vimentin as an EGCG-binding Protein-
Proteins from JB6 Cl41 cell lysates were subjected to affinity chromatography using EGCG-Sepharose 4B (Fig. 2). Fractions containing proteins binding with EGCG were analyzed by twodimensional electrophoresis and MALDI-TOF-MS to identify proteins that directly bind with EGCG. Experimental conditions were first established to obtain a reproducible pattern of spots by two-dimensional electrophoresis (Fig. 3).
The two-dimensional electrophoresis was performed with different quantities of protein from the affinity column (data not shown) to determine optimal conditions for two-dimen- Lysates from JB6 Cl41 cells were applied to an EGCG-Sepharose 4B affinity column equilibrated in T buffer. The proteins that did not bind with EGCG were eluted with T buffer; the proteins binding with EGCG were eluted with T1 buffer. The arrows indicate the elution with T and T1 buffers, respectively. Fractions eluted with T1 buffer were analyzed by two-dimensional electrophoresis. sional electrophoresis, and each protein fraction was run in triplicate. The results indicate that three spots were consistently observed (Fig. 3) by two-dimensional electrophoresis stained with SYPRO Ruby protein gel stain. The respective spots were removed and digested with trypsin, and the resulting peptides were subjected to analysis by MALDI-TOF-MS (Table I). A search of the NCBI data base revealed that all three spots corresponded to vimentin, an intermediate filament protein.
Western Blot Confirmation of Vimentin as an EGCG-binding Protein-Further evidence for the binding of vimentin with EGCG was demonstrated with immunoblot analysis. Analysis was done using JB6 cell lysates and EGCG-Sepharose 4B. To create a two-dimensional image of vimentin binding with EGCG, samples subjected to the affinity column were analyzed by two-dimensional electrophoresis under the same conditions as described above, and the proteins were transferred onto nitrocellulose membranes. The spots were confirmed to be vimentin by Western analysis using a mouse monoclonal (VI-01) vimentin antibody (Fig. 4A). These results strongly indicate that vimentin binds with EGCG.
To directly test this possibility, we used a combination of a pull-down assay with EGCG-Sepharose 4B and an immunoprecipitation assay with protein G-Sepharose 4B and the vimentin antibody. The lysate from JB6 Cl41 cells was directly applied to the EGCG-Sepharose 4B column, and following elution, Western blot analysis revealed the presence of a 54-kDa protein that was reactive with the vimentin antibody (VI-01) (Fig. 4B, lane  2), confirming the presence of vimentin in the EGCG-bound fraction. Lane 1 shows vimentin from lysate as the control. The lysates were applied to protein G-Sepharose 4B beads and immunoprecipitated with the vimentin antibody (Fig. 4C, lane  2) or were applied to the EGCG-Sepharose 4B column and Sepharose 4B column and eluted (Fig. 4C, lanes 3 and 4, respectively). The results clearly indicate that vimentin is present in mouse cells and that EGCG binds with vimentin in vivo. To characterize the binding interaction, we measured the binding affinity of GST-vimentin and [ 3 H]EGCG using a pull-down assay. The results indicate that vimentin displayed a high affinity for binding with [ 3 H]EGCG (Fig. 4D). GST-Sepharose 4B alone was used only as a negative control, and no binding occurred (data not shown). The K d value for the binding of EGCG to vimentin was 3.3 nM.

EGCG Inhibits Phosphorylation of Vimentin at Ser 50 and
Ser 55 in a Dose-dependent Manner-JB6 Cl41 cells were employed to analyze the EGCG inhibition of phosphorylation of vimentin at serine 50 and serine 55 (Fig. 5A). The dose course study shows that phosphorylation of vimentin gradually decreases following 72 h of treatment with increasing amounts of EGCG (1, 5, 10, or 20 M) (Fig. 5B), with no effect on total vimentin protein levels (Fig. 5B). These results indicate that phosphorylation of vimentin at serine 50 and serine 55 is inhibited by EGCG in a dose-dependent manner.
EGCG Specifically Inhibits the Phosphorylation of Vimentin by Cdc2-Cdc2 kinase activity was analyzed by in vitro kinase assays using GST-vimentin and histone H1 as substrates. First we determined the effect of EGCG on the total phosphorylation of vimentin by Cdc2. As shown in Fig. 5C (top panel), up to 5 M EGCG did not affect Cdc2 kinase activity, whereas 10 or 20 M induced a gradual decrease in kinase activity. The results indicate that increasing concentrations of EGCG inhibit Cdc2 kinase-mediated phosphorylation of vimentin.
The next question to be answered was whether the inhibition was due to an interference of EGCG with the phosphorylation of vimentin directly or indirectly via Cdc2 kinase. In a second series of experiments, the kinase assay was carried out using GST-vimentin or histone H1 as a substrate for Cdc2 kinase. The results confirm that the addition of EGCG to the kinase mixture decreased the phosphorylation of vimentin by Cdc2 (Fig. 5C, bottom panel; IC 50 ϭ 17 M EGCG). On the other hand, Cdc2 kinase activity was not affected by EGCG with histone H1 as substrate (Fig. 5C, bottom panel). These results clearly indicate that EGCG specifically inhibits Cdc2 kinase activity toward vimentin but not its activity toward histone H1. Furthermore, when purified Cdc2 kinase was applied to the EGCG-Sepharose 4B column and eluted with T1 buffer, no proteins bands were detected (data not shown). These data strongly support our contention that EGCG interacts directly with vimentin and not with Cdc2 kinase.
This observation prompted us to investigate the effect of EGCG analogues, including EC, ECG, and EGC (Fig. 1) on Cdc2 kinase activity. The results indicate that besides EGCG, only ECG (20 M) inhibited Cdc2 kinase activity (31%) (Fig.  5D). Both compounds (EGCG and ECG) that inhibited Cdc2 kinase activity contain the gallate group, suggesting that this moiety may be critical for binding with vimentin. FIG. 3. Two-dimensional electrophoresis of proteins captured from EGCG-Sepharose 4B beads. Cell lysates were applied to an EGCG-Sepharose 4B column, and ECGG-binding proteins were used for analysis by twodimensional electrophoresis. Spots pertaining to vimentin are indicated in the dotted ring and enlarged in the inset. The numbered protein spots were digested with trypsin, and the resulting peptides were analyzed by mass spectrometry (Table I).
EGCG Decreases PKA Activity-PKA activity was analyzed by in vitro kinase assays using GST-vimentin or kemptide as substrate. Using vimentin as substrate for PKA, the dose course study showed that kinase activity of PKA for vimentin decreases after treatment with increasing concentrations of EGCG (Fig. 5E, top panel). Using GST-vimentin or kemptide as a substrate for PKA kinase, the results indicated that EGCG inhibited PKA-mediated phosphorylation of vimentin more effectively than phosphorylation of kemptide (IC 50 ϭ 2 and 8 M, respectively; Fig. 5E, bottom panel). Thus, EGCG inhibits PKAmediated phosphorylation of vimentin in vitro.
EGCG Inhibits the Proliferation of JB6 Cl41 Cells but Not Vimentin Knockdown Cells-IFs, such as vimentin, have an important functional involvement in cell division and proliferation (15). Therefore, to verify that vimentin will affect cell proliferation, we used control (GFP siRNA) and vimentin knockdown (vimentin siRNA) JB6 Cl41 stably transfected cells (Fig. 6, A and B). The results indicated that control cells grew significantly more quickly than vimentin knockdown cells (*, p Ͻ 0.05; **, p Ͼ 0.005; Fig. 6C). EGCG has been reported to inhibit cell proliferation of a variety of cell lines (4,(31)(32)(33)(34)(35). The cells were treated with different doses of EGCG for 72 h, and proliferation was determined as described under "Experimental Procedures. " Proliferation was inhibited in control (GFP siRNA) cells in a dose-dependent manner (*, p Ͻ 0.05; **, p Ͼ 0.005; Fig. 6D). The IC 50 was 13 M. Furthermore, proliferation of cells expressing vimentin siRNA was not affected by EGCG treatment (Fig. 6D). These results strongly suggested that EGCG binds with vimentin, and the association can have a regulatory role in controlling cell proliferation. DISCUSSION Green tea is widely used as a beverage in China, Japan, and other Asian countries, whereas black tea is more popular in Western countries (4,6). In recent years many animal studies and several epidemiological studies have reported the anticarcinogenic effects of tea (1, 2). EGCG is the major polyphenol component of green tea and the most active anticancer factor found in tea (6). Although several mechanisms explaining the anti-carcinogenic effects of tea have been presented, tea components target specific cell signaling pathways responsible for regulating cell proliferation and cell transformation (1,5,7,21,22,31). We have identified an important and novel molecular target of EGCG. This is the first report to show that EGCG binds with a type III IF protein, vimentin.
IFs are relatively static polymers, and their functions are restricted to establishing and maintaining the mechanical integrity of cells (36,37). However, results from recent studies (17) using live imaging with GFP-tagged IF proteins demonstrate that IFs actually form dynamic, motile networks. These newly elucidated properties of IFs have important functional implications for their involvement in cell division, protein trafficking, cellular motility, intracellular signaling, and the regional control of cytoplasmic architecture. In support of this, accumulations of vimentin are frequently noted as a pathological hallmark in a wide range of human diseases, including giant axonal neuropathy in which vimentin aggregates are found in skin fibroblasts of these patients (38). Vimentin aggregates are also observed in the spheroids in amyotrophic lateral sclerosis (39) and in patients with fetal alcohol syndrome (40).
In this study, using affinity chromatography, two-dimensional electrophoresis, and MALDI-TOF analysis, we have provided clear evidence that EGCG directly binds with vimentin. We have further shown the vimentin and EGCG interact in vitro using binding assays. This result raises questions about the physiological significance of the EGCG binding with vimentin. Difficulty in interpreting these data lies with the inability to clearly establish a functional role for the intermediate filament vimentin. Although a number of theories have been proposed for the function of this network, the data are not conclusive. Many diverse cellular processes, including differentiation, motility, signal transduction, cell division, cytoskeletal stability, and vesicular trafficking, have been associated with alterations in the dynamics of the IFs (19, 41).
The association of EGCG with vimentin may serve different functions, including activation or inhibition of kinase activities.  tains two phosphorylation sites of vimentin, serine 50 and serine 55, that are important for binding with EGCG. The most dramatic changes in IF structure, which are accompanied by an increased level of phosphorylation, are observed when cells enter mitosis. Phosphopeptide analysis provided clear evidence that Cdc2 kinase directly phosphorylates vimentin at serine 55 during mitosis (50). An amino-terminally mutated vimentin (S55A) significantly decreases the fraction of transfected mitotic cells with a disassembled IF network, strongly suggesting that phosphorylation by Cdc2 is essential for IF dissembly in vivo (47). We examined the effect of EGCG on phosphorylation of vimentin at serine 55. Our results in vivo (Fig. 5B) demonstrated that EGCG inhibited the phosphorylation of vimentin at serine 55 in a dose-dependent manner. The data from the in vitro Cdc2 kinase assay suggest that EGCG inhibits Cdc2 kinase-mediated phosphorylation of vimentin.
In addition to Cdc2, the well characterized PKA is also capable of phosphorylating vimentin in vitro. PKA is involved in cellular trafficking, motility, and reorganization of vimentin. Microinjection of PKA resulted in hyperphosphorylation and either reorganization or disassembly of the IF network (47). One of the vimentin sites phosphorylated by PKA is serine 50 and our results in vivo suggest that phosphorylation of vimentin at serine 50 is inhibited by EGCG in a dose-dependent manner. The data further indicate that EGCG inhibits in vitro PKA-mediated phosphorylation of vimentin (Fig. 5B, top panel).
As far as we know, phosphorylation of IF proteins by individual protein kinases is mostly associated with the loss of filamentous structure. Based on the data regarding vimentin phosphorylation in vivo and from in vitro kinase assays, EGCG clearly inhibits phosphorylation of vimentin, and thus, EGCG can regulate vimentin phosphorylation of Cdc2 and PKA by its binding with vimentin.
These results raise questions about the physiological significance of the EGCG inhibition of vimentin phosphorylation. Cdc2 kinase directly phosphorylates vimentin at serine 55 during mitosis, and the proliferation or survival of many cell lines has been shown to be inhibited by EGCG (20 -25). We examined the effect of EGCG on the cell proliferation of control JB6 Cl41 cells (GFP siRNA) and vimentin knockdown JB6 Cl41 cells (vimentin siRNA). We found that control GFP siRNA cells grew significantly faster than vimentin knockdown JB6 Cl41 cells.
This is the first report to show that EGCG specifically inhibits cell proliferation by binding to an intracellular structural protein and preventing its phosphorylation. Our data have increased the understanding of the molecular and biochemical mechanisms of the anticancer effects of tea polyphenols and provide good validation for those effects in cancer prevention and treatment.