Resveratrol (trans-3,4′,5-trihydroxystilbene), a polyphenolic phytoalexin found in grapes, fruits, and root extracts of the weed Polygonum cuspidatum, exhibits anti-inflammatory, cell growth-modulatory, and anticarcinogenic effects. How this chemical produces these effects is not known, but it may work by suppressing NF-κB, a nuclear transcription factor that regulates the expression of various genes involved in inflammation, cytoprotection, and carcinogenesis. In this study, we investigated the effect of resveratrol on NF-κB activation induced by various inflammatory agents. Resveratrol blocked TNF-induced activation of NF-κB in a dose- and time-dependent manner. Resveratrol also suppressed TNF-induced phosphorylation and nuclear translocation of the p65 subunit of NF-κB, and NF-κB-dependent reporter gene transcription. Suppression of TNF-induced NF-κB activation by resveratrol was not restricted to myeloid cells (U-937); it was also observed in lymphoid (Jurkat) and epithelial (HeLa and H4) cells. Resveratrol also blocked NF-κB activation induced by PMA, LPS, H2O2, okadaic acid, and ceramide. The suppression of NF-κB coincided with suppression of AP-1. Resveratrol also inhibited the TNF-induced activation of mitogen-activated protein kinase kinase and c-Jun N-terminal kinase and abrogated TNF-induced cytotoxicity and caspase activation. Both reactive oxygen intermediate generation and lipid peroxidation induced by TNF were suppressed by resveratrol. Resveratrol’s anticarcinogenic, anti-inflammatory, and growth-modulatory effects may thus be partially ascribed to the inhibition of activation of NF-κB and AP-1 and the associated kinases.
Resveratrol (trans-3,4′,5-trihydroxystilbene)3 is a polyphenol found in various fruits and vegetables and is abundant in grapes. The root extracts of the weed Polygonum cuspidatum, an important constituent of Japanese and Chinese folk medicine, is also an ample source of resveratrol (for references, see 1). In plants, resveratrol functions as a phytoalexin that protects against fungal infections (2, 3). Several studies within the last few years have shown that resveratrol exhibits cardioprotective and chemopreventive effects (1 and references therein). This constituent may account for the reduced risk of coronary heart disease in humans that has been associated with moderate wine consumption (4, 5). A constituent of the skin of grapes, its concentration reaches 10–20 μM in red wine, but is absent in white wines (5). How exactly resveratrol exerts its cardioprotective effects is not understood, but they have been ascribed to its ability to block platelet aggregation (6, 7), inhibit oxidation of low density lipoprotein (8, 9), and induce NO production (10). Resveratrol’s ability to inhibit ribonucleotide reductase and DNA polymerase and to suppress cell growth have also been suggested to play a role in cardioprotection (10, 11, 12, 13).
In 1997, resveratrol was reported to be one of the most potent chemopreventive agents able to block all three phases of tumor development that includes initiation, promotion, and progression, induced by the aryl hydrocarbon DMBA (14). How resveratrol exerts its anticarcinogenic effects is only partially understood. This polyphenol has been shown to inhibit the growth of a wide variety of tumor cells, including leukemic, prostate, breast, and endothelial cells (10, 15, 16, 17). The ability of resveratrol to induce the expression of CD95L (also called FasL), p53, and p21 may contribute to its growth-inhibitory effects (10, 15). The suppression of cyclooxygenase-2 (COX-2), cytochrome p450, and c-fos expression by resveratrol may account for its ability to inhibit tumor promotion (18, 19, 20). Recently, the drug was reported to be a phytoestrogen that behaves as superagonist of estrogen receptor and thereby an inducer of tumor cell proliferation (21). Its structural similarity with estrogen may also account for its cardioprotective effects.
Because the carcinogenic, inflammatory, and growth-modulatory effects of many chemicals are mediated by NF-κB, we hypothesized that the suppression of NF-κB activation pathway accounts for resveratrol’s activities. Numerous lines of evidence suggest this possibility. For example, various agents that promote tumorigenesis are known to activate NF-κB (for references, see 22), including phorbol ester, okadaic acid, and TNF. Experiments in TNF-deficient mice showed that TNF is required for tumor promotion (23). In addition, several genes that are involved in tumorigenesis, metastasis, and inflammation are regulated by NF-κB (22). A critical role for NF-κB in cellular transformation has also been reported (24).
Most agents that activate NF-κB also activate another transcription factor, AP-1 (25). That AP-1 activation mediates tumorigenesis and invasiveness has also been described (26 and references therein). The activation of NF-κB and AP-1 is regulated by several protein kinases that belong to the mitogen-activated protein kinase (MAPK) family (27). The activation of NF-κB and AP-1 and its associated kinases is in most cases dependent on the production of reactive oxygen species (28, 29, 30, 31).
In this study, we tested the hypothesis that the anti-inflammatory and anticarcinogenic effects of resveratrol are mediated through its modulation of NF-κB and AP-1 activation, members of the MAPK, and caspase-mediated apoptosis. We demonstrated that resveratrol was a potent inhibitor of NF-κB and AP-1 activation. It also inhibited TNF-induced c-Jun N-terminal protein kinase (JNK) and MAPK kinase (MEK) activation and caspase-induced apoptosis. Both reactive oxygen intermediate (ROI) generation and lipid peroxidation induced by TNF were also suppressed by resveratrol.
Materials and Methods
Resveratrol, penicillin, streptomycin, RPMI 1640 medium, and FCS were obtained from Life Technologies (Grand Island, NY). Glycine, PMA, LPS, ceramide, NaCl, and BSA were obtained from Sigma (St. Louis, MO). A 5 mM solution of resveratrol (m.w. 228.2) was prepared in H2O and used directly at this concentration. Bacteria-derived human rTNF, purified to homogeneity with a sp. act. of 5 × 107 U/mg, was kindly provided by Genentech (South San Francisco, CA). Abs against IκBα and p65 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Poly(ADP-ribose) polymerase (PARP) Ab was purchased from PharMingen (San Diego, CA). Phospho-IBα (Ser32) Ab was purchased from New England Biolabs (Beverly, MA). The rat MDR1bCAT plasmid −243RMICAT containing the chloramphenicol acetyltransferase (CAT) gene with either wild-type or mutated NF-κB binding site was kindly supplied by Dr. M. Tien Kuo (University of Texas M. D. Anderson Cancer Center, Houston, TX). The characterization of these plasmids has been described previously in detail (32).
The cell lines used in this study were as follows: U-937 (human histiocytic lymphoma), HeLa (human epithelial cells), H4 (glioma cells), and T-Jurkat (T cells); they were obtained from American Type Culture Collection (Manassas, VA). Cells were cultured in RPMI 1640 medium supplemented with 10% FBS, penicillin (100 U/ml), and streptomycin (100 μg/ml). All cells were free from mycoplasma, as detected by Gen-Probe mycoplasma rapid detection kit (Fisher Scientific, Pittsburgh, PA).
Isolation of PBL
Freshly drawn human blood was incubated with 2.5% gelatin in saline in 1:1 ratio for 30 min at 37°C. The supernatant was layered on Histopaque 1077 (from Sigma) and centrifuged at 1500 rpm for 30 min at room temperature. The cells were then collected from the top layer of Histopaque, diluted with Dulbecco’s PBS and centrifuged at 2000 rpm for 10 min. To get rid of mixed reticulocytes, pellet was suspended in 0.2% NaCl for 1 min, immediately diluted with equal volume of 1.6% NaCl, and centrifuged at 1000 rpm. The pellet was then suspended in RPMI 1640 medium supplemented with 10% FBS and cultured for 2 h at 37°C in a CO2 incubator in a petri dish to remove macrophages by adherence. Then the lymphocytes were harvested from the medium by centrifugation at 1000 rpm.
NF-κB activation assays
To determine NF-κB activation, EMSA were conducted essentially as described (33, 34). Briefly, nuclear extracts prepared from TNF-treated cells (2 × 106/ml) were incubated with 32P end-labeled 45-mer double-stranded NF-κB oligonucleotide (4 μg protein with 16 fmol DNA) from the HIV-LTR, 5′-TTGTTACAAGGGACTTTCCGCTGGGGACTTTCCAGGGAGGCGTGG-3′ (bold indicates NF-κB binding sites) for 15 min at 37°C, and the DNA-protein complex formed was separated from free oligonucleotide on 6.6% native polyacrylamide gels. A double-stranded mutated oligonucleotide, 5′-TTGTTACAACTCACTTTCCGCTGCTCACTTTCCAGGGAGGCGTGG-3′, was used to examine the specificity of binding of NF-κB to the DNA. The specificity of binding was also examined by competition with the unlabeled oligonucleotide. The dried gels were visualized, and radioactive bands were quantitated by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) using ImageQuant software.
AP-1 activation assay
The activation of AP-1 was determined as described (28). Briefly, 6 μg of nuclear extract prepared as indicated above was incubated with 16 fmol of the 32P end-labeled AP-1 consensus oligonucleotide 5′-CGCTTGATGACTCAGCCGGAA-3′ (bold indicates AP-1 binding site) for 15 min at 37°C and analyzed by using 6% native polyacrylamide gel. The specificity of binding was examined by competition with unlabeled oligonucleotide. Visualization and quantitation of radioactive bands were done as indicated above.
Western blot for IκBα and p65
To determine the levels of IκBα, postnuclear (cytoplasmic) extracts were prepared (33) from TNF-treated cells and resolved on 10% SDS-polyacrylamide gels. To determine the levels of NF-κB protein p65, nuclear and postnuclear extracts were prepared from TNF-treated cells and were resolved on 10% SDS-PAGE. After electrophoresis, the proteins were electrotransferred to nitrocellulose filters, probed with rabbit polyclonal Abs against IκBα or p65, and detected by chemiluminescence (ECL; Amersham, Arlington Heights, IL). The bands obtained were quantitated using Personal Densitometer Scan v1.30 using Imagequant software version 3.3 (Molecular Dynamics).
Oct-1 and Sp1 binding
The effect of resveratrol on the binding of Oct-1 and Sp1 was determined by incubating 6 μg of nuclear extracts with 16 fmol of the 32P end labeled with either Oct-1 consensus oligonucleotide 5′-TGTCGAATGCAAATCACTAGAA-3′ (bold indicates Oct-1 binding site) or Sp1 consensus oligonucleotide 5′-ATTCGATCGGGGCGGGGCGAGC-3′ for 15 min at 37°C and analyzed by using 6% native polyacrylamide gel. Visualization and quantitation of radioactive bands were done as indicated above.
Immunoprecipitation of p65 from orthophosphate-labeled cells
To determine the phosphorylation of p65 subunit of NF-κB, U-937 cells were labeled with [32P]orthophosphate (Amersham) in phosphate-free medium for 1 h at 37°C, and then resveratrol (5 μM) was added and incubation continued for another 2 h at 37°C. Then cells were washed and suspended with same medium. Cells were then treated with 0.1 nM TNF for 30 min at 37°C. The cells were washed with Dulbecco’s PBS and then lysed on ice for 15 min with buffer containing 20 mM Tris-Cl, pH 7.9, 2 mM EDTA, 250 mM NaCl, 0.1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, 1 mM DTT, and 1 mM sodium orthovanadate. An 800-μg protein was immunoprecipitated with 0.5 μg anti-p65 polyclonal Ab (Santa Cruz Biotechnology) overnight at 4°C. Immune complexes were collected by incubation with protein A/G Sepharose beads for 1 h at 4°C. The beads were extensively washed with lysis buffer (4 × 400 μl) and wash buffer (2 × 400 μl: 20 mM HEPES, pH 7.4, 1 mM DTT, 25 mM NaCl). Washed beads were then boiled with 15 μl of 2× SDS sample buffer for 5 min, and subjected to SDS-PAGE (9%). The gel was dried, exposed to phospho-screen, and analyzed by a PhosphorImager (Molecular Dynamics). To determine equal loading, 50-μg protein was resolved on 10% SDS-PAGE, electrotransferred to nitrocellulose filters, and probed with the anti-p65 Ab, and the bands were detected by chemiluminescence (ECL; Amersham).
The TNF-induced cytotoxicity was measured by the MTT assay (35). Briefly, cells (10,000 cells/well) were incubated in the presence or absence of the indicated test sample in a final volume of 0.1 ml for 72 h at 37°C. Thereafter, 0.025 ml of MTT solution (5 mg/ml in PBS) was added to each well. After a 2-h incubation at 37°C, 0.1 ml of the extraction buffer (20% SDS, 50% dimethylformamide) was added. After an overnight incubation at 37°C, the OD at 590 nm were measured using a 96-well multiscanner autoreader (Dynatech MR 5000, Chantilly, VA), with the extraction buffer as a blank.
Immunoblot analysis of PARP degradation
TNF-induced apoptosis was examined by proteolytic cleavage of PARP (35). Briefly, cells (2 × 106/ml) were treated with different concentrations of resveratrol at 37°C for 1 h and then stimulated with 1 mM TNF with cycloheximide (2 μg/ml) for 2 h at 37°C. The cells were then washed and extracted by incubation for 30 min on ice in 0.05 ml buffer containing 20 mM HEPES, pH 7.4, 2 mM EDTA, 250 mM NaCl, 0.1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, 1 mM DTT, and 1 mM sodium vanadate. The lysate was centrifuged and the supernatant was collected. Cell extract protein (50 μg) was resolved on 7.5% SDS-PAGE, electrotransferred onto a nitrocellulose membrane, blotted with mouse anti-PARP Ab, and then detected by chemiluminescence (ECL; Amersham). Apoptosis was represented by the cleavage of 116-kDa PARP into a 85-kDa product (36).
Activation of MEK was assayed as described (37). U-937 cells, treated with different concentrations of resveratrol for 1 h and then stimulated with 1 nM TNF for 30 min at 37°C, were washed with Dulbecco’s PBS and then lysed on ice for 15 min with buffer containing 20 mM HEPES, pH 7.4, 2 mM EDTA, 250 mM NaCl, 0.1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, 1 mM DTT, and 1 mM sodium orthovanadate. A 50-μg aliquot of protein was resolved on 10% SDS-PAGE, electrotransferred to nitrocellulose filters, and probed with the phospho-specific anti-p44/42 MAPK (Thr202/Tyr204) Ab (New England Biolabs) raised in rabbits (1/3000 dilution). Then the membrane was incubated with peroxidase-conjugated anti-rabbit IgG (1/3000 dilution), and the bands were detected by chemiluminescence (ECL; Amersham).
c-Jun kinase assay
The c-Jun kinase assay was performed by a modified method, as described earlier (31). Briefly, after treatment of cells (3 × 106/ml) with TNF for 10 min, cell extracts were prepared by lysing cells in buffer containing 20 mM HEPES, pH 7.4, 2 mM EDTA, 250 mM NaCl, 1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, and 1 mM DTT. Cell extracts (150 μg/sample) were immunoprecipitated with 0.3 μg anti-JNK Ab for 60 min at 4°C. Immune complexes were collected by incubation with protein A/G Sepharose beads for 45 min at 4°C. The beads were extensively washed with lysis buffer (4 × 400 μl) and kinase buffer (2 × 400 μl: 20 mM HEPES, pH 7.4, 1 mM DTT, 25 mM NaCl). Kinase assays were performed for 15 min at 30°C with GST-Jun 1–79(1–79) as a substrate in 20 mM HEPES, pH 7.4, 10 mM MgCl2, 1 mM DTT, and 10 μCi [γ-32P]ATP. Reactions were stopped by the addition of 15 μl of 2× SDS sample buffer, boiled for 5 min, and subjected to SDS-PAGE (9%). GST-Jun 1–79(1–79) was visualized by staining with Coomassie blue, and the dried gel was analyzed by a PhosphorImager (Molecular Dynamics).
Transient transfection and CAT assay
To determine TNF-induced NF-κB-mediated reporter gene transcription, U-937 cells were transiently transfected by the calcium phosphate method with the plasmids 243RMICAT (contains wild-type NF-κB binding site) and −243 RMICAT-km (mutated binding site), according to the instructions supplied by the manufacturer (Life Technologies). After 12 h of transfection, the cells were stimulated with different concentrations of TNF for 2 h, washed, and examined for CAT activity, as described (38).
Determination of lipid peroxidation
TNF-induced lipid peroxidation was determined by detection of thiobarbituric acid-reactive malondialdehyde (MDA), an end product of the peroxidation of polyunsaturated fatty acids and related esters, as described (39). Results were normalized with the amount of MDA equivalents/mg of protein and expressed as a percentage of thiobarbituric acid-reactive substances above control values. Untreated cells showed 0.571 ± 0.126 nmol of MDA equivalents/mg of protein.
Measurement of ROI
The production of ROI upon treatment of cells with TNF was determined by flow cytometry, as described (39).
In this study, we examined the effect of resveratrol on TNF-induced signal transduction. The chemical structure of resveratrol is shown in Fig. 1. It is a highly water-soluble compound. For most studies, U-937 cells were used because these cells express both types of TNF receptor, and TNF-induced responses in this cell type are well characterized in our laboratory. The concentration of resveratrol and its time of exposure had no effect on TNF receptors or on cell viability (data not shown).
Resveratrol inhibits TNF-induced NF-κB activation
U-937 cells were pretreated for 4 h with different concentrations of resveratrol and then stimulated with 100 pM TNF for 30 min. Nuclear extracts were prepared and assayed for NF-κB by EMSA. As shown in Fig. 2,A, TNF induced 10-fold activation of NF-κB, and resveratrol inhibited this activation in a dose-dependent manner; full inhibition occurred at 5 μM resveratrol. Resveratrol even at 25 μM by itself did not activate NF-κB. We next examined the effect of changes in the length of incubation with resveratrol on NF-κB activation by TNF. Cells were incubated with 5 μM resveratrol for different times and then stimulated with 0.1 nM TNF for 30 min and assayed for NF-κB. The results in Fig. 2 B show that resveratrol inhibited TNF-induced NF-κB activation with increased time of incubation. At 4 h, complete inhibition was observed.
Previous studies from our laboratory have shown that a high concentration of TNF (10 nM) can activate NF-κB within 5 min, and this induction is higher in its intensity than that obtained with cells using a 100-fold lower concentration of TNF for longer times (40). To determine the effect of resveratrol on NF-κB activation at even higher concentrations, both untreated and resveratrol-pretreated cells were incubated with various concentrations of TNF (0–10,000 pM) for 30 min and then assayed for NF-κB by EMSA. Although the activation of NF-κB by 10,000 pM TNF was strong (Fig. 2,C), resveratrol completely inhibited it as efficiently as it did at 0.1 nM concentration. These results show that resveratrol is a very potent inhibitor of NF-κB activation. We also examined the effect of resveratrol on the kinetics of TNF-induced NF-κB activation. Both untreated and resveratrol-pretreated cells were incubated with TNF (100 pM) for different times and then assayed for NF-κB. In untreated cells, TNF activated NF-κB in a time-dependent manner with almost maximum activation at 15 min (Fig. 2,D). In resveratrol-pretreated cells, however, a little activation of NF-κB was detected after TNF exposure of up to 60 min (Fig. 2 D).
Activated NF-κB inhibited by resveratrol consists of p50 and p65 subunits
Various combinations of Rel/NF-κB proteins can constitute an active NF-κB heterodimer that binds to specific sequences in DNA. To show that the retarded band visualized by EMSA in TNF-treated cells was indeed NF-κB, we incubated the nuclear extracts from TNF-activated cells with Ab to either p50 (NF-κBI) or p65 (Rel A) subunits and then conducted EMSA. Abs to either subunit of NF-κB shifted the band to a higher m.w. (Fig. 3 A), thus suggesting that the TNF-activated complex consisted of p50 and p65 subunits. Neither preimmune serum nor such irrelevant Abs as anti-c-Rel or anticyclin DI had any effect on the mobility of NF-κB. Excess cold NF-κB (100-fold) almost completely eradicated the band, indicating the specificity of NF-κB. Further specificity is indicated by the observations that the oligonucleotide probe with labeled mutated NF-κB binding site failed to bind the NF-κB protein.
Resveratrol does not interfere with the DNA-binding ability of NF-κB proteins
It has been shown that N-tosyl-l-Phe-chloromethylketone (TPCK), a serine protease inhibitor, and herbimycin A, a protein tyrosine kinase inhibitor, and caffeic acid phenylethyl ester down-regulate NF-κB activation by chemical modification of the NF-κB subunits, thus preventing its binding to DNA (41, 42, 43). To determine whether resveratrol also directly modifies the ability of NF-κB proteins to bind to the DNA, we incubated the cytoplasmic extracts with deoxycholate (DOC) (0.8%) for 15 min at room temperature. The DOC treatment has been shown to dissociate the IκBα subunit, thus releasing NF-κB for binding to the DNA. DOC-treated cytoplasmic extracts were then exposed to various concentrations of resveratrol and assayed for DNA binding by EMSA. As shown in Fig. 3 B, resveratrol had no effect on the binding of NF-κB to the DNA.
Whether resveratrol modifies the nuclear fraction of NF-κB in TNF-treated cells was also examined. The nuclear extracts from TNF-pretreated cells were treated with various concentrations of resveratrol and then examined for DNA-binding activity by EMSA. Our results in Fig. 3 C show that resveratrol did not modify the DNA-binding ability of NF-κB proteins prepared from TNF-treated cells either. Therefore, resveratrol inhibits NF-κB activation through a mechanism different from that of TPCK, herbimycin A, and caffeic acid phenylethyl ester (41, 42, 43).
Whether resveratrol suppresses the DNA binding of other transcription factors, such as Oct-1 and Sp1, was also examined. As shown in Fig. 3, D and E, resveratrol has no effect on Oct-1 or Sp1, respectively, indicating that the effects of resveratrol are specific to NF-κB.
Inhibition of NF-κB activation by resveratrol is not cell type specific
That distinct signal transduction pathways could mediate NF-κB induction in epithelial and lymphoid cells has been demonstrated (44). All the effects of resveratrol described above were conducted with U-937, a myeloid cell line. In another set of experiments, we found that resveratrol blocks TNF-induced NF-κB activation in T cells (Jurkat) and epithelial (HeLa) and glioma (H4) cells (Fig. 4). NF-κB binding in all three cell lines was abrogated by a 25-fold molar excess of unlabeled oligonucleotide. An almost complete inhibition in all the cell types suggests that this effect of resveratrol is not restricted to myeloid cells.
Resveratrol blocks phorbol ester-, LPS-, okadaic acid-, ceramide-, and H2O2-mediated activation of NF-κB
Besides TNF, NF-κB is also activated by various other tumor promoters and inflammatory agents, including phorbol ester, H2O2, LPS, okadaic acid, and ceramide (22), but by different signal transduction pathways (44, 45, 46). We found that these five agents activated NF-κB and that resveratrol completely blocked the activation of NF-κB induced by all five inducers (Fig. 5 A). These results suggest that resveratrol may act at a step in which all these agents converge in the signal transduction pathway leading to NF-κB activation.
All the experiments described were performed with the tumor cell lines. Whether resveratrol also affects NF-κB in normal cells was examined. As shown in Fig. 5 B, PMA, LPS, PHA, and TNF, all agents activated NF-κB in normal human PBL, and the pretreatment with resveratrol (5 μM) abolished the activation.
Resveratrol does not inhibit TNF-dependent phosphorylation and degradation of IκBα
The translocation of NF-κB to the nucleus is preceded by the phosphorylation, ubiquitination, and proteolytic degradation of IκBα (22). To determine whether the inhibitory action of resveratrol was due to an effect on IκBα degradation, the cytoplasmic level of IκBα proteins was examined by Western blot analysis. IκBα degradation started 5 min after TNF treatment of U937 cells and was complete within 10 min. The band reappeared by 30 min owing to NF-κB-dependent IκBα resynthesis. The presence of resveratrol had no significant effect on the TNF-induced IκBα degradation (Fig. 6 A).
To determine whether resveratrol modulates TNF-induced IκBα phosphorylation, cells were treated with the proteosome inhibitor N-acetylleucyl-leucylnorleucinal (42) for 1 h and then assayed by Western blot with Abs against either the serine-phosphorylated (Fig. 6,B, upper panel) or nonphosphorylated form of IκBα (Fig. 6,B, lower panel). Resveratrol had neither any effect on the TNF-induced phosphorylation of IκBa (upper panel), nor on the migration of the hyperphosphorylated form of IκBα, which appeared as a slow-migrating band on SDS-PAGE (Fig. 6 B, lower panel) in TNF-treated cells.
Resveratrol inhibits TNF-dependent phosphorylation and nuclear translocation of p65 subunit of NF-κB
Whether resveratrol affects the TNF-induced nuclear translocation of the p65 subunit of NF-κB was also examined by Western blot analysis. As shown in Fig. 6 C, upon TNF treatment, p65 disappeared from the cytoplasm, and resveratrol prevented the disappearance. In the nuclear fraction, however, p65 appeared after TNF treatment and resveratrol inhibited the appearance. Resveratrol alone had no effect in these experiments. These results indicate that resveratrol blocks the nuclear translocation of NF-κB.
Recently, it was reported that mesalamine inhibits IL-1-induced NF-κB activation by blocking the phoshorylation of p65 subunit (47). Whether resveratrol affects the TNF-induced phosphorylation of the p65 subunit of NF-κB was also examined by metabolic labeling of cells with [32P]orthophosphate, followed by immunoprecipitation of p65 from labeled cells treated with either TNF or resveratrol or combination. As shown in Fig. 6 D, TNF induced the phosphorylation of the p65 subunit and resveratrol inhibited it. Resveratrol alone had no effect in these experiments. These results indicate that resveratrol also blocks the phosphorylation of p65 subunit of NF-κB.
Resveratrol represses TNF-induced NF-κB-dependent reporter gene expression
Although we have shown by EMSA that resveratrol blocks the NF-κB activation and blocks the phosphorylation and nuclear translocation of p65, DNA binding alone does not always correlate with NF-κB-dependent gene transcription, suggesting the role of additional regulatory steps (48). To determine the effect of resveratrol on TNF-induced NF-κB-dependent reporter gene expression, we transiently transfected resveratrol-pretreated or untreated cells with the CAT reporter construct and then stimulated with TNF. An almost 6-fold increase in CAT activity over the vector control was noted upon stimulation with TNF (Fig. 7). The CAT gene reporter construct with mutated NF-κB could not be activated by TNF, suggesting specificity of action. TNF-induced CAT activity was almost completely abolished when the cells were pretreated with resveratrol. These results demonstrate that resveratrol also represses NF-κB-dependent reporter gene expression induced by TNF.
Resveratrol inhibits TNF-induced c-Jun kinase and MEK activation
TNF is one of the most potent activators of various kinases of the MAPK family (49). There are also reports that some of the kinases of this family are required for TNF-induced NF-κB activation (50). And TNF is known to be a potent activator of JNK (51). Whether resveratrol affects any of these kinases was also examined. The U-937 cells were pretreated with different concentrations of resveratrol for 4 h and then stimulated with TNF (1 nM) for 10 min. About a 17-fold activation of c-jun kinase was detected with 1 nM TNF. This activation gradually decreased with increasing concentrations of resveratrol, and at 5 μM resveratrol the activation of JNK by TNF was completely inhibited (Fig. 8 A).
The activation of JNK is regulated by an upstream dual specificity kinase, referred to as MAPK kinase (also called MEK). To determine whether resveratrol inhibits this kinase, U-937 cells were pretreated with different concentrations of resveratrol for 4 h and then stimulated with 1 nM TNF for 30 min. The phosphorylated form of MAPK was then assayed. We found that resveratrol inhibited the TNF-induced activation of MEK in a dose-dependent manner, with maximum suppression occurring at 5 μM resveratrol (Fig. 8 B).
Resveratrol inhibits TNF-induced AP-1 activation
The activation of JNK causes the activation of AP-1. TNF is also a potent activator of AP-1 (52). TNF induced AP-1 expression by 7-fold in myeloid cells at 1 nM concentration. The activation of AP-1 was completely inhibited by resveratrol in a dose-dependent manner, with maximum suppression occurring at 5 μM (Fig. 9,A). Supershift analysis with specific Abs against c-fos and c-jun indicated that TNF-induced AP-1 consists of c-fos and c-jun (data not shown). Lack of supershift by unrelated Abs and disappearance of the AP-1 band by competition with cold oligo show the specificity. In untreated cells, TNF activated AP-1 in a dose-dependent manner, but in resveratrol-treated cells, no AP-1 activation was observed (Fig. 9 B).
Resveratrol blocks TNF-induced cytotoxicity and caspase activation
Among all the cytokines, TNF is one of the most potent inducers of apoptosis (for references, see 53). Whether resveratrol modulates TNF-induced apoptosis was also investigated. U-937 cells were treated with variable concentrations of TNF for 72 h either in the absence or presence of resveratrol and then examined for cytotoxicity by the MTT method. Results in Fig. 10 A show that the cytotoxic effects of TNF in U-937 cells were dose dependent, with almost 70% killing occurring at 5 nM concentration of the cytokine. This cytotoxicity was completely inhibited by treatment of cells with 5 μM resveratrol.
Because the cytotoxic effects of TNF are mediated through the activation of caspases, we also examined the effect of resveratrol on TNF-induced caspase activation. Activated caspase-2, -3, and -7 are known to cleave PARP protein. As shown in Fig. 10 B, TNF induced complete cleavage of PARP, and this cleavage was inhibited in a dose-dependent manner by treatment of cells with resveratrol, with maximum effect at 3 μM concentration. Thus, resveratrol also blocks TNF-induced apoptosis.
Resveratrol blocks TNF-induced ROI generation and lipid peroxidation
Previous reports have shown that TNF activates NF-κB, AP-1, JNK, and apoptosis through generation of ROI (28, 29, 30, 31, 53, 54). Whether resveratrol mediates its effects through suppression of ROI production was examined by flow cytometry. As shown in Fig. 11,A, TNF induced ROI generation in a time-dependent manner, but this was suppressed by pretreatment of cells with resveratrol. Because lipid peroxidation has also been implicated in TNF-induced NF-κB activation and cytotoxicity (53, 55), we also examined the effect of resveratrol on TNF-induced lipid peroxidation. Results in Fig. 11 B show that TNF induced lipid peroxidation in U-937 cells, and this was completely suppressed by resveratrol. Thus, it is quite likely that resveratrol blocks TNF signaling through suppression of ROI generation and of lipid peroxidation.
Because several in vitro and in vivo activities assigned to resveratrol require suppression of NF-κB activation, we tested the hypothesis that resveratrol directly blocks NF-κB activation. We found that resveratrol is indeed a potent inhibitor of TNF-induced activation of NF-κB, and this inhibition is not cell type specific. The suppression is observed in both normal and tumor cells. Besides TNF, resveratrol also blocked NF-κB activation induced by a wide variety of other inflammatory agents. NF-κB-dependent reporter gene transcription was also suppressed by resveratrol. Besides NF-κB, resveratrol blocked activation of AP-1 and the associated kinases MEK and JNK. TNF-induced cytotoxicity and caspase activation were also down-regulated by resveratrol. Resveratrol’s ability to block both ROI generation and lipid peroxidation induced by TNF may account for its effects on transcription factors and the associated kinases.
Recent evidence indicates that different inflammatory agents may activate NF-κB through mechanisms that consist of some overlapping and some nonoverlapping steps (44, 45, 46). How resveratrol blocks NF-κB activation by TNF is not clear. Its suppression of NF-κB activation by a wide variety of agents suggests that resveratrol must act at a step common to all agents. Most inhibitors of NF-κB activation, such as curcumin and silymarin, mediate their effects through suppression of phosphorylation and degradation of IκBa (39, 56, 57). Resveratrol, however, blocked neither the phosphorylation nor the degradation of IκBa. These results are similar to that described for caffeic acid phenethyl ester or mesalamine, which also block NF-κB activation without any effect on IκBα phosphorylation or degradation (43, 47). Caffeic acid phenethyl ester, however, modifies the NF-κB protein so that it can no longer bind to DNA. Resveratrol had no effect on the binding of NF-κB proteins to the DNA, but it did block the TNF-induced translocation of NF-κB’s p65 subunit and reporter gene transcription. These results are similar to those described recently for mesalamine, which inhibits cytokine-induced and NF-κB-dependent gene expression without degrading IκBα (47). Egan et al. (47) reported that mesalamine did not suppress nuclear translocation of p65. In contrast, resveratrol did block p65 translocation, which may explain how it suppresses reporter gene expression.
We found that resveratrol blocks TNF-induced phosphorylation of p65, which is in agreement with the results of Egan et al. (47), who showed suppression of IL-1-induced phosphorylation of p65 by mesalamine. Several kinases have been implicated that could phosphorylate p65, including protein kinase A and IKK (47, 58 and references therein). Because IKK that phosphorylates IκBα can also phosphorylate p65 (58) and IκBα phosphorylation is unaffected, our results indicate that IKK is not inhibited by resveratrol.
Resveratrol also blocked TNF-induced AP-1 activation. The mechanism of activation of NF-κB and AP-1 is very similar. Most agents that activate NF-κB also activate AP-1. Similarly, agents that suppress NF-κB also suppress AP-1 (28, 29). The activation of AP-1 requires the activation of JNK and the upstream kinase MEK. Both of these kinases were inhibited by resveratrol, which may explain the mechanism of suppression of AP-1.
TNF-induced cyotoxicity and caspase activation were also blocked by resveratrol. Because NF-κB activation has been shown to play an antiapoptotic role (59), the suppression of apoptosis by resveratrol may seem paradoxical. However, NF-κB activation does not block apoptosis induced by all the agents (28). The overexpression of the antioxidant enzymes manganous superoxide dismutase or γ-glutamyl cysteinyl synthetase has been shown to suppress TNF-induced apoptosis and NF-κB (28, 29), suggesting that the mechanisms of activation of apoptosis and NF-κB are very similar. Our discovery that resveratrol blocks TNF-induced ROI generation and lipid peroxidation explains the mechanism by which resveratrol exerts its effects. The antioxidant properties of resveratrol have been previously reported (8, 9).
Resveratrol also blocked TNF-induced NF-κB-mediated gene transcription. Previously, it has been shown that PMA-induced COX-2 is blocked by resveratrol (16). This gene is known to be regulated by NF-κB activation (60, 61). NO synthase gene is also regulated by NF-κB (62). Thus, it is possible that resveratrol suppresses COX-2 and NO synthase expression by inhibiting of NF-κB activation. Besides COX-2, various other genes, including those for matrix metalloproteinase-9 (MMP-9) and cell surface adhesion molecules (e.g., ICAM-1, endothelial leukocyte adhesion molecule 1 (ELAM-1), and VCAM-1), are also regulated by NF-κB (63, 64, 65). Urokinase-type plaminogen activator, whose gene is regulated by NF-κB (66), is also involved in tumor growth and metastasis (67). All these proteins have been implicated in carcinogenesis (68). It is possible that the anticarcinogeneic properties assigned to resveratrol (14, 15) are due to the suppression of NF-κB-mediated expression of the genes for these enzymes and adhesion molecules. For instance, high COX-2 expression has been associated with cancer progression and inhibition of apoptosis, and antioxidants reduce COX-2 expression, prostaglandin production, and proliferation in colorectal cancer cells (69). Due to its ability to suppress COX-2 through NF-κB, aspirin is beneficial for preventing colon cancer (70). This suggests that resveratrol may also prove to be beneficial for colon cancer. By using TNF-deficient mice, it was shown that TNF is required for tumor promotion (23), thus suggesting its role in carcinogenesis, the role of JNK in TNF-induced cellular transformation, has been documented (71). Thus, resveratrol’s ability to suppress TNF-induced NF-κB, JNK, AP-1, and other cellular responses may provide the molecular basis for the anticarcinogenic properties of resveratrol. Recently, resveratrol was also found to inhibit the expression and function of androgen receptors in prostate cancer cells (72). In addition, adenovirus-enforced overexpression of mitochondrial superoxide dismutase gene therapy has been used to treat ischemia/reperfusion injury of the liver through the down-regulation of NF-κB and AP-1 activation (73). Our results indicate that suppressive effects of resveratrol on NF-κB and AP-1 activation and on other TNF-mediated cellular responses may also explain its protective effects on liver and against cardiovascular diseases.
Our results indicate that 5 μM resveratrol is sufficient to suppress most of the TNF-mediated cellular responses by greater than 90%. Previous studies have shown that to block progression of carcinogenesis and to induce terminal differentiation by 50%, 19 μM resveratrol is required (14). Similarly, 98% inhibition of DMBA plus phorbol ester-induced skin tumors in mice occurred by a topical application of 25 μM resveratrol (14). Thus, concentrations used in our studies are comparable with that used in animal studies. Considering that each gram of fresh grape skin contains 50–100 μg (200–400 μM) resveratrol and the red wine has 1.5–3 mg/L (5, 14), this suggests that resveratrol concentration used in our studies is achievable in vivo by consumption of grapes or wine.
This research was conducted by the Clayton Foundation for Research.
Abbreviations used in this paper: resveratrol, trans-3,4′,5-trihydroxystilbene; DMBA, dimethylbenz(a)anthracene; CAT, chloramphenicol acetyltransferase; COX, cyclooxygenase; DOC, deoxycholate; ECL, enhanced chemiluminescence; IκB, inhibitory subunit of NF-κB; IKK, IκB kinase; JNK, c-Jun N-terminal protein kinase; MAPK, mitogen-activated protein kinase; MDA, malondialdehyde; MEK, MAPK kinase; PARP, poly(ADP-ribose) polymerase; ROI, reactive oxygen intermediate; TPCK, N-tosyl-l-Phe-chloromethylketone.