Activation of the vascular endothelium, increased adhesion of mononuclear cells to the injured endothelial layer, and their subsequent extravasations into the vessel wall are initial events in atherogenesis. Endothelial cells recruit leucocytes by expressing adhesion molecules, such as vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1 (ICAM-1) and endothelial leucocyte adhesion molecules (E-selectin; Cybulsky & Gimbrone, Reference Cybulsky and Gimbrone1991). Several inflammatory cytokines including IL-1, TNF and interferon produced by activated monocytes and macrophages may stimulate the endothelium to up-regulate genes encoding chemokines, other cytokines and adhesion molecules (Ross, Reference Ross1999).
Previous studies have indicated that NF-κB/Rel transcription factors may play an important role in the development of atherosclerosis (Collins, Reference Collins1993; Qwarnstrom et al. Reference Qwarnstrom, Ostberg, Turk, Richardson and Bomsztyk1994). Activation of NF-κB by inflammatory stimuli has been demonstrated in cultures of endothelial cells using electrophoretic mobility shift assays (Collins, Reference Collins1993). A variety of genes induced in the atherosclerotic lesion have been shown to be regulated by NF-κB proteins, including the genes encoding TNF-α (Baeuerle & Henkel, Reference Baeuerle and Henkel1994), IL-1β (Hiscott et al. Reference Hiscott, Marois and Garoufalis1993), VCAM-1 (Neish et al. Reference Neish, Williams, Palmer, Whitley and Clollins1992) and ICAM-1 (Poston et al. Reference Poston, Haskard, Coucher, Gall and Johnson-Tidey1992).
It is well established that dietary polyphenolic compounds play significant roles in the prevention of atherosclerosis and CVD (Gaziano et al. Reference Gaziano, Manson, Buring and Hennekens1992; Gey et al. Reference Gey, Stahelin and Eichholser1993). Polyphenolic compounds affect the development of atherosclerosis not only through modulation of serum lipids but also by influencing the immune and inflammatory processes associated with the development of this disease. Previous studies have indicated that polyphenolic compounds such as vitamin E or tea flavonoid may exert their effects through modulation of cytokines, adhesion molecules and interaction of immune cells with endothelial cells (Martin et al. Reference Martin, Foxall, Blumberg and Meydani1997; Islam et al. Reference Islam, Devaraj and Jialal1998; Ludwig et al. Reference Ludwig, Lorenz, Grimbo, Steinle, Meiners, Bartsch, Stangl, Baumann and Stangl2004). Ellagic acid (EA) is a phenolic compound found in fruits including grape juice (10·2 mg/100 g), grape wine (5·6 mg/100 g), blueberries (0·9 mg/100 g), blackberries (42·4 mg/100 g), raspberries (17·9 mg/100 g) and strawberries (19·8 mg/100 g) (de Ancos et al. Reference de Ancos, Gonzalez and Cano2000; Sellappan et al. Reference Sellappan, Akoh and Krewer2002; Mertens-Talcott et al. Reference Mertens-Talcott, Talcott and Percival2003). Previous studies have indicated that EA scavenges both oxygen and hydroxyl radicals, and inhibits lipid peroxidation (Cozzi et al. Reference Cozzi, Ricordy, Bartolini, Ramadori, Perticone and De Salvia1995; Laranjinha et al. Reference Laranjinha, Vierira, Almeida and Madeira1996; Iino et al. Reference Iino, Nakahara, Miki, Kiso, Ogawa, Kato and Takeuchi2001). In our laboratory, we found that EA reduced oxidative stress and atherosclerosis in a hyperlipidaemic rabbit model (Yu et al. Reference Yu, Chang, Wu and Chiang2005). Therefore, the present study was designed to examine the effect of EA on monocyte adhesion to cultured human endothelial cells and the expression of adhesion molecules (VCAM-1, ICAM-1 and E-selectin) and to elucidate its possible mechanism.
Human umbilical vein endothelial cells (HUVEC) were isolated by collagenase type II (Biochrom KG, Berlin, Germany) digestion of human umbilical veins by standard techniques and cultured in EC medium (MCDB 131; Gibco-BRL, Life Technologies GmbH, Karlsruhe, Germany) at 37°C in a humidified atmosphere of 5 % CO2 and 95 % air as described previously (Stangl et al. Reference Stangl, Gunther, Jarrin, Bramlage, Moobed, Staudt, Baumann and Stangl2001). All experiments were performed with HUVEC from passages one to three. HUVEC were seeded at 1 × 104 cells/well in ninety-six-well plates. After 3 d, the medium was replaced by fresh EC medium before treatment.
Cell viability assay using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide
Cell viability was measured by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay (Chen et al. Reference Chen, Lin, Chen, Ku and Chen2002). The principle of this assay is that mitochondria dehydrogenase in viable cells reduces MTT to a blue formazan. Briefly, cells were grown in ninety-six-well plates and incubated with various concentrations of EA (which was dissolved in dimethyl sulphoxide) for 24 h; 100 μl MTT (0·5 mg/ml) were then added to each well and incubation continued at 37°C for an additional 4 h. The medium was then carefully removed, so as not to disturb the formazan crystals which had formed. Dimethyl sulphoxide (100 μl), which solubilizes formazan crystals, was added to each well and the absorbance of the solubilized blue formazan was read at 530 nml/l (reaction) and 690 nml/l (background) using a DIAS Microplate Reader (Dynex Technologies, Chantilly, VA, USA). The reduction in optical density caused by EA was used as a measurement of cell viability, normalized to cells incubated in control medium, which were considered 100 % viable.
Estimation of the production of reactive oxygen species
The production of intracellular reactive oxygen species (ROS) induced by IL-1β was estimated by a fluorometric assay using 2′,7′-dichlorofluorescein-diacetate (DCFH-DA) as a probe according to the method reported by Bass et al. (Reference Bass, Parce, Decharelet, Szejda, Seeds and Thomas1983). HUVEC (1 × 106) were incubated with IL-1β and EA, suspended in PBS containing 2 % fetal calf serum, and then incubated again with 5 mmol/l 2′,7′-dichlorofluorescein-diacetate for 30 min at 37°C. The formation of 2′,7′-dichlorofluorescein was determined by flow cytometry. The excitation wavelength was 488 nm, and green fluorescence collected through a 530 nm band-pass filter was measured on a logarithmic scale. The formation of ROS was expressed as relative fluorescence intensity.
Real-time PCR for vascular cell adhesion molecule-1, intercellular adhesion molecule-1 and endothelial leucocyte adhesion molecule
The real-time PCR assay for adhesion molecules was conducted according to the method reported by Li & Wang (Reference Li and Wang2002). Total cellular RNA was isolated from samples (HUVEC) using the Trizol reagent according to the manufacturer's instructions (Gibco BRL). RT reactions were carried out for each RNA sample in thin-welled PCR tubes using the First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA). Each reaction tube contained 4·2 μg total RNA in a volume of 21 μg containing 1 × RT buffer, 5·5 mmol/l MgCl2, 500 μmol/l of each dNTP, 2·5 μmol/l of oligo-d(T)12–18 primers, 40 U/μl RNase inhibitor, 2 U/μl Escherichia coli RNaseH and 50 U/μl of SuperScript II RT. RT reaction was carried out at 65°C for 5 min, 42°C for 50 min and 70°C for 15 min. The RT reaction mixture was then placed at 4°C for immediate PCR amplification or stored at − 20°C for later use. Real-time PCR was performed in optical real-time PCR tubes. The following primers were used: VCAM-1: forward 5′-AAGCGGAGACAGGAGACAC-3′, reverse 5′-TGGCAGGTATTATTAAGGAGGATG-3′; ICAM-1: forward 5′-TGGTTCACAGGTTCAGATTAC-3′, reverse 5′-GACAAGAGGACAAGGCATAGC-3′; E-selectin: forward 5′-TGTGAGATGCGATGCTGTC-3′, reverse 5′-AACCTCTTCTGTCCATTGTCC-3′; glyceraldehyde-3-phosphate dehydrogenase: forward 5′-CCCACTCCTCCACCTTTG-3′, reverse 5′-CTTCCTCTTGTGCTCTTGC-3′.
Each tube contained 1 μl of each RT product (200 ng total RNA), 5·5 mmol/l MgCl2, 400 μm-dNTP, 500 nmol/l primer (forward and reverse), 0·005 U/μl iTaq DNA polymerase and 20 nmol/l SYBR Green I in a total volume of 25 μl. Amplification conditions were 3 min at 95°C for activation, then run for forty cycles at 95°C for 15 s and 60°C for 1 min. All reactions were performed in the Bio-Rad iCycle Sequence Detection System using the iCycle V3.1 program. The threshold cycle (Ct) and melting point (Mt) were obtained during each reaction. The relative quantification was calculated based on its 2− ΔCt value, ΔCt = Ct (sample) – Ct (control).
Measurement of NF-κB activity
Nuclear extracts were prepared as described previously (Dschietzig et al. Reference Dschietzig, Richter, Pfannenschmidt, Bartsch, Laule, Baumann and Stangl2001). Bradford reagent determined protein concentrations. For analysis of NF-κB activity, a TransAM NF-κB Family kit was used (Active Motif, Rixensart, Belgium). In this assay, ninety-six-well plates were coated with an oligonucleotide containing the consensus binding sequence for NF-κB 5′-GGGACTTTCC-3′. Specific primary antibodies included in the kit detected the binding of NF-κB family transcription factors to their consensus sequence. Experiments were analysed by an ELISA-based assay. A total of 10 μg nuclear extract was used in each experiment and processed according to the manufacturer's protocol. Briefly, nuclear extracts were incubated with the oligonucleotide-coated wells for 60 min. Where indicated a competitor for NF-κB binding (NF-κB wild-type consensus oligonucleotide) was added in molar excess prior to the probe. The wells were then washed and incubated with the primary antibodies for p65, p50, c-Rel, p52 and RelB for 60 min. After incubation with a horseradish peroxidase-conjugated secondary antibody, a substrate was added to produce blue colour and then for quantitation by a standard ELISA reader. The absorbance was read at 450 nm and the blanks were subtracted from all measurements. The data presented are the result of three independent experiments.
Monocyte-endothelial cell adhesion
HUVEC (2 × 105) were distributed into six-well plates and allowed to reach confluence. They were then incubated for 18 h with medium supplemented with EA at concentrations of 25 and 50 μmol/l according to the MTT test, followed by incubation for 6 h with 10 ng/ml IL-1β in the continued presence of EA. U937 cells, originally derived from a human histiocytic lymphoma and used for the monocyte-endothelial cell adhesion assay, were grown in RPMI-1640 medium (Gibco, New York, USA) containing 10 % fetal bovine serum and subcultured at a 1 : 5 ratio three times per week, labelled for 30 min at 37°C with calcein AM (10 nmol/l; Molecular Probe; Invitrogen) in RPMI-1640 medium and washed with PBS to remove free dye, and then resuspended in 10 % M-199 medium. Labelled U937 cells (1 × 106) were added to each HUVEC-containing well and incubated for 1 h. Non-adherent cells were removed by two gentle washes with PBS. Then, adherent U937 cells were determined by a fluorescence plate reader at an excitation wavelength of 485 nm and emission at 530 nm; HUVEC cell monolayers served as the blank.
Results are presented as means and standard deviations. Statistical significance was determined by one-way ANOVA. Differences were considered significant at P < 0·05.
Concentrations of ellagic acid for human umbilical vein endothelial cells
Cell viability was assayed by the MTT test. After 24 h incubation with 10, 25, 50, 75, 100 and 150 μmol/l EA, cell viability was 125·2 (sd 4·9), 122·3 (sd 4·4), 106·7 (sd 2·7), 91·3 (sd 1·8), 75·8 (sd 1·4) and 78·4 (sd 1·4) %, respectively, of control levels, the three highest concentrations causing a significant reduction in cell viability. Therefore, according to the MTT test we chose 25 and 50 μmol/l to do all the experiments.
Ellagic acid inhibits IL-1β-induced reactive oxygen species production in human umbilical vein endothelial cells
Ellagic acid inhibits IL-1β-induced cell surface expression of vascular cell adhesion molecule-1 and endothelial leucocyte adhesion molecule but not expression of intercellular adhesion molecule-1 in human umbilical vein endothelial cells
The effects of EA on IL-1β-induced VCAM-1, ICAM-1 and E-selectin expression by HUVEC were studied by pretreating HUVEC for 18 h with 25 or 50 μmol/l EA before addition of 10 ng/ml IL-1β. This resulted in reduced cell surface expression of VCAM-1 and E-selectin, but had no effect on cell surface expression of ICAM-1 (Fig. 2(A–C)).
Ellagic acid attenuates activation of NF-κB expression and nuclear translocation of NF-κB p65 and p50 in IL-1β-stimulated human umbilical vein endothelial cells
To examine whether the inhibitory effect of EA on the cytokine-induced expression of adhesion molecules is medicated via NF-κB, we measured the nuclear translocation of p65 and p50 protein of the NF-κB family of transcription factors. Incubation of IL-1β (10 ng/ml) for 6 h induced the nuclear translocation of p65 and p50 (Fig. 3(A, B)). Preincubation of HUVEC with 50 μmol/l EA prior to IL-1β stimulation did significantly prevent the nuclear translocation of p65 and p50 (Fig. 3(A, B)).
Ellagic acid inhibits adhesion of U937 cells to IL-1β-stimulated human umbilical vein endothelial cells
To explore the effects of EA on endothelial cell leucocyte interactions, we examined the adhesion of U937 cells to cytokine-activated HUVEC. Control confluent HUVEC showed minimal binding to U937 cells, but adhesion increased when the HUVEC were treated with IL-1β (Fig. 4(A, B)). Pretreatment of HUVEC with 50 μmol/l EA reduced the number of U937 cells adhering to IL-1β-stimulated HUVEC (Fig. 4(A, B)).
An early stage in atherosclerosis is the adhesion of monocytes to the arterial wall, followed by their infiltration and differentiation into macrophages. This key stage is mediated by the interaction of monocytes with adhesion molecules expressed by endothelial cells. In the present study, we found that 50 μmol/l EA treatment (50 μmol/l EA is equivalent to the dietary intake of approximately 200 g blackberries or 350 g strawberries; Walgren et al. Reference Walgren, Walle and Walle1998; Mertens-Talcott et al. Reference Mertens-Talcott, Talcott and Percival2003; Whitley et al. Reference Whitley, Stoner, Darby and Walle2003) effectively blocked VCAM-1 and E-selectin expression in IL-1β-induced HUVEC. It significantly reduced the binding of human monocytic cell line U937 to IL-1β-induced HUVEC. Previous studies also showed that other polyphenolic compounds, such as vitamin E (40 μmol/l), probucol (50 μmol/l) or tea flavonoid (60 μmol/l epigallocatechin-3-gallate), reduce cytokine-induced adhesion molecule expression and monocyte adhesion to endothelial cells (Islam et al. Reference Islam, Devaraj and Jialal1998; Zapolska-Downar et al. Reference Zapolska-Downar, Zapolski-Downar, Markiewski, Diechanowicz, Kaczmarczyk and Naruszewicz2001; Ludwig et al. Reference Ludwig, Lorenz, Grimbo, Steinle, Meiners, Bartsch, Stangl, Baumann and Stangl2004). In the present study, EA reduced cytokine-induced expression of VCAM-1 and E-selectin but not ICAM-1. A similar result was seen when HUVEC were pretreated with probucol; probucol reduced IL-1β-induced VCAM-1 surface protein and mRNA expression, but not ICAM-1 expression (Zapolska-Downar et al. Reference Zapolska-Downar, Zapolski-Downar, Markiewski, Diechanowicz, Kaczmarczyk and Naruszewicz2001). Previous studies indicated that VCAM-1, but not ICAM-1, plays a critical role in the initiation of atherosclerosis (Cybulsky et al. Reference Cybulsky, Iiyama, Li, Zhu, Chen, Liyama, Davis, Gutierrez-Ramos, Connelly and Milstone2001). VCAM-1 is expressed in vascular lesions in early atherosclerosis and has been found to be elevated in serum from patients with early atherosclerosis, suggesting that this adhesion protein is one of the key molecules involved in the atherogenic process (Cybulsky & Gimbrone, Reference Cybulsky and Gimbrone1991; Rohde et al. Reference Rohde, Lee, Rivero, Jamacochian, Arroyo, Briggs, Rifai, Libby, Creager and Ridker1998).
The NF-κB family controls the expression of genes involved in the inflammation and immune response (Baeuerle, Reference Baeuerle1991). In the cytoplasm, inactive NF-κB exists as a heterodimeric complex of subunits p50 and p65 that binds to a cytoplasmic protein, IκB (Baeuerle & Henkel, Reference Baeuerle and Henkel1994). Upon activation, IκB is rapidly degraded, and the p50/p65 heterodimer is translocated from the cytoplasm into the nucleus where the dimer interacts with regulatory κB elements in promoters and enhancers, thereby controlling gene transcription (Baeuerle & Baltimore, Reference Baeuerle and Baltimore1988; Grilli et al. Reference Grilli, Chiu and Lenardo1993; Chenbg et al. Reference Chenbg, Cant, Moll, Hofer-Warbinek, Wagner, Birnstiel, Bach and de Martin1994). NF-κB is activated by a multitude of stimuli, including inflammatory cytokines and reactive oxygen intermediates (Baeuerle & Baltimore, Reference Baeuerle and Baltimore1988; Grilli et al. Reference Grilli, Chiu and Lenardo1993; Chenbg et al. Reference Chenbg, Cant, Moll, Hofer-Warbinek, Wagner, Birnstiel, Bach and de Martin1994; Muller et al. Reference Muller, Rupec and Baeuerle1997), which are activated in atherosclerotic lesions (Brand et al. Reference Brand, Page, Rogler, Bartsch, Brandl, Knuechel, Page, Kalschmidt, Baeuerle and Neumeier1996; Barnes & Karin, Reference Barnes and Karin1997; D'Acquisto et al. Reference D'Acquisto, May and Ghosh2002). In the present study, we demonstrated that EA reduced cytokine-induced expression of VCAM-1 and E-selectin and prevented the nuclear translocation of p65 and p50 in endothelial cells. The present results suggest that the inhibitory mechanisms of EA might interrupt a signalling cascade involving VCAM transcription-mediated activation of NF-κB.
Several studies have indicated that ROS are implicated in the activation of NF-κB (Muller et al. Reference Muller, Rupec and Baeuerle1997). The current study shows that the ROS production stimulated by IL-1β was decreased by EA pretreatment (Fig. 2(A–C)). Based on the present result, we propose that the inhibitory effect of EA on VCAM-1 expression and NF-κB activation may be due to its antioxidant properties and that it may act by directly scavenging free radicals. In one of our previous studies, we found that EA is approximately 2–3-fold more potent than Trolox in antioxidative ability. Our previous results showed that it scavenged α-α-diphenol-β-picrylhydrazyl (DPPH), alkoxyl radical (RO°) and peroxyl radical (ROO°) and inhibited LDL oxidation (Yu et al. Reference Yu, Chang, Wu and Chiang2005). Since atherosclerosis is a chronic inflammatory disease associated with increased oxidative stress in the vascular endothelium, it would be conceivable that the anti-atherogenic effects of EA might due to its antioxidative properties. The inhibition of cytokine-induced VCAM-1 expression has been described for other substances with antioxidant properties such as tea flavonoid epigallocatechin-3-gallate, probucol, magnolol, protocatechuic aldehyde and other flavonoids (Zapolska-Downar et al. Reference Zapolska-Downar, Zapolski-Downar, Markiewski, Diechanowicz, Kaczmarczyk and Naruszewicz2001; Chen et al. Reference Chen, Lin, Chen, Ku and Chen2002; Ludwig et al. Reference Ludwig, Lorenz, Grimbo, Steinle, Meiners, Bartsch, Stangl, Baumann and Stangl2004; Zhou et al. Reference Zhou, Liu, Miao and I Wang2005).
In conclusion, EA inhibits IL-1β-induced VCAM-1 and E-selectin expression in HUVEC through a mechanism that involves NF-κB. It reduces the binding of human monocytic cell line U937 to IL-1β-induced HUVEC, which might be due to its antioxidant properties.
The research was supported by grants from the National Science Counsel of Taiwan (NSC 93-2320-B-039-017) and China Medical University (CMU 93-NT-01).