Human endothelial cells were cultured in medium 199 containing fetal bovine serum (FBS) and antibiotics (penicillin/streptomycin) prepared by Gibco; Thermo Fisher Scientific, Inc. (Waltham, MA, USA). Anti-human VCAM-1 (1:200; cat no. SC-13160) and anti-b-actin (1:200; cat no. SC-47778) antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). The antibodies targeting phosphorylated (p-)c-Jun N-terminal kinase (JNK; 46 kDa, 1:1,000; cat no. MAB1205), p-extracellular signal-regulated kinase (ERK)1/2 (44,42 kDa, 1:1,000; cat no. MAB18251), p-p38 (38 kDa, 1:500; cat no. MAB8691), JNK, ERK and p38 antibodies were from R&D Systems, Inc. (Minneapolis, MN, USA). Antibodies against p-IKK α/β (87, 85 kDa, 1:1,000; cat no. #2697) p-IκB-α (40 kDa, 1:2,000; cat no. #9246), anti-p65 (65 kDa, 1:1,000; cat no. #8242) and all non-phosphorylated antibodies were prepared and acquired from Cell Signaling Technology, Inc. (Beverly, MA, USA). Signal inhibitors AG490 (a Janus kinase 2 inhibitor), U0126 (an MAPK kinase/ERK inhibitor) and SB203580 (a p38 inhibitor) were purchased from Cell Signaling Technology, Inc., and PDTC (an NF-κB inhibitor) was from EMD Millipore (Billerica, MA, USA). The primers used for polymerase chain reaction (PCR) analysis were prepared from Bioneer Corporation (Daejeon, Korea). Other chemicals were purchased commercially from Sigma-Aldrich; Merck Millipore (Darmstadt, Germany).

The HUVECs (BMS, Seoul, Korea) were grown over three to six passages in cell culture medium at 37°C under a humidified 95 and 5% (v/v) mixture of air and CO₂ supplemented with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, 3 ng/ml bFGF and 5 U/ml heparin. The U937 human monocyte-like cell line (American Type Culture Collection, Manassas, VA, USA; CRL-1593.2™) was also cultured in complete RPMI-1640 cell culture medium under the same conditions as the HUVECs.

The HUVECs (1×10⁴ cells/well) were removed from the cell growth medium and incubated in serum-free medium for 18 h. These cells were then incubated with or without LPS (1 µg/ml) in the presence of each indicated CA concentration (0, 5, 10, 20 and 40 µM) for 48 h at 37°C Cell viability was measured by the addition of 50 µg/ml 3-(4,5-dimetnythiazol-2-yl)-2,5-diphenyl-thetazolium bromide (MTT). Following incubation for 2 h, purple formazan had formed, the supernatants were removed from each well and 100 µl DMSO was added to completely dissolve the formazan crystals. Data was quantified by measuring the absorbance at 540 nm using a spectrophotometric multi-well microplate reader (Multiskan MS; Thermo Fisher Scientific, Inc.).

Total RNA from HUVECs was isolated according to the manufacturer's protocol using an RNA-Bee isolation kit (Tel-Test, Inc., Friendswood, TX, USA). The cDNA was prepared by reverse transcription using M-MuLV reverse transcriptase (Fermentas; Thermo Fisher Scientific, Inc.) and PCR amplification was performed according to the manufacturer's protocol using an AccuPower PCR PreMix (Bioneer Coporation; Daejeon, Korea). Using 250 ng of cDNA and 10 pmole/µl of forward and reverse primers, PCR was performed as follows: Intercellular adhesion molecule (ICAM)-1 (556 bp, annealing temp. 60°C, 25 cycles), forward 5′-CAG TGA CCA TCT ACA GCT TTC CGG-3′ and reverse 5′-GCT GCT ACC ACA GTG ATG ATG ACA A-3′; VCAM-1 (742 bp, annealing temp. 62°C, 32 cycles), forward 5′-AAT TTA TGT GTG TGA AGG AG-3′ and reverse 5′-GCA TGT CAT ATT CAC AGA A-3′; IL-1β (264 bp, annealing temp. 62°C, 33 cycles), forward 5′-GGA TAT GGA GCA ACA ACT GG-3′ and reverse 5′-ATG TAC CAG TTG GGG AAC TG-3′; IL-6 (497 bp, annealing temp. 54°C, 30 cycles), forward 5′-TGA CAA ACA AAT TCG GTA CAT CC-3′ and reverse 5′-ATC TGA GGT GCC CAT GCT AC-3′; IL-8 (292 bp, annealing temp. 63°C, 28 cycles), forward 5′-ATG ACT TCC AAG CTG GCC GTG GCT-3′ and reverse 5′-TCT CAG CCC TCT TCA AAA ACT TCT C-3′; MCP-1 (261 bp, annealing temp. 62°C, 30 cycles), forward 5′-TCT GTG CCT GCT GCT CAT AG-3′ and reverse 5′-TTT GCT TGT CCA GGT GGT CC-3′; and TNF-α (444 bp, annealing temp. 62°C, 32 cycles), forward 5′-GAC TGA CAA GCC TGT AGC CCA TGT TGT TGT AGC A-3′ and reverse 5′-GCA ATG ATC CCA AAG TAG ACC TGC CCA GAC-3′; β-actin (482 bp, annealing temp. 58°C, 28 cycles), forward 5′-TGA GAC CTT CAA CAC CCC AG-3′ and reverse 5′-CAC TGT GTT GGC GTA CAG GT-3′. Each band intensity was detected for all bands and standardization relative β-actin and measured using ImageJ software (Ver.1.51e; National Institutes of Health, Bethesda, MD, USA).

The HUVECs were treated with cell lysis buffer to obtain cell extracts. Protein quantity was determined using a Bradford assay kit according to the manufacturer's protocol (Thermo Fisher Scientific, Inc.) and equal quantities of protein (30 µg) were separated by performing electrophoresis on 6-15% SDS-PAGE gels. The proteins were then transferred onto a nitrocellulose membrane, which was gently agitated and washed with 3–5% non-fat dry milk in Tris-buffered saline-Tween-20 buffer, containing 0.5 mol/l Tris-HCl (pH 7.5), 0.15 mol/l NaCl and 1 g/l Tween-20. The membrane was incubated with each primary antibody overnight at 4°C according to manufacturer's protocol followed by the secondary antibodies (1:2,000) 2 h at room temperature which was then visualized through enhanced chemiluminescence and exposure to X-ray films. The intensity of each band was measured as aforementioned using ImageJ software.

VCAM-1 promoter region information was obtained by searching human VCAM-1 gene sequences (www.ncbi.nlm.nih.gov/), following which the pVCAM-1/Luc plasmid construct, containing the −1,350 to +45 bp promoter region in the pGL3-basic vector (Promega, Madison, WI, USA), was constructed. The HUVECs were transiently transfected with empty vector (pGL3-basic), constructed pVCAM-1/Luc or pNF-κB/Luc (Stratagene; Agilent Technologies, Inc., Santa Clara, CA, USA) using Lipofectamine™ 2000 reagent (Stratagene; Agilent Technologies, Inc.). Following 24 h of transfection, cell extracts were prepared and luciferase activity was detected using the Renilla luciferase assay kit according to the manufacturer's protocol (Promega). Data was quantified by measuring the fluorescence intensity at 480 nm (excitation) and 560 nm (emission) using a fluorescence multi-well microplate reader (Wallac 1420 Victor 2; PerkinElmer, Inc., Waltham, MA, USA).

Growth medium was removed from the HUVECs and serum-free medium containing CA was stimulated with or without LPS (1 µg/ml) for 2 h. The HUVECs were incubated overnight at 4°C with the FITC-labeled NF-κB p65 antibody [(F-6) Alexa Fluor^R488; 65 kDa; 1:100; cat. no. SC-8008; Santa Cruz Biotechnology, Inc.], and then treated with 3.7% paraformaldehyde solution to preserve the cell state and then washed in PBS. The fluorescence intensities were analyzed for the distribution and translocation of p65 protein. The fixed cells were stained with mounting solution containing 4′–6-diamidino-2-phenylindole to indicate the location of nuclei in the cells.

The U937 cell line has the monocyte phenotype and can therefore be used in monocyte adhesion experiments, which were performed as described in a previous study (24). The HUVECs were cultured to confluence in 6-well plates and pre-treated with each signal inhibitor (10 µM) with CA at the indicated concentrations for 12 h. The U937 cells were cultured in 1% FBS and 5 µg/ml calcein-AM (Invitrogen; Thermo Fisher Scientific, Inc.) RPMI medium at 37°C for 30 min for fluorescent labeling. The fluorescence-labeled U937 cells (1×10⁷) were re-suspended in HUVEC culture medium and then incubated for 30 min with the pre-cultured HUVECs. The non-adherent fluorescent U937 cells were washed with PBS and the fluorescence intensity of attached cells was quantitated using a Wallace Victor2 1420 Multi-label counter (PerkinElmer, Inc.).

Unless otherwise stated, all experiments were performed with triplicate samples and repeated at least three times. The data are presented as the mean ± standard deviation and statistical comparisons between groups were performed using one-way analysis of variance followed by Student's t-tests using SigmaPlot software (version 11; Systat Software Inc., San Jose, CA, USA). P<0.05 was considered to indicate a statistically significant differences.

The present study examined cell viability using an MTT assay to determine the in vitro experimental concentration of CA. As shown in Fig. 1B and C, the cells were treated with CA for 24 h in the presence or absence of LPS. It was confirmed that treatment of the HUVECs with CA was not toxic up to a concentration of 40 µM CA. Based on this information, the mRNA expression levels of ICAM-1 and VCAM-1 were examined in the LPS-induced HUVECs at the indicated concentrations of CA. The results showed that CA did not affect the mRNA expression of ICAM-1, however, the mRNA expression of VCAM-1 was decreased in a concentration-dependent manner in the LPS-induced HUVECs (Fig. 1D). Subsequently, the results of western blot analysis showed that the protein expression levels of ICAM-1 and VCAM-1 were significantly increased by LPS stimulation in the HUVECs. Similar to the mRNA expression levels, CA markedly inhibited the protein expression of VCAM-1 and had no effect on the protein expression level of ICAM-1 in the LPS-stimulated HUVECs (Fig. 1E).

The present study subsequently examined the activity of CA on the mRNA expression levels of IL-1β, IL-6, IL-8, MCP-1 and TNF-α in HUVECs. In the LPS-stimulated HUVECs, the gene expression levels of the pro-inflammatory cytokines IL-1β, IL-6, IL-8, MCP-1 and TNF-α increased ~6–9-fold, compared with those in the untreated control. In the CA-treated HUVECs, the mRNA level of each pro-inflammatory cytokine was significantly decreased in a CA concentration-dependent manner (Fig. 2).

To reveal the molecular mechanism for the inhibitory activity of CA on the expression of VCAM-1 and pro-inflammatory cytokines, the present study analyzed the phosphorylation of MAPKs in the LPS-induced HUVECs with or without CA exposure. The effects of CA and commercial signal inhibitors on the mRNA expression levels of VCAM-1 and pro-inflammatory cytokines were examined in the LPS-activated HUVECs. The HUVECs were treated with AG490 (a Janus kinase inhibitor), U0126 (an MEK/ERK inhibitor), SB203580 (a p38 inhibitor), PDTC (an NF-κB inhibitor) and at the indicated concentration of CA. After 1 h, the incubated HUVECs in the commercial signal inhibitors and CA were stimulated with LPS. Our data showed that, with the exception of AG490, all the commercial signal inhibitors inhibited the mRNA and protein levels of VCAM-1 in LPS-induced HUVECs. CA inhibited the mRNA and protein expression levels of VCAM-1 to levels similar to those in the U0126, SB203580 and PDTC treatment groups (Fig. 3A). This result indicated that key signal inhibitors, including U0126, SB203580 and PDTC, inhibited the expression of VCAM-1 and the upstream signaling molecules of VCAM-1 in the LPS-stimulated HUVECs.

Subsequently, the present study examined the effect of CA on cellular signaling molecules activated by LPS-stimulation in HUVECs. The inhibitory activity of CA on the phosphorylation of the most well known MAPK pathways, including JNK, ERK1/2 and p38 proteins, were then confirmed in LPS-induced HUVECs. The results indicated that CA decreased the phosphorylation of ERK1/2 and p38 proteins in the LPS-stimulated HUVECs in a concentration-dependent manner, whereas CA had minimal effect on the phosphorylation of JNK (Fig. 3B). These results suggested that the inhibitory activity of CA on the expression of VCAM-1 may be associated with inhibition of the phosphorylation of ERK1/2 and p38 kinase in LPS-stimulated HUVECs.

To analyze the effect of CA on MAPK downstream signaling pathways, the present study confirmed the activity of CA on the phosphorylation of IKK/IκB-α signal transduction. The NF-κB p65 subunit is an important transcription factor for the expression of VCAM-1 in LPS-stimulated HUVECs (25). The nuclear translocation of the p65 subunit is activated by the phosphorylation of IκB-α. Therefore, the effect of CA on the phosphorylation of IKK was examined. The results demonstrated that CA significantly inhibited the phosphorylation of IKK in LPS-stimulated HUVECs. In addition, it was found that CA inhibited the phosphorylation of IκB-α protein and suppressed the degradation of IκB-α protein in the cell cytosolic fraction (Fig. 3C). These results suggested that CA significantly inhibited IKK/IκB-α signal transduction. The protein levels of p65 in the nucleus of cells treated with CA in the LPS-stimulated HUVECs were examined. In the nuclei of resting state cells, a basic level of p65 protein was identified, and in the cells treated with LPS alone, the nuclear translocation of p65 subunit was significantly increased. In cells stimulated with LPS in the presence of CA, it was confirmed that the protein expression of p65 was decreased in a concentration-dependent manner. The reduced protein expression of p65 was similar to that in the non-stimulated cells (Fig. 3D). The inhibitory activity of CA on the expression of VCAM-1 and pro-inflammatory cytokines may be due to the decrease in nuclear p65 translocation through inhibition of the phosphorylation of MAPK/IKK/IκB-α and degradation of IκB-α.

The present study investigated whether CA regulates the transcriptional activity of the VCAM-1 gene using VCAM-1 luciferase reporter gene constructs. The transfected HUVECs were treated with or without CA and then activated with LPS. Stimulation of the transfected cells with LPS promoted luciferase activity by ~5–6-fold over that in the mock transfected cells. CA significantly reduced the LPS-activated induction of VCAM-1 in the transfected HUVECs at the transcriptional level (Fig. 4A). The regulatory activity of CA on transcriptional activation was examined using the established pNF-κB/Luc reporter gene construct by LPS stimulation. As shown Fig. 4B, the results demonstrated that CA reduced transcriptional activity by LPS in a dose-dependent manner in the HUVECs transfected with the pNF-κB/Luc reporter gene construct. To analyze the activity of CA on the translocation of the p65 subunit into the nucleus, fluorescence microscopy was used to analyze the HUVECs treated with or without CA. It was confirmed that the fluorescence intensity in the nucleus was increased in cells treated with LPS alone, and a large quantity of p65-FITC antibody was present in the nucleus. By contrast, CA decreased fluorescence intensity in the nucleus by LPS stimulation, and a similar result was observed in the cells treated with PDTC, which is an NF-κB specific signaling inhibitor (Fig. 4C).

The overproduction of cell adhesion molecules, including VCAM-1, by LPS likely increases the adherence of monocytes to endothelial cells in the early stages of vascular inflammation. Therefore, in order to demonstrate the effect of CA on U937 cell adhesion to endothelial cells, the present study compared cell adhesion activity with various signal inhibitors in the presence of CA in the LPS-induced HUVECs. In the case of cells treated with LPS alone, the adhesion activity of the U937 cells was markedly increased compared with that in the control cells. The data showed that U0126, SB203580 and PDTC significantly inhibited U937 adhesion in the LPS-induced HUVECs, whereas AG490 had no effect. The LPS-activated HUVECs were treated with CA (20 or 40 µM), and it was shown that CA inhibited U937 adhesion to the HUVECs (Fig. 5). The data showed that CA inhibited the LPS-induced expression of VCAM-1 and suppressed U937 adhesion to HUVECs.