Fig. 1. Chemical structure of prodelphinidin B2 3,3′ di-O-gallate (PDGG).

Fig. 2. PDGG caused a dose-dependent inhibition of COX-2 protein (a) and PGE₂ release (b) in RAW264 cells. (a) COX-2 detection. After RAW264 cells (1 × 10⁶ cells) were starved in serum-free medium for 2.5 h, the cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min, and then exposed to 40 ng/ml LPS for 12 h. COX-2 and COX-1 were detected by Western blotting analysis with their antibodies, respectively. Histograms show the densitometric fold of COX-2 protein normalized to COX-1. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS. (b) PGE₂ measurement. After RAW264 cells (5 × 10⁵ cells) were starved in serum-free medium for 2.5 h, the cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min, and then exposed to 40 ng/ml LPS for 12 h. The amount of PGE₂ in medium was measured. Each value represents mean ± SD of triplicate tests. ^*P < 0.05; ^**P < 0.01 vs. LPS.

Fig. 3. PDGG suppresses COX-2 mRNA expression (a) and promoter activity (b). (a) RT-PCR. RAW264 cells (1 × 10⁶ cells/6 ml) were starved in serum-free medium for 2.5 h, the cells were then treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min before exposure to 40 ng/ml LPS for 6 h. RNA was extracted with ISOGEN RNA isolation kit and mRNA expressions of COX-2 and COX-1 were detected by RT-PCR technique. The RT-PCR products were separated on 2% agarose gel, and digitally imaged after staining with ethidium bromide. Quantification of the bands was performed using Imager Gauge Software (Fuji Photo Film). Histograms show the densitometric fold of COX-2 mRNA normalized to COX-1 mRNA. Each value represents mean ± SD of three to four separate experiments. ^*P < 0.05 vs. LPS. (b) Report gene assay. RAW264 cells (1 × 10⁵ cells) seeded into each wells of 12-well plates were transfected with 0.5 mg of COX-2 promoter (−327/+59)-luciferase reporter constructs, and 0.12 mg of CMV-β-galactosidase plasmid. After 5 h incubation, the medium was replaced with complete medium, and cultured for another 20.5 h. The cells were then treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min before they were exposed to 40 ng/ml LPS for 6 h. The values of luciferase activity were normalized to transfection efficiency monitored by β-galactosidase expression, and COX-2 promoter activity is expressed as fold induction to the (−327/+59) construct without LPS treatment. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS.

Fig. 4. Effects of PDGG on the transcriptional factors regulating COX-2 expression. (a) phosphorylation of c-Jun and CREB. Cell culture and Western blotting were done as described in Fig. 2a. RAW264 cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min, and then exposed to 40 ng/ml LPS for 30 min. Phosphorylated c-Jun and CREB were detected with their antibodies, respectively. Histograms show the densitometric fold of phosphorylated c-Jun and CREB normalized to total c-Jun and CREB, respectively. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS. (b) Nuclear translocation of C/EBPδ and C/EBPβ. The cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min, and then exposed to 40 ng/ml LPS for 4 h. Nuclear protein was extracted and nuclear C/EBPδ and β were detected with their antibodies. Histograms show the densitometric fold of C/EBPδ and C/EBPβ. Each value represents mean ± SD of three to four separate experiments. ^*P < 0.05; ^**P < 0.01 vs. LPS. (c) IκB-α degradation. To identify the time of LPS-induced degradation of IκB-α protein, RAW264 cells were treated with 40 ng/ml LPS from 10 to 120 min. To determine the effect of PDGG on the degradation of IκB-α protein, the cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min before exposure to 40 ng/ml LPS for 30 min. IκB-α protein was detected with its antibody. Histograms show the percentage of IκB-α protein to non-LPS treatment after normalizing to α-tubulin. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS. (d) Nuclear translocation of p65. RAW264 cells were treated with 0, 25 or 50 μM of PDGG for 30 min before exposure to 40 ng/ml LPS for 30 min. Nuclear p65 were detected with p65 antibody. Histograms show the densitometric fold of p65 in nuclear lysates. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS.

Fig. 5. Effects of PDGG on MAPK phosphorylation induced by LPS. Cell culture and Western blotting were done as described in Fig. 2a. RAW264 cells were treated with 0, 12.5, 25 or 50 μM of PDGG for 30 min, and then exposed to 40 ng/ml LPS for 30 min. Total or phosphorylated MAPK were detected with their antibodies, respectively. Histograms show the densitometric fold of phosphorylated MAPK normalized to total MAPK, respectively. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS.

Fig. 6. Effects of green tea proanthocyanidins on COX-2 protein and PGE₂ release in LPS-activated RAW264 cells. (a) Chemical structures of tea proanthocyanidins (also see Fig. 1). (b) COX-2 detection. After RAW264 cells (1 × 10⁶ cells) were starved in serum-free medium for 2.5 h, the cells were treated with 50 μM of the indicated proanthocyanidins for 30 min, respectively, and then exposed to 40 ng/ml LPS for 12 h. Western blotting was done as described in Fig. 2a. Histograms show the densitometric fold of COX-2 protein normalized to COX-1 protein. Each value represents mean ± SD of three to four separate experiments. ^**P < 0.01 vs. LPS. (c) PGE₂ measurement. Cell culture and PGE₂ measurement are done as described in Fig. 2b. After RAW264 cells (5 × 10⁵ cells) were starved in serum-free medium for 2.5 h, the cells were treated with 50 μM of the indicated proanthocyanidins for 30 min, and then exposed to 40 ng/ml LPS for 12 h. The amount of PGE₂ in medium was measured. Each value represents mean ± SD of triplicate tests. ^*P < 0.05; ^**P < 0.01 vs. LPS.