Fig. 1

The effect of high glucose and high mannitol on COX-2 expression in endothelial cells. Immunoblotting with an antibody against COX-2 of proteins extracted from serum-starved subconfluent human aortic endothelial cells (HAECs) incubated with control D-glucose (5.5 mmol/L, control), high glucose (HG, 30.5 or 50.5 mmol/L), high mannitol (HM, 5.5 mmol/L glucose + 25 or 45 mmol/L mannitol) for 24 h (a, b) and for 3 days (c, d). After COX-2 analyses, proteins were stripped and restained with an anti-β-actin antibody. Lipopolysaccharide (LPS, 1 μg/mL) was used as positive control for COX-2 expression (a, d). e, f Immunoblotting with an antibody against COX-2 of proteins extracted from serum-starved subconfluent human microvascular endothelial cells (HMVECs) incubated with stimuli (12.5 and 25 mmol/L glucose or mannitol) for 3 days. Representative immunoblots are shown. The results of scanning densitometry (n = 3 independent experiments) are expressed as arbitrary units in panel g. Columns and bars represent the mean ± SD (*, P < 0.05 vs control cells; **, P < 0.01 vs control cells). Equal loading/equal protein transfers were verified by protein band detection with the anti-β-actin antibody. h, i The effect of high D-glucose, high l-glucose, high mannitol and NaCl on COX-2 expression in endothelial cells. Immunoblotting with an antibody against COX-2 of proteins extracted from serum-starved subconfluent human aortic endothelial cells (HAECs) incubated with control glucose concentration (5.5 mmol/L, control), high glucose (30.5 mmol/L), high mannitol (5.5 mmol/L glucose + 25 mmol/L mannitol), or sodium chloride (30.5 mmol/L), for 24 h. After COX-2 analyses, proteins were stripped and restained with an anti-β-actin antibody. Representative immunoblots from three different experiments are shown. i Columns and bars represent the mean ± SD of scanning densitometry of panel H (*, P < 0.05 vs control cells; **, P < 0.01 vs control cells)