Fig. 3

The involvement of AQP1, COX-2 and TonEBP/NFAT5 in glucose-induced in vitro angiogenic activity. A, C, D The effect of gene silencing to AQP1 or to TonEBP/NFAT5 on high glucose- and high mannitol-induced angiogenic activity of endothelial cells. B The effect of the COX-2 inhibitor NS-398 on high glucose- and high mannitol-induced angiogenic activity of endothelial cells. E The effect of TonEBP/NFAT5 gene silencing on high glucose- and high mannitol-induced endothelial cell migration. Serum-starved human aortic endothelial cells (HAECs) were transfected with Cy5-coniugated-siRNA to AQP1 or with FITC-coniugated-siRNA to TonEBP/NFTA5 and/or siRNA with scrambled sequence for 24 h (D), and then treated with stimuli for 24 h with high glucose or high mannitol. At the end of treatments, transfection efficiency was checked by observing the plates under fluorescence microscopy (D), while TonEBP/NFAT5 gene knockdown efficiency was detected at the protein level by Western analysis (D). At the end of incubations cells were enzymatically harvested, plated on Matrigel-coated 96 well-plates for the in vitro angiogenesis assay (A–C, E) or cell migration assay (F), and further incubated with stimuli for 24 h. In parallel experiments (B), serum starved HAECs were treated with high glucose or high mannitol (both at 25 mmol/L) in the presence or absence of 5 µg/mL NS-938 (NS) for 48 h and, after enzymatic detachment, plated on Matrigel-coated 96 well-plates, and further incubated with stimuli for 24 h. After incubation, images to document angiogenic activity were taken and analyzed as indicated in Methods. Upper panels (A–E): representative images. Lower panels (C, F): quantitation of angiogenic activity (network area, C) and for cell migration (F). *, P < 0.05 mannitol- or glucose-treated vs control HAECs; §, P < 0.05 vs without AQP1-siRNA or without TonEBP/NFAT-siRNA; # P < 0.05 vs without NS-938. RFU: relative fluorescence units