Differential roles of eNOS in late effects of VEGF-A on hyperpermeability in different types of endothelial cells

Abstract

Vascular endothelial growth factor (VEGF)-A induces endothelial hyperpermeability, but the molecular pathways remain incompletely understood. Endothelial nitric oxide synthase (eNOS) regulates acute effects of VEGF-A on permeability of endothelial cells (ECs), but it remains unknown whether and how eNOS regulates late effects of VEGF-A-induced hyperpermeability. Here we show that VEGF-A induces hyperpermeability via eNOS-dependent and eNOS-independent mechanisms at 2 days after VEGF-A stimulation. Silencing of expression of the eNOS gene (NOS3) reduced VEGF-A-induced permeability for dextran (70 kDa) and 766 Da-tracer in human dermal microvascular ECs (HDMVECs), but not in human retinal microvascular ECs (HRECs) and human umbilical vein ECs (HUVECs). However, silencing of NOS3 expression in HRECs increased permeability to dextran, BSA and 766 Da-tracer in the absence of VEGF-A stimulation, suggesting a barrier-protective function of eNOS. We also investigated how silencing of NOS3 expression regulates the expression of permeability-related transcripts, and found that NOS3 silencing downregulates the expression of PLVAP, a molecule associated with trans-endothelial transport via caveolae, in HDMVECs and HUVECs, but not in HRECs. Our findings underscore the complexity of VEGF-A-induced permeability pathways in ECs and the role of eNOS therein, and demonstrate that different pathways are activated depending on the EC phenotype.

Figure 1

VEGF-A-induced hyperpermeability is partially facilitated by eNOS in HDMVECs, but not in HRECs and HUVECs. (a) Relative NOS3 mRNA levels in control and siNOS3-treated HDMVECs, HRECs and HUVECs at 72 h after transfection, in n = 3–4 independent experiments. (b) Representative cropped images of western blots of eNOS and β-actin (loading control) expression in HDMVECs, HRECs and HUVECs transfected with siNT or siNOS3. Original blots are presented in Supplementary Figs. S1–S3. (c) eNOS protein levels were significantly reduced in siNOS3-treated HDMVECs, HRECs and HUVECs at 72 h after siRNA transfection, in n = 3–4 independent experiments. Permeability for 70 kDa dextran-FITC (d), 766 Da Cy3 tracer (e) and 67 kDa BSA-FITC (f) in siNT- and siNOS3-treated HDMVECs, HRECs and HUVECs in the presence and absence of exogenous basolateral VEGF-A (25 ng/mL) stimulation. n ≥ 4 independent experiments (each dot represents 1 experiment). Results are normalized to the unstimulated siNT control. Data are represented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001 (one-sample t-test or Student’s t-test).

Figure 2

Effect of NOS3 silencing on tracer uptake in HDMVECs and HRECs. (a) Representative images of dextran-Texas Red uptake in HDMVECs and HRECs treated with siNT and siNOS3. Cells were left unstimulated or treated with VEGF-A (25 ng/mL) for 48 h prior to the addition of the tracer. Quantification of dextran uptake is shown in (b), in n = 4 independent experiments. (c) Representative images of BSA-FITC uptake in HDMVECs and HRECs treated with siNT and siNOS3. (d) Quantification of BSA-FITC uptake in images shown in (d), n = 3 independent experiments. All results are normalized to the unstimulated siNT control. Scale bars: 20 µm. Data are represented as mean ± SEM.

Figure 3

PLVAP expression is regulated by eNOS in HDMVECs and HUVECs, but not in HRECs. (a) PLVAP mRNA levels in VEGF-A-stimulated (25 ng/mL for 48 h) siNT- and siNOS3-treated HDMVECs, HRECs and HUVECs in n = 3–4 independent experiments. (b) Representative confocal images of PLVAP immunofluorescence staining in siNT- and siNOS3-treated HDMVECs and HRECs. Cells were cultured in the presence or absence of VEGF-A (25 ng/mL for 48 h). Scale bars: 20 µm. (c) Quantification of PLVAP immunofluorescence staining data shown in (b), n = 3 independent experiments. (d) PLVAP mRNA levels in control, L-NAME pretreated (100 µM, 30 min), VEGF-A-stimulated (25 ng/mL for 48 h) or a combination of L-NAME pretreated and VEGF-A-stimulated HDMVECs, HRECs, HUVECs and BRECs, in n = 3–4 independent experiments. (e) PLVAP mRNA levels in HDMVECs, HRECs and HUVECs after repeated treatment with NO donor SNAP (50 µM, 100 µM or 200 µM) over a 48 h incubation period (cells were stimulated 3 times for 4 h per day with freshly prepared SNAP), in n = 3–4 independent experiments. Data are represented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant (one-sample t-test or Student’s t-test).

Figure 4

VEGFR2 expression is regulated by eNOS in HUVECs, but not in HDMVECs and HUVECs. (a) Relative KDR mRNA levels in control and siNOS3-treated HDMVECs, HRECs and HUVECs at 72 h after transfection, in n = 3–4 independent experiments. (b) Representative cropped images of western blots of VEGFR2 and β-actin (loading control) expression in HDMVECs, HRECs and HUVECs transfected with siNT or siNOS3. Cells were cultured in the presence or absence of VEGF-A (25 ng/mL for 48 h). Original blots are presented in Supplementary Figs. S2, S7 and S8. (c) VEGFR2 protein levels in control and in siNOS3-treated HDMVECs, HRECs and HUVECs at 72 h after siRNA transfection, in n = 3–4 independent experiments. (d) KDR mRNA levels in HDMVECs, HRECs and HUVECs after repeated treatment with NO donor SNAP (50 µM, 100 µM or 200 µM) over a 48 h incubation period (cells were stimulated 3 times for 4 h per day with freshly prepared SNAP), in n = 3–4 independent experiments. Data are represented as mean ± SEM. *p < 0.05; **p < 0.01 (one-sample t-test or Student’s t-test).

Figure 5

Silencing of NOS3 expression has limited or no effects on the expression of various transcellular transport-related, paracellular transport-related and VEGF family-related transcripts in HDMVECs, HRECs and HUVECs. mRNA levels of various transcellular transport-related, paracellular transport-related and VEGF family-related transcripts in siNOS3-treated HDMVECs (a), HRECs (b) and HUVECs (c), in n = 3–4 independent experiments. Results are expressed as fold change relative to the siNT control. Data are represented as mean ± SEM. *p < 0.05; **p < 0.01 (one-sample t-test).