Magnitude-dependent effects of cyclic stretch on HGF- and VEGF-induced pulmonary endothelial remodeling and barrier regulation

Abstract

Mechanical ventilation at high tidal volumes compromises the blood-gas barrier and increases lung vascular permeability, which may lead to ventilator-induced lung injury and pulmonary edema. Using pulmonary endothelial cell (ECs) exposed to physiologically [5% cyclic stretch (CS)] and pathologically (18% CS) relevant magnitudes of CS, we evaluated the potential protective effects of hepatocyte growth factor (HGF) on EC barrier dysfunction induced by CS and vascular endothelial growth factor (VEGF). In static culture, HGF enhanced EC barrier function in a Rac-dependent manner and attenuated VEGF-induced EC permeability and paracellular gap formation. The protective effects of HGF were associated with the suppression of Rho-dependent signaling triggered by VEGF. Five percent CS promoted HGF-induced enhancement of the cortical F-actin rim and activation of Rac-dependent signaling, suggesting synergistic barrier-protective effects of physiological CS and HGF. In contrast, 18% CS further enhanced VEGF-induced EC permeability, activation of Rho signaling, and formation of actin stress fibers and paracellular gaps. These effects were attenuated by HGF pretreatment. EC preconditioning at 5% CS before HGF and VEGF further promoted EC barrier maintenance. Our data suggest synergistic effects of HGF and physiological CS in the Rac-mediated mechanisms of EC barrier protection. In turn, HGF reduced the barrier-disruptive effects of VEGF and pathological CS via downregulation of the Rho pathway. These results support the importance of HGF-VEGF balance in control of acute lung injury/acute respiratory distress syndrome severity via small GTPase-dependent regulation of lung endothelial permeability.

Fig. 1.

Effect of VEGF on pulmonary endothelial cell (EC) barrier function. A: ECs were plated on gold microelectrodes. At time indicated by arrow, ECs were treated with either vehicle or VEGF (50, 100, 200, or 500 ng/ml) and used for measurements of transendothelial electrical resistance (TER). B and C: cells were treated with VEGF (200 ng/ml) for indicated periods of time. B: Rho activity was measured using an in vitro activation assay described in materials and methods. C: levels of phosphorylated myosin light chain (MLC) in the total lysates were determined by Western blot analysis using di-phospho-MLC antibodies. Results represent 3–5 independent experiments.

Fig. 2.

Effect of hepatocyte growth factor (HGF) on pulmonary EC barrier function. A: ECs were grown on gold microelectrodes followed by treatment with either vehicle or HGF (10, 20, or 30 ng/ml, indicated by arrow), and TER was measured over the time. B and C: cells were treated with HGF (30 ng/ml) for indicated periods of time. B: Rac activity was measured using a Rac-GTP pull-down assay. C: levels of p21-activated kinase PAK autophosphorylation in total lysates were determined by Western blot analysis with phospho-PAK1 antibodies. Results represent 3–7 independent experiments. RDU, relative densitometry units.

Fig. 3.

Effects of HGF on VEGF-induced EC permeability. A: at time indicated by first arrow, confluent pulmonary ECs were preincubated with HGF (30 ng/ml) followed by VEGF (200 ng/ml) challenge, indicated by second arrow, and TER changes were monitored over time. B: pulmonary EC monolayers grown in Transwell plates were incubated with VEGF (200 ng/ml) with or without HGF (30 ng/ml) pretreatment, and the amount of dextran that crossed the EC monolayer was determined by measurements of FITC-labeled dextran fluorescence values in bottom chambers. Data are means ± SD of 3 independent experiments (*P < 0.05).

Fig. 4.

Effect of HGF on VEGF-induced EC barrier dysfunction. Endothelial monolayers were pretreated with HGF (30 ng/ml, 15 min) and stimulated with VEGF (200 ng/ml at 30, 45, or 60 min). A: Rho activity after 30 min of VEGF challenge was measured using the Rho-GTP pull-down assay described in materials and methods. Bottom: total Rho content in EC lysates. B–D: phosphorylation of myosin-associated phosphatase type (MYPT; B), MLC (C), or PAK1 (D) in lung endothelial EC pretreated with HGF followed by VEGF challenge was detected by Western blot with phospho-specific antibodies. MLC and PAK1 phosphorylation was determined after 30 min of VEGF treatment. E: effect of HGF (30 ng/ml, 15 min) on VEGF-induced (200 ng/ml, 30 min) cytoskeletal remodeling and MLC phosphorylation. Double immunofluorescence staining was performed using di-phospho-MLC antibodies and Texas red phalloidin to detect F-actin. VEGF-induced stress fibers and HGF-mediated peripheral actin accumulation are marked by arrows. Results represent 3–5 independent experiments.

Fig. 5.

Effect of high magnitude CS on VEGF-induced EC barrier compromise. Pulmonary ECs grown to confluence on Flexcell plates were exposed to pathological (18%) CS for 2 h and then stimulated with VEGF (200 ng/ml, 30 min). A: immunofluorescence staining using Texas red phalloidin was performed to detect F-actin. Paracellular gaps are marked by arrows. Right: VEGF-induced gap formation was attenuated by the Rho kinase inhibitor Y27632 (2.5 μM, 30 min). B: quantitative analysis of CS and VEGF-induced paracellular gap formation. Data are means ± SD of 7 independent experiments (*P < 0.05). C: phosphorylated MLC was detected by immunoblotting with di-phospho-MLC specific antibodies. Equal protein loading was confirmed by reprobing of membranes with antibodies to nonphosphorylated protein. D: ECs were transfected with Rho-specific or nonspecific (ns) small interfering (si) RNA for 48 h before CS experiments. Cytoskeletal remodeling in control and VEGF-treated monolayers was analyzed by inmmunifluorescence staining of F-actin. Paracellular gaps are marked by arrows.

Fig. 6.

Effect of low magnitude CS on HGF-induced EC barrier protection. A: confluent ECs were exposed to physiological (5%) CS for 2 h followed by HGF (30 ng/ml, 15 min) stimulation and immunofluorescence staining for F-actin. Areas of peripheral F-actin accumulation are marked by arrows. B: quantitative analysis of CS- and HGF-induced cortical actin rim formation. Data are means ± SD of 4 independent experiments (*P < 0.05). C: Western blot analysis of HGF-induced PAK1 autophosphorylation in static cultures and ECs exposed to low magnitude CS (5%, 2 h). Equal protein loading was confirmed by reprobing of membranes with antibodies to nonphosphorylated protein. Results represent 3–5 independent experiments.

Fig. 7.

Effects of physiological CS preconditioning on VEGF-mediated EC barrier dysfunction. A: EC monolayers grown to confluence on Flexcell plates were stimulated with VEGF (200 ng/ml, 30 min) with or without HGF pretreatment (30 ng/ml, 15 min) under static conditions. F-actin was visualized by immunofluorescence staining with Texas red phalloidin. B: EC monolayers grown on Flexcell plates were preconditioned at physiological (5% CS) or pathological (18% CS) levels of CS for 2 h and stimulated with VEGF (200 ng/ml, 30 min) with or without HGF pretreatment (30 ng/ml, 15 min). F-actin was visualized by immunofluorescence staining with Texas red phalloidin. Paracellular gaps are marked by arrows. A and B: VEGF-induced paracellular gaps are partially inhibited by HGF pretreatment of static ECs and completely inhibited by HGF in 5% CS-preconditioned ECs. Insets: high magnification of selected areas. C and D: quantitative analysis of paracellular gap formation induced by VEGF challenge of pulmonary ECs exposed to 5% CS with or without HGF pretreatment (C) or to 18% CS and 5% CS (D). Data are means ± SD of 4 independent experiments (*P < 0.05). E: phosphorylated MLC was detected by Western blot with specific antibodies. Equal protein loading was confirmed by reprobing of membranes with antibodies to nonphosphorylated MLC.

Fig. 8.

Rac knockdown attenuates protective effects of physiological CS and HFG against VEGF-induced EC barrier disruption. A and B: human pulmonary ECs were transfected with Rac-specific or nonspecific siRNA. After 48 h of transfection, cells were preconditioned at physiological (5% CS, 2 h) followed by VEGF (200 ng/ml, 30 min) challenge alone (A) or with HGF pretreatment (30 ng/ml, 15 min) before VEGF stimulation (B). Cytoskeletal remodeling in control and VEGF-treated EC monolayers was analyzed by inmmunifluorescence staining for F-actin. Paracellular gaps are marked by arrows. Results represent 3 independent experiments.