Hepatocyte growth factor as a downstream mediator of vascular endothelial growth factor-dependent preservation of growth in the developing lung

Abstract

Impaired vascular endothelial growth factor (VEGF) signaling contributes to the pathogenesis of bronchopulmonary dysplasia (BPD). We hypothesized that the effects of VEGF on lung structure during development may be mediated through its downstream effects on both endothelial nitric oxide synthase (eNOS) and hepatocyte growth factor (HGF) activity, and that, in the absence of eNOS, trophic effects of VEGF would be mediated through HGF signaling. To test this hypothesis, we performed an integrative series of in vitro (fetal rat lung explants and isolated fetal alveolar and endothelial cells) and in vivo studies with normal rat pups and eNOS(-/-) mice. Compared with controls, fetal lung explants from eNOS(-/-) mice had decreased terminal lung bud formation, which was restored with recombinant human VEGF (rhVEGF) treatment. Neonatal eNOS(-/-) mice were more susceptible to hyperoxia-induced inhibition of lung growth than controls, which was prevented with rhVEGF treatment. Fetal alveolar type II (AT2) cell proliferation was increased with rhVEGF treatment only with mesenchymal cell (MC) coculture, and these effects were attenuated with anti-HGF antibody treatment. Unlike VEGF, HGF directly stimulated isolated AT2 cells even without MC coculture. HGF directly stimulates fetal pulmonary artery endothelial cell growth and tube formation, which is attenuated by treatment with JNJ-38877605, a c-Met inhibitor. rHGF treatment preserves alveolar and vascular growth after postnatal exposure to SU-5416, a VEGF receptor inhibitor. We conclude that the effects of VEGF on AT2 and endothelial cells during lung development are partly mediated through HGF-c-Met signaling and speculate that reciprocal VEGF-HGF signaling between epithelia and endothelia is disrupted in infants who develop BPD.

Keywords: bronchopulmonary dysplasia; endothelial nitric oxide synthase; hepatocyte growth factor; lung development; lung hypoplasia; nitric oxide; vascular endothelial growth factor.

Fig. 1.

JNJ-38877605 (JNJ) dose response was determined with concentrations of 4, 10, 50, 100, 500, and 1,000 nM. JNJ doses of 4 and 10 nM failed to decrease hepatocyte growth factor (HGF) response whereas 100, 500, and 1,000 nM caused concentration-related apoptosis in pulmonary artery endothelial cells (PAECs).

Fig. 2.

Treatment with recombinant human VEGF (rhVEGF) improves lung structure in fetal lung explants of endothelial nitric oxide synthase (eNOS)^−/− mice. Lung explants were collected from eNOS^+/+, eNOS^+/−, and eNOS^−/− fetal mice on embryonic day 13 and cultured for 3 days (A). Morphometric analysis showed decreased terminal lung bud formation in eNOS^−/− lungs after 3 days of culture (B). Lung explants of eNOS^−/− mice showed improvement in gross appearance (C) and terminal bud formation (D) following rhVEGF treatment. Data expressed as means ± SE.

Fig. 3.

Neonatal eNOS^−/− mice are susceptible to hyperoxia-induced changes in lung structure. Neonatal lung morphology was examined following 10 days of exposure to normoxia (21% O₂) or hyperoxia (75% O₂). No difference in lung structure was observed in eNOS^+/+ lungs (A and B), while eNOS^−/− lungs showed reduced septation (C and D). Morphometric analysis demonstrated reduced radial alveolar counts in both eNOS^+/− and eNOS^−/− lungs (E). Data expressed as means ± SE.

Fig. 4.

Treatment with rhVEGF improves lung structure in neonatal eNOS^−/− mice. Mice were exposed to 10 days of hyperoxia (75% O₂) followed by 7 days of treatment with rhVEGF or control. No difference in lung structure was observed in eNOS^+/+ lungs following rhVEGF treatment (A and B). Lungs of eNOS^−/− mice showed increased septation and smaller distal air spaces with rhVEGF treatment (C and D). Both eNOS^+/− and eNOS^−/− lungs demonstrated increased radial alveolar counts after rhVEGF treatment (E). Data expressed as means ± SE.

Fig. 5.

HGF and c-Met protein expression in eNOS^−/− lungs. Protein expression was determined by Western blot analysis for HGF and c-Met protein in neonatal lung tissue after exposure to hyperoxia (75% O₂). HGF protein expression trended lower in the lungs of eNOS^−/− mice (A). c-Met protein expression was decreased (B) and the phospho-c-Met (p-cMet)-to-c-Met ratio trended toward reduction when standardized to β-actin in eNOS^−/− lungs (C). Data expressed as means ± SE.

Fig. 6.

Treatment with rhVEGF increases HGF and c-Met expression in eNOS^−/− lungs. Protein expression was determined by Western blot analysis for HGF and c-Met protein in neonatal eNOS^−/− lung tissue after exposure to hyperoxia (75% O₂) and treatment with rhVEGF. Lung tissue showed an increase in both HGF (A) and c-Met (B) protein expression when standardized to β-actin following rhVEGF treatment. Data expressed as means ± SE.

Fig. 7.

Paracrine secretion of HGF from MCs mediates rhVEGF-induced AT2 cell proliferation in coculture. AT2 cell proliferation was examined after treatment with rhVEGF and rHGF in both monoculture and coculture with MCs. AT2 cell proliferation was increased with rhVEGF treatment only in coculture, while proliferation was increased in both monoculture and coculture following rHGF treatment (A). Addition of an anti-HGF antibody to cocultures attenuated rhVEGF-mediated AT2 cell proliferation (B). Data expressed as means ± SE.

Fig. 8.

Treatment with JNJ, a c-Met inhibitor, attenuates rhVEGF-induced PAEC proliferation and tube formation. PAEC proliferation and tube formation were examined following treatment with rhVEGF and JNJ. Treatment with rhVEGF caused an increase in both PAEC proliferation (A) and tube formation (B), which was partially attenuated with rhVEGF/JNJ combination treatment in both assays. Data expressed as means ± SE.

Fig. 9.

Treatment with rHGF increases PAEC proliferation and tube formation in hyperoxia. PAEC proliferation and tube formation were examined under normoxia (21% O₂) and hyperoxia (75% O₂) following treatment with rHGF and SNAP, a NO donor. Under normoxic conditions, all treatment groups showed an increase in PAEC proliferation (A) and tube formation (C). When exposed to hyperoxia, only treatment with rHGF and rHGF/SNAP in combination caused increased PAEC proliferation (B) and tube formation (D). Data expressed as means ± SE.

Fig. 10.

Treatment of neonatal rats with rHGF improves lung structure in the presence of VEGF-R inhibitor. Histological sections of lung tissue were prepared from rat pups treated with a vehicle control, the VEGF-R inhibitor SU-5416, or SU-5416/rHGF in combination (A–C). Lungs of animals treated with SU-5416 showed a decrease in both radial alveolar counts (D) and vessel density (E) compared with the control, whereas the SU-5416/rHGF treatment group demonstrated no significant difference in radial alveolar count or vessel density. Data expressed as means ± SE.

Fig. 11.

Angiocrine mechanisms during fetal lung development. In this study, we examined the role of HGF as a downstream mediator of the VEGF signaling pathway utilizing SU-5416, a VEGF-R inhibitor, eNOS^−/− mice, and JNJ, a selective c-Met antagonist. We speculate that low fetal oxygen tension increases hypoxia inducible factor 1α (HIF-1α) activity and vascular endothelial growth factor (VEGF) expression within alveolar type II (AT2) cells of the fetal lung. VEGF is secreted in a paracrine fashion and binds the VEGF-receptor (VEGF-R) of PAEC, inducing the production of HGF and nitric oxide (NO). The autocrine action of NO stimulates PAEC proliferation and angiogenesis. HGF acts in both autocrine and paracrine fashions via the c-Met receptor, stimulating PAEC growth and angiogenesis, as well as AT2 cell proliferation and enhancing alveolarization. Exposure to hyperoxia disrupts this cellular communication pathway between AT2 cells and PAEC by downregulating VEGF production perhaps through downregulation of HIF-1α activity.