Hepatocyte growth factor mediates angiopoietin-induced smooth muscle cell recruitment

Abstract

Communication between endothelial cells (ECs) and mural cells is critical in vascular maturation. Genetic studies suggest that angiopoietin/Tie2 signaling may play a role in the recruitment of pericytes or smooth muscle cells (SMCs) during vascular maturation. However, the molecular mechanism is unclear. We used microarray technology to analyze genes regulated by angiopoietin-1 (Ang1), an agonist ligand for Tie2, in endothelial cells (ECs). We observed that hepatocyte growth factor (HGF), a mediator of mural cell motility, was up-regulated by Ang1 stimulation. We confirmed this finding by Northern blot and Western blot analyses in cultured vascular endothelial cells. Furthermore, stimulation of ECs with Ang1 increased SMC migration toward endothelial cells in a coculture assay. Addition of a neutralizing anti-HGF antibody inhibited Ang1-induced SMC recruitment, indicating that the induction of SMC migration by Ang1 was caused by the increase of HGF. Interestingly, Ang2, an antagonist ligand of Tie2, inhibited Ang1-induced HGF production and Ang1-induced SMC migration. Finally, we showed that deletion of Tie2 in transgenic mouse reduced HGF production. Collectively, our data reveal a novel mechanism of Ang/Tie2 signaling in regulating vascular maturation and suggest that a delicate balance between Ang1 and Ang2 is critical in this process.

Figure 1.

Ang1 induced HGF expression in HUVECs. (A) Total RNAs were extracted from corresponding adenoviral vector–infected HUVECs for 48 hours. HGF mRNA was analyzed by Northern blot. 18S rRNA, visualized with ethidium bromide and UV, was used as a loading control. Images are representative of 3 separate experiments. (B) Cells were lysed from corresponding adenoviral vector–infected HUVECs for 48 hours. Cell lysates were analyzed by Western blotting and probed with an anti-HGF antibody. (C) Cell-conditioned media were collected, concentrated, and analyzed for protein expression by Western blotting. Filters were immunoblotted with anti-Ang1–, anti-Ang2–, and anti-VEGF–specific antibodies, respectively.

Figure 2.

Ang1 induced HGF production independently of Akt, MAPK, and p38 MAPK activation. (A) Total RNAs were extracted from corresponding adenoviral vector–infected HUVECs at 16, 24, and 48 hours, respectively. HGF mRNA was analyzed by Northern blot. 18S rRNA, visualized with ethidium bromide and UV, was used as a loading control. Images are representative of 3 separate experiments. (B) Inactivation of dominant-negative Akt was confirmed by Akt kinase assay using cells treated in the same manner as in panel A. (C) HUVECs were treated with PD98059 at 20 μM for 30 minutes before adenoviral-Ang1* infection for 16, 24, and 48 hours, respectively. Northern blot analysis was performed using total RNAs. Images are representative of 3 separate experiments. (D) Effectiveness of PD98059 was tested in the cells treated as in panel C. As a positive control, FGF was used. Extracted protein was analyzed by Western blot. (E) HUVECs were treated with SB203580 at 10 μM for 30 minutes before adenoviral-Ang1* infection for 16, 24, and 48 hours, and Northern blot was performed.

Figure 3.

Stimulation of endothelial cells with Ang1 induced SMC migration through HGF up-regulation. (A) Coculture assay using Transwell was developed to evaluate SMC migration. Both sides of the filters were coated with Matrigel. ECs were seeded underneath the filter, and SMCs were seeded in the upper chamber. (B) HUVECs were infected with various viral vectors expressing genes of interest for 48 hours, and SMCs were labeled with RFP. Then the coculture assay was set up as shown in panel A. Migrated SMCs on the other side of the filter were counted through a microscope in randomly selected high-power fields. (C) Effects of neutralization of HGF function in EC-SMC recruitment were evaluated with the coculture assay. Neutralizing HGF antibody or control IgG was added 1 hour before HASMCs were added to the wells. Experiments were performed at least 3 times in more than 2 wells for each treatment. Five fields were counted for each filter in each experiment. *P < .01 compared with control.

Figure 4.

Ang2 inhibited Ang1-induced HGF expression in endothelial cells. (A) HUVECs were infected with different viral vectors expressing genes of interest for 48 hours. Total RNAs were extracted from the cells. HGF mRNA was analyzed by Northern blotting and was probed with HGF cDNA. 18S rRNA, visualized with ethidium bromide and UV, was used as a loading control. (B) Cell-conditioned media were collected, concentrated, and analyzed for protein expression by Western blotting. Filters were immunoblotted with anti-Ang1– and anti-Ang2–specific antibodies, respectively. Representative images are shown. Experiments were performed at least 3 times.

Figure 5.

Ang2 antagonized Ang1 stimulation of endothelial cell–induced SMC migration. HUVECs were infected with a control vector, AdAng1, and AdAng1 plus AdAng2 for 48 hours. SMC migration was evaluated with the EC-SMC coculture assay. Migrated SMCs were counted in randomly selected fields. Experiments were repeated at least 3 times in more than 2 wells for each treatment. Five fields were counted for each filter in each experiment. *P < .01 compared with control. **P < .01 compared with Ang1 group.

Figure 6.

Quantification of HGF levels in Tie2 KO embryos. (A) Real time RT-PCR analysis on HGF levels was performed on wild-type and Tie2 knockout embryo yolk sacs harvested at day 9.5. Gene expression levels were normalized according to the expression of the housekeeping gene GAPDH. HGF expression was decreased 4.26-fold in Tie2-null yolk sacs compared with wild-type controls. Three pairs of embryo tissues were used in the study. RT-PCR was performed in triplicate and with 3 different dilutions of cDNA. *P < .01 compared with control. (B) RT-PCR analysis on HGF was also performed on endothelial cells isolated from the wild-type yolk sac (C166). Two sets of primers for HGF were used to increase specificity. HGF expression was confirmed (lane 2, 300 bp; lane 3, 500 bp). GAPDH was used as a control (lane 1, 440 bp).

Figure 7.

Working model based on results with HUVECs and HASMCs. Pericytes/SMCs expressed Ang1, which up-regulated HGF expression in surrounding endothelial cells and resulted in pericyte/SMC recruitment toward endothelial cells. Ang2 antagonized Ang1-induced HGF expression and inhibited Ang1-induced SMC recruitment.