Vegf mediates the apoptotic activity of tgf-β1 in vivo
To assess the biological significance of our findings in vivo we tested the effect of neutralizing antibody to VEGF on TGF-β1 induction of endothelial cell apoptosis in the chicken embryo choriollantoic membrane (CAM). For this purpose we treated chicken embryo CAMs with TGF-β1 in the absence or presence of neutralizing VEGF antibody. We used a TGF-β1 concentration that induces angiogenesis in the CAM (Yang and Moses, 1990). After 6 h incubation we analyzed CAM sections by TUNEL and immunostaining with antibody to CD31, an endothelial cell marker. Analysis of the sections by confocal microscopy (Fig. 6) showed that TGF-β1 strongly upregulated the number of TUNEL-positive (apoptotic) cells in the CAM with an effect comparable to that of TNF-α, a potent inducer of endothelial cell apoptosis (Ferrari et al., 2006). Virtually all apoptotic cells colocalized with CD31-positive (endothelial) cells, consistent with previous reports that TGF-β1 selectively induces apoptosis in endothelial cells but not in smooth muscle cells or pericytes (Pollman et al., 1999b). VEGF antibody strongly downregulated the number of TUNEL-positive cells induced by TGF-β1, whereas control n.i. IgG had no such effect (Fig. 6). Thus, these results showed that TGF-β1 induction of endothelial cell apoptosis in vivo requires VEGF signaling.
Figure 6
Inhibition of vegf signaling blocks tgf-β1 induction of endothelial cell apoptosis in vivo
Vegf mediates tgf-β1 induction of angiogenesis in vivo
We then tested the effect of neutralizing anti-VEGF antibody on TGF-β1 induction of blood vessel formation in the CAM. As described (Yang and Moses, 1990), TGF-β1 induced formation of new capillaries from pre-existing CAM vessels with an effect comparable to that of FGF-2. The vessels formed in the presence of TGF-β1 were relatively large and had few branches, whereas those induced by FGF-2 were thin and had many branches (Fig. 7 A and B) (Yang and Moses, 1990). To confirm that our VEGF antibody neutralized chicken VEGF, we tested if it blocked angiogenesis induction in the CAM by FGF-2. We have previously shown that FGF-2 upregulates endothelial cell expression of VEGF, and that antibody to VEGF blocks FGF-2 induction of angiogenesis in vivo (Ferrari et al., 2006; Seghezzi et al., 1998). VEGF antibody blocked the angiogenic effect of FGF-2 in the CAM, showing that it neutralized chicken VEGF (Fig. 7 A and B). We therefore tested the effect of VEGF antibody on TGF-β1 induction of angiogenesis. TGF-β1 increased the number of CAM vessels approximately 2.5-fold, an effect comparable to that obtained with FGF-2 and consistent with previous findings (Yang and Moses, 1990). Antibody to VEGF completely abrogated this effect, showing that TGF-β1 induces angiogenesis in vivo through the autocrine and/or paracrine action of VEGF.
Figure 7
A and B. Inhibition of VEGF signaling blocks TGF-β1 induction of angiogenesis in vivo
Because VEGF mediates the apoptotic activity of TGF-β1 on the CAM endothelial cells, these results indicated that the rapid apoptotic effect mediated by VEGF is required for TGF-β1 induction of angiogenesis in vivo. Therefore, we treated CAMs with TGF-β1 in the presence or absence of z-VAD, a general caspase inhibitor that blocks in vitro angiogenesis (Fig. 2, panel q). As shown in Figure 7 C and D, z-VAD completely abrogated the angiogenic activity of TGF-β1 with an effect comparable to that of VEGF antibody. In our assays the growth factors were administered to the CAM in the presence of hydroxycortisone acetate (2.5 mg/ml) as described (Brooks et al., 1999), to block the inflammatory response. Therefore, under our experimental conditions, endothelial cell apoptosis did not result from the inflammatory response, and VEGF derived not from inflammatory cells but from the epithelial and/or endothelial cells of the CAM. Thus, these results showed that rapid and transient upregulation of VEGF- VEGFR2-mediated endothelial cell apoptosis is required for TGF-β1 induction of angiogenesis in vivo.
Go to: