Angiotensin II Induces Expression of the Tie2 Receptor Ligand, Angiopoietin-2, in Bovine Retinal Endothelial Cells

Primary BREC cultures were isolated by homogenization and a series of filtration steps, as previously described (29). Primary BRECs were grown on fibronectin (Sigma, St. Louis, MO)-coated dishes (Iwaki Glass, Tokyo) containing Dulbecco’s modified Eagle’s medium (DMEM) with 5.5 mmol/l glucose, 10% plasma-deprived horse serum (PDHS) (Wheaton, Pipersville, PA), 50 mg/l heparin, and 50 U/l endothelial cell growth factor (Boehringer Mannheim, Indianapolis, IN). Bovine aortic endothelial cells (BAECs) were also isolated from bovine aorta and cultured in DMEM containing 5% calf serum and 10% PDHS. The cells were cultured in 5% CO₂ at 37°C. Endothelial cell homogeneity was confirmed by immunoreactivity with anti-factor VIII antibodies analyzed by confocal microscopy. For AII receptor–antagonist studies, BRECs were pretreated with AT1 antagonist, losartan (Merck Research Laboratories, Rahway, NJ), or AT2 antagonist PD123319 (Research Biochemicals International, Natick, MA) for 15 min followed by stimulation with 10 nmol/l AII for 4 h. To determine the roles of protein kinase C (PKC) and mitogen-activated protein kinase (MAPK) on AII-stimulated Ang2 mRNA expression, BRECs were pretreated with 10 μmol/l GF109203X (GFX; Calbiochem-Novabiochem, La Jolla, CA) or 10 μmol/l PD98059 (Research Biochemicals International), respectively, followed by stimulation with 10 nmol/l AII or 160 nmol/l phorbol 12-myristate 13-acetate (PMA) (Sigma).

Total RNA was isolated from individual tissue culture plates using guanidium thiocyanate. Northern blot analysis was performed on 20 μg total RNA after 1% agarose-2 mol/l formaldehyde gel electrophoresis and subsequent capillary transfer to Biodyne nylon membranes (Pall BioSupport, East Hills, NY), and ultraviolet crosslinking was performed using a FUNA-UV-LINKER (FS-1500; Funakoshi, Tokyo). Radioactive probes were generated using Amersham Megaprime labeling kits and ³²P-labeled dATP (Amersham, Arlington Heights, IL). Blots were prehybridized, hybridized, and washed in 0.5 × SSC, 5% SDS, at 65°C with four changes over 1 h in a rotating hybridization oven (TAITEC, Koshigaya, Japan). All signals were analyzed using a densitometer (BAS-2000II; Fuji Photo Film, Tokyo), and lane-loading differences were normalized using a 36B4 cDNA probe, which hybridizes to acidic ribosomal phosphoprotein PO (30). cDNA probes for human Ang1 and Ang2 were synthesized by reverse transcriptase–polymerase chain reaction (PCR) (31). For Ang1 and Ang2 cDNAs, a standard PCR was performed (PCR optimizer kit; Invitrogen) using 5′-AGA ACC ACA CGG CTA CCA TGC T-3′ (Ang1 sense primer corresponding to nucleotides 1671–1692), 5′-TGT GTC CAT CAG CTC CAG TTG C-3′ (Ang1antisense primer), 5′-AGC TGT GAT CTT GTC TTG GC-3′ (Ang2 sense primer corresponding to nucleotides 1377–1396), and 5′-GTT CAAGTC TCG TGG TCT GA-3′ (Ang2 antisense primer corresponding to nucleotides 1802–1821).

To determine whether the increase in Ang2 mRNA was caused by increased mRNA stability, BRECs were exposed to 4 μmol/l actinomycin-d (Wako, Osaka, Japan) after 4 h incubation with vehicle or 10 nmol/l AII. The total RNA was then extracted, and Northern blot analyses were performed.

BRECs were treated with vehicle or 10 nmol/l AII for 4 h, then the cells were lysed in solubilizing buffer (10 mmol/l Tris-HCl, 10 mmol/l NaCl, 3 mmol/l MgCl₂, and 0.5% NP-40), and the nuclei were isolated. ATP CTP, GTP (500 mmol/l each), and 3.7 MBq of ³²P-labeled UTP (110 Tbq/mmol; Amersham) were added to the nuclear suspension (100 μl) and incubated for 30 min. The samples were extracted with phenol/chloroform and precipitated. cDNA probes (Ang2 and 36B4, 10 μg each) were slot blotted onto nitrocellulose filters (Schleicher & Schuel, Keene, NH) and hybridized with the precipitated samples of equal counts per minute per milliliter in hybridization buffer at 45°C for 48 h. The filters were washed, and the radioactivity was measured using the densitometer (BAS-2000II; Fuji Photo Film).

BRECs were treated with 10 nmol/l AII, 160 nmol/l PMA, or vehicle for 5 h, and the culture media were replaced with labeling media (DMEM without methionine and cysteine: 100 μCi ³⁵S-labeled methionine and cysteine) supplemented with AII, PMA, or vehicle. After a 3-h incubation, the cells were lysed in solubilizing buffer (50 mmol/l Hepes, pH 7.4, 10 mmol/l EDTA, 100 mmol/l NaF, 10 mmol/l Na pyrophosphate, 1% Triton X-100, 10 mmol/l NaVO₄, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 2 mmol/l phenyl methylsulfonyl fluoride) at 4°C for 1 h. To clear the protein extract, protein A-Sepharose (20 μl of a 50% suspension; Pharmacia Biotech, Uppsala, Sweden) was added to the cell lysate and incubated for 1 h, followed by centrifugation and collection of the supernatant. A specific antibody to Ang2 (50 ng/ml; Santa Cruz Biotechnology, Santa Cruz, CA) was added to the protein samples (500 μg) and rocked at 4°C for 1.5 h; 10 μg protein A-Sepharose was then added, and the sample was rocked another 1.5 h at 4°C. Protein A-Sepharose antigen-antibody conjugates were separated by centrifugation, washed five times, and boiled for 3 min in Laemmli sample buffer for denaturation. The samples were separated by 7.5% SDS-polyacrylamide gel (Bio-Rad Laboratories, Richmond, CA), and the gel was vacuum dried. Results were visualized and quantified by a densitometer (Fuji Photo Film). To obtain specificity of the signals, we made a control in which antibodies had been preincubated for 1 h with a fivefold molar excess of the respective blocking peptides (Santa Cruz Biotechnology).

To identify the effect of AII in Ang2 expression in vivo, we made an AII-induced corneal angiogenesis model as previously reported (32). We made a slow-release formulation polyacrylamide gel containing AII 10 umol/l and implanted it in a pocket made in the corneal stroma of pigmented Long-Evans rats. For controls, polyacrylamide gel alone was implanted. Neovascularization was assessed on day 14 by direct examination with a surgical microscope. The eyes were enucleated after scarifying the animal. The eyes were fixed in 4% paraformaldehyde solution. A frozen block of the specimen was made using Tissue Tek (Sakura, Torrance, CA), and serial frozen sections were cut from the specimens in a cryostat (LeicaReichert, Vienna, Austria) at −20°C and mounted on slides. The sections were fixed in acetone at 4°C for 10 min, and immunohistochemical staining was performed according to the manufacturers protocol (Elite ABC kit; Vector Laboratories, Burlingame, CA). The specimens were incubated overnight at 4°C with goat polyclonal anti–angiopoietin-2, 1:500 dilution (Santa Cruz Biotechnology, Santa Cruz, CA), and then a standard indirect immunoperoxidase procedure using the ABC kit was performed with AEC (3-amino-9-ethylcarbazole) (DAKO, Carpenteria, CA) as the substrate. For the negative control, the primary antibody preincubated with the immunizing peptide (Santa Cruz Biotechnology) was used. Other staining procedures were the same as previously described.

Determinations were performed in triplicate, and experiments were performed at least three times. Results are expressed as means ± SE, unless otherwise indicated. For multiple treatment groups, a factorial analysis of variance followed by Fisher’s least-significant difference test was performed. Results were considered significant when P was <0.05.

To investigate the effect of AII on Ang1 and Ang2 mRNA expression, BRECs were treated with 10 nmol/l AII for the indicated times, and Northern blot analysis was performed. From several independent experiments, the effect of AII on Ang2 mRNA expression was time-dependent, with a maximal 2.6 ± 0.5-fold increase at 4 h (P < 0.01) that progressively diminished up to 12 h (Fig. 1A). AII also stimulated Ang2 mRNA expression in a dose-dependent manner with an EC₅₀ of ∼3 nmol/l and a maximal 2.4 ± 0.2-fold (P < 0.01) increase at a concentration of 10 nmol/l (Fig. 1B). In both experiments, AII did not affect the expression of Ang1 mRNA (Figs. 1A and B). We also examined the Ang1 and Ang2 mRNA expression in BAECs. The expression of Ang2 mRNA in BAECs was very faint, and no changes could be detected after AII stimulation (Fig. 1C).

To characterize the AII receptor subtype that is responsible for Ang2 mRNA induction, Northern blot analyses were performed using the AT1 antagonist losartan or the AT2 antagonist PD123319 for 15 min before stimulation with 10 nmol/l AII (Fig. 2). In contrast with PD123319, losartan inhibited AII-induced Ang2 mRNA expression (Fig. 2). The observed response suggests that AII stimulates Ang2 induction mainly through the AT1 receptor.

We investigated whether the AII-induced increase of the Ang2 mRNA level is mediated through regulation of mRNA stability or transcription. To determine whether AII affects the half-life of Ang2 mRNA, we performed Northern blot analysis using a de novo gene transcription inhibitor, 4 μmol/l actinomycin-d with or without 10 nmol/l AII. AII did not change the stability and half-life of Ang2 mRNA (Fig. 3A). However, 10 nmol/l AII increased the rate of Ang2 gene transcription by 2.1-fold in BRECs that were examined by nuclear run-on analysis (Fig. 3B). These data suggest that AII stimulated the increase of the Ang2 mRNA level mainly through an increase of the transcriptional rate.

To determine the role of PKC and MAPK in AII-induced Ang2 mRNA expression, BRECs were pretreated with a highly selective PKC inhibitor, bisindolylmaleimide (GFX), or a specific MAPK inhibitor (PD98059) followed by treatment with 10 nmol/l AII and PMA, a direct PKC stimulator. PMA (0–500 nmol/l) increased the expression of Ang2 mRNA in a dose-dependent manner (data not shown), and the effect (160 nmol/l) was completely inhibited by 10 μmol/l GFX (Fig. 4A). The same concentration of GFX inhibited 10 nmol/l AII-induced Ang2 mRNA expression completely (Fig. 4A). Treatment with PD98059 reduced AII-induced Ang2 mRNA expression in a dose-dependent manner (0–50 μmol/l), and maximal inhibition (69.4 ± 15.6%) at 10 μmol/l was observed (Fig. 4B). PD98059 (10 μmol/l) also inhibited PMA (160 nmol/l)-induced Ang2 expression by 76.8 ± 8.1% (Fig. 5). These data suggest that AII upregulates Ang2 mRNA expression totally through the PKC-dependent pathway, and MAPK is also involved in the downstream PKC pathway.

Because VEGF upregulates Ang2 expression without affecting Ang1 expression in the same cell type (31), VEGF induction by AII might be involved in the observed AII-dependent regulation of Ang1 and Ang2. To investigate this possibility, the effect of AII on VEGF expression was determined. Northern blot analysis revealed that 10 nmol/l AII did not alter the mRNA level of VEGF during 2–24 h in BRECs. The data suggest that VEGF induction is not involved in the Ang1 and Ang2 mRNA expression (Fig. 6).

To determine whether the increase in Ang2 mRNA expression was accompanied by an increase in new protein synthesis, we precipitated the ³⁵S-labeled cell lysates of BRECs with anti-Ang2 antibody and analyzed them by SDS-PAGE. The molecular mass of Ang2 protein was reported to range from 55 to 70 kDa because of glycosylation (13). In our experiments, several bands that agree with the molecular mass were detected by immunoprecipitation of a rabbit anti-human Ang2 antibody (Fig. 7). Specificity controls were made with nonstarved cells and performed by incubating the antibodies with an excess of the blocking peptide before detection (Figs. 7A and B), and a band ∼55 kDa was detected as Ang2 protein. The level of the band was increased by adding AII as well as PMA (Figs. 7D and E). These data indicate that AII increased protein synthesis of Ang2 in BRECs.

To determine whether the AII-induced Ang2 expression is involved in neovascularization in vivo, we made a rat corneal angiogenesis model. AII significantly stimulates new vessel formation (Fig. 8A); in contrast, polyacrylamide gel alone failed to stimulate any significant new vessel formation (Fig. 8B). Immunohistochemical study revealed that Ang2 was strongly positive in these AII-induced vessels in the corneal stroma, whereas no staining was observed in the control cornea. The data suggest that AII induces Ang2 in vivo.

Recent in vitro studies demonstrated that Ang1 induces potent chemotaxis, weak but positive mitogenesis, and capillary sprouts in endothelial cells, which confirm the critical role of Ang1 in angiogenesis (33). Although Ang2 has been shown to be a natural antagonist to Ang1 in developmental angiogenesis and cultured endothelial cells (12,34), when co-administered with VEGF in vivo, Ang2 can potentiate VEGF-induced angiogenesis, probably by enhancing the initiation of neovascularization, whereas Ang1 may potentiate vascular network maturation (17). In the present study, we demonstrated that AII stimulates Ang2 expression, but has no significant effect on Ang1 expression in retinal microvascular endothelial cells. This finding might implicate a substantial effect of AII in promoting pathologic angiogenesis in diabetic retinopathy and other ischemic neovascular diseases in which VEGF-dependent angiogenesis is a predominant pathologic change (35–37).

Previous studies have implicated RAS as a key factor in the development of diabetic retinopathy. The severity of retinopathy correlates with RAS activity, and the treatment of diabetic patients with an ACE inhibitor reduces both the rate of progression of nonproliferative retinopathy and the development of proliferative changes (25). However, it remains unclear what mechanism underlies the interaction of RAS and the development of diabetic retinopathy. The present findings might implicate a novel role of AII in the development of diabetic retinopathy, i.e., AII may enhance pathological angiogenesis by a selective induction of Ang2 in diabetic retinopathy.

Ang2 expression has recently been demonstrated in a subset of angiogenic vessels in glioblastoma and hypervascularized specimens of hepatocellular carcinoma, suggesting its biologic importance in tumor-associated angiogenesis (38,39). Furthermore, recent studies also demonstrated that Ang2 is upregulated by angiogenic stimuli, such as hypoxia, VEGF, and basic fibroblast growth factor, and is downregulated by transforming growth factor β-1, which is associated with vessel stabilization (31,40). These data suggest that Ang2 has a prominent role in pathologic angiogenesis. The present data showing selective upregulation of Ang2 by AII agrees with previously published studies and may strongly define the role of Ang2 in retinal angiogenesis in diabetic retinopathy.

Because VEGF regulates both angiopoietins in the similar way (31,40), one could imagine that VEGF possibly mediates the observed AII-dependent regulation of Ang1 and Ang2 mRNA expression. However, this is not the case, because AII did not alter the VEGF mRNA level in BRECs. Furthermore, the time-course study demonstrated that Ang2 induction by AII started at 2 h and peaked at 4 h and returned to the basal level at 12 h. The observed prompt induction of Ang2 suggests that VEGF production is not associated with the AII-stimulated Ang2 induction.

The dose-response study demonstrated an EC₅₀ of ∼3 nmol/l and a maximal increase at 10 nmol/l AII stimulation. The AII concentrations used in our experiments were considerably higher than those found in the plasma and vitreous of diabetic patients (41). However, the presence of angiotensinogen, renin, and ACE has been reported in ocular tissues (42,43), and the AII retinal concentration in ocular tissues is much higher than that in plasma (44). These data support the presence of intraocular RAS. AII may act as an autocrine or paracrine factor in retinal tissues.

Two major angiotensin receptor subtypes have been defined: the AT1 and AT2 receptors (45,46). Most of the actions of AII, including its growth-promoting effect, are mediated by the AT1 receptor (18). The actions of the AT2 receptor have not been well defined (18), but recent studies suggest that it is involved in both apoptosis and regeneration (47,48). In our study, AII-induced Ang2 mRNA expression was mediated through the AT1 receptor, and the AT2 receptor was not involved.

Nuclear run-on assays and experiments using actinomycin-d indicated that the primary effect of AII is to increase transcription of the Ang2 gene. These data might suggest that transcriptional regulation of this gene is mediated through a transacting transcription factor. Further studies are necessary to elucidate in detail the mechanism of transcriptional regulation of Ang2, including cis- and trans- activation.

The major signaling events activated by AII through the AT1 receptor include production of diacylglycerol and inositol 1,4,5-triphosphate by phospholipase C–dependent hydrolysis of phosphatidylinositol 4,5-bisphosphate (18). Diacylglycerol activates PKC. AII also activates MAPK, which is thought to play a critical role in cellular proliferation (49–51). In many receptor types, including RTK and G protein–coupled receptors, MAPK activation is induced through PKC activation (52). In adrenal glomerulosa cells, AII was reported to act as a mitogen in its primary culture through the G protein–coupled AT1 receptor, and it activates MAPK via PKC and Ras/Raf-1 kinase (53). In the present study, we examined the involvement of PKC and MAPK in AII-stimulated Ang2 expression. PMA induced Ang2 mRNA expression in a dose-dependent manner (the maximal effect was observed at 160 nmol/l), and this effect was completely inhibited by adding 10 μmol/l GFX (Fig. 4A). The same GFX concentration completely inhibited AII-induced Ang2 mRNA expression, suggesting that the AII-induced increase of Ang2 mRNA is mediated totally by a PKC-dependent signaling pathway (Fig. 4A). MAPK inhibitor also blocked AII-induced Ang2 mRNA expression in a dose-dependent manner, with a maximal inhibition of 69.4 ± 15.6% (Fig. 4B). MAPK inhibitor blocked PMA-induced Ang2 mRNA expression by 76.8 ± 8.1% (Fig. 5); this inhibitory effect was well correlated with that in AII-induced Ang2 mRNA expression. These data suggest that MAPK is also involved in the downstream pathway of PKC; most AII-induced Ang2 expression is PKC-activated and MAPK-dependent, and the remainder is PKC-dependent without MAPK activation.

The effect of AII on Ang2 protein synthesis was studied by immunoprecipitation of ³⁵S-labeled cell lysates of BRECs. In our experiments, a rabbit anti-human Ang2 antibody detected several bands from 55 to 70 kDa. Because a band at 55 kDa was reduced when antibodies with an excess of the blocking peptide were used before detection, it can be concluded that that band represents Ang2 protein and the other bands probably result from cross-reactivity of the antibody. This is further confirmed by the finding that the signal of the band is increased by PMA stimulation, as previously reported in bovine microvascular endothelial cells (40). The signal of the band is increased by AII stimulation, which indicates that AII stimulates Ang2 protein synthesis. The AII effect on Ang2 protein synthesis was also confirmed in vivo by immunohistochemistry in a model of corneal angiogenesis.

In our experiments, Ang2 expression in BAECs was undetectable using a similar Northern blot analysis. Because the expression of Tie2 was similar in both cell types (data not shown), the difference in Ang2 expression might suggest that Ang2 regulation is more important in retinal microvasculature than in large vasculature. However, further investigations, including histologic studies and tests of other cell types, are needed to confirm this theory.

In addition to our previous findings that AII promotes VEGF-dependent angiogenesis (27,28), the data presented here demonstrate that RAS upregulates both the VEGF/VEGF receptor system and Ang2 expression in retinal microvascular cells. Such upregulation of an angiogenesis-promoting environment most likely contributes to the development of VEGF-dependent angiogenesis in ischemic neovascular diseases, such as diabetic retinopathy. AII-induced upregulation of these molecules is similarly mediated through the AT1 receptor. AT1 blockade, as well as the already proven ACE inhibitor, may thus be an effective modality to prevent the development of diabetic retinopathy.

Support for this study was provided by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of the Japanese Government and by the Japan Association for Inhibition of Blindness.

We gratefully acknowledge the expert technical assistance of Dr. Tsutomu Yasukawa in corneal angiogenesis model.