Molecular Vision 2004; 10:341-350 <http://www.molvis.org/molvis/v10/a43/> Received 24 October 2003 | Accepted 7 May 2004 | Published 17 May 2004

Download Reprint

Hypoxia activates matrix metalloproteinase expression and the VEGF system in monkey choroid-retinal endothelial cells: Involvement of cytosolic phospholipase A₂ activity

Paulo Ottino,^1,2 Joelle Finley,^1,2 Eileen Rojo,³ Anna Ottlecz,³ George N. Lambrou,³ Haydee E. P. Bazan,^1,2 Nicolas G. Bazan^1,2
 
 

¹Neuroscience Center of Excellence and ²Department of Ophthalmology, Louisiana State University Health Sciences Center, New Orleans, LA; ³Novartis IBR, Basel, Switzerland

Correspondence to: Haydee E. P. Bazan LSU Neuroscience Center, 2020 Gravier Street, Suite D, New Orleans, LA, 70112; Phone: (504) 599-0877; FAX: (504) 568-5801; email: hbazan1@lsuhsc.edu

Abstract

Purpose: To determine whether the gene expression of matrix metalloproteinases (MMPs) as well as that of the pro-angiogenic cytokine vascular endothelial growth factor (VEGF) and its receptors change in response to hypoxic exposure in a primate choroid-retinal endothelial cell line, and furthermore, whether cytosolic phospholipase A₂ (cPLA₂) plays a role in this process.

Methods: Rhesus macaque choroid-retinal endothelial (RF/6A) cells were incubated under hypoxic conditions for 1, 2, 4, or 8 h prior to RNA extraction. In some experiments cells were pretreated with the cPLA₂ inhibitor AACOCF3 (10 μM) for 30 min prior to hypoxia. Changes in gene expression were determined by RT-PCR and quantified by real-time PCR for urokinase plasminogen activator (uPA), collagenase-1 (MMP-1), membrane type-1 metalloproteinase (MT1-MMP), gelatinases A and B (MMP-2, MMP-9), tissue inhibitor-2 (TIMP-2), VEGF and its receptors, Flt-1 (VEGFR-1), KDR (VEGFR-2), and neuropilin-1 (NP-1). MMP-2 secreted by the cells was evaluated by zymography. VEGF release was measured by ELISA. In tube-formation studies, endothelial cells (EC) were seeded into collagen gel, exposed to hypoxia for 4 h, then incubated under normoxic conditions for 72 h.

Results: Hypoxia triggered a three fold increase in the gene expression of MT1-MMP, MMP-2, and TIMP-2, and a ten fold increase in MMP-2 levels. Moreover it also induced tube formation in EC. Expression of uPA, MMP-1, and MMP-9 mRNA was not detected. Pretreatment with AACOCF3 abolished hypoxia-induced tube formation and MT1-MMP, MMP-2, and TIMP-2 transcription. Furthermore, hypoxia produced a significant, sustained increase in the gene expression and release of VEGF-165, the only VEGF-A isoform detected in these cells. AACOCF3 reduced the hypoxia-induced VEGF release at 8 h of hypoxia. VEGF receptors KDR and NP-1 were constitutively expressed in EC and up-regulated under hypoxic conditions.

Conclusions: In monkey choroid-retinal EC, hypoxia selectively induces MMP-2 activity. This induction is preceded by MT1-MMP, MMP-2, and TIMP-2 mRNA expression, as well as that of the VEGF-165 isoform and its receptors KDR and NP1. These increases possibly result from hypoxia-induced activation of cPLA₂ and subsequent release of arachidonic acid and its conversion to prostaglandins. These molecular changes in EC could, in part, contribute to the angiogenic response that occurs in the development of ischemic retinopathies and choroidal neovascularization.

Introduction

Pathoangiogenesis participates in the development and progression of diabetic retinopathy, exudative age-related macular degeneration (AMD), and retinopathy of prematurity [1,2]. While diabetic retinopathy is the major cause of severe vision loss in middle-aged patients, choroidal neovascularization (CNV) is the reason for the same in elderly patients. Thus, there is an increased interest in understanding what factors are important in regulating the retinal angiogenic process, so that therapeutic intervention can be developed to treat these conditions.

The retinal angiogenic process is driven by the growth factor VEGF (vascular endothelial growth factor), adhesion molecules, matrix metalloproteinases (MMPs), lipid messengers, and inflammatory cytokines [3-6]. Pathoangiogenesis involves vascular basement membrane degradation, endothelial cell (EC) migration, EC proliferation, and the formation of new capillary tubes [7]. The retina is very sensitive to hypoxia, since this neural tissue has the highest blood flow rate of the body. Retinal hypoxia stimulates hypoxia-inducible factor (HIF)-1αactivation and enhances transcription and secretion of VEGF [8]. Others have shown that the retina and vitreous isolated from patients with ischemic retinopathies have increased levels of VEGF [5,8,9]. Choroidal neovascularizaton, a consequence of abnormalities in the Bruch's membrane and retinal pigment epithelial cells, correlates with VEGF expression [10]. It was initially thought that under hypoxic conditions, retinal pigment epithelial (RPE) cells regulate choroidal EC changes through the release of pro-angiogenic factors, such as VEGF [8]. However, more recently it has been suggested that choroidal circulation is altered in patients with AMD, thus resulting in a hypoxic environment surrounding the endothelium, which could change the levels of pro-angiogenic factors, such as VEGF [11]. VEGF-A is the major isoform of the VEGF family involved in the regulation of pathoangiogenesis. The VEGF-165 and VEGF-121 isoforms are the most abundant spliced variants of VEGF-A [12]. The targets of VEGF are two distinct and homologous tyrosine kinase receptors, the fms (feline McDonough strain)-like tyrosine kinase receptor Flt-1 (VEGFR-1) and the fetal liver kinase-1 receptor Flk-1 (VEGFR-2), also referred to as KDR (kinase domain region) [13,14]. More recently, neuropilin-1 (NP-1), a receptor that potentiates the interactions between KDR and VEGF-165, has been identified [15]. Increased expression of these receptors occurs under pathologic conditions in which hypoxia is a main feature [16].

During the initiation of the angiogenic response, recruited vascular ECs promote the degradation of the extracellular matrix (ECM) through the regulated release of MMPs, which allows the cells to proliferate and migrate, producing endothelial cords that eventually form new blood vessels [17]. MMPs are a group of tightly regulated extracellular matrix-degrading enzymes implicated in the development of neovascularization [18]. These enzymes aid in the migration of vascular EC into the interstitial space, where they proliferate and eventually form new blood vessels. In situ hybridization studies have detected MMP-2 expression in the vascular EC from subfoveal fibrovascular membranes isolated from the human eye [19]. The localization of MMP-2 to areas of new vessel formation suggests an important role for this enzyme in the progression of choroidal neovascularization [20].

Hypoxia also increases PLA₂ expression, which leads to the synthesis of prostaglandins (PGs) [21,22]. PGE₂ regulates the induction of HIF-1α, one of whose functions is to induce VEGF mRNA expression by binding to hypoxia-response elements in the VEGF promoter [8,23]. Moreover PGs increase the expression of MMPs and of their inhibitors in the cornea [24]. In this study we used RF/6A cells, a spontaneously immortalized choroid-retinal EC line derived from a rhesus macaque fetus. Primate EC lines are evolutionarily close to those derived from humans and therefore are an attractive model for studies targeting neovascularization. Here we report that hypoxia stimulates a selective pattern of MMP expression and of VEGF-165 isoform and its receptors KDR and NP1 in primate choroid-retinal EC. Furthermore, we demonstrate that cytosolic PLA₂ (cPLA₂) is involved in hypoxia-induced tube formation and in the expression of MMPs and the release of VEGF.

Methods

Materials

The choroid-retinal EC line RF/6A was obtained from ATCC (Manassas, VA; CRL-1780). The endothelial origin of these cells has been corroborated by morphology, growth patterns, and the presence of factor VIII-related antigen (von Willebrand factor) [25]. We also used p-aminophenylmercuric acetate (APMA), an organomercurial activator of matrix metalloproteinases and collagen type-I (Sigma Chemical Co. St. Louis, MO); VEGF ELISA kit (R & D Systems, Minneapolis, MN); arachidonyl trifluoromethyl ketone (AACOCF3) and human granulocyte MMP-9 standard (Calbiochem La Jolla, CA); SYBR Green PCR Master Mix, TaqMan Reverse Transcription, TaqMan Ribosomal RNA control reagents (18S RNA), deoxynucleotides (dNTPs), and Ampli-Taq Gold (Perkin Elmer Branchburg, NJ); agarose, ethidium bromide, and DNA Mass Ladder (100 bp; Life Technologies, Gibco BRL Grand Island, NY); SV Total RNA Isolation system (Promega, Madison, WI); iCycler iQ optical-quality sealing tape, PCR plates, 10% zymogram ready gel containing gelatin, 1X Tris/glycine/SDS running buffer, zymogram renaturation buffer, and zymogram development buffer (Bio-Rad Hercules, CA); Vitrogen 100 solution (Collagen Corp., Cambridge, UK); and human MMP-2 standard (Chemicon, Temecula, CA).

EC culture

RF/6A cells were grown in Ham's-F12K medium (Gibco), supplemented with 5% fetal calf serum (FCS), 100 U/ml penicillin, and 0.1 mg/ml streptomycin at 37 °C in a humidified atmosphere of normal air. In these studies, cells from passages 35-55 were used. Cells (5 X 10⁵) were seeded onto Petri dishes (60 mm diameter) coated with collagen type-I (Sigma Chemical Co.). When cells were 95% confluent, the medium was replaced with fresh F12K medium containing 1% FBS and cells were incubated overnight. Fresh normoxic or hypoxic (degassed) medium containing 1% FBS was added to the cells and the cultures were placed in a Therma Forma Series II water-jacketed CO₂ incubator (Meriette, OH) and perfused with 1% O₂; 94% N₂; 5% CO₂ or incubated under normoxic conditions at 37 °C for 1, 2, 4, or 8 h. For MMP-2 activity studies, the EC were subjected to hypoxic conditions for 2 or 4 h followed by incubation under normoxic conditions for an additional 12 h.

For the ELISA experiments, RF/6A cells were seeded at a density of 7.0x10⁴ cells/ml onto 24 well Falcon tissue-culture plates precoated with collagen and were grown in a humidified cell-culture incubator (95% air; 5% CO₂) until cells were 95% confluent. For cPLA₂ inhibitor AACOCF3 studies, the drug was prepared by serial dilution in sterile F12K culture medium to obtain a concentration of 10 μM AACOCF3 in 0.01% ethanol. Afterwards, the medium of the cells was removed by aspiration, and the cells were then replenished with F12K medium plus 1% FBS with or without 10 μM AACOCF3. Control cultures contained 0.01% ethanol. The cells were pre-incubated for 30 min with inhibitor prior to hypoxic exposure.

RNA extraction

RNA was extracted as described in the SV Total RNA Isolation kit (Promega). Briefly, 175 μl lysis buffer was added to each dish. RF6/A cells were scraped and immediately transferred to RNAse-free tubes. Tissue samples were gently vortexed and mixed with 350 μl SV total RNA dilution buffer. After heating for 3 min at 70 °C, extracts were centrifuged at 14,000x g for 10 min. Supernatant was removed and mixed with 200 μl 95% ethanol, followed by transfer to spin baskets and centrifugation at 14,000x g for 1 min. Final RNA extracts were eluted from the spin columns in a total volume of 30 μl nuclease-free water. The concentration and purity of RNA were determined by spectrophotometry. Typical yields of RNA varied from 5-10 μg per sample. All RNA preparations had an OD₂₆₀:OD₂₈₀ ratio of 1.8 to 2.0.

Reverse transcription and PCR

RNA (2.5 μg) from RF/6A cells was reverse-transcribed, using a TaqMan Reverse-Transcription kit from PerkinElmer (Branchburg, NJ), in a total reaction volume of 25 μl, containing 1X first-strand buffer, 1 unit of RNase, 4 mM dNTPs, 2.5 μM random hexamers, 0.01 mM DTT, and 400 Units of MMLV-reverse transcriptase. The mixture was heated at 25 °C for 10 min, followed by 30 min at 42 °C and a final cycle at 95 °C for 5 min. For PCR analysis, aliquots (5 μl) of cDNA were added to 25 μl reaction mixture containing 1X PCR buffer (10 mM Tris-HCL; pH 8.3; 50 mM KCl; 1.5 mM MgCl₂; 0.001% (w/v) gelatin), 0.4 μM of each dNTP, and 0.4 μM of sense and anti-sense primers, and amplified in a GeneAmp 9600 series PCR machine (Perkin Elmer, Norwalk, CT). Table 1 lists the primer sequences, PCR conditions, and product size for each gene studied. Duplicate reactions lacking reverse transcriptase and RNA were included as negative controls. Amplification products were resolved on a 2% agarose gel containing 1 μg/ml ethidium bromide, and the product size was determined by comparison to a 100 bp ladder run on the gel.

Real-Time PCR

A 2.5 μg sample of total RNA was transcribed into cDNA as described above. Amplification was conducted in a total reaction volume of 25 μl containing 5 μl cDNA, 12.5 μl 2X SYBR Green PCR Master Mix, and 0.25 μl of each primer (10 μM; Table 1). The primers were identical to those used for RT-PCR. Samples of the internal control 18S rRNA and the gene of interest were run, respectively, in duplicate and triplicate on the same plate. Amplifications were performed on an iCycler IQ Multi-color Real-Time PCR Detection System (Biorad). All quantitations were normalized to the 18S rRNA endogenous control, and changes in gene expression were reported as fold increases relative to untreated controls as previously described [26]. Controls lacking an RNA template and reverse transcriptase were set up to determine whether fluorescent contaminants were present in the sample. The absence of nonspecific amplification products was confirmed by agarose gel electrophoresis and melt-curve analysis.

Gelatin zymography assay

Conditioned medium (10 μl) from control and hypoxia treated (2 or 4 h) cells was collected at 12 h and diluted 1:1 with zymography sample loading buffer (Biorad). Electrophoresis was carried out on a 10% zymogram ready gel (Biorad) containing gelatin for 1 h at 100 volts in a 1X Tris/glycine/SDS running buffer (Biorad). This was followed by incubation in zymogram renaturation buffer (Biorad) for 1 h at room temperature. The buffer was then replaced with 1X zymogram development buffer (Biorad) and incubation continued overnight. The following day, gels were stained with Coomassie blue and the positions of the enzyme were visualized as clear bands against a uniformly dark-stained background. The identity of the bands was confirmed by comparison to APMA-activated human MMP-2 (Chemicon). Images were recorded on a Biorad Gel Doc-1000 fitted with a white-light conversion screen. To verify that proteolytic activities detected were due to MMPs, duplicate zymograms were developed in zymogram renaturation buffer containing 10 mM of MMP inhibitor 1,10-phenanthroline [24].

Enzyme-Linked immunosorbent assay (ELISA)

Cell culture supernatant (200 μl) harvested at each time point was evaluated for VEGF release according to the manufacturer's instructions (R&D Systems). Prior to ELISA analysis, samples were centrifuged at 14,000 rpm for 1 min. Three ELISA experiments were conducted, with each sample performed in duplicate. Colorimetric analysis was performed on a Perkin-Elmer microtiter plate reader with Soft Pro computer hardware (Bucher Biotec, Basel, Switzerland). The optical density of the samples was measured at 450 nm and 570 nm to account for possible background interference in the values. A standard curve was generated from which the concentrations (pg/ml) of VEGF were obtained.

EC tube formation in collagen gel culture

For these studies, a collagen gel mixture was prepared by using 8 volumes of collagen type 1 (Sigma Chem. Co), 1 volume of 0.2 N NaOH, 200 mmol/l HEPES, and one volume of 10X RPMI (Sigma Chem. Co.) containing 5 μg/ml fibronectin and 5 μg/ml laminin and then was added to a 24 well plate (400 μl/well). The plates were incubated at 37 °C for 3 h. RF/6A cells in F12K medium containing 5% FBS were seeded at 7.5x10⁴ cells/well at 37 °C for 24 h. Then the medium was removed by aspiration and cells were covered by collagen mixture (120 μl/well). After 1 h incubation at 37 °C, 1 ml fresh medium containing 1% FBS was added to each well. For hypoxia experiments, medium was replaced with hypoxic (degassed) medium containing 1% FBS and perfused with 1% O₂; 94% N₂; 5% CO₂ for 4 h at 37 °C. This was followed by an additional incubation under normoxic conditions for a total of 72 h. For cPLA₂ inhibitor studies, cells were pre-incubated for 30 min with 10 μM AACOCF3 in F12K medium containing 1% FBS and 0.01% ethanol, prior to hypoxic exposure. Then the cells were exposed to hypoxia for 4 h followed by further incubation under normoxic conditions for 72 h. In control dishes, medium was replaced with F12K medium supplemented with 1% FBS and incubated at 37 °C in a 5% CO₂; 95% air mixture for 72 h. In all experiments, five fields per sample were photographed at 100x magnification by phase-contrast microscopy using a Nikon Optiphot-2 (Nikon Laboratories, Melville, NY).

Statistical analysis

The results are expressed as means obtained from three to five separate experiments. All statistical procedures were performed with the programs and procedures in the SAS language (SAS v.5, SAS Institute, Cary, NC).

Results

MT1-MMP, MMP-2, and TIMP-2 expression is stimulated by hypoxia

Because MMP expression and activity contribute significantly to tube formation [17], we determined the effect of hypoxia on the gene expression of MMPs and their inhibitors in RF/6A cells by RT-PCR using the primers listed in Table 1. Under normoxic conditions, the cells expressed MT1-MMP, MMP-2, and TIMP-2 mRNA; however, MMP-1, MMP-9, and uPA expression was not observed (data not shown). Hypoxia triggered similar kinetics of expression for MMP-2, TIMP-2, and MT1-MMP mRNA, with an increase in expression occurring at 2 h, followed by a decrease to basal levels by 4 h (Figure 1A). PCR amplification using total RNA samples prior to the RT-step (RT-minus) revealed no genomic DNA contamination. We used real-time PCR to compare hypoxic induction of MMP-2, TIMP-2, and MT1-MMP to that in normoxic controls (expressed as fold increases) [26]. There was a 2.7 fold increase in MMP-2 (p<0.01) gene expression and a three fold increase in MT1-MMP (p<0.02) and TIMP-2 (p<0.05) mRNA expression after 2 h hypoxia (Figure 1B).

Hypoxia selectively induces MMP-2 activity

Choroid-retinal EC were found to constitutively express low levels of MMP-2 gelatinase activity, whereas MMP-9 activity was undetected (Figure 2). Hypoxic treatment of EC for 2 or 4 h followed by reoxygenation for 12 h produced a ten fold increase in MMP-2 levels when compared to normoxic control.

Hypoxia selectively induces VEGF-165 isoform expression

Recent studies in our laboratory using primers designed from the mouse VEGF-165 coding region demonstrated the presence of VEGF expression in RF/6A cells [23]. Here we used primer sets designed from monkey VEGF-165 mRNA [27] and confirmed the presence of low levels of this isoform in the EC (Figure 3A). By using oligonucleotide primers specific for all VEGF-A isoforms (VEGF-121, 165, 189, 206) and designed from a highly conserved sequence in intron 4 and the 3' untranslated region of the VEGF gene (Table 1), we found that VEGF-165 was the only isoform consistently present in these cells (Figure 3B). Furthermore, we showed by real-time PCR analysis that hypoxic treatment of EC for 1-8 h significantly increased VEGF expression in these cells. At 1 h of hypoxia, a 1.9 fold increase (p<0.001) in VEGF mRNA expression was observed (Figure 3C). This increase was maintained at 2 (p<0.001) and 4 h (p<0.05) of hypoxia. At 8 h of hypoxia, a 2.6 fold increase (p<0.01) in VEGF gene expression occurred.

Hypoxia selectively induces KDR and co-receptor neuropilin-1 expression

The angiogenic process might be regulated not only through VEGF induction but also through a concomitant change in receptor expression. We therefore investigated the role of hypoxia on the expression of Flt-1, KDR, and its co-receptor NP1 in EC. Flt-1 expression was consistently not detected in these cells. In low-passage-number (35 to 40) cells, low constitutive levels of Flt-1 mRNA were detected, while at high passage numbers (45 to 55), no receptor expression was found by RT-PCR. Moreover, hypoxia produced no changes in Flt-1 mRNA expression (data not shown). EC subjected to 1-2 h hypoxia showed increased KDR expression when compared to normoxic controls. At 4 and 8 h hypoxia, no changes in KDR expression occurred (Figure 4A). The co-receptor of KDR, NP-1, was also expressed in these cells and its mRNA was induced at 2-4 h hypoxia, followed by a decline to basal levels at 8 h of hypoxia.

Real-time PCR analysis of KDR expression at 1 and 2 h of hypoxia showed significant 1.72 (p<0.001) and 2.6 (p<0.05) fold increases, respectively, as compared to normoxic controls (Figure 4B). The NP-1 receptor expression was increased 1.6 fold (p<0.02) at 2 h and 2.4 fold (p<0.05) at 4 h of hypoxia, respectively.

Inhibition of cPLA₂ blocks the hypoxic induction of MMPs

Hypoxia increases arachidonic acid (AA) release from EC through PLA₂ activation and induces cyclooxygenase-2 expression (COX-2), which catalyzes the formation of PGs from AA [21,22]. PGs derived from COX-2 activation are inducers of MMPs in corneal epithelial cells [24]. We hypothesized that, under hypoxic conditions, PLA₂ activity could be involved in the induction of MMP gene expression in EC. To test this hypothesis, experiments were conducted under hypoxic conditions in the presence of the cPLA₂ inhibitor AACOCF3. Results from these experiments showed a significant (p<0.01) inhibition of MT1-MMP, MMP-2, and TIMP-2 mRNA expression (Figure 5).

Inhibition of cPLA₂ blocks VEGF release from EC

Results from the ELISA experiments revealed an approximate 3.5 fold increase in VEGF release at 8 h of hypoxia (p<0.001). This increase was significantly inhibited in the presence of 10 μM AACOCF3 (p<0.05), which decreases VEGF release by approximately 30% when compared to hypoxia plus vehicle treated cells (Figure 6). Inhibitor treated controls showed no inhibition when compared to vehicle treated controls (data not shown). In conclusion, this result suggests that hypoxia-induced VEGF secretion is dependent on PLA₂ activation.

Hypoxia induces tube formation in EC in culture

To demonstrate the effect of hypoxia on tube formation, RF/6A cells were grown in collagen gels and incubated under hypoxic or normoxic conditions for 4 h followed by a further incubation under normoxic conditions for 72 h. Cells grown in the standard normoxic incubator produced a small number of cellular aggregates and short, tube-like structures (Figure 7). The extent of tube formation was determined by oxygen tension. The hypoxic challenge applied for 4 h intensively increased the tube formation. The onset of tubular structure formation became apparent after 1 day, and reached a plateau at 3 days post-hypoxia. The presence of the cPLA₂ inhibitor significantly decreased tube formation.

Discussion

Studies to date have focused on retinal neovascularization using in vivo models of laser-induced photocoagulation (for choroidal neovascularization) [20,28], hypoxia (for retinal neovascularization) [29], or isolation of bovine or mouse vascular EC [30,31]. In our studies we have employed a primate choroid-retinal cell line to investigate the effect of hypoxia on the induction of several angiogenic factors and their receptors as well as the role of inflammation in this process. With this approach we eliminated the contribution of the retinal pigment epithelium to the effects of hypoxia on the EC. The VEGF-165 isoform, a mitogenic and chemotactic factor capable of inducing angiogenesis, was the only VEGF isoform expressed in these cells, and its gene induction was up-regulated between 1 and 8 h of hypoxia. These findings are in agreement with previous studies using experimental models of laser-induced choroid neovascularization in monkeys that show increased VEGF mRNA expression in the choroidal vascular endothelium [32]. We also found that a significant increase in VEGF release occurred at 8 h of hypoxia, probably the result of early induction at the mRNA level.

VEGF-165 binds to two homologous tyrosine kinase receptors, Flt-1 and KDR. [13,14]. It is believed that hypoxia-dependent local factors might regulate angiogenesis not only via VEGF induction but also by a concomitant increase in specific VEGF receptor expression [33]. Our results show that Flt-1 is weakly expressed in EC and that its expression is dependent on the cell-passage number. KDR and NP-1 mRNA expressions, on the other hand, are induced by hypoxia. Flt-1 has a higher affinity for VEGF than does KDR but KDR possesses a higher tyrosine kinase activity and is the major mediator of EC proliferation [14]. Over-expression of Flt-1 could lead to decreased binding of VEGF-165 to KDR, thereby reducing signaling through low-affinity, high-tyrosine-kinase-activity KDR receptors. Down-regulation or absence of Flt-1 expression during hypoxia would redirect VEGF signaling towards the KDR receptor.

Our results are in agreement with a very recent study in which bovine choroidal EC express Flt-1 and KDR mRNA [30]. Furthermore, when compared to retinal microvascular EC, the choroidal EC expressed higher KDR and lower Flt-1 mRNA levels [30]. Although in our cell line we cannot distinguish between the choroidal and retinal EC, these findings suggest that the responses observed in our cell line could be a consequence of choroidal EC. Hypoxia enhanced NP-1 expression in these cells at 4 h of exposure, a phenomenon that correlates with the induction of VEGF-165. Recent studies have demonstrated NP-1 immunostaining in choroid neovascular membranes of patients with AMD [34]. One possibility is that NP-1 receptor up-regulation in choroidal cells promotes a complex formation with KDR, thereby augmenting VEGF-165 signaling through this receptor [35]. Based on the results above, we propose that hypoxia triggers in EC the induction of VEGF-165 expression, which, through an autocrine mechanism, binds to the up-regulated receptor KDR and co-receptor NP-1, and promotes cell proliferation (Figure 8).

We showed that under hypoxic conditions, there is no detectable uPA, MMP-1, or MMP-9 mRNA expression. The lack of uPA expression could be attributed to the predominantly choroidal nature of these cells, since previous studies have shown that choroid is not a significant source of uPA [36,37]. Inflammatory cells are thought to be the predominant suppliers of MMP-9 [38]. In fact, in a recent study using a mouse model of laser-induced choroidal neovascularization, MMP-9 expression correlated with the appearance of inflammatory cells [39]. With respect to MMP-1 expression, a previous study using fibrovascular membranes isolated from patients with age-related macular degeneration (AMD) showed no expression of this enzyme [19]. Hypoxia triggers the expression of MT1-MMP, TIMP-2, and MMP-2 mRNA in the EC. In addition, we showed that the activity of MMP-2 increased. Since MMP-2 is essential for angiogenesis [20,40] it is reasonable to assume that up-regulation of this protease facilitates capillary basement membrane breakdown and the development of CNV during AMD. It has been suggested that MT1-MMP expressed by EC can activate MMP-2 [41,42]. Studies have shown that MT1-MMP forms a tri-molecular complex with MMP-2 and TIMP-2 at the cell surface. This complex brings the MT1-MMP active site into proximity with the pro-form of MMP-2, resulting in its activation (Figure 8) [43]. Therefore, the presence of MT1-MMP, MMP-2, and TIMP-2 mRNA in choroidal and retinal EC and their up-regulation by hypoxia suggest that this complex could be involved in the activation of MMP-2, degradation of the ECM, and the growth of vessels during angiogenesis.

One important finding was that cPLA₂ inhibition prevented a hypoxia-induced increase in MT1-MMP, MMP-2, and TIMP-2 gene expression. Hypoxia-stimulated PLA₂ activity and arachidonic acid (AA) release in HUVEC leads to increased prostaglandin synthesis [21,22]. Moreover, previous studies in our laboratory have demonstrated that (a) prostaglandins can induce the expression of proteolytic enzymes involved in ECM catabolism [24] and (b) secretory PLA₂ inhibition in EC prevents COX-2 activation and PGE₂ release as well as tube formation [23]. These findings taken together suggest that in response to hypoxia, an increase in MMP expression is mediated in EC, in part through the activation of PLA₂ enzymes, AA release, and PG synthesis. Further studies are required to determine the role of each PLA₂ in the activation of these MMPs.

We also demonstrated that under hypoxic conditions, inhibition of cPLA₂ interferes with the release of VEGF by EC. This finding agrees with previous observations that HIF-1αactivity is modulated by PGs such as PGE₂, whose expression is up-regulated upon PLA₂ activation [44]. Interestingly, HIF-1α-binding sites have been found in the promoter region of VEGF and are important activators of VEGF gene expression [23,45]. The role of HIF-1αin this scenario may be multi-faceted: VEGF release may not only be dependent on direct HIF-1αregulation at its promoter, but also at the level of prostaglandins. Our results strongly suggest that the role of cPLA₂ in the neovascularizing phenotype may involve its ability to directly influence blood-vessel growth and EC migration in a hypoxia-dependent manner. In retrospect, hypoxia might regulate angiogenesis through various means: VEGF induction, a concomitant increase in the expression of specific VEGF receptors on EC, and increased MMP expression and a subsequent rise in ECM degradation.

In conclusion, we have demonstrated that hypoxia, a common trigger of ischemic retinopathies and CNV in AMD patients [1-5], induces the gene expression and release of VEGF, its receptors, KDR and NP-1, as well as MT1-MMP, MMP-2, and TIMP-2 in monkey choroid-retinal EC in culture. These results collectively suggest that the increased expression of these proteins may play key roles in facilitating the growth and invasiveness of abnormal blood vessels. Moreover, the hypoxia-induced increase in MMP expression was reversed by pre-treatment with the PLA₂ inhibitor AACOCF3, which supports the concept that hypoxia-induced MMP expression and tube formation are mediated, in part, by an inflammatory response that triggers PLA₂ activation and AA release. Figure 8 is a schematic diagram summarizing our results. These findings may be important in understanding the pathogenesis of CNV and ischemic retinopathies, which may ultimately lead to the development of effective therapeutic strategies.

Acknowledgements

This work was supported by United States Public Health Service grant EY04928 from the National Eye Institute, National Institutes of Health, Bethesda, Maryland (to HEPB), and by a grant from Novartis Ophthalmics, Basel, Switzerland.

References

1. Garner A. Ocular angiogenesis. Int Rev Exp Pathol 1986; 28:249-306.

2. Bressler SB, Silva JC, Bressler NM, Alexander J, Green WR. Clinicopathologic correlation of occult choroidal neovascularization in age-related macular degeneration. Arch Ophthalmol 1992; 110:827-32.

3. D'Amore PA. Mechanisms of retinal and choroidal neovascularization. Invest Ophthalmol Vis Sci 1994; 35:3974-9.

4. Pierce EA, Avery RL, Foley ED, Aiello LP, Smith LE. Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neovascularization. Proc Natl Acad Sci U S A 1995; 92:905-9.

5. Casey R, Li WW. Factors controlling ocular angiogenesis. Am J Ophthalmol 1997; 124:521-9.

6. Leibovich SJ, Polverini PJ, Shepard HM, Wiseman DM, Shively V, Nuseir N. Macrophage-induced angiogenesis is mediated by tumour necrosis factor-alpha. Nature 1987; 329:630-2.

7. Folkman J. How is blood vessel growth regulated in normal and neoplastic tissue? G.H.A. Clowes memorial Award lecture. Cancer Res 1986; 46:467-73.

8. Pe'er J, Shweiki D, Itin A, Hemo I, Gnessin H, Keshet E. Hypoxia-induced expression of vascular endothelial growth factor by retinal cells is a common factor in neovascularizing ocular diseases. Lab Invest 1995; 72:638-45.

9. Kliffen M, Sharma HS, Mooy CM, Kerkvliet S, de Jong PT. Increased expression of angiogenic growth factors in age-related maculopathy. Br J Ophthalmol 1997; 81:154-62.

10. Campochiaro PA, Hackett SF. Ocular neovascularization: a valuable model system. Oncogene 2003; 22:6537-48.

11. Ross RD, Barofsky JM, Cohen G, Baber WB, Palao SW, Gitter KA. Presumed macular choroidal watershed vascular filling, choroidal neovascularization, and systemic vascular disease in patients with age-related macular degeneration. Am J Ophthalmol 1998; 125:71-80.

12. Houck KA, Ferrara N, Winer J, Cachianes G, Li B, Leung DW. The vascular endothelial growth factor family: identification of a fourth molecular species and characterization of alternative splicing of RNA. Mol Endocrinol 1991; 5:1806-14.

13. de Vries C, Escobedo J, Ueno H, Houck K, Ferrara N, Williams LT. The fms-like tyrosine kinase, a receptor for vascular endothelial growth factor. Science 1992; 255:989-91.

14. Terman BI, Dougher-Vermazen M, Carrion ME, Dimitrov D, Armellino DC, Gospodarowicz D, Bohlen P. Identification of the KDR tyrosine kinase as a receptor for vascular endothelial cell growth factor. Biochem Biophys Res Commun 1992; 187:1579-86.

15. Brown LF, Berse B, Jackman RW, Tognazzi K, Guidi AJ, Dvorak HF, Senger DR, Connolly JL, Schnitt SJ. Expression of vascular permeability factor (vascular endothelial growth factor) and its receptors in breast cancer. Hum Pathol 1995; 26:86-91.

16. Tuder RM, Flook BE, Voelkel NF. Increased gene expression for VEGF and the VEGF receptors KDR/Flk and Flt in lungs exposed to acute or to chronic hypoxia. Modulation of gene expression by nitric oxide. J Clin Invest 1995; 95:1798-807.

17. Moses MA. The regulation of neovascularization of matrix metalloproteinases and their inhibitors. Stem Cells 1997; 15:180-9.

18. Kvanta A, Shen WY, Sarman S, Seregard S, Steen B, Rakoczy E. Matrix metalloproteinase (MMP) expression in experimental choroidal neovascularization. Curr Eye Res 2000; 21:684-90.

19. Steen B, Sejersen S, Berglin L, Seregard S, Kvanta A. Matrix metalloproteinases and metalloproteinase inhibitors in choroidal neovascular membranes. Invest Ophthalmol Vis Sci 1998; 39:2194-200.

20. Berglin L, Sarman S, van der Ploeg I, Steen B, Ming Y, Itohara S, Seregard S, Kvanta A. Reduced choroidal neovascular membrane formation in matrix metalloproteinase-2-deficient mice. Invest Ophthalmol Vis Sci 2003; 44:403-8.

21. Michiels C, Renard P, Bouaziz N, Heck N, Eliaers F, Ninane N, Quarck R, Holvoet P, Raes M. Identification of the phospholipase A(2) isoforms that contribute to arachidonic acid release in hypoxic endothelial cells: limits of phospholipase A(2) inhibitors. Biochem Pharmacol 2002; 63:321-32.

22. Michiels C, Arnould T, Knott I, Dieu M, Remacle J. Stimulation of prostaglandin synthesis by human endothelial cells exposed to hypoxia. Am J Physiol 1993; 264:C866-74.

23. Lukiw WJ, Ottlecz A, Lambrou G, Grueninger M, Finley J, Thompson HW, Bazan NG. Coordinate activation of HIF-1 and NF-kappaB DNA binding and COX-2 and VEGF expression in retinal cells by hypoxia. Invest Ophthalmol Vis Sci 2003; 44:4163-70.

24. Ottino P, Bazan HE. Corneal stimulation of MMP-1, -9 and uPA by platelet-activating factor is mediated by cyclooxygenase-2 metabolites. Curr Eye Res 2001; 23:77-85.

25. Lou DA, Hu FN. Specific antigen and organelle expression of a long-term rhesus endothelial cell line. In Vitro Cell Dev Biol 1987; 23:75-85.

26. Ottino P, Taheri F, Bazan HE. Platelet-activating factor induces the gene expression of TIMP-1, -2, and PAI-1: imbalance between the gene expression of MMP-9 and TIMP-1 and -2. Exp Eye Res 2002; 74:393-402.

27. Shima DT, Gougos A, Miller JW, Tolentino M, Robinson G, Adamis AP, D'Amore PA. Cloning and mRNA expression of vascular endothelial growth factor in ischemic retinas of Macaca fascicularis. Invest Ophthalmol Vis Sci 1996; 37:1334-40.

28. Rakic JM, Lambert V, Munaut C, Bajou K, Peyrollier K, Alvarez-Gonzalez ML, Carmeliet P, Foidart JM, Noel A. Mice without uPA, tPA, or plasminogen genes are resistant to experimental choroidal neovascularization. Invest Ophthalmol Vis Sci 2003; 44:1732-9.

29. Ferrara N. Vascular endothelial growth factor. The trigger for neovascularization in the eye. Lab Invest 1995; 72:615-8.

30. Brylla E, Tscheudschilsuren G, Santos AN, Nieber K, Spanel-Borowski K, Aust G. Differences between retinal and choroidal microvascular endothelial cells (MVECs) under normal and hypoxic conditions. Exp Eye Res 2003; 77:527-35.

31. Paradis H, Liu CY, Saika S, Azhar M, Doetschman T, Good WV, Nayak R, Laver N, Kao CW, Kao WW, Gendron RL. Tubedown-1 in remodeling of the developing vitreal vasculature in vivo and regulation of capillary outgrowth in vitro. Dev Biol 2002; 249:140-55.

32. Ishibashi T, Hata Y, Yoshikawa H, Nakagawa K, Sueishi K, Inomata H. Expression of vascular endothelial growth factor in experimental choroidal neovascularization. Graefes Arch Clin Exp Ophthalmol 1997; 235:159-67.

33. Brogi E, Schatteman G, Wu T, Kim EA, Varticovski L, Keyt B, Isner JM. Hypoxia-induced paracrine regulation of vascular endothelial growth factor receptor expression. J Clin Invest 1996; 97:469-76.

34. Cui JZ, Hinz BJ, Greve MD, Potter MJ, Hornan D, Samad A, To E, Matsubara JA. Expression of neuropilin-1 in choroidal neovascular membranes. Can J Ophthalmol 2003; 38:41-5.

35. Mamluk R, Gechtman Z, Kutcher ME, Gasiunas N, Gallagher J, Klagsbrun M. Neuropilin-1 binds vascular endothelial growth factor 165, placenta growth factor-2, and heparin via its b1b2 domain. J Biol Chem 2002; 277:24818-25.

36. Wang Y, Gillies C, Cone RE, O'Rourke J. Extravascular secretion of t-PA by the intact superfused choroid. Invest Ophthalmol Vis Sci 1995; 36:1625-32.

37. Hackett SF, Campochiaro PA. Modulation of plasminogen activator inhibitor-1 and urokinase in retinal pigmented epithelial cells. Invest Ophthalmol Vis Sci 1993; 34:2055-61.

38. Coussens LM, Tinkle CL, Hanahan D, Werb Z. MMP-9 supplied by bone marrow-derived cells contributes to skin carcinogenesis. Cell 2000; 103:481-90.

39. Lambert V, Munaut C, Jost M, Noel A, Werb Z, Foidart JM, Rakic JM. Matrix metalloproteinase-9 contributes to choroidal neovascularization. Am J Pathol 2002; 161:1247-53.

40. Itoh T, Tanioka M, Yoshida H, Yoshioka T, Nishimoto H, Itohara S. Reduced angiogenesis and tumor progression in gelatinase A-deficient mice. Cancer Res 1998; 58:1048-51.

41. Cao J, Sato H, Takino T, Seiki M. The C-terminal region of membrane type matrix metalloproteinase is a functional transmembrane domain required for pro-gelatinase A activation. J Biol Chem 1995; 270:801-5.

42. Lichte A, Kolkenbrock H, Tschesche H. The recombinant catalytic domain of membrane-type matrix metalloproteinase-1 (MT1-MMP) induces activation of progelatinase A and progelatinase A complexed with TIMP-2. FEBS Lett 1996; 397:277-82.

43. Axelrad TW, Deo DD, Ottino P, Van Kirk J, Bazan NG, Bazan HE, Hunt JD. Platelet-activating factor (PAF) induces activation of matrix metalloproteinase 2 activity and vascular endothelial cell invasion and migration. FASEB J 2004; 18:568-70.

44. Semenza GL. HIF-1: mediator of physiological and pathophysiological responses to hypoxia. J Appl Physiol 2000; 88:1474-80.

45. Iyer NV, Kotch LE, Agani F, Leung SW, Laughner E, Wenger RH, Gassmann M, Gearhart JD, Lawler AM, Yu AY, Semenza GL. Cellular and developmental control of O2 homeostasis by hypoxia-inducible factor 1 alpha. Genes Dev 1998; 12:149-62.

46. Lafleur MA, Hollenberg MD, Atkinson SJ, Knauper V, Murphy G, Edwards DR. Activation of pro-(matrix metalloproteinase-2) (pro-MMP-2) by thrombin is membrane-type-MMP-dependent in human umbilical vein endothelial cells and generates a distinct 63 kDa active species. Biochem J 2001; 357:107-15.

47. Kim I, Ryan AM, Rohan R, Amano S, Agular S, Miller JW, Adamis AP. Constitutive expression of VEGF, VEGFR-1, and VEGFR-2 in normal eyes. Invest Ophthalmol Vis Sci 1999; 40:2115-21. Erratum in: Invest Ophthalmol Vis Sci 2000; 41:368.

48. Gerber H-P, Condorelli F, Park J, Ferrara N. Differential transcriptional regulation of the two vascular endothelial growth factor receptor genes. J Biol Chem 1997; 272:23659.

49. Nayak NR, Critchley HO, Slayden OD, Menrad A, Chwalisz K, Baird DT, Brenner RM. Progesterone withdrawal up-regulates vascular endothelial growth factor receptor type 2 in the superficial zone stroma of the human and macaque endometrium: potential relevance to menstruation. J Clin Endocrinol Metab 2000; 85:3442-52.

50. Ishida S, Shinoda K, Kawashima S, Oguchi Y, Okada Y, Ikeda E. Coexpression of VEGF receptors VEGF-R2 and neuropilin-1 in proliferative diabetic retinopathy. Invest Ophthalmol Vis Sci 2000; 41:1649-56.