The Urokinase/Urokinase Receptor System in Retinal Neovascularization: Inhibition by Å6 Suggests a New Therapeutic Target

Abstract

purpose. The objective of the study was to determine the role of urokinase (uPA) and the urokinase receptor (uPAR) in retinal angiogenesis, and whether loss of uPAR or the inhibition of uPA/uPAR interactions could suppress the extent of retinal neovascularization in an animal model of ischemic retinopathy.

methods. Retinal neovascularization was induced by exposing newborn mice to 75% oxygen on postnatal day 7 for 5 days, followed by exposure to room air on days 12 to 17. The expression of uPAR in the retina was investigated by RT-PCR and immunohistochemistry. The role of uPAR in ischemic retinopathy was investigated by quantitating the extent of retinal neovascularization in the uPAR^−/− mouse. The effects of inhibiting the uPA/uPAR interaction on the development of retinal neovascularization were studied in this animal model with a uPA-derived peptide, Å6. Animals were treated with an intraperitoneal injection of Å6 at a dose of 5, 10, or 100 mg/kg once a day on days 12 to 16. Control animals included oxygen-exposed mice treated with similar amounts of PBS only on days 12 to 16. The effect of Å6 on the expression of uPAR in the retina was examined by real-time RT-PCR.

results. The expression of uPAR mRNA was upregulated in experimental animals during the period of angiogenesis and was localized to endothelial cells in the superficial layers of the retina. The uPAR^−/− mouse demonstrated normal retinal vascular development; however, the absence of functional uPAR resulted in a significant reduction in the extent of retinal neovascularization. Histologic analysis of mice treated with Å6 peptide showed significant inhibition of retinal neovascularization, and the response was dose dependent. The RT-PCR analysis of the retinas of the Å6-treated animals showed a greater than twofold decrease in uPAR expression.

conclusions. Expression of the urokinase receptor uPAR is essential to the development of retinal neovascularization. Inhibition of the activity of uPAR suppresses retinal neovascularization, possibly through a reduction in cell-associated proteolytic activity, cell signaling, or cell-matrix adhesion necessary for cell migration during angiogenesis. The uPA/uPAR interaction may be an important therapeutic target in the management of proliferative retinopathies.

Diseases of the retina such as diabetic retinopathy and retinopathy of prematurity are characterized by the formation of abnormal new blood vessels. These proliferative retinopathies are postulated to result from hypoxic insults to the retina that stimulate the angiogenic response and may lead to complications that include vitreous hemorrhage, retinal detachment, and eventual blindness. ^{1 2}  

The process of angiogenesis in the retina and other tissues is characterized by distinct phases or activities including an initial response to locally produced angiogenic factors and signals. This event is followed by a rapid upregulation of extracellular proteinases that facilitate the breakdown of the capillary basal lamina and the migration of endothelial cells into and through the surrounding extracellular matrix. After the proliferation of endothelial cells, new capillary tubes are formed and stabilized through the action of specific growth factors and interactions with surrounding pericytes. ^{3 4 5}  

The main focus of our laboratory has been the role of urokinase (uPA) and other extracellular proteinases in facilitating the development of new vessels in the retina and the potential for these enzymes to serve as targets for the development of new antiangiogenic therapies. ^{6 7} We have reported an increased level of urokinase in the ocular tissues of patients with proliferative diabetic retinopathy. Epiretinal neovascular membranes obtained from these patients at the time of surgery showed significantly elevated levels of this proteinase. ⁷ A similar finding was reported in the retinal tissue of mice with ischemia-induced retinal neovascularization. ⁶ The importance of urokinase in the regulation of cell migration has been shown in numerous studies and is partially dependent on its localization at the cell surface by the urokinase receptor, uPAR (CD87). ^{8 9 10} The objectives of this study were to determine the role of uPAR in the development of retinal neovascularization and the ability of Å6, a peptide derived from the non-receptor-binding region of urokinase, to inhibit this process. The Å6 peptide inhibits the interaction of uPA with uPAR in a noncompetitive manner, inhibits tumor cell invasion in vitro, and has antiangiogenic and antitumor activity in vivo. ^{11 12 13}  

Materials and Methods

Animal Models

Specific pathogen-free C57BL/6J mice were bred at the University of New Mexico Animal Research Facility. All experiments were consistent with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Experimental Animals.

Seven-day-old C57BL/6J mice were placed into an oxygen chamber maintained at 75% oxygen until postnatal day (P)12. ¹⁴ Mice were removed from the chamber on P12 and maintained in room air until P17. By P17, retinal neovascularization was present in 100% of the experimental animals. 

Control Animals.

Newborn mice exposed only to room air served as the control. Mice without a functional uPAR gene (uPAR^−/− mice) ¹⁵ were obtained from the Jackson Laboratory (Bar Harbor, ME), and a colony is maintained at the University of New Mexico. In some experiments, these mice underwent the identical angiogenesis-inducing oxygen protocol as did the C57BL/6J mice. 

Histologic Quantitation of Retinal Neovascularization

Eyes from control and experimental animals were fixed overnight in 10% neutral buffered formalin and embedded in paraffin. Serial sections (6 μm) parallel to the optic nerve were mounted on gelatin-coated slides, stained with Griffonia simplicifolia lectin I-fluorescein isothiocyanate (GSA-FITC; Vector Laboratories, Burlingame, CA) and coverslipped with mounting medium containing diamidinophenylindole (DAPI) (Vectashield; Vector Laboratories). Every third section was examined with a fluorescence microscope, and nuclei of GSA-positive cells on the vitreous side of the inner limiting membrane of the retina were counted manually. Sections containing the optic nerve were excluded because of epiretinal vasculature that may be mistaken for neovascular nuclei. Data are expressed as the number of nuclei per section. Two-sample t-tests were used to make comparisons between experimental and control animals. 

Immunostaining

Eyes were collected from experimental (n = 4) and control (n = 2) animals and fixed overnight in zinc fixative (IHC; BD PharMingen, San Diego, CA) after removal of the cornea and lens. The eye cups were processed and embedded in paraffin and sectioned at 6 μm. Sections were pretreated with hydrogen peroxide, blocked with 10% normal goat serum and incubated with either an anti-mouse uPAR antibody (R&D Systems, Minneapolis, MN) or anti-mouse CD31 antibody (BD PharMingen). Sections were washed, incubated with a peroxidase-labeled secondary antibody, and reacted with diaminobenzidine. 

ADPase Staining

The retinal vasculature was examined in day 17 C57BL/6J (n = 3) and uPAR^−/− mice (n = 3) by a modification of the lead sulfide technique. ¹⁶ Eyes were enucleated and the retinas were removed and fixed overnight in 10% formalin. The retinas were incubated in an adenosine diphosphatase (ADPase)-containing solution and cleared with ammonium sulfide. The retinas were viewed as wholemounts, and six random images of each retina were collected with a digital camera. The images were analyzed on computer (MetaMorph; Universal Imaging Corp., Downingtown, PA) for the percentage of retinal area occupied by vascular profiles. 

Å6 Treatment

In some studies, experimental C57BL/6 mice were injected intraperitoneally twice daily from P12 to P16 with increasing doses (10, 50, or 100 mg/kg) of the Å6 peptide (Ångstrom Pharmaceuticals, San Diego, CA; n = 4 mice per dose of Å6 peptide). As a control, some experimental animals received an equal volume of PBS (n = 4 mice). Eyes were collected at P17 and analyzed histologically for the extent of retinal neovascularization. In some studies, experimental mice were treated with Å6 (100 mg/kg) from P12 to P14 and the retinas collected on P15 (n = 2 mice). The retinal mRNA was then extracted for PCR analysis. 

Semiquantitative and Real-Time RT-PCR

Total RNA was extracted from control and experimental retinas and used to generate first-strand cDNA using reverse transcriptase (Superscript; Gibco, Grand Island, NY). PCR was performed with primers specific for the uPAR mRNA (5′-TTCCACCGAATGGCTTCCAG-3′ and 5′-AGGCAATGAGGCTGAGTTGAGC-3′). In some cases, the relative level of the specific product was standardized to a coamplified invariant internal standard (18s ribosomal RNA). Ten microliters of each reaction was examined by agarose gel electrophoresis and ethidium bromide staining and band densities quantitated on computer (Alpha Imager 2200 software package; Alpha Innotech Corp., San Leandro, CA). For real-time RT-PCR analysis, cDNA samples were analyzed using a fluorescently-labeled probe and primers specific for mouse uPAR and 18s RNA. Primers were designed on computer (Primer Express; Applied Biosystems) and purchased from Integrated DNA Technologies (Coralville, IA). Reactions were run in triplicate, and mean values were characterized by comparing threshold cycle (C_t) values. ¹⁷ Transcripts of the 18s RNA gene were used as an endogenous control, with each unknown sample normalized to 18s content. 

Results

Expression of uPAR during Retinal Neovascularization

We have shown that uPA activity is significantly upregulated in the retinas of animals with retinal neovascularization. ⁶ To gain a better understanding of the role of the uPA/uPAR system in this process, we examined the retinas of mice with active neovascularization by RT-PCR analysis. The angiogenic process begins shortly after the return to 20% oxygen on P12 and is maximal by P17. ¹⁴ Expression of uPAR mRNA was detected in experimental mice at P17 during the active angiogenic period. This was in contrast to that seen in normal control mice, which demonstrated undetectable expression of uPAR at the same age (Fig. 1) . 

Localization of uPAR during Retinal Neovascularization

To identify the cell type(s) responsible for the expression of uPAR in the retina, we next examined control and experimental mice by immuohistochemistry for the localization of uPAR protein. Adjacent sections of the retina from P17 mice were reacted with antibodies specific for mouse uPAR or mouse CD31 to identify endothelial cells. The uPAR protein was localized to vessel profiles within the superficial portion of the retina. Some of these uPAR-expressing cells extended to the vitreous side of the inner limiting membrane. Control P17 mice showed no uPAR staining, consistent with the absence of uPAR mRNA. Adjacent sections stained with an endothelial-specific antibody confirmed that uPAR was expressed exclusively by endothelial cells of the retina (Fig. 2) . 

Retinal Neovascularization in uPAR^−/− Mice

We next investigated the role of uPAR in facilitating the formation of new vessels by comparing the extent of oxygen-induced retinal neovascularization in normal C57BL6 mice and homozygous uPAR^−/− mice. ¹⁵ Normal C57BL6 mice exposed to 75% oxygen from P7 to P12 showed development of extensive neovascularization by P17, as determined by the presence of nuclei on the vitreous side of the inner limiting membrane (Fig. 3) . This was in dramatic contrast to that observed in the uPAR^−/− mice treated with the same experimental protocol. Retinal neovascularization in the uPAR^−/− mice was quantitatively reduced by 73% (14.52 ± 0.86 vs. 53.4 ± 1.14 neovascular nuclei per section per eye) compared with normal mice. 

The loss of uPAR activity and its effect on vascularization appeared to be limited to abnormal neovascularization and not the normal development of the retinal vasculature. Retinal wholemounts from C57BL6 and uPAR^−/− mice were stained for ADPase activity to label the vasculature. Both the C57BL6 and uPAR^−/− mice exhibited a dense retinal vascular pattern that abruptly ended at the peripheral avascular retina. There were no quantitative differences in the percentage of the retinal area occupied by ADPase-positive vascular profiles in the uPAR^−/− mice compared with the C57BL6 (25.15% ± 1.23% and 28.41% ± 2.48%, respectively; Fig. 4 ). 

Effect of Å6 on Retinal Neovascularization

The extent of retinal neovascularization was quantitated in mice treated with increasing concentrations of the Å6 peptide. The mice received an intraperitoneal injection of either the Å6 peptide (10, 50, or 100 mg/kg) or PBS twice daily until P17. Neovascularization was inhibited to a small degree at the lower doses of Å6 (10 and 50 mg/kg), although the inhibition was not significant. At the highest dose (100 mg/kg), there was significant inhibition of the extent of retinal neovascularization in the mice. The inhibition by Å6 approached the extent of reduction in the uPAR^−/− mice (63% with Å6 treatment vs. 73% in uPAR^−/− mice; Fig. 5 ). 

Treatment of experimental mice with the Å6 peptide resulted in a decrease in the level of uPAR mRNA in the retinal tissues, as detected by semiquantitative RT-PCR. This result was subsequently confirmed using real-time PCR in which a 2.3-fold decrease in uPAR mRNA was detected in the Å6-treated animals (Fig. 6) . 

Discussion

Numerous extracellular enzymes have been shown to play important roles in the process of angiogenesis. One in particular, urokinase, is a serine proteinase produced as a single-chain proenzyme that is converted to a two-chain active enzyme. The active form of the enzyme is capable of converting plasminogen to plasmin, which can degrade many components of the extracellular matrix. ¹⁸ Through the activation of plasminogen, uPA is also capable of initiating a proteolytic cascade resulting in the activation of members of the matrix metalloproteinase family that act on additional components of the matrix. ¹⁹ The efficiency by which uPA converts plasminogen is enhanced when uPA is bound to its cell surface receptor uPAR. ²⁰ The interaction of uPA with uPAR occurs within a region of uPA known as the growth factor or epidermal growth factor (EGF)-like domain, and this binding may be supported by additional domains within the A-chain of the enzyme. ^{11 21} Numerous studies suggest that the localization of uPA activity to the cell surface significantly affects the motile activity of both normal and neoplastic cells. ^{22 23 24 25 26} The uPA/uPAR interaction has also been shown to influence cell behavior independent of any proteolytic activity. These nonenzymatic activities of the uPA/uPAR complex include changes in gene expression, signaling events, and cell-matrix adhesion. ^{27 28 29}  

In the present study, uPAR was significantly upregulated in the retina during the process of angiogenesis in experimental mice. The uPAR protein was localized to vascular endothelial cells within the superficial layers of the retina and on tufts of new capillaries on the surface of the inner limiting membrane that extended into the vitreous cavity. Control P17 mice demonstrated no uPAR expression, as evidenced by both immunostaining and RT-PCR. This result was initially surprising, because most normal retinal vasculature development takes place after birth, and uPAR may be expected to play a role. The absence of uPAR expression and staining may be explained by the fact that the migration of endothelial cells to the peripheral avascular region of the retina is complete or nearly complete by P17. ³⁰ It should also be noted that the formation of the superficial retinal vasculature is postulated to occur by the process of vasculogenesis, as opposed to angiogenesis that may depend on or use different cellular mechanisms. ³⁰  

The uPAR^−/− mouse was used as a model to explore further the role of this protein in retinal angiogenesis. The uPAR^−/− mouse was generated by elimination of the third exon of the uPAR gene that encodes a portion of the uPA-binding domain. ¹⁴ Disruption of the uPAR gene in this way resulted in the production of mutant mRNA transcripts in the uPAR^−/− mice that did not give rise to any detectable or functional uPAR protein. Analysis and comparison of normal C57BL6 and uPAR^−/− mice revealed no difference in the pattern or extent of retinal vascularization in the P17 mouse, based on ADPase staining and quantitation. These results together with the absence of expression of uPAR suggest that uPAR may not be necessary for normal vasculogenesis of the retina or that some level of compensation by other proteins exists. 

The requirement of uPAR in abnormal or pathologic retinal angiogenesis was confirmed in studies in which the extent of experimentally induced angiogenesis was significantly reduced in uPAR^−/− mice when compared with normal C57BL6 mice. uPAR may be necessary to facilitate the formation of new vessels from the existing superficial retinal vasculature, or it may be required for the migration of other cell types contributing to the new vessels. Recent studies have suggested that some endothelial cells of vessels that form during pathologic angiogenesis are derived from a population of bone marrow stem cells. ^{31 32} It is unclear at the present time whether the model used in these studies uses a population of stem cells as part of the mechanism for oxygen-induced retinal neovascularization. However, if this turns out to be the case, then the absence of functional uPAR may prevent the mobilization of endothelial stem cells from the bone marrow, thus contributing to a decrease in the extent of angiogenesis. 

The current treatment for the proliferative retinopathies is panretinal laser photocoagulation, which in many cases is effective but not optimal. In addition to possible recurrence and progression of the disease requiring repeated laser treatments, there are also significant side effects that include the loss of peripheral and night vision. Several experimental approaches have been undertaken to develop new alternative therapies designed to curtail the development and/or progression of retinal neovascularization without the side effects. These approaches include targeting growth factors, ^{33 34 35 36 37 38 39} cell surface receptors, ⁴⁰ and proteinases. ^{6 41} Many of these approaches may be useful, either in conjunction with the current therapy or as alternatives to the laser treatment. 

Our studies further explored the possibility that the Å6 peptide may provide an efficient means of disrupting the uPA/uPAR system, so as to inhibit the extent of retinal angiogenesis in the mouse model. Experimental animals treated with Å6 peptide demonstrated a significant reduction in the extent of retinal angiogenesis compared with PBS-treated animals. The reduction of neovascularization was nearly the same as that in the uPAR^−/− mice suggesting that disruption of uPAR function may be an efficient and effective means of regulating the extent of abnormal new vessel formation. Å6 has been shown to inhibit the interaction of uPA with uPAR in a noncompetitive, allosteric manner. ¹¹ The peptide has been administered to animals and shown to inhibit tumor growth and metastasis with no detectable toxic side effects. ¹¹ A recent study ⁴² demonstrated that the administration of Å6, in a syngeneic model of breast cancer, inhibited tumor growth through a decrease in blood vessel density and increased tumor-cell death. The authors further demonstrated that a mammary adenocarcinoma cell line treated with Å6 showed decreased TGF β activity and expression of the VEGF receptor flk-1. Although this was not specifically demonstrated in the animals, it may be a direct or indirect result of inhibition of the uPA/uPAR interaction, leading to the inhibition of new vessel formation. 

In the present study, the level of uPAR mRNA was quantitatively reduced in experimental mice treated with Å6. This reduction in uPAR mRNA may simply reflect an overall reduction in the number of uPAR-expressing endothelial cells in the treated tissues or could indicate a direct or indirect effect on the expression of uPAR itself. Further studies of Å6-treated cultured cells may help resolve this issue. 

Taken together, these studies demonstrate that the uPA/uPAR system is important in facilitating the development of abnormal new vessels in the retina of mice. The uPA/uPAR interactions may represent a new target for the development of antiangiogenic therapies based on the results showing that the Å6 peptide was effective in inhibiting the development of abnormal new vessels in the retina with no observable toxic side effects. This peptide may be useful alone or in combination with other current therapies to inhibit the progression or the recurrence of the proliferative retinopathies. 

 Presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2002.

 Supported by National Eye Institute Grant R01-EY12604-04 (AD).

 Submitted for publication November 14, 2002; revised January 8, 2003; accepted January 21, 2003.

 Disclosure: P.G. McGuire, None; T.R. Jones, Ångstrom Pharmaceuticals (E, F); N. Talarico, None; E. Warren, None; A. Das, Ångstrom Pharmaceuticals (C, R)

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Retinal expression of uPAR mRNA. Representative gel of results from RT-PCR analysis of retinas from P17 control (lanes 1–6) and P17 experimental (lanes 7–12) mice. uPAR mRNA was present in the retina of three experimental littermates with active neovascularization (lanes 7–9) and was undetectable in control animals (lanes 1–3). 18s RNA product was detected in all samples (lanes 4–6 and 10–12).

Figure 2.

 

Localization of uPAR expression during retinal neovascularization. Representative adjacent sections of the retina from a P17 experimental mouse retina stained with antibody to uPAR (A) or CD31 (B) or with no primary antibody (C). The expression of uPAR was restricted to endothelial cells within the superficial layer of the retina and neovascular tufts on the surface of the inner limiting membrane (A, B, arrows).

Figure 3.

 

Absence of the urokinase receptor uPAR reduced the extent of retinal neovascularization in the mouse. (A) Representative section of the retina from an experimental oxygen-treated P17 C57BL6 mouse demonstrating numerous neovascular tufts on the surface of the retina (arrows.) (B) A similar section from an experimental oxygen-treated P17 uPAR^−/− mouse with many fewer vascular tufts (arrow). (C) Quantitation of neovascularization in C57BL6 and uPAR^−/− mice. The uPAR^−/− mice demonstrated 73% less neovascularization compared with the normal C57BL6 mice. Values are the mean ± SEM for n = 4 mice in each group (eight eyes, 15–20 sections/eye). *Significantly less than in C57BL6 mice, P < 0.01.

Figure 4.

 

The retinal vasculature of the uPAR^−/− mouse developed normally. Representative areas of the retina from P17 uPAR^−/− (A) and C57BL6 (B) mice demonstrating the vasculature stained for ADPase activity. In both mice, the normal vessels extended to the avascular peripheral portion of the retina. The overall pattern appeared the same in both cases, and quantitation of the area of the retina occupied by vessels was nearly identical (28.41% ± 2.48% vs. 25.15% ± 1.23%). n = 3 mice in each group (six wholemounts and six random areas from each).

Figure 5.

 

The Å6 peptide inhibits retinal neovascularization in a dose-dependent manner. Experimental oxygen-treated mice were injected with Å6 at the indicated concentrations from P12 to P16. Neovascularization was quantitated and compared with both PBS injected mice and uPAR^−/− mice treated with the oxygen protocol. Increasing inhibition occurred with increasing doses of Å6 peptide (10 mg/kg, 6%; 50 mg/kg, 16%; 100 mg/kg, 63%). Data are the mean ± SEM of results in four mice for each treatment. *Significantly less than in PBS-treated mice, P < 0.01.

Figure 6.

 

The expression of uPAR mRNA in retina is reduced in response to Å6 treatment. (A) RT-PCR analysis of retinas from P15 experimental mice treated from days P12 to P14 with either Å6 peptide (100 mg/kg) (lanes 2 and 5) or PBS (lanes 3 and 6). The amount of uPAR mRNA product was reduced in the Å6 treated tissues (lane 5) compared with those treated with PBS (lane 6). 18s RNA product was detected in all samples (lanes 2 and 3). (B) Real-time RT-PCR analysis of uPAR mRNA in retinas of Å6- and PBS-injected mice. A 2.3-fold decrease in the level of uPAR mRNA is seen in the Å6-treated animals. Data are the mean ± SEM of values derived using the comparative C_t method (2^−ΔΔCt) of triplicate reactions in two mice in each group. *Significantly less than in PBS-treated mice, P < 0.01