Cysteine-rich 61, a Member of the CCN Family, as a Factor Involved in the Pathogenesis of Proliferative Diabetic Retinopathy

Abstract

purpose. Cysteine-rich 61 (Cyr61/CCN1) is reported to mediate angiogenesis. In this study, its role in ocular angiogenesis and proliferative diabetic retinopathy (PDR) was investigated.

methods. The effects of Cyr61 were evaluated by determining proliferation and chemotaxis and in an assay of capillary tube formation in synthetic matrix by chorioretinal endothelial cells (RF/6A). In the same cells, Cyr61 expression under hypoxic conditions was then investigated. Interactions between Cyr61 and vascular endothelial growth factor (VEGF) were examined using endothelial cell chemotaxis, tube-formation assay, and cross-stimulation assay. A mouse model of oxygen-induced retinopathy (OIR) and a rat model of streptozocin-induced diabetes were used to evaluate Cyr61 expression in the retina. Cyr61 levels were also measured and chemotactic effects were evaluated in vitreous samples from patients with PDR.

results. Cyr61 significantly induced proliferation, migration, and synthetic matrix tube formation of RF/6A cells. Hypoxia significantly induced Cyr61 mRNA and protein expression. Cyr61 induced expression of VEGF and vice versa. Anti-Cyr61 or anti-VEGF could inhibit the effects of both Cyr61 and VEGF. Intravitreal injection of anti-Cyr61 antibody significantly inhibited retinal neovascularization in the mouse OIR model. Cyr61 mRNA and protein were significantly expressed in the retina of streptozocin-induced diabetic rats. Vitreous levels of Cyr61 were elevated in patients with PDR when compared with nondiabetic patients.

conclusions. Cyr61 acts as an angiogenic mediator of ocular neovascularization in vitro and in vivo. It may interact with VEGF in a synergetic manner. Vitreous levels of Cyr61 are elevated in PDR, and it may play an important role in the disease’s pathogenesis.

Proliferative diabetic retinopathy (PDR) is the most common cause of visual loss in middle-aged adults. ¹ Its pathophysiology involves microvascular dysfunction of the retina, followed by neovascularization of the optic disc and retina due to retinal ischemia. If untreated, it may lead to exacerbation of retinal fibrovascularization, causing vitreous hemorrhage, tractional retinal detachment, neovascularization glaucoma, and eventual blindness. ²  

Vascular endothelial growth factor (VEGF) is a primary angiogenic factor that mediates ischemia-induced retinal neovascularization. Vitreous VEGF levels are elevated in patients with PDR. ^{3 4} The anti-VEGF antibody bevacizumab (Avastin; Genentech/Roche, South San Francisco, CA) is used for diabetic retinopathy treatment. ^{5 6 7 8} Despite its potent anti-VEGF properties, it may not completely inhibit retinal fibrovascularization. In fact, tractional retinal detachment after intravitreal injection of bevacizumab, in cases of severe PDR, has been reported. ⁹ It is possible that several other factors participate in the angiogenic and fibrotic processes involved in diabetic retinopathy. The candidate factors include basic fibroblast growth factor (bFGF), insulin-like growth factor (IGF)-I, and CCN family proteins. ^{10 11 12}  

Cysteine-rich 61 (Cyr61) is the first cloned member of the CCN family, which comprises Cyr61/CCN1, connective tissue growth factor (CTGF/CCN2), nephroblastoma overexpressed (Nov/CCN3), Wisp-1/elm1 (CCN4), Wisp-2/rCop1 (CCN5), and Wisp-3 (CCN6). ^{13 14 15} Most members share a uniform modular structure and are involved in various cellular functions such as cell division, chemotaxis, apoptosis, adhesion, motility, and ion transport. ^{16 17} Connective tissue growth factor (CTGF), one CCN family member, was found to be involved in the angiogenic and fibrotic mechanisms of PDR. ^{18 19} Cyr61 is known to mediate cell adhesion, stimulate chemotaxis, augment growth factor-induced DNA synthesis, foster cell survival, and enhance angiogenesis. ^{20 21} Further, Cyr61 may induce fibrosis. ²² Studies have shown that patients with PDR have increased levels of total or latent vitreous TGF-β, which in turn can induce Cyr61expression. ^{23 24} However, the role of Cyr61 in the regulation of retinal angiogenesis and fibrosis in PDR remains unclear. 

We hypothesized that Cyr61 not only takes part in ocular angiogenesis but also promotes retinal angiogenesis in PDR. To investigate this hypothesis, we studied the angiogenic effects of Cyr61 in the cell culture model. In addition, the expression of Cyr61 in the retina was evaluated with a mouse model of oxygen-induced retinopathy (OIR) and a rat model of streptozocin (STZ)-induced diabetes. We further measured the vitreous levels of Cyr61 in patients with PDR and tried to clarify its role in the pathogenesis of PDR. 

Materials and Methods

C57BL/6 mice and Wistar rats were obtained from the National Taiwan University (Taipei, Taiwan) College of Medicine animal resource center. All experiments were in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Patients with PDR or nondiabetic ocular diseases who received pars plana vitrectomy at our center were enrolled. The study was conducted in accordance with The Declaration of Helsinki, and approval was obtained from the National Taiwan University Hospital ethics committee. To ensure reliability, we performed in vitro experiments in triplicate or quadruplicate. All subjective measurements were performed by an independent investigator from our group, without the knowledge of previous sampling sequences. 

Cell Culture, Proteins, Antibodies, and Reagents

Monkey chorioretinal vessel endothelial cells (RF/6A) were obtained from the American Type Culture Collection (Manassas, VA, and Rockville, MD). All cells were grown in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal bovine serum (FBS), 100 μg/mL streptomycin, and 100 U/mL penicillin (all from Invitrogen-Gibco, Carlsbad, CA). The cells were maintained at 37°C in a humidified 5% CO₂ atmosphere. Recombinant Cyr61 (rCyr61) was obtained from the Abnova Company (Taipei, Taiwan). Recombinant (r)VEGF, recombinant (r)IGF-I, monoclonal antibodies against β-actin, VEGF, and Cyr61 were from R&D Systems Inc. (Minneapolis, MN). Rabbit anti-Cyr61polyclonal antibody, rabbit anti-VEGF polyclonal antibody and mouse anti-hypoxia-inducible factor (HIF)-1α antibody were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Synthetic matrix (Matrigel) was obtained from BD Biosciences (Franklin Lakes, NJ), and STZ was acquired from Sigma-Aldrich (St. Louis, MO). 

Cell Proliferation Assay

DNA synthesis was assessed using a bromodeoxyuridine (BrdU) cell-proliferation assay kit (Calbiochem, EMD Biosciences, Darmstadt, Germany) after manufacturer protocols. RF/6A cells were seeded in 96-well plates (at 2 × 10⁴ cells/well) in complete culture medium and starved in serum-free basal medium overnight. rCyr61 (0.5–125 nM) or rVEGF (0.01–2.62 nM) was added to the wells, together with basal assay medium containing BrdU. After 24 hours of incubation, the cells were fixed in the fixative/denaturing solution. Incorporated BrdU was quantified using enzyme-linked immunoabsorbent assay (ELISA) according to standard protocols. The reaction product amount was determined by measuring absorbance at 450 and 540 nm using a plate reader (Titer-tek Multiscan MCC/340; ICN, Tokyo, Japan). From the dose–response curve, EC₅₀s of rCyr61 and rVEGF were calculated. In the cross-stimulation experiments, rCyr61 (0.5–125 nM)/EC₅₀ of rVEGF combinations or rVEGF (0.01–2.62 nM)/EC₅₀ of rCyr61 combinations were added to the wells. Each sample was tested in quadruplicate. 

Endothelial Cell Chemotaxis Assay

Chemotaxis was performed in 96-well, blind-well chemotaxis chambers with gelatin-coated, 8-μm pore size, polycarbonate membranes (Neuro Probe, Gaithersburg, MD). RF/6A cells (1 × 10⁴) in 100 μL of DMEM containing 0.1% FBS was added to the bottom wells. Chambers were inverted and incubated for 4 hours at 37°C, which allowed cell attachment to the membrane. rCyr61 (10 nM) was preincubated with 25 μg/mL of either rabbit anti-Cyr61 polyclonal antibody (pAb), rabbit anti-VEGF pAb, or control IgG; rVEGF (2 nM) was preincubated with 25 μg/mL of either rabbit anti-Cyr61 pAb or control IgG for 1 hour at 37°C. rCyr61 (10⁻² to 10² nM), rCyr61/pAb combinations, rVEGF/pAb combinations, vehicle control phosphate balance solution (PBS), or positive control rVEGF (2 nM) were added to the top wells, followed by incubation of the chambers for 6 hours at 37°C. Membranes were removed, fixed in methanol, and stained with Coomassie blue. The number of cells that had migrated through the filter pores was counted per three high-power fields. Each test group was tested in quadruplicate. 

Endothelial Cell Tube-Formation Assay

The synthetic matrix (Matrigel; BD Biosciences) assay was performed according to the method described by Gately et al. ²⁵ with minor modifications. The matrix was thawed on ice to prevent premature polymerization; 50 μL was plated into individual wells of a 96-well chamber and allowed to polymerize at 37°C for 30 to 60 minutes. RF/6A cells were removed from culture by trypsinization and resuspended at a concentration of 5 × 10⁴ cells/mL in DMEM containing 2% FBS. rCyr61 (10 nM) was preincubated with 25 μg/mL of either rabbit anti-human Cyr61 pAb, rabbit anti-VEGF pAb, or control IgG; rVEGF (2 nM) was preincubated with 25 μg/mL of either rabbit anti-Cyr61 pAb or control IgG for 1 hour at 37°C. PBS acted as a vehicle control. The cell suspension (100 μL) was plated and incubated with test substances for 8 to 12 hours at 37°C in a 5% CO₂ humidified atmosphere. Each chamber was photographed at a final magnification of 100×. The tube branch count was quantified by a blinded observer according the methods described by Gately et al. ²⁵ All groups were performed in triplicate. 

RNA Isolation and Reverse Transcription–Polymerase Chain Reaction (RT-PCR) Amplification

Total RNA was extracted (Invitrogen; TRIzol Reagent) and first-strand cDNA was synthesized with oligo-dT–primed Moloney murine leukemia virus (MMLV) reverse transcriptase (RT). The primer sequences used in RF/6A cells were as follows: Cyr61 sense, 5′-GAG GGC AGA CCC TGT GAA TA-3′, and antisense, 5′-TGG TCT TGC TGC ATT TCT TG-3′; VEGF sense, 5′-ACG AAG TGG TGA AGT TCA TGG-3′, and antisense, 5′-TCA CAT CTG CAA GTA CGT TCG-3′; β-actin sense, 5′-GGT GGC TTT TAG GAT GGC AAG-3′, and antisense, 5′-ACT GGA ACG GTG AAG GTG ACA G-3′. Primer sequences used in rat retina were as follows: Cyr61 sense, 5′-GGG CAG TGC TGT GAA GAG-3′, and antisense, 5′-TTT GGG CCG GTA TTT CTT-3′; VEGF sense, 5′-CAG AAA GCC CAT GAA GTG-3′, and antisense, 5′-TTT GAC CCT TTC CCT TTC-3′; β-actin sense, 5′-TCT CTT CCA GCC TTC CTT-3′, and antisense, 5′-AGT TCC GCC TAG AAG CAT T-3′. PCR cycling conditions were as follows: 5 minutes at 95°C, followed by 28 cycles of 95°C for 30 seconds and 55°C for 60 seconds for Cyr61 and VEGF, or 25 cycles of 95°C for 30 seconds and 55°C for 60 seconds for β-actin, and 72°C for 90 seconds. PCR products were separated on 2% agarose gels containing ethidium bromide (0.5 g/mL), then visualized, and photographed. Quantification of band intensity was performed by densitometry scanning (GS-800 Densitometer; Bio-Rad Laboratory, Hercules, CA). 

Western Immunoblot Analysis and Immunohistochemistry

Proteins were extracted from cell lysates and tissue homogenates. For Western blot analysis, the protein samples were fractionated in a 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel and transferred to nitrocellulose membrane. The analysis was performed with anti-Cyr61, anti-VEGF, anti–HIF-1α, or anti–β-actin antibodies. Immunodetection was performed by enhanced chemiluminescence (Pierce Biotechnology, Rockford, IL), in line with the manufacturer’s instructions. Protein levels were determined from the analysis by densitometry scanning of protein bands. 

Formalin fixed, paraffin embedded 6-μm eye tissue sections were placed on slides, deparaffinized in xylenes, and rehydrated by incubation in graded ethanol baths in PBS. Endogenous peroxidase was blocked with 0.3% hydrogen peroxide in methanol. Sections were then treated with 5% normal horse serum and incubated overnight with anti-Cyr61 antibody at 4°C. Thereafter, a biotinylated horse secondary antibody against rabbit IgG and an avidin-biotinylated peroxidase complex were used with 3,3′diaminobenzidine as a peroxidase substrate. Sections were counterstained with hematoxylin, dehydrated, and mounted. Isotype control IgG was used as the primary antibody in the negative control experiments. 

ELISA of Cyr61 and VEGF

Cyr61 levels were determined by designed sandwich ELISA under combined mouse anti-Cyr61 mAb, rCyr61 protein, and rabbit anti-Cyr61 pAb. This method has successfully determined vitreous levels of growth factors. ¹⁹ Standard solution (100 μL) and the sample (100 μL) were added to a 96-well plate coated with polyclonal antibody. After incubation, the plate was washed, and a detection antibody was added. Later, the plate was washed and an enzyme-linked secondary antibody was added. After further incubation, the plate was washed again and the substrate added. The reaction was stopped by adding a blocking solution after color emerged. Optical density was determined at 450 and 540 nm using an absorption spectrophotometer (Titer-tek Multiscan MCC/340; ICN). VEGF levels were measured by ELISA using kits for human rVEGF (R&D Systems). The procedure was performed according to the manufacturer’s instructions. Instead of the detecting antibody and the enzyme-linked secondary antibody, biotin-labeled detecting antibody and streptavidin-HRP were used for VEFG ELISA. A standard curve was plotted from measurements made with standard Cyr61 and VEGF solutions (from 2 to 400 ng/mL for Cyr61 and 12.5 to 4000 pg/mL for VEGF) and was used to determine the concentration of Cyr61 or VEGF in the sample. The standard curve for Cyr61 concentration showed an R ² of 0.985. 

Expression of Cyr61 under the Stimulation of VEGF and IGF-I

The effects of VEGF- and IGF-I-induced Cyr61 expression and Cyr61-induced VEGF expression were investigated. We treated RF/6A cells with 10 nM of rIGF-I, 10 nM of rCyr61, or 2 nM of rVEGF for 2 hours and then total RNA and total protein were extracted. RT-PCR and Western blot analysis were performed to determine Cyr61 expression under VEGF or IGF-1 stimulation. At the same time, VEGF expression under Cyr61 stimulation was evaluated using mRNA and protein levels. 

Expression of Cyr61 in RF/6A Cells under Hypoxic Conditions

RF/6A cells were placed in serum-free medium. One milliliter of cells (1 × 10⁵ cells/well) was plated into one well of a six-well culture plate. Hypoxic cultures were transferred for various time periods (1% O₂/5% CO₂/94% N₂ labeled hypoxia) in a hypoxic incubator (BioSpherix, Redfield, NY). The cells were harvested in hypoxic conditions to avoid reoxygenation artifacts. At the same time, the media were collected and stored at −80°C for ELISA assay. For each analysis, three replicates were taken using 100 μL of supernatant each. Parallel cultures were kept in normal oxygen levels (labeled as normoxia). Cells were harvested at various times. Total RNA and total protein were extracted. RT-PCR and Western blot were used for mRNA, and protein expression. 

Oxygen-Induced Retinopathy Model and Inhibition by Anti-Cyr61 Antibody

The oxygen-induced retinopathy (OIR) mouse model used a previously published method. ²⁶ Seven-day-old pups and their mother were housed in sealed chambers that contained 75% ± 5% O₂ and 2% CO₂, using an O₂-producing machine. Gas levels in the chamber were monitored daily by gas analyzer and a chart recorder. Mice remained in the chamber for 5 days (hyperoxic period, postnatal day [P]7–P12) and then in room air for an additional 5 days (hypoxia-induced angiogenic period P12–P17). Anti-Cyr61 antibody was administered at a dose of 1 μL (200 μg/mL) to the pups (n = 10) by intravitreal injection in the right eye, and the control left eye was injected with rabbit IgG (200 μg/mL) on P12. The mice were deeply anesthetized with ketamine for all procedures. The lid fissure was opened with a no. 11 scalpel blade and the eye was proptosed. Intravitreal injections were performed by first entering the eye with an 8-0 suture needle (Ethicon, Piscataway, NJ) at the posterior limbus. A 32-gauge needle and syringe (Hamilton, Reno, NV) were used to deliver 1 μL of antibody solution through the existing entrance site. The eye was then repositioned and the lids were approximated over the cornea. Erythromycin ointment was applied after the procedure. During the experiment, the mothers were provided with water and standard mice food and were exposed to normal 12-hour light/dark cycles. The pups received nutrition from their mothers. They were killed on P19, and the eyeballs were enucleated and fixed in formalin. Quantitation of neovascularization was performed using a modified technique originating from Smith et al. ²⁷ Briefly, 6-μm-thick serial sections, each separated by at least 40 μm, were taken from around the region of the optic nerve. The hematoxylin and eosin–stained sections were examined in a blinded fashion for the presence of neovascular tufts projecting into the vitreous from the retina. The neovascular score was defined as the mean number of neovascular tufts per section found in 16 sections per eye. Immunohistochemical staining of anti-Cyr61 antibody was performed to investigate sites of Cyr61 expression. 

Streptozocin-Induced Diabetic Rat Model

Rats were treated with STZ to produce type 1 diabetes. ²⁸ Male Wistar rats weighing 250g were given a single intraperitoneal injection of STZ at 50 mg/kg body weight, dissolved in 50 mM citrate buffer (pH 5). Control rats (n = 8) of similar age and body weight were injected with vehicle buffer in comparable amounts. Body weight and blood glucose levels were measured before injection, and 2 days and 2 weeks after. STZ rats were considered diabetic (n = 8) if they had blood glucose levels >250 mg/dL at 2 days after STZ injection. Blood glucose levels were more than 300 mg/dL 2 weeks after STZ injection. Three months later, control and STZ rats were killed, and retinal tissue samples from one eye each were harvested. Each tissue sample was divided into two pieces and quickly frozen for later preparation of total RNA and protein. RT-PCR and Western blot analysis were used for analysis of expression of mRNA and protein. Immunohistochemical staining of anti-Cyr61 antibody was performed to investigate the degree and the location of Cyr61 expression. 

Vitreous Levels of Cyr61 and VEGF and Immunodepletion of Cyr61 in PDR Vitreous Samples for RF/6A Cell Chemotaxis Assays

Patients with PDR or nondiabetic ocular diseases who received pars plana vitrectomy at our center were enrolled. Samples of undiluted vitreous fluid were harvested from the eyes of participating patients who had PDR or nondiabetic ocular disease. Active PDR was defined as PDR with perfused, multibranching preretinal capillaries. Quiescent PDR was defined as PDR with fully regressed active proliferation or with only nonperfused, gliotic vessels. Fifty-six subjects with active PDR, 19 subjects with quiescent PDR, and 25 control subjects, including 8 with idiopathic epiretinal membranes and 17 with idiopathic macular holes, were enrolled. Samples of undiluted vitreous fluid were harvested at the start of vitrectomy. Vitreous levels of Cyr61 and VEGF were measured in triplicate by ELISA. 

Vitreous samples of 10 patients with active PDR were preincubated with 25 μg/mL of rabbit anti-human Cyr61 antibody or rabbit IgG control for 1 hour at 37°C. On completion of this neutralization period, the PDR vitreous sample–antibody combinations were tested in the RF/6A cell chemotaxis assay, as described herein. Vitreous samples in 10 patients in the nondiabetic control group and 10 patients in the active PDR group were tested in the same assay as the negative and the positive controls, respectively. Each assay was tested in triplicate. 

Statistical Analysis

Data in the text and figure legends are expressed as the mean ± SD. Differences between the means of experimental and respective control groups were calculated by Mann-Whitney U test. Group vitreous levels were analyzed by one-way ANOVA and the Bonferroni post hoc test was used for pair-wise comparisons. For evaluation of in vivo retinal angiogenesis, paired t-test or Wilcoxon signed rank test was used for quantitative data. All reported P-values are two-sided (P < 0.05). 

Results

CCN1/Cyr61 Promotes Proliferation of RF/6A Endothelial Cells

As endothelial cell proliferation is important in angiogenesis, we measured BrdU incorporation of endothelial cells in response to Cyr61. Serum-starved RF/6A cells were treated for 24 hours with various concentrations of rVEGF (0–2.5 nM) or rCyr61 (0–125 nM). In the dose–response curve, EC₅₀ of the proliferation assay was 5.2 nM for rCyr61 and 0.21 nM for rVEGF (Figs. 1A 1B) . To study the relationship of Cyr61 and VEGF, we added the EC₅₀ concentration of rVEGF (0.21 nM) into different concentrations of rCyr61 in the endothelial cell proliferation assay. Results showed that the EC₅₀ of the combinations (2.17 nM) was less than half the EC₅₀ of rCyr61(5.2 nM), suggesting a synergetic effect (Fig. 1C) . In a similar manner, and EC₅₀ concentration of rCyr61 (5.2 nM) was added into different concentrations of rVEGF, and a similar synergetic effect was noted (0.068 nM vs. 0.21 nM; Fig. 1D ). 

Cyr61-Induced RF/6A Cell Chemotaxis and Cell Tube Formation in Synthetic Matrix

To demonstrate the angiogenic potential of Cyr61, endothelial cell chemotaxis and tube formation were performed. Results showed that Cyr61 induced RF/6A cell chemotaxis in a concentration-dependent manner (P < 0.05; Fig. 2A ). Cyr61 significantly induced endothelial cell chemotaxis in comparison with PBS control (P < 0.05). Incubation with 25 μg/mL anti-Cyr61 antibody significantly inhibited endothelial cell migration (P < 0.05; Fig. 2B ). 

In the endothelial cell tube-formation assay, we counted the number of tubes in the experimental wells (Cyr61 group) and the positive control (VEGF group), and the vehicle control (PBS group) wells. Cyr61 induced significantly more tube formation in RF/6A than PBS (32 ± 2.8 vs. 6 ± 1.6 tubes/well; P < 0.05). Similar results were obtained from VEGF (35.8 ± 3.1 vs. 6 ± 1.6 tubes/well; P < 0.05). There was no statistically significant difference between the Cyr61 and VEGF groups. Anti-Cyr61 antibody significantly inhibited Cyr61-induced RF/6A tube formation (P < 0.05; Fig. 2C ). To investigate the interaction of Cyr61 and VEGF, we added anti-VEGF and anti-Cyr61 the to Cyr61 and VEGF solutions, respectively. In both the endothelial cell chemotaxis assay and endothelial cell tube-formation assay, anti-VEGF antibody inhibited Cyr61-induced effects when compared with the combination of Cyr61 and control IgG (P < 0.05) and anti-Cyr61 antibody also inhibited VEGF-induced effects (P < 0.05; Figs. 2B 2C ). 

Effect of Hypoxia on Cyr61 mRNA and Protein Expression

As hypoxia plays a central role in PDR pathogenesis, we examined Cyr61 expression under hypoxia in RF/6A cells. Cyr61 mRNA levels were induced at the first hour after hypoxia and further thereafter. The degree of kinetics for expression was observed for VEGF, with sharp increase after 2 hours of hypoxia (Fig. 3A) . Both Cyr61 and VEGF showed increased expression at the protein level, which correlated well with their expression seen at the mRNA level (Fig. 3B) . To investigate the expression of Cyr61 as a secreted factor, Cyr61 levels in culture media were measured by ELISA. They were significantly elevated after 2 hours of hypoxia (Fig. 3C) . 

Interaction of Cyr61 and VEGF

To further investigate the interaction of Cyr61 and other growth factors, we studied the expression of Cyr61 under the stimulation of VEGF and IGF-I. Results showed that under IGF-I (10 nM) and VEGF (2 nM) stimulation, the mRNA and protein expression of Cyr61 was significantly induced in RF/6A cells (Fig. 4A) . Cyr61 (10 nM) also significantly induced VEGF expression in terms of mRNA and protein (Fig. 4B) . 

Retinal Neovascularization and Cyr61 Expression in Mice of Oxygen-Induced Retinopathy

The OIR model of mice was designed to investigate the role of Cyr61 in retinal angiogenesis. C57BL/6J mice exposed to 75% O₂ from P7 to P12 experience extensive retinal capillary obliteration. When these mice were returned to room air on P12, the inner retina became relatively hypoxic. Retinal angiogenesis developed from P19 to P22. Intraocular injection of 1 μL (200 μg/mL) of anti-Cyr61 antibody significantly reduced retinal neovascularization as evaluated by histology (P < 0.01; Figs. 5A 5C ), compared with injection of control IgG in the contralateral eye (Fig. 5B) . No retinal toxicity or inflammation was apparent under light microscopy. Mice with OIR showed strong expression of Cyr61 in the neovascular vessels, as shown by retinal immunochemical staining for Cyr61 (Fig. 5D)in comparison with isotype IgG staining (Fig. 5E) . Control sections showed only background staining (Fig. 5F) . 

Expression of Cyr61 in the Retina of STZ-Induced Diabetic Rats

The STZ-induced diabetes rat model was used to demonstrate the role of Cyr61 in diabetic retinopathy. An early and significant increase in Cyr61 mRNA levels when compared with control eyes was observed 3 months after STZ induction of diabetes in rats. (Fig. 6A)Western blot analysis showed that Cyr61 protein levels were substantially higher in the retina of STZ-induced diabetic rats than in control rats (Fig. 6B) . Immunohistochemistry was performed to investigate the localization and grade of Cyr61 expression in the retina. Retinas of nondiabetic rat showed slight staining of Cyr61 in the ganglial cell layer (Fig. 6C) . Retinas incubated with isotype IgG in the controls and STZ-induced diabetic rats did not show any staining (Figs. 6D 6F) . A strong staining for Cyr61 has found in the ganglion cell layer of retinas in STZ-induced rats (Fig 6E) . 

Increase in Vitreous Cyr61 Levels in Patients with PDR and Effect of Immunodepletion of Cyr61 on Chemotactic Activity in RF/6A Cells in Vitreous Samples from PDR Patients

We measured the vitreous levels of Cyr61 to determine its expression in patients with PDR. Vitreous Cyr61 levels were significantly higher in patients with PDR (204.2 ± 95.7 ng/mL) than in control patients (65.0 ± 32.2 ng/mL, P < 0.0001). Cyr61 levels were 232.7 ± 100.5 ng/mL in patients with active PDR and 128.4 ± 67.9 ng/mL in patients with quiescent PDR (P < 0.001; Fig. 7A ). Vitreous levels of VEGF were 2048.1 ± 991.6 pg/mL in patients with active PDR, 936.6 ± 291.3 pg/mL in patients with quiescent PDR, and 464.3 ± 372.1 pg/mL in control patients (P < 0.01; Fig. 7B ). Mean Cyr61 plasma levels were 48.5 ± 7.5 ng/mL in the PDR group, slightly lower than that in the nondiabetic group (44.5 ± 8.2 ng/mL; data not shown). 

Compared with those of control patients, vitreous samples of patients with PDR significantly induced RF/6A cell chemotaxis. To determine the biological relevance of Cyr61 in a disease characterized by angiogenesis, vitreous samples of patients with PDR were immunodepleted by anti-Cyr61 antibody and tested for their RF/6A cell chemotactic activity. Although PDR vitreous samples were potently chemotactic for RF/6A cells, immunodepletion of Cyr61 significantly decreased this effect (P < 0.01; Fig. 7C ). 

Discussion

In the present study, Cyr61 induced angiogenesis in vivo and in vitro. Furthermore, results showed that Cyr61 mRNA and protein expression in the retina increased in the diabetic animal model. Vitreous samples of patients with PDR had higher Cyr61 concentrations than did nondiabetic control subjects. The study also suggested that Cyr61 and VEGF may act in a synergetic manner in angiogenesis. These experimental results strongly indicate that Cyr61 may play an important role in retinal angiogenesis, and may be a relevant factor in the pathogenesis of PDR. 

Previous studies have shown CCN1/Cyr61 to be a potent angiogenic factor. Potent proangiogenic properties of Cyr61 were demonstrated in a rat cornea model and in a rabbit ischemia hind limb model. ^{20 29} In addition, Cyr61 can also regulate the expression of genes involved in angiogenesis and matrix remolding, including VEGF-A, VEGF-C, type-I collage, matrix metalloproteinase (MMP)-1, MMP-3, and tissue inhibitors of metalloproteinase (TIMPs). ³⁰ Thus, CCN1/Cyr61 may induce angiogenesis both directly and indirectly. In the experiments using chorioretinal endothelial RF/6A cells, Cyr61 induced endothelial cell proliferation in a dose-dependent manner and induced cell chemotaxis in a concentration-dependent manner from 10⁻² to 10² nM. Cyr61 significantly induced RF/6A cells to form capillary tubes in synthetic matrix. Furthermore, anti-Cyr61 antibody significantly inhibited Cyr61-induced endothelial cell migration and capillary tube formation. These results are consistent with previous studies and clearly show the potent angiogenic effect of Cyr61. In addition to angiogenic activity, Cyr61 may have other tissue effects. A recent study showed that Cyr61, which is expressed in endothelial cells of capillaries and smooth muscle cells of small vessels in arteriosclerotic lesions, may play a role in both angiogenic and fibrogenic pathways. ²² It is similar to CTGF in the development and progression of PDR. ¹⁹ Thus, Cyr61 may mediate tissue postangiogenic fibrosis in addition to angiogenesis. 

In the present study Cyr61 mRNA and protein expression increased under hypoxia in vitro and in vivo. Under hypoxia, expression of Cyr61 mRNA and protein significantly increased. In the mouse OIR model, immunohistochemical studies showed Cyr61 expression in the neovascular tufts of the retina and retinal angiogenesis was inhibited by intravitreal injection of anti-Cyr61 antibody. Hypoxia is a major factor that induces retinal angiogenesis. Our results showed that VEGF and Cyr61 were induced by hypoxia. VEGF is upregulated through the activation of the transcription factor HIF-1α under hypoxia. It have been reported that hypoxia induced Cyr61 expression via cooperation of HIF-1α and c-Jun/AP-1 pathways in melanoma cells. ³¹ Therefore, both of Cyr61and VEGF possibly were induced by activating the HIF-1α-dependent pathway under hypoxia. 

To further investigate the relevance of Cyr61 in diabetic retinopathy, studies have been performed in rats with STZ-induced diabetes. A previous study showed that after 6 weeks of diabetes, Cyr61 expression levels were increased more than threefold in retinas of STZ-induced diabetes rats. ³² Our results showed a higher level of Cyr61 mRNA and protein expression in retinas of diabetic rats when compared with controls. In immunohistochemistry studies, we found Cyr61 to be significantly expressed in the ganglion cell layer of the retina. In our study, hyperglycemia induced expression of Cyr61 in vivo, which suggests that Cyr61 is involved in diabetic retinopathy. 

Previous studies have demonstrated the importance of VEGF in ocular neovascularization. ^{33 34} Vitreous levels of VEGF are significantly elevated in patients with PDR. ^{4 34} Recently, erythropoietin has been measured in vitreous samples from patients with PDR. ³⁵ The present study indicates that vitreous Cyr61 levels in PDR patients, especially those with active PDR, are significantly higher than those in nondiabetic patients. Patients with more severe PDR had higher vitreous levels. Moreover, immunodepletion of vitreous samples from patients with PDR with anti-Cyr61 antibody significantly reduced angiogenic activity in the endothelial cell chemotaxis assay. This phenomenon may be due to a complex interaction between different angiogenic mediators which may act in a synergetic manner. If this is true, immunodepletion of an individual factor may have a major impact on the angiogenic response, and Cyr61 may be a new potential target for disease treatment. 

VEGF is a major player in diabetic retinopathy and was used as the positive control in our studies. ³³ Understanding the interaction of Cyr61 and VEGF is crucial. We showed that the EC₅₀ of rCyr61 is 2.4 times that of rCyr61 when added with rVEGF, suggesting a synergetic interaction between Cyr61 and VEGF. It has been reported that several growth factors induce Cyr61 expression. ^{36 37 38} In our study, both IGF-I and VEGF induce expression of Cyr61. Similarly, Cyr61 also induces VEGF expression in terms of mRNA and protein in retinal endothelial cells. Inhibition of either Cyr61 or VEGF resulted in inhibition of the effects of the other in endothelial cells chemotaxis and endothelial cell tube-formation assays. Cyr61 and VEGF exert their effects through different and multiple signal pathways via their receptors. ^{39 40} The capability of Cyr61 to induce VEGF or VEGF to induce Cyr61 is cell-type specific. ^{30 37 38 41 42 43} Previous reports have demonstrated that VEGF can upregulate Cyr61 expression. ^{38 41 42} In addition, it has been reported that Cyr61 induces VEGF expression in different types of cells, such as fibroblasts and smooth muscle cells. ^{30 43} In this report, we show that Cyr61 could also induce VEGF expression in endothelial cells. The interaction between Cyr61 and VEGF might develop not only through interaction of downstream signal molecules, but also through interaction either with the cell itself or surrounding cells, such as pericytes and fibroblasts, in an autocrine–paracrine manner. Previous studies have suggested Cyr61 is an extracellular matrix-associated protein that can act on endothelial cells, fibroblasts, macrophages, and platelets. ^{12 21 44} It may be synthesized by both endothelial cells and fibroblasts. Thus, Cyr61 not only acts directly on endothelial cells to promote angiogenesis but also stimulates fibroblasts to produce VEGF to further enhance angiogenesis. ¹⁶ Further studies should clarify the relationship between Cyr61 and VEGF. 

In summary, Cyr61, a member of the CCN family, induces endothelial cell proliferation and chemotaxis. Hypoxia induces the expression of Cyr61. Inhibition of the effect of Cyr61 reduces retinal neovascularization in a mouse model of OIR. Cyr61 is significantly expressed in the STZ-induced diabetic rat model. The concentration of Cyr61 is elevated in the vitreous of patients with PDR, especially those with active PDR. These results suggest that Cyr61 is involved in the pathogenesis of PDR and may be a useful target for PDR treatment. 

 Supported by Grant 9605 from the Department of Health, Executive Yuan, Republic of China (JJY), Grant NSC 96-2628-B-002-032-MY3 from the National Science Council, Executive Yuan, Republic of China (CMY).

 Submitted for publication July 23, 2008; revised November 26, 2008, and January 14, 2009; accepted May 14, 2009.

 Disclosure: J.-J. You, None; C.-H. Yang, None; M.-S. Chen, None; C.-M. Yang, None

 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.

 

Assessment of Cyr61 on RF/6A cell proliferation and the interaction of Cyr61 and VEGF by BrdU cell proliferation assay. (A, B) After RF/6A cells (seeded in 96-well plates, at 2 × 10⁴ cells/well) were exposed to various concentrations of rCyr61 or rVEGF in media containing BrdU for 24 hours, BrdU incorporation into RF/6A cells was measured. Both Cyr61 and VEGF had a significant effect on RF/6A cell proliferation in a concentration-dependent manner. EC₅₀ of Cyr61 was 5.2 nM and EC₅₀ of VEGF was 0.21 nM. (C, D) To investigate the interaction of Cyr61 and VEGF, EC₅₀ concentration of rVEGF was added to various concentrations of rCyr61; EC₅₀ of the combination was calculated to be 2.17 nM (C). Next, EC₅₀ concentration of rCyr61 was added to different concentrations of rVEGF; EC₅₀ of the combination was calculated to be 0.068 nM (D). Data are shown as the mean ± SD of one of three independent experiments performed in quadruplicate. Values of cell proliferation are expressed as absorbance (450–540 nm).

Figure 2.

 

Evaluation of Cyr61 on RF/6A cell migration and tube formation and of the interaction between Cyr61 and VEGF. (A) rCyr61 induced chemotaxis in a concentration-dependent manner with cell chemotaxis assay. (B) rCyr61 and rVEGF significantly induced endothelial cell chemotaxis when compared with controls. In addition, whereas anti-Cyr61 antibody inhibited Cyr61-induced endothelial cell migration, cross-inhibition assays showed that the anti-Cyr61 antibody significantly reduced VEGF-induced endothelial cell migration, and anti-VEGF antibody also significantly reduced Cyr61-induced endothelial cell migration. (C) rCyr61 and rVEGF significantly induced endothelial cell tube formation when compared with the negative control. In addition, whereas anti-Cyr61 antibody inhibited Cyr61-induced endothelial cell tube formation, cross-inhibition assays showed anti-Cyr61 antibody significantly reduced VEGF-induced endothelial cell tube formation, and anti-VEGF antibody also significantly reduced Cyr61-induced endothelial cell tube formation. Results represent mean cells or tubes per well ± SD of one representative assay of three independent experiments performed in quadruplicate in chemotaxis assay and in triplicate in tube-formation assay.*P < 0.05, significantly different from PBS control. †P < 0.05, significantly different from rabbit IgG control. ‡P < 0.05, significantly different from isotype IgG control in cross-inhibition assays.

Figure 3.

 

Hypoxia-mediated induction of Cyr61 mRNA and Cry61 protein expression. (A) RF/6A cells were exposed to hypoxia (1% O₂/5% CO₂/94% N₂) for 0, 1, 2, 4, 8, 16, and 24 hours. Expression of Cyr61 and VEGF mRNA was measured by RT-PCR. Cyr61 mRNA levels were induced at the first hour after hypoxia and increased further thereafter. The same kinetic of expression was observed for VEGF, with a sharp increase in expression after 2 hours of hypoxia. (B) Western blot analysis was used to examine the expressions of Cyr61, VEGF, and HIF-1α. β-Actin was used as an internal control. Both Cyr61 and VEGF demonstrated increased expression at the protein level. The values of mRNA and protein expression were analyzed, using β-actin as the internal control. Data are presented as the mean ± SD of three independent experiments. (C) One milliliter of 1 × 10⁵ cells/mL RF/6A cells was plated on each well of six-well culture plates. Cyr61 levels of collected media on exposure to hypoxia for different periods were measured by ELISA. The secretory Cyr61 levels were significantly elevated after 2 hours of hypoxia. Data are stated as the mean ± SD of one representative assay of three independent experiments performed in triplicate. *P < 0.05 compared with the control.

Figure 4.

 

Interaction of Cyr61 and VEGF. The experiment was performed using RF/6A cells. After growth factors stimulation, cells were harvested to extract mRNA and protein. (A, B) After cells were incubated with rIGF-I (10 nM) or VEGF (2 nM) for 2 hours, PCR (A) and Western blot (B) showed enhanced expression of Cyr61. mRNA (A) and protein (B) expressions of VEGF were also significantly induced after 2-hour stimulation of rCyr61 (10 nM). The values of mRNA and protein expressions were recorded with β-actin as the internal control. Data are presented as the mean ± SD of three independent experiments. *P < 0.05, significant difference from nonstimulated control.

Figure 5.

 

Effect of Cyr61 on retina neovascularization in the OIR model of retinopathy. (A, B) Representative histopathological pictures from OIR mice retinas. Intraocular injection of anti-mouse Cyr61 antibody (1 μL at 200 μg/mL), performed as mice were returned to room air, significantly decreased retinal neovascularization (arrow) compared with equivalent injection of control IgG in the contralateral eye at P19. (C) Anti-mouse Cyr61 treatment resulted in decreased nuclei number in retinal endothelial cells, shown as mean ± SD of number of neovascular tufts in the retina (n = 10). (D–F) Representative pictures of retinal immunohistochemistry in formalin-fixed, paraffin-embedded tissues. Mice with OIR showed slight Cyr61 staining in the nerve fiber layer, the outer plexiform layer and the inner nuclear layers, as well as, strong Cyr61 expression in vascular ECs in the sensory retina and in the neovascular vessels (arrow, D); IgG staining of OIR mice was used as a negative control (E). Normal mice showed staining of Cyr61 in the nerve fiber layer, the outer plexiform layer, and the inner nuclear layers (F). Original magnification: (A, B) ×200; (D–F) ×200; bar: (A, B) 75 μm; (D–F) 70 μm.

Figure 6.

 

Retinal expression of Cyr61 in STZ-induced diabetic rats. (A, B) Cyr61 and VEGF, in terms of mRNA (A) and protein levels (B), were significantly increased in the retina of STZ-induced rats when compared with control rats. The values of mRNA and protein expression were analyzed, using β-actin as the internal control. Data are expressed as the mean ± SD of three independent experiments. *P < 0.05 when compared with nondiabetic controls. (C–F) Representative pictures of Cyr61 retinal immunohistochemistry study. Control rats (C) revealed slight staining in the ganglial cells layer and inner nuclear cell layer compared with the negative staining control (D). STZ-induced diabetic rats (E) showed significant expression of Cyr61 in the ganglial cell layer (arrow) when compared with the negative staining control (F). Original magnification, ×200; bar, 60 μm.

Figure 7.

 

Vitreous levels of Cyr61 and VEGF and effects of immunodepletion of Cyr61 on chemotactic activity in RF/6A cells in vitreous samples from patients with PDR. (A, B) Vitreous levels of Cyr61 and VEGF were determined by ELISA. Vertical-point figures show levels of Cyr61 (A) and VEGF (B) for the patients with active PDR (n = 56), quiescent PDR (n = 19), or nondiabetic ocular diseases (n = 25). Units for vitreous Cyr61 and VEGF levels are nanograms per milliliter and picograms per milliliter, respectively. *P < 0.01, significantly different when compared with control group. Each assay was in triplicate. † P < 0.01; significantly different between active and quiescent PDR groups. (C) Chemotactic assay showed anti-Cyr61 antibody inhibited PDR vitreous-induced endothelial cell migration. Results of a chemotactic assay represented as the mean number of cells per well ± SD of one of three independent experiments performed in quadruplicate, measured in three high-power fields. Each sample was tested in four wells. Anti-Cyr61 antibody (25 μg/mL) for neutralization of PDR migratory properties was determined as the percent suppression of migration compared with isotype control antibody. † P < 0.05, significantly different from rabbit IgG control. Magnification: ×400. (Note: average Cyr61 levels in the vitreous samples used in chemotaxis assay were 54.0 ± 17.5 ng/mL in the control group, 302.3 ± 74.4 ng/mL in the active PDR group, and 210.1 ± 93.2 ng/mL in the active PDR group vitreous pretreated with either rabbit IgG group or anti-Cyr61 antibody. VEGF levels were 308.0 ± 172.4 pg/mL in the control group, 1863.7 ± 788.3 pg/mL in the active PDR group, and 1976.6 ± 1102.7 pg/mL in the active PDR group vitreous pretreated with either rabbit IgG or anti-Cyr61 antibody. Vitreous samples of patients with active PDR were used as the positive control, and those of nondiabetic control patients were used as the negative control. *P < 0.05, significantly different from control patients).