Modulation of Tissue Plasminogen Activator and Plasminogen Activator Inhibitor-1 by Transforming Growth Factor-β in Human Retinal Glial Cells

Abstract

purpose. The serine proteases tissue plasminogen activator (t-PA) and urokinase plasminogen activator (u-PA) and their inhibitor, plasminogen activator inhibitor (PAI)-1, regulate a variety of processes involved in tissue morphogenesis and differentiation. There is much evidence that plasminogen activator-mediated extracellular matrix degradation is an important step in the development of ocular neovascular diseases. The authors investigated whether expression of t-PA, u-PA, and PAI-1 in human retinal glial cells (HRGCs) is influenced by exposure to transforming growth factor (TGF)-β, a cytokine that regulates the proliferation and differentiation of cells.

methods. The extracellular release of t-PA, u-PA, and PAI-1 was measured by enzyme-linked immunosorbent assay (ELISA) in the supernatant of HRGC cultures, under basal conditions and after stimulation with TGF-β at various concentrations (2, 5, 10, or 20 ng/mL). Reverse transcription-polymerase chain reaction (RT-PCR) was used to analyze mRNA levels. Smad2 phosphorylation was detected by Western blot analysis.

results. Under basal conditions, HRGCs secreted considerable amounts of t-PA and PAI-1. Stimulation with TGF-β resulted in increased synthesis of t-PA and PAI-1 protein in a time- and dose-dependent manner. Moreover, an increased expression of t-PA and PAI-1 mRNA after supplementation with TGF-β was observed, with maximum expression at 12 hours. In contrast, HRGCs did not respond to TGF-β with any change of u-PA production, although there were detectable amounts of u-PA mRNA and protein. Phosphorylation of Smad2 was increased after addition of TGF-β. This effect was partially reversible after treatment with interferon-γ.

conclusions. The production of plasminogen activators and PAI-1 by HRGCs reflects the potential role of these cells in the progression of neovascular ocular diseases. Furthermore, the finding that t-PA and PAI-1 synthesis by HRGCs is mediated by TGF-β and its downstream effector Smad2 confirms the importance of the TGF-β signaling pathway in the regulation of interactions between retinal cells and the extracellular matrix.

Cell matrix degradation is an important prerequisite for the migration and proliferation of cells in numerous disease states. Dissolution of the basement membrane requires production of several enzymes, including plasmin, which is generated locally from plasminogen, a leaked plasma protein, by the action of plasminogen activators. ^{1 2} Observations of tumor invasion and vascularization have suggested that the serine proteases tissue-plasminogen activator (t-PA) and urokinase-plasminogen activator (u-PA) play a central role in the regulation of this process. ^{2 3} Moreover, there is mounting evidence that plasminogen activators are involved in neovascular diseases of the retina and choroid. ^{4 5 6 7} Previous studies from our laboratory suggest the involvement of plasminogen activators and their inhibitors in diabetic retinopathy. We found that concentrations of t-PA and plasminogen activator inhibitor (PAI)-1 in the vitreous of patients with proliferative diabetic retinopathy are higher than control levels, and levels of both t-PA and PAI-1 correlate with vitreous concentrations of the angiogenic cytokine vascular endothelial growth factor (VEGF) in such patients. ⁵ Additional supporting evidence for this postulate includes the observation that human diabetic neovascular membranes contain high levels of proteolytic enzymes. ⁶  

Several studies have demonstrated that human retinal or microvascular endothelial cells produce t-PA, u-PA, and PAI-1 under basal conditions or after stimulation with cytokines. ^{8 9 10} However, although most investigators agree that the endothelial cell alone is capable of expressing all the information required for forming a new vascular network, it is likely that various cell types contribute to the complex mechanisms involved in the progression of neovascular ocular disorders. ^{1 11 12} It has been shown that several key angiogenic factors are produced primarily by nonvascular retinal cells, including ganglion cells, glial cells, and retinal pigmented epithelial cells. ^{13 14 15 16} This finding suggests that proliferative retinal changes may be the consequence of angiogenic and growth-promoting compounds in the neural retina, acting secondarily on the microvasculature, and supports the theory that proliferative disorders of the posterior segment may not be a primary vascular disease. 

In the present study, we considered the capability of human retinal glial cells (HRGC) to secrete plasminogen activators in vitro. To investigate possible mechanisms of interaction, we examined the effect of transforming growth factor (TGF)-β for its ability to modulate the production of t-PA, u-PA, or PAI-1 in HRGCs. TGF-β has been implicated in the regulation of cellular proliferation and differentiation, embryonic development, wound healing, and angiogenesis. ^{17 18 19} In particular, TGF-β has been linked to choroidal and vitreoretinal proliferative diseases. ^{20 21 22 23} Because recent studies have shown that TGF-β signals through a receptor kinase that phosphorylates and activates the transcription factors Smad2 and Smad3, ^{24 25} further investigation into the regulation of plasminogen activator mRNA and protein included the detection of Smad2 phosphorylation. 

Methods

Cell Culture

Human retinal Müller glial cells (HRGCs) were isolated from freshly enucleated bulbi for corneal transplantation, as described elsewhere, ²⁶ with slight modifications; tenets of the Declaration of Helsinki were observed. In brief, sensory retinas were removed from hemisected eyes, using forceps and scissors. Pars plana remnants and adherent vitreous tissue were cut away from the detached retinas, and the retinas were rinsed with Ca⁺²-Mg⁺²-free PBS and incubated in 0.1% trypsin solution for 30 to 90 minutes. PBS containing 20% FBS was then added to the cells, and the cells were dispersed by gentle pipetting. The cells were removed by centrifugation and resuspended in culture medium (M500; Cascade Biologics, Portland, OR), containing 20% FCS and glial-derived neurotropic factor (R&D Sytems, Wiesbaden-Nordenstadt, Germany) at a concentration of 5 ng/mL. Culture dishes and flasks were coated with synthetic basement membrane (0.6 mg/mL; Matrigel; Becton Dickinson, Heidelberg, Germany). The glial identity of the cells was confirmed by positive staining, using monoclonal antibody against fibrillary acidic protein and polyclonal antibody against cellular retinaldehyde-binding protein. 

Passage 7 to 13 HRGCs were used in the study. Cultures maintained for up to 13 passages did not show observable morphologic changes or loss of mitotic vigor. Cytotoxicity was monitored by using a commercially available cytotoxicity detection kit (Roche Molecular Biochemicals, Mannheim, Germany). 

Enzyme-Linked Immunosorbent Assay

Levels of t-PA, u-PA, and PAI-1 antigen in the supernatant of HRGC cultures, with and without stimulation with TGF-β, were determined by using commercial enzyme immunoassay kits (Biopool, Umeå, Sweden). Immunoassays for the measurement of concentrations of t-PA antigen were performed in 96-well test plates coated with anti t-PA antibody. The microtiter plates were incubated with 20 μL of undiluted and centrifuged samples for 60 minutes, after which the wells were washed four times with phosphate-buffered saline. In the next step, 50 μL of conjugate (horseradish peroxidase [HRP]-labeled anti-t-PA antibody) was added and incubated for 15 minutes. The amount of conjugated antibodies was detected by addition of 100 μL of substrate (tetramethylbenzidine) and incubated for 15 minutes. The reaction was stopped by addition of 100 μL of 1.5 M sulfuric acid. Absorbance was measured at 492 nm. The detection limit was 0.05 ng/mL. 

For the detection of u-PA antigen, microtiter well plates coated with anti-u-PA antibody were incubated with 50 μL undiluted and centrifuged samples, together with 50 μL anti-u-PA conjugate for 1 hour. After this, antibodies and samples were removed and the plates were washed. Substrate (200 μL ortho-phenylene-diamine) was then added, and the plate was incubated for 15 minutes at room temperature. Absorbance was measured at 492 nm. The detection limit for the determination of u-PA was 0.1 ng/mL. 

Detection of PAI-1 antigen was performed in 96-well plates containing immobilized monoclonal antibodies against PAI-1. Volumes of 20 μL of standard or diluted samples together with 50 μL anti PAI-1 conjugate were incubated for 2 hours. The wells were then washed, followed by incubation with 200 μL of substrate (ortho-phenylene-diamine). After 20 minutes, the reaction was stopped, and absorbance was measured at 492 nm. Assay sensitivity was 0.5 ng/mL. 

Activity Assay

Activity of secreted t-PA and PAI-1 was measured by using commercially available kits (Chromolize tPA Assay Kit: Biopool; Actibind PAI-1 Assay: WAK Chemie Medical GmbH, Bad Soden, Germany). 

The t-PA assay was performed as follows: 100 μL of each sample and standards were incubated for 20 minutes on a 96-well plate coated with monoclonal antibody against t-PA. Thereafter, the plate was washed and 50 μL of substrate reagent (lyophilized H-D-But-CHT-Lys-pNA and poly-d-lysine) and 50 μL plasminogen reagent (lyophilized plasminogen and fibrin degradation products [TDP]) were added. After 90 minutes’ incubation, the reaction was stopped, and the absorbance was measured at 405 nm. 

For the determination of PAI-1 activity the 96-well plate containing a monoclonal t-PA antibody was washed, and 100 μL of the diluted samples and standards were added to the wells. After an incubation period of 1 hour, the wells were washed, and an HRP-labeled monoclonal anti-PAI-1 antibody was added for 30 minutes. Again, the plate was washed and 100 μL substrate (8-(N,N′-diethylamino)-n-octyl-3,4,5-trimethoxybenzoate [TMB] in substrate buffer containing H₂O₂) was added. After 20 minutes, the reaction was terminated, and the absorbance was read at 450 nm (with a 492-nm reference filter). 

Reverse Transcription-Polymerase Chain Reaction

Total RNA was isolated from HRGCs after supplementation with vehicle (control experiments) or TGF-β at 4, 8, 12, and 24 hours with extraction reagent (TRIzol; Sigma, Deisenhofen, Germany). The RT-reaction was performed with 1 μg of RNA of each sample, with reverse transcriptase (Superscript II; Gibco BRL, Eggenstein, Germany) and oligo(dT)_12-18 primers. For determination of t-PA, the reverse transcript was incubated with Taq-polymerase and primers t-PA forward (5′-GGA TTC GTG ACA ACA TGC GAC-3′), which matches with the coding strand of human t-PA cDNA at positions 1740 to 1760 (GenBank accession number D01096; GenBank is provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD, and is available at http://www.ncbi.nlm.nih.gov/Genbank) and t-PA reverse (5′-TTT GAG GAA CAT GAC GGG CCA-3′), which matches with the reverse strand at positions 2245 to 2265, for 27 cycles (denaturing, 1 minute 94°C; annealing, 1 minute 56°C; and elongation, 1 minute 72°C), followed by a 72°C cycle (5 minutes). A PAI-1 fragment was amplified by using the primer-pair PAI-1 forward (5′-TCA TGG GCC AAG TGA TGG AAC-3′; positions 1136 to 1156) and PAI-1 reverse (5′-CAT GCA CAC TGT TTC TGG GGA-3′; positions 1778 to 1798; GenBank accession number X04744), for 26 cycles with an annealing temperature of 56°C. The RT product was also incubated for 23 cycles (annealing: 1 minute, 58°C) with a primer pair for GAPDH (sense: 5′-CCA TCA CCA TCT TCC AGG AG-3′, positions 131 to 150 and anti-sense: 5′-CCT GCT TCA CCA CCT TCT TG-3′, positions 705 to 724; GenBank accession number M17701). Electrophoresis of PCR-products and the control was performed on a 1% agarose gel containing ethidium bromide. 

Competitive PCR

To obtain an unspecific competitor DNA that fits the t-PA primers but yields a fragment of different size compared with the t-PA-specific RT-PCR product, a low-stringency RT-PCR (annealing temperature: 30°C for 27 cycles) using the t-PA primer-pair and cDNA from unstimulated cells was performed. The PCR-products were then separated on a 1% agarose gel, and a band of suitable size (approximately 800 bp) was cut out and cloned in the EcoRV site of a vector (Bluescript KS⁺; Stratagene, LaJolla, CA). For the experiments, the cDNA mimic was serially diluted and added to the usual PCR master mix containing 1 μL of template per reaction tube. 

Western Blot Analysis

Smad2 phosphorylation was detected by Western blot analysis. Stimulated cells were lysed in homogenization buffer (20 mM Tris-HCl [pH 7.5], 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2 mM dithiothreitol, 50 μg/mL leupeptin, 1 mM phenylmethylsulfonyl fluoride [PMSF]) and left on ice for 15 minutes. Samples were sonicated three times and centrifuged at 13,000 rpm for 2 minutes. The protein concentration of the lysate was determined using the Bradford protein assay (Bio-Rad, Munich, Germany). Total protein (40 μg) from each sample was subjected to SDS-PAGE (10% acrylamide). The protein was blotted to a nylon membrane and incubated overnight with a rabbit polyclonal anti-phospho-Smad2 antibody (Ser465/467; Upstate-Biomol, Hamburg, Germany). Blots were washed and incubated with a HRP-linked secondary antibody. Antibody-antigen complexes were detected using enhanced chemiluminescence Western blot analysis signal detection reagent (Amersham, Freiburg, Germany). 

Statistical Analysis

Results are expressed as the mean ± SD. The significance of differences between corresponding groups of observations was evaluated by one-way analysis of variance (ANOVA). The observed significance levels from multiple comparisons were adjusted using the Tukey test. Acceptable significance was recorded at P < 0.05. 

Results

Production of t-PA, u-PA, and PAI-1 Protein

We examined human HRGCs for their ability to produce t-PA, u-PA, and PAI-1 and their responsiveness to stimulation with TGF-β. Confluent cells were stimulated with vehicle (the control) and TGF-β at various concentrations. Levels of t-PA antigen in the supernatant of stimulated and unstimulated cells were determined by ELISA after 24, 48, and 72 hours. Under basal conditions (control), HRGC exhibited a time-dependent production of t-PA protein (Fig. 1A) and a considerable synthesis of PAI-1 protein (Fig. 1B) . Levels of both t-PA and PAI-1 continued to increase throughout the 72-hour assay period. 

Exposure to TGF-β at all concentrations tested resulted in a significantly increased production of t-PA and PAI-1 protein compared with the control. Levels of t-PA and PAI-1 antigen accumulated in a time- and dose-dependent manner (Figs. 1A 1B) . 

Cells exposed to vehicle (control) exhibited a measurable basic production of u-PA antigen (1.025 ± 0.17 ng/mL) in the supernatant of HRGC cultures. As for control cultures, u-PA antigen was detectable in cell culture supernatant from cells exposed to stimulation with TGF-β, at all concentrations tested, throughout the 72-hour period. However, after 72 hours, no significant increase in u-PA-antigen levels was observed (data not shown). 

Measurement of t-PA and PAI-1 Activity

After stimulation with TGF-β, a time- and dose-dependent increase in t-PA and PAI-1 activity was observed. This increase was consistent with the accumulation of t-PA and PAI-1 protein in the supernatant of HRGCs (Fig. 2A 2B) . 

Expression of t-PA, u-PA, and PAI-1 mRNAs in HRGCs

The presence of specific plasminogen activator mRNA was determined by semiquantitative RT-PCR in both control and TGF-β-stimulated cultures (Fig. 3A) . Based on the protein data, a TGF-β concentration of 10 ng/mL was selected as a suitable stimulus for further experiments. Levels of t-PA, u-PA, and PAI-1 mRNA were measured after stimulation with TGF-β at 4, 8, 12, and 24 hours. Densitometric analysis corrected for GAPDH revealed that TGF-β (10 ng/mL) treatment for 12 hours increased t-PA mRNA expression 2.2-fold (Fig. 3B) and PAI-1 mRNA expression 2.7-fold, compared with untreated control cells (Fig. 3C) . These differences were statistically significant (P < 0.05 and P < 0.01). Control cultures expressed basal levels of u-PA mRNA. We observed no increase in u-PA mRNA levels after stimulation with TGF-β (data not shown). 

To verify the experimental setup for semiquantitative RT-PCR, we also performed competitive PCR for t-PA with the same RT reactions as used for the 12-hour level (Fig. 4) . This method revealed a similar increase in t-PA mRNA as determined by semiquantitative RT-PCR. 

Phosphorylation of Smad2

Treatment of HRGCs with TGF-β resulted in phosphorylation of Smad2, with a maximum response after 30 minutes (Fig. 5A) . The stimulatory effect of TGF-β was inhibited after pretreatment with IFN-γ in a dose-dependent manner (Fig. 5B) . 

Effects of IFN-γ Treatment on t-PA and PAI-1 Expression

To investigate whether blocking of Smad2 phosphorylation by IFN-γ affects the expression of t-PA and PAI-1 mRNA, cDNA, and supernatant from HRGCs exposed to TGF-β and IFN-γ were used for further experiments. RT-PCR analysis revealed that pretreatment with 20 ng/mL IFN-γ completely reversed the upregulation of t-PA mRNA induced by TGF-β (Fig 6) . By contrast, IFN-γ exhibited no inhibitory effect on the increase of PAI-1 mRNA. 

The effect of IFN-γ on t-PA protein levels was less marked, with a significant reduction occurring only after preincubation with IFN-γ at a concentration of 20 ng/mL (Fig. 7A) . In accordance with the mRNA data, we found no reduction in PAI-1 protein levels after pretreatment with IFN-γ (Fig 7B) . 

Discussion

To date, there has been a wealth of evidence that proteolytic enzymes are involved in tissue destruction or remodeling and angiogenesis. In particular, the plasminogen-plasmin system has been linked to neoplastic spread and ocular proliferative diseases. ^{1 2 3 4 5 6 7 27 28 29} Results from studies using human vitreous or aqueous humor indicate that ocular tissues produce and release substantial amounts of plasminogen activators in vivo. ^{5 30} Supportive evidence for the production of plasminogen activators within the eye has also been provided by several immunohistochemical studies. Positive staining for t-PA in the posterior segment was observed in the inner retina, peripheral vitreous, and all vascular endothelia, and u-PA was localized in peripheral vitreous and retinal pigment epithelium. ^{31 32} Thus far, in vitro analysis of human retinal pigmented epithelial and endothelial cells has revealed that these cells are capable of generating t-PA, PAI-1, or u-PA and that the synthesis of plasminogen activators by these cells is modulated by various cytokines. ^{8 9 33 34}  

The present study identifies retinal glial cells as a potential source of plasminogen activators and plasminogen activator inhibitor within the eye microenvironment. It is generally believed that Müller glial cells play an important role in maintaining the integrity and normal function of the retina. It seems plausible that these cells are among the nonvascular retinal cells that may be involved in the development of proliferative diseases of the posterior segment. Although the endothelial cell alone is capable of expressing all the information required for the development and regression of new vessel sprouts, there is much evidence that nonvascular cells participate by producing cytokines or proteolytic enzymes. ^{13 14 15 16} Studies using tissue specimens of patients with proliferative diabetic vitreoretinopathy have provided evidence to support this hypothesis. Immunolabeling of human epiretinal membranes identified one of the cellular components in these membranes as Müller glia. ^{35 36} Furthermore, retinal glial cells have been shown to generate angiogenesis-regulating factors under normoxic and hypoxic conditions. ^{11 37}  

In this study, we observed a considerable production of t-PA and PAI-1 antigens in the supernatant of HRGCs under basal conditions. Moreover, stimulation with TGF-β resulted in a significantly increased synthesis of both t-PA and PAI-1 in a time- and dose-dependent manner. By contrast, little u-PA antigen was detected in the conditioned medium, and HRGCs did not respond to TGF-β with any change of u-PA production. These data suggest that although HRGCs are capable of synthesizing both types of plasminogen activators under conditions of confluence, the predominant plasminogen activator released by the cells is t-PA. 

It is widely accepted that cytokines have a central role in the progression of various ocular diseases associated with proliferation and neovascularization. In our study, we observed an increased expression of t-PA and PAI-1 mRNA after supplementation of TGF-β. It has been demonstrated that human retinal glial cells express TGF-β₁, TGF-β₂, and TGF-β₃, as well as TGF-β-receptor types I and II. ¹⁴ TGF-β is one of the most potent regulators of the production and deposition of extracellular matrix and has been implicated in a wide variety of cellular processes. ^{17 18 19} It stimulates the production and affects the adhesive properties of the extracellular matrix by two mechanisms. First, TGF-β stimulates various cells to produce extracellular matrix proteins and cell adhesion proteins. Second, TGF-β modulates the production of enzymes that degrade the extracellular matrix. ¹⁹ In mice, targeted deletion of TGF-β receptors results in decreased vasculogenesis associated with defective differentiation of capillary endothelium and inadequate capillary tube formation. ³⁸  

However, depending on the target cell and the conditions used, TGF-β may have stimulatory or inhibitory effects. ^{17 18 19 39} In an experimental study, Yoshimura et al. ⁴⁰ demonstrated that photocoagulation of cultured human retinal pigmented epithelial cells results in the production of an inhibitor of vascular endothelial cell proliferation. This inhibitor was identified as TGF-β₂. In a recent study, they were able to demonstrate that panretinal photocoagulation increases the synthesis of TGF-β in the retina at the mRNA and protein level. Based on their results, they hypothesized that upregulation of this cytokine may play an important role in the regression of ocular neovascularization after panretinal photocoagulation. ⁴¹ Other investigators have demonstrated that levels of TGF-β are increased in the vitreous of patients with proliferative diabetic retinopathy or proliferative vitreoretinopathy. ^{21 22} Moreover, expression of TGF-β mRNA has been shown to be upregulated in experimental choroidal neovascularization. ²⁰  

It is important to note that both the secretion of t-PA or PAI-1 protein into the medium and t-PA or PAI-1 activity exhibited a similar increase after stimulation with TGF-β. This strongly indicates a transcriptional mechanism of t-PA and PAI-1 upregulation in human retinal glial cells. Moreover, our data suggest that TGF-β-induced upregulation of t-PA mRNA and protein is mediated by Smad2, a pivotal downstream effector of TGF-β family members. An interesting observation was the inhibitory effect on the induction of t-PA expression of pretreatment with IFN-γ. This is consistent with the recent finding that IFN-γ induces the expression of Smad7, thereby interfering with the activation of Smad3 by TGF-β. ⁴² However, we found no effect of IFN-γ on the expression of PAI-1, indicating that other signaling pathways may be involved in the differential expression of plasminogen activators and their inhibitors by HRGCs. Furthermore, from the current data, it is not possible to state whether the release of other mediators stimulated by IFN-γ may influence the signaling downstream of the TGF-β- Smad cascade. 

Based on the findings presented herein, it can be speculated that glial cells constitute a potential retinal source of proteinases that may play a critical role in the development of neovascular diseases of the eye. However, because this was an in vitro study, it is not possible to draw definitive conclusions regarding the synthesis of proteinases within the eye microenvironment in vivo. Accordingly, it is not possible to state whether the generation of plasminogen activators by retinal glial cells exerts an inhibitory or stimulatory effect on intraocular proliferative changes. 

Because growth of the microvasculature is regulated by a balance between tissue destruction and remodeling, there is a reasonable likelihood that the expression of proteolytic enzymes is part of a sequence of events that contributes to ocular neovascularization. Experiments to determine whether plasminogen activator synthesis predominates in relation to health or disease were beyond the scope of the current studies. However, such studies would provide clinically relevant data and would be worth pursuing in future investigations. 

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2000.

 Supported by the August-Scheidel-Stiftung.

 Submitted for publication December 12, 2001; revised March 21, 2002; accepted March 29, 2002.

 Commercial relationships policy: N.

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2000.

 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.

 

Effect of TGF-β on t-PA (A) and PAI-1 (B) protein levels in the supernatant of HRGCs after 24, 48, and 72 hours, as determined by ELISA. Cells were supplemented with vehicle (control) or TGF-β at 2, 5, 10, or 20 ng/mL. Exposure to TGF-β at all concentrations tested resulted in a significantly increased production of t-PA antigen. The experiments were performed at least in triplicate with triple values each. *P < 0.05; **P < 0.01 versus the control.

Figure 2.

 

Activity of t-PA (A) and PAI-1 (B) from supernatants of TGF-β-treated HRGC cultures. Activity assays were performed in parallel with ELISA assays with the same samples. *P < 0.05; **P < 0.01 versus the control.

Figure 3.

 

Temporal expression of t-PA and PAI-1 mRNA in cultured human retinal glial cells exposed to TGF-β at a concentration of 10 ng/mL. Levels of mRNA were determined by RT-PCR at 4, 8, 12, and 24 hours. Each bar represents the mean (±SD) of three independent experiments. (A) PCR products of t-PA, PAI-1, and GAPDH. (B) Densitometric analysis of the t-PA signals corrected for GAPDH. (C) Densitometric analysis of the PAI-1 signals corrected for GAPDH. *P < 0.05; **P < 0.01 versus the control.

Figure 4.

 

Competitive PCR of the control and after stimulation of HRGCs with TGF-β (10 ng/mL) at 12 hours. For each reaction, the PCR master mix was supplemented with 1 μL of RT product of treated or untreated cells, plus decreasing concentrations of the DNA mimic.

Figure 5.

 

Effects of TGF-β and IFN-γ on the phosphorylation of Smad2 in HRGCs. Total protein of treated and untreated cells was subjected to Western blot analysis. Signals of phosphorylated Smad2 were detected after stimulation with TGF-β for various periods (A) and after pretreatment with IFN-γ at various concentrations (B).

Figure 6.

 

RT-PCR of stimulated HRGCs and the control after 8 hours. To analyze the effects of IFN-γ on the expression of t-PA and PAI-1 mRNA, cells were pretreated with 20 ng/mL IFN-γ for 30 minutes before the addition of TGF-β.

Figure 7.

 

Levels of t-PA (A) and PAI-1(B) protein in the supernatants of HRGCs as determined by ELISA. Treatment with various concentrations of IFN-γ was performed 30 minutes before the addition of TGF-β. *P < 0.05; **P < 0.01 versus the control.

The authors thank Jindrich Cinatl, Department of Microbiology, Johann Wolfgang Goethe University Hospital, Frankfurt, Germany, for the kind donation of human retinal glial cells.