Molecular Vision 2005; 11:397-413 <http://www.molvis.org/molvis/v11/a47/> Received 1 September 2004 | Accepted 31 May 2005 | Published 10 June 2005

Download Reprint

Changes in retinal gene expression in proliferative vitreoretinopathy: glial cell expression of HB-EGF

Margrit Hollborn,¹ Solveig Tenckhoff,¹ Karsten Jahn,¹ Ianors Iandiev,² Bernd Biedermann,² Ute E. K. Schnurrbusch,¹ G. Astrid Limb,³ Andreas Reichenbach,² Sebastian Wolf,¹ Peter Wiedemann,¹ Leon Kohen,¹ Andreas Bringmann¹
 
 

¹Department of Ophthalmology, University Eye Hospital and ²Paul Flechsig Institute of Brain Research, University of Leipzig, Leipzig, Germany; ³Institute of Ophthalmology and Moorfields Eye Hospital, London, UK

Correspondence to: M. Hollborn, PhD, Department of Ophthalmology, Eye Hospital, University of Leipzig Medical Faculty, Liebigstrasse 10-14, D-04103 Leipzig, Germany; Phone: (49) 341-97-21-561; FAX: (49) 341-97-21-659; email: hollbm@medizin.uni-leipzig.de

Abstract

Purpose: To compare the gene expression pattern of control postmortem retinas with retinas from patients with proliferative vitreoretinopathy (PVR), to determine the expression of the heparin binding epidermal growth factor-like growth factor (HB-EGF) by glial cells in fibroproliferative membranes, and to examine whether cells of the human Müller cell line, MIO-M1, respond to HB-EGF with proliferation, migration, and secretion of the vascular endothelial growth factor (VEGF).

Methods: To identify genes that were differently expressed in PVR and control retinas, the RNA from the neural retinas of seven postmortem donors and of two patients with PVR were analyzed for differential gene expression, by hybridization of labeled cRNA probes to an Affymetrix human genome microarray set. The results were validated by real time PCR experiments investigating RNA from 6 postmortem retinas and 4 PVR retinas. Epiretinal PVR membranes were immunohistochemically stained for colocalization of HB-EGF and the glial cell marker, glial fibrillary acidic protein (GFAP). The HB-EGF evoked proliferation of cultured Müller cells was investigated by a bromodeoxyuridine immunoassay, chemotaxis was assessed with a migration assay, and the release of VEGF was evaluated by ELISA.

Results: Out of the 12,600 genes and expressed sequence tags investigated, the levels of 80 showed an increased expression, and 21 were expressed at decreased levels, in the retinas of PVR patients compared to the control retinas. The upregulated signals include genes for nuclear and cell cycle related proteins, extracellular secretory proteins, cytosolic signaling proteins, and proteins of the membrane and the extracellular matrix. The genes of the hepatocyte growth factor and of HB-EGF were found to be expressed in PVR retinas but not in control retinas. In epiretinal membranes of patients with PVR, HB-EGF immunoreactivity partially colocalized with GFAP. In cultured Müller cells, HB-EGF stimulated both proliferation and chemotaxis, and the secretion of VEGF, via activation of the extracellular signal regulated kinases 1 and 2 and of the phosphatidylinositol-3 kinase.

Conclusions: The development of PVR is accompanied by complex changes of the gene expression in the neural retina, with an upregulation of genes that support cell proliferation, cell signaling, cell motility, and extracellular matrix remodeling. HB-EGF is one of the factors that are significantly upregulated in PVR retinas. HB-EGF expression in fibroproliferative tissue and its stimulatory effect on glial cell proliferation, chemotaxis, and VEGF secretion suggest that HB-EGF may be a factor mediating glial cell responses during PVR.

Introduction

Proliferative vitreoretinopathy (PVR) is considered to represent a dysregulated wound healing process of the retina in which retinal cells are overstimulated by growth factors and cytokines [1-3]. Together with fibroblasts, retinal pigment epithelium (RPE), and inflammatory cells, retinal glial cells represent a major cellular constituent of fibroproliferative tissue associated with PVR [4,5]. Similar to RPE cells, retinal glial cells are immediately activated by experimental retinal detachment, and begin to proliferate within hours of detachment [6,7]. Gliotic Müller cells from PVR retinas display significant alterations of various physiological and morphological features such as cell hypertrophy and immunoreactivity for glial fibrillary acidic protein (GFAP), and a changed responsiveness of cell surface receptors. One class of receptors that are upregulated in their functional responsiveness in glial cells from PVR retinas, are purinergic P2 receptors [8,9]. However, a comprehensive analysis of altered cellular functions in retinas of PVR patients is missing until today. Therefore, we performed a gene expression analysis using the Affymetrix human genome microarray set, in order to detect genes that were up or down regulated in retinas from PVR patients as compared to retinas from postmortem control donors. We found that, in addition to genes that support cell proliferation, cell signaling, cell motility, and extracellular matrix remodeling, the mRNA levels of two growth factors, hepatocyte growth factor (HGF) and heparin binding epidermal growth factor-like growth factor (HB-EGF), were elevated in retinas of PVR patients. While HGF has been implicated in the development of proliferative retinopathies, as it induces scattering, migration, and proliferation of RPE cells [10,11] and has effects on retinal glial cells [12], HB-EGF is a factor not yet described as being involved in proliferative diseases of the retina.

In various cell lines, the proliferation-stimulating effect of growth factor and G protein coupled receptors depends on the transactivation of receptor tyrosine kinases, especially of those of the EGF receptor [13]. One major pathway of this transactivation is via autocrine/paracrine shedding of HB-EGF [14]. Recently, this mitogenic signaling cascade has been suggested to be activated in primary cultures of retinal glial (Müller) cells of the guinea pig when Müller cell proliferation was stimulated by purinergic agonists [15]. In many types of cells and tissues, the HB-EGF precursor is expressed as a membrane bound cell surface molecule, and the soluble form of HB-EGF is formed by metalloproteinase-mediated shedding after cell activation [16]. In the rat brain, neuroblasts, mature neurons, and glial cells express HB-EGF [17,18], and this expression was found to be elevated under ischemic/hypoxic conditions [19]. HB-EGF supports the survival of hippocampal neurons after excitotoxic insults, probably via activation of glial cells [20], and it stimulates the proliferation and migration of brain astrocytes [21,22]. Recently, a role of HB-EGF shedding and EGF receptor activation was suggested in corneal epithelial wound healing [23]. However, nothing is presently known about a possible role of this factor in mediating wound healing processes and cell proliferation in disorders of the human retina. Therefore, the second aim of the present study was to investigate whether HB-EGF may represent a factor involved in glial cell proliferation during PVR. We studied the presence of the factor in fibroproliferative membranes surgically excised from eyes of patients with PVR, and examined the proliferation and migration stimulating effect of this factor in a human Müller cell line known as MIO-M1 [24]. Additionally, the effect of HB-EGF on the secretion of vascular endothelial growth factor (VEGF) by cells of the Müller cell line was determined.

Methods

Materials

HB-EGF was purchased from R&D Systems (Minneapolis, MN). 6,7-Dimethoxy-3-phenylquinoxaline (Thyrphostin AG1296), 4-[(3-Bromophenyl)amino]-6-(methylamino)-pyrido[3,4-d]pyridimine (PD158780), 2'-Amino-3'-methoxyflavone (PD98059), and 2-(4-Morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002) were obtained from Calbiochem (Bad Soden, Germany). [4-(3-Chloranilino)]-6,7-dimethoxyquinazoline (Thyrphostin AG1478) was obtained from Alexis (Grünberg, Germany). 1,4-Diamino-2,3-dicyano-1,4-bis[2-aminophenylthio]butadiene (UO126) and 4-[5-(4-Fluorophenyl)]-2-[(4-methbylsulfonyl)phenyl]-1[H-imidazol-4-yl] pyridine (SB203580) were from Tocris (Ellisville, MO), and Hoechst 33258 was from Molecular Probes (Eugene, OR). All other reagents were obtained from Sigma (Deisenhofen, Germany), unless otherwise indicated. The following antibodies were used: a neutralizing goat anti-human HB-EGF (1:100; R&D Systems), a rabbit anti-bovine GFAP (1:1000; Dako A/S, Glostrup, Denmark), a rabbit anti-human p44/p42 MAPK (New England Biolabs, Frankfurt/M., Germany; 1:1000), a rabbit anti-phosphorylated MAPK (New England Biolabs, 1:1000), a rabbit anti-human Akt (New England Biolabs, 1:1000), a rabbit anti-phosphorylated Akt (New England Biolabs, 1:1000), a Cy3-tagged donkey anti-goat IgG (1:400; Dianova, Hamburg, Germany), a Cy2 coupled donkey anti-rabbit IgG (1:400; Dianova), and an anti-rabbit IgG conjugated with alkaline phosphatase (Chemicon; Hofheim, Germany, 1:2000).

Ocular tissues

All tissue was used in accordance with applicable laws and with the Declaration of Helsinki for research involving human tissue, and was approved by the Ethics Committee of the Leipzig University Medical School.

Total RNA preparation

Immediately after surgery, PVR retinas were removed from the vitrectomy waste, and collected by centrifugation at 4 °C, washed twice with chilled phosphate buffered saline to remove blood cells, and used for total RNA preparation. The preparation was carried out using 1 ml Trizol reagent (Gibco BRL, Paisley, UK). RNA was additionally purified by using an RNeasy Mini Kit (Qiagen, Hilden, Germany). The total RNA samples were analyzed by agarose gel electrophoresis to assure that the ribosomal RNA was intact. The A₂₆₀/A₂₈₀ ratio was measured by using a GeneQuant_pro spectrophotometer (Pharmacia, Uppsala, Sweden), and the ratio was between 1.9 and 2.1 for all samples, indicating that the total RNA samples were of acceptable quality.

RT-PCR

For the reverse transcriptase polymerase chain reaction (RT-PCR), the total RNA of 5 postmortem retinas were used (one woman, 4 men; age 46.4±25.30 years; range, 17-70 years). Additionally, the total RNAs from retinal fragments of 5 patients undergoing 360° retinectomies for PVR (grades CP4, CA4 and above) were analyzed (4 women, 1 man, age, 71.8±2.95 years; range, 68-76 years). cDNA was synthesized from 1 μg total RNA using the First-Strand cDNA Synthesis Kit for RT-PCR (Roche, Mannheim, Germany). For PCR, the Taq PCR Master Mix Kit (Qiagen, Hilden, Germany) was used. An aliquot (1 μl) of the first-strand mixture and 1 μM of each gene specific sense and antisense primer were used for the amplification reaction in a final volume of 25 μl. The following primer pairs were used: HB-EGF (accession number M60278), sense 5'-CCG TGG TGG CTG TGG TGC TGT-3', antisense 5'-GCA GTC CCC AGC CGA TTC CTT-3', producing a 173 bp amplicon; β-actin (accession number M10277), sense 5'-ATG GCC ACG GCT GCT TCC AGC-3', antisense 5'-CAT GGT GGT GCC GCC AGA CAG-3', producing a 237 bp amplicon. Thermocycling was carried out by using a PTC-200 Thermal Cycler (MJ Research, Watertown, MA) according to the following protocol: denaturation at 94 °C for 3 min, followed by 40 cycles of denaturation at 94 °C for 30 s, annealing at 60 °C for 1 min, and polymerization at 72 °C for 2 min. A final extension step was made at 72 °C for 10 min. The amplified products were separated on a 1.7% agarose gel containing 10 ng/ml ethidium bromide, and were visualized using an ultraviolet transilluminator (Fluor-S-Imager; BioRad, Munich, Germany).

Quantitative real-time PCR

Total RNAs from retinal fragments of 4 patients with PVR (2 women, 2 men; age 64.5±14.15 years; range, 44-76 years) or of 6 postmortem donors (3 women, 3 men; age 59.83±18.25 years; range, 33-86 years) were used. The RNA of two specimens of this 6 post mortem retinas were used for the RT-PCR in parallel. The total RNA preparations from control and PVR retinas were analyzed in duplicate. The contaminating genomic DNA was eliminated by using 1 unit of DNase I (Roche, Mannheim, Germany) per μg of total RNA. First-strand cDNA synthesis was performed with 1 μg of total RNA by using the iScript cDNA Synthesis Kit (BioRad, Hercules, CA) in a final volume of 20 μl. The following primer pairs were used: HB-EGF: sense 5'-CTG TCT CCC CGT GTC CTC TCC-3', antisense 5'-GCT CCA ATG TTC CCT GGT CCT-3', producing a 252 bp amplicon; and GAPDH (accession number M33197), sense 5'-GCA GGG GGG AGC CAA AAG GGT-3', antisense 5'-TGG GTG GCA GTG ATG GCA TGG-3', producing a 219 bp amplicon. Real-time PCR was performed with the MyiQ^TM Single-Color Real-Time PCR Detection System (BioRad). The PCR solution contained 1 μl cDNA, specific primer set (1 μM each), and 10 μl of iQ^TM SYBR® Green Supermix (BioRad) in a final volume of 20 μl. The PCR parameters were initial denaturation, one cycle at 95 °C for 3 min; amplification and quantification, 40 cycles at 95 °C for 30 s, 58 °C for 30 s, and 72 °C for 1 min; melting curve, 55 °C with the temperature gradually increased (0.5 °C) up to 95 °C. The amplified products were separated on a 1.7% agarose gel. The relative quantification of real-time PCR results was performed by using the mathematical model of Pfaffl [25]. The changes in HB-EGF gene expression were calculated using the average values of two independent experiments; the HB-EGF mRNA expression was normalized to the levels of GAPDH mRNA. The degree of change in HB-EGF mRNA was calculated and compared with the control retinas; results are expressed as means±SEM.

Oligonucleotide microarray analysis

The total RNA from retinal fragments of two patients (one man, one woman; age 68 and 73 years) were used in two experiments, respectively. One experiment was performed using pooled total RNA (2 μg of each) of seven postmortem retinas of human subjects (postmortem time up to 24 h) with no reported history of eye disease (2 women, 5 men; age 69.3±12.8 years; range, 46-86 years).

The expression of different types of mRNA in the total RNA preparations was analyzed by using the Human Genome U95Av2 Array of the Affymetrix Genechip System (Affymetrix, Santa Clara, CA), according to the manufacturer's instructions. This microarray contains probes representing approximately 12,600 full length genes and expressed sequence tags. Each gene analyzed was represented by 16 to 20 different oligonucleotide probe pairs, with each probe pair consisting of a match and a mismatch oligonucleotide. The mismatch probes, which served as controls for the determination of background and of nonspecific hybridization signals, included a single base substitution which inhibited the hybridization with the mRNA of the target gene. The mRNA expression levels were evaluated using the Affymetrix Microarray Suite 5.0 software, which calculated three detection levels (present, marginal, absent) and the size of the messages, by considering both the intensities of the signals that were emitted from the probe sets and the number of probe pairs in which the perfect match was specific. The RNA abundance was determined based on the average of the differences between perfect match and mismatch intensities for each probe family. Gene induction or repression was considered as significant if the change in average difference intensity was above three fold. A comparison analysis was carried out, which evaluated the relative change in abundance for each transcript between a baseline (control retinas) and an experimental sample (each PVR retina). The integrity of the used cRNA samples was tested by checking the 5'-to-3' ratios of the housekeeping genes.

Cell culture

The culture experiments were carried out by using a spontaneously immortalized human Müller cell line, MIO-M1 [24]. The cells were cultured in tissue culture flasks (Greiner, Nürtingen, Germany) in Dulbecco's modified Eagle's medium (Invitrogen, Paisley, UK) containing 10% fetal bovine serum (FBS), glutamax II, and gentamycin in 19.5% O₂/5% CO₂ at 37 °C. After reaching confluency the cells were used for further experiments.

DNA synthesis rate

The rate of incorporation of bromodeoxyuridine (BrdU) into the genomic DNA during cell growth was used as an indirect parameter to determine the proliferative activity of the cells. To perform the proliferation experiments, the cells were seeded at 3x10³ cells per well in 96 well flat bottom microtiter plates (Greiner) containing medium supplemented with 10% serum, and were allowed to attach for 48 h. Thereafter, the cells were growth arrested in medium without serum for 16 h, and subsequently, the test substances were added to the culture medium for 24 h. The DNA synthesis rate was assayed by using the Cell Proliferation ELISA BrdU Kit (Roche, Mannheim, Germany) according to the manufacturer's instructions. BrdU (10 μM) was added to the culture medium for 5 h. After fixation of the cells, a peroxidase (POD) labeled antibody against BrdU (1:200) was added for 4 h. After washing 3 times with PBS, the POD substrate 3,3',5,5'-Tetramethylbenzidine (TMB; Calbiochem) was added, and the absorbance in the samples was determined using a microplate reader (SpectraMax 250; Molecular Devices, Sunnyvale, CA) at 405 nm.

Chemotaxis

Measurements of chemotaxis were performed with a modified Boyden chamber assay. Suspensions of MIO-M1 cells (100 μl; 5x10⁵ cells/ml serum free medium) were seeded onto polyethylene terephthalate filters (pore size 8 μm; Becton Dickinson, Heidelberg, Germany), coated with fibronectin (50 μg/ml) and gelatin (0.5 mg/ml). Within 4 h after seeding, the cells attached to the filter and formed a semiconfluent monolayer. The medium was then changed to medium without additives in the upper well and medium containing HB-EGF at different concentrations in the lower well. After incubation for 6 h, the inserts were washed with buffered saline, fixed with Karnofsky's reagent, and stained with hematoxylin. Nonmigrated cells were removed from the filters by gentle scrubbing with a cotton swab. The migrated cells were counted, and the results were expressed relative to cell migration without HB-EGF.

VEGF ELISA

MIO-M1 cells were cultured at 3x10³ cells per well in 96 well microtiter plates (0.1 ml culture medium per well). After about 80% confluency was achieved, the cells were cultured in serum free medium for 16 h. Subsequently, the culture medium was changed, and the cells were stimulated with HB-EGF (50 ng/ml), in the absence and presence of blocking substances always in triplicate. The supernatants were collected at 6 h (supernatants of three wells were pooled), and levels of VEGF-A₁₆₅ in the supernatants were determined by ELISA (R&D Systems).

Immunohistochemistry

Six surgically excised epiretinal PVR membranes, obtained from consenting patients during vitrectomy surgery (4 women, 2 men; age 72.3±8.9 years; range, 49-87 years), were used for immunohistochemical staining. Epiretinal membranes were fixed in acetone, and stored at -80 °C. The membranes were immunostained as free floating wholemounts. After incubation in 10% normal goat or donkey serum plus 0.3% Triton X-100 in saline for 1 h, the wholemounts were incubated in primary antibody overnight at 4 °C. After washing in 1% bovine serum albumin in saline, the secondary antibodies were applied for 4 h at room temperature. The labeling was visualized by means of a confocal laser scanning microscope LSM 510 (Zeiss, Oberkochen, Germany).

Western immunoblotting

MIO-M1 cells were seeded in 6 well plates (5x10⁵ cells/well) and were cultured until 90% of confluency was reached in medium with 10% FBS. Then the cells were growth arrested in serum free medium for 16 h. The cultures were preincubated with the blockers for 30 min and then stimulated with HB-EGF for 10 min to cell harvest. After the stimulation the medium was removed, the cells were washed twice with cold phosphate buffered saline (pH 7.4; Invitrogen, Paisley, UK), and the monolayer was scraped into 200 μl of lysis buffer (Mammalian Cell Lysis-1 Kit, Sigma). The total cell lysates were centrifuged at 10,000x g for 10 min, and the supernatants were analyzed by immunoblot. Equal amounts of protein (30 μg) were separated by 10% SDS-polyacrylamide gel electrophoresis. Immunoblots were probed with primary and secondary antibodies, and immunoreactive bands were visualized by using 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Roth, Karlsruhe, Germany).

Statistics

Bromodeoxyuridine (BrdU) incorporation, migration rates, and VEGF content of the supernatants are expressed as percent of untreated control. Real time PCR results were compared using the nonparametric Mann-Whitney U test. Data are given as means±SEM; statistical significance (Student's-t-test) was accepted at p<0.05.

Results

Gene expression analysis

To obtain data on differences in retinal gene expression between postmortem control retinas and retinas from patients with PVR, a genechip microarray analysis was carried out. The total RNAs of 7 control retinas of individuals without reported eye disease and two retinal fragments removed from patients with PVR were used. The difference in the expression level of the vast majority of the 12,600 genes and expressed sequence tags investigated were less than three fold; however, there were 101 genes that showed stronger differences and were, therefore, considered to be up or down regulated in retinas of PVR patients. The results of two independent DNA microarray analyses are summarized in Table 1, Table 2, and Table 3. The majority of genes that were found to be expressed at elevated levels in PVR retinas, or to be significantly present only in PVR retinas, could be classified into five groups: nuclear and cell cycle related genes (26% of the identified upregulated genes), genes for cytosolic signaling proteins (13%), genes for extracellular secretory proteins (12%), genes for membrane proteins (11%), and genes for extracellular matrix proteins (5%; Table 1, Table 2).

Among the genes that were upregulated in their expression in PVR retinas, genes for c-fos and Jun B (Table 1) belong to the immediate early genes; ras related rho (RhoA) and the G protein β-3 subunit (Table 2) are involved in transmembrane signal transduction, conveying signals from activated receptors to an intracellular response. Upregulation of the retinal expression of Jun B has been linked to glial cell proliferation [26] while upregulation of cyclin D1 (Table 2) has been associated with neuronal apoptosis [27]. Different genes for proteins associated with the actin cytoskeleton (e.g., the genes for β-actin) were upregulated in their expression (Table 1) and for proteins that regulate the cytoskeleton, such as filamin A and the LIM domain protein (Table 2). The small guanosine triphosphatase RhoA links extracellular signals to dynamic changes in the actin cytoskeleton, and is associated with the formation of actin stress fibers and focal adhesions [28]. The strong expression of the genes for fibronectin and for tissue inhibitors of metalloproteinase (TIMP)-1 and -3 (Table 2) may be involved in extracellular matrix remodeling in PVR retinas, and may support cell shape changes during migration. The strong upregulation of the expression of cellular fibronectin (Table 2), which has been also described to be present in epiretinal membranes [29], may be related to changes in actin cytoskeleton, via connection to transmembrane integrins, and may transmit cell motility signals. Fibronection has been implicated not only in contraction but also in the stimulation of cell proliferation [30]. Immunorelated genes, such as IgG Fc binding protein (Table 2), complement component 4, and histocompatibility antigens (Table 1) were also upregulated. Activation of the complement system is one feature of neuroinflammatory responses that also involves an increased expression of major histocompatibility complex (MHC) class I and II antigens, cytokines, and cell adhesion molecules. Class I MHC molecules are expressed by retinal ganglion cells [31] and may play a role in inflammatory and immune responses and in the remodeling of neural tissue [32]. β2 Microglobulin (Table 1) is a cosubunit of class I MHC. Apolipoprotein E (Table 1), which may be synthesized by Müller glial cells and possibly by ganglion cells [33], has been implicated in the wound healing process after retinal detachment [34]. The prostate tumor-overexpressed gene 1 (PTOV1; Table 1) recently was found to be expressed in prostate carcinomas [35], and high levels of PTOV1 in tumors correlates with the proliferative index. PTOV1 promotes entry into the S phase of the cell division cycle, and this action is accompanied by greatly increased levels of cyclin D1 protein [36]. The insulin-like growth factor binding proteins (IGFBPs, Table 2) may stimulate cell proliferation in an IGF independent manner, or may inhibit proliferation by suppressing the action of IGFs [37], and may positively or negatively modulate apoptosis induction [38,39]. The bone morphogenetic protein-4 (Table 2) has been shown to either inhibit or enhance the proliferation and apoptosis of retinal cells [40-42]. The downregulation of the gene for rhodopsin (Table 1) may reflect a dysfunction of photoreceptor cells, likely by degeneration due to retinal detachment. Similarly, the enhanced expression of αA-crystallin (Table 1) may reflect photoreceptor cell degeneration [43]. The decreased expression of neurofilament (Table 1) may be caused by a degeneration of retinal ganglion cells [44].

Among the genes for growth factors and cytokines suggested to be involved in proliferative and migratory cell responses during PVR, the gene products for HGF and HB-EGF were found to be absent in control retinas but were significantly expressed in PVR retinas (Table 2). The genes for several other factors such as tumor necrosis factor-αand basic fibroblast growth factor (bFGF) showed only a doubling of their expression (data not shown). Other members of the EGF family of growth factors (EGF, amphiregulin, β-cellulin, epiregulin), platelet derived growth factor (PDGF), and IGF were either expressed at low levels in the retinas and, therefore, were not detected by array hybridization, or showed no significant alteration in their expression (data not shown). The microarray did not identify transforming growth factor-α. Until now, HB-EGF was not considered to be involved in cellular responses during proliferative diseases of the retina. Therefore, further experiments were conducted in order to validate the microarray results regarding the enhanced expression of HB-EGF in PVR retinas, to determine a possible glial cell expression of HB-EGF in fibroproliferative retinal material, and to investigate the physiological responses of cultured Müller cells to stimulation with HB-EGF.

HB-EGF mRNA expression in retinas and Müller cells

In order to confirm the expression of mRNA for HB-EGF in human retinas, an RT-PCR analysis was carried out using the total RNA extracted from the retinas of 4 patients with PVR. As shown in Figure 1A, all retinas investigated expressed the mRNA for HB-EGF. To validate the cDNA microarray results obtained for HB-EGF mRNA in PVR retinas compared to postmortem control retinas, quantitative real-time RT-PCR was carried out. With this method, a 2.32±0.45 fold elevated expression level of HB-EGF mRNA was observed in PVR retinas (Figure 1B,C). The increased expression level of the mRNA for HB-EGF in PVR retinas suggests that the factor may be involved in alterations of retinal cell physiology during PVR. By using RT-PCR, it was revealed that MIO-M1 cells express mRNA for HB-EGF and that the mRNA levels coding for this factor are increased after stimulation of the cells with different growth factors such as transforming growth factor-β and bFGF (Figure 1D). The quantification of this observation were determined by qRT-PCR and the results are shown in Figure 1E. A strong significant upregulation of the HB-EGF mRNA expression after TGF-β1 or TGF-β2 stimulation in MIO-M1 was found, whereas PDGF showed no influence and bFGF had a slight, but not significant, upregulating effect.

HB-EGF immunoreactivity in epiretinal membranes

To determine whether HB-EGF immunoreactivity is expressed by glial cells present in epiretinal membranes from patients with PVR, wholemounts of surgically excised specimens were co-labeled for HB-EGF and the glial cell marker GFAP. All 6 PVR membranes investigated showed both HB-EGF and GFAP immunoreactivity. In addition to glial cell bodies (Figure 2A,B), the predominant glial structures observed in epiretinal PVR membranes consisted of networks of fibrillary, long and thin GFAP positive structures which passed through certain regions of the membranes and which resembled processes of hypertrophic glial cells (Figure 2C). These thin GFAP positive structures were, at certain sites, double stained with HB-EGF, whereas they were partially devoid of HB-EGF protein at other sites. Similarly, some glial cell bodies in PVR membranes expressed HB-EGF immunoreactivity, while others did not stain for this factor (Figure 2A). Control stainings (omitting the anti-HB-EGF antibody) revealed no immunoreactivity (data not shown). The results indicate that at least a subpopulation, or subcellular regions, of glial cell bodies in epiretinal PVR membranes expressed HB-EGF immunoreactivity. There were also HB-EGF expressing structures in PVR membranes that were not double labeled for GFAP (Figure 2B), representing non-glial cells expressing HB-EGF protein, or secreted HB-EGF protein attached to the extracellular matrix.

Proliferation of Müller cells

In order to determine a possible physiological significance of HB-EGF in Müller cells, the HB-EGF evoked proliferation of cells of the human Müller cell line, MIO-M1, was investigated. Addition of HB-EGF to the culture medium evoked a dose dependent increase in the proliferation rate of the cells, with an apparent EC₅₀ of about 1 ng/ml and a maximal effect near 10 ng/ml (Figure 3A). The mitogenic effect of HB-EGF was inhibited in the presence of a neutralizing anti-HB-EGF antibody (Figure 3B). The HB-EGF evoked Müller cell proliferation was dependent on the activation of the EGF receptor tyrosine kinase, as indicated by the inhibiting effects of two different selective blockers, tyrphostin AG1478 [45] and PD158780, respectively (Figure 3C). The specificity of these effects is supported by the observation that a selective blocker of PDGF receptor tyrosine kinase, tyrphostin AG1296 [46], did not inhibit the HB-EGF effect (Figure 3C), and that the proliferation evoked by PDGF-BB (100 ng/ml) was inhibited by AG1296 but not by blockers of the EGF receptor tyrosine kinase (data not shown).

The HB-EGF evoked proliferation of Müller cells was dependent on the activation of the extracellular signal regulated kinases 1 and 2 (ERK-1, ERK-2), as indicated by the decreased effect of HB-EGF in the presence of the MAPK kinase 1 and 2 (MEK1, MEK2) inhibitor U0126 (Figure 3D). The stimulating effect of HB-EGF on the tyrosine phosphorylation of ERK-1 and ERK-2 was confirmed by using western blotting (Figure 4). HB-EGF caused a dose dependent increase in the phosphorylation level of ERK-1 and ERK-2, with the maximal effect at 10 ng/ml (Figure 4A). The stimulating effect of HB-EGF on ERK-1 and ERK-2 phosphorylation was fully prevented by co-incubation with the selective blocker of the EGF receptor tyrosine kinase, tyrphostin AG1478, whereas the blocker of the PDGF receptor tyrosine kinase, tyrphostin AG1296, had no effect (Figure 4B). Activation of the p38 mitogen activated protein kinase (p38 MAPK) was not necessary for the mitogenic effect of HB-EGF, as suggested by the absence of an effect of the p38 MAPK blocker, SB203580 (Figure 3D, Figure 4B). The inhibitor of the phosphatidylinositol-3 kinase (PI3K), LY294002, decreased both the control and growth factor evoked proliferation (Figure 3D). However, the inhibiting effect of LY294002 on ERK-1 and ERK-2 activation was very small in comparison to the HB-EGF evoked stimulation (Figure 4B). In contrast to the mitogenic effect of HB-EGF on the Müller cell line, addition of VEGF (100 ng/ml) did not significantly stimulate the rate of DNA synthesis (102.6±6.0%, an average of two experiments).

Chemotaxis of Müller cells

To test for a possible chemotactic effect of HB-EGF on Müller glial cells, MIO-M1 cells were stimulated with different concentrations of HB-EGF, and their migration rate was measured after 6 h. As shown in Figure 5, HB-EGF induced a dose dependent increase of Müller cell migration, at doses above 0.1 ng/ml.

HB-EGF evoked release of VEGF

Retinal hypoxia is a consistent feature of proliferative retinopathies [47], and Müller cells release VEGF in response to hypoxic stimuli [48-50]. Since HB-EGF is expressed by glial cells in fibroproliferative epiretinal membranes (Figure 2), we investigated whether HB-EGF may be a factor causing VEGF release by Müller cells and whether this effect is mediated by activation of the EGF receptor tyrosine kinase, ERK-1, ERK-2, or PI3K. MIO-M1 cells were cultured for 6 h in the presence of HB-EGF, and the VEGF content in the supernatants was measured by ELISA. VEGF is constitutively released by the cells into the culture medium, and HB-EGF (50 ng/ml) evoked a doubling of the basic VEGF release (Figure 6A). The HB-EGF evoked release of VEGF was significantly diminished in the presence of a neutralizing anti-HB-EGF antibody (Figure 6B) and completely inhibited in the presence of the EGF receptor tyrosine kinase inhibitor, tyrphostin AG1478, but remained unaffected by the PDGF receptor tyrosin kinase inhibitor, tyrphostin AG1296 (Figure 6A). The effect was significantly decreased by the MEK inhibitor PD98059 but was not affected by the p38 MAPK blocker, SB203580 (Figure 6A). The blocker of the PI3K, LY294002, significantly suppressed the effect of HB-EGF on VEGF release. The results suggest that VEGF secretion evoked by HB-EGF is mediated by activation of the EGF receptor tyrosine kinase and by stimulation of the ERK-1, ERK-2, and PI3K signaling pathways. The results regarding the PI3K signaling pathway were supported by western blot experiments detecting the activation of the Akt protein kinase, an enzyme downstream of PI3K (Figure 7). HB-EGF increased the phosphorylation level of Akt in a dose dependent manner. A stimulating effect was obtained at about 1 ng HB-EGF/ml (Figure 7A).

Discussion

PVR is considered to be an uncontrolled wound healing process of the retina. Given the strong proliferative, migratory, and secretory activity of the cells involved, retinal gene expression should be significantly modified during PVR development. To study global differences in retinal gene expression patterns between postmortem controls and patients with PVR, we used oligonucleotide microarray expression profiling. Due to the rarity of appropriate surgical material, retinal samples of not more than two PVR patients were available for this study. However, given the uniform direction of the gene expression changes, and the similar magnitudes of these differences observed in the two retinal samples, it appears to be reasonable to assume that the observed changes are representative for retinas of PVR patients. As controls, we used postmortem retinas of human subjects with no reported history of eye disease. Although we cannot rule out the possibility that gene expression patterns and mRNA integrity may be altered during the postmortem interval, (1) the relatively low number of genes that differed in their expression between control and PVR retinas and (2) the selectivity of the observed expression differences both argue against any overall degeneration or alteration of the postmortem retinas used.

The analysis of retinal gene expression revealed that, among the 12,600 genes and expressed sequence tags investigated, 101 identified genes showed more than a three-fold change of expression in retinectomy material from PVR patients compared to postmortem controls. Upregulated genes were classified into at least five groups, including genes that support cell proliferation, cell signaling, cell motility, and extracellular matrix remodeling. A significant fraction of these genes are associated with inflammatory and immune responses, and with alterations in the actin cytoskeleton. The decreased expression levels of several genes, such as rhodopsin and neurofilaments, may reflect neuronal degeneration in PVR retinas. Potential new targets for an adjunctive therapy of PVR as revealed by the present study, may include inhibitors of RhoA, which controls cell motility and tissue contraction, and PTOV1, which is involved in proliferation signaling, and of IGFBPs.

Altered amounts of growth factors and cytokines within the subretinal space and in the vitreous are involved in the pathological processes that underlie the development of PVR. These factors are implicated in the stimulation of scattering, proliferation, and migration of retinal cells and in the secretion of extracellular matrix components. In this study we found that the genes for HGF and for HB-EGF were significantly upregulated in their expression in PVR retinas. Since HB-EGF has not yet been described to be involved in proliferative retinal disorders, we focussed our attention on this factor. HB-EGF is a potent fibroblast and epithelial cell mitogen and motogen, involved in wound healing processes of the skin [51] and the corneal epithelium [23]. As proliferative vitreoretinal diseases are thought to be uncontrolled, over-stimulated wound healing processes of the retina, we investigated the possible involvement of HB-EGF in PVR. In various types of cells and tissues, both the membrane bound cell surface precursor and the soluble form of HB-EGF are involved in various cellular processes such as tumor formation, cell migration, extracellular matrix formation, cell adherence, and survival [16]. The shedding of HB-EGF from the membrane anchored precursor protein is mediated by matrix metalloproteinases, and TIMPs prevent this process [52]. In the brain, HB-EGF contributes to proliferation and differentiation processes during ontogenetic development [18,22], and is involved in tissue responses to cytotoxic and ischemic injuries [19,20]. HB-EGF stimulates the proliferation and migration of brain astrocytes [21,22], and mediates the expression of VEGF by heart endothelial cells and, therefore, cardiac angiogenesis [53]. The present results of (1) an increased expression of the mRNA for HB-EGF in PVR retinas (Figure 1B), (2) the expression of this factor in fibroproliferative tissues of human retinas (Figure 2), and (3) the proliferation and migration stimulating effect of HB-EGF on cultured human Müller cells (Figure 3 and Figure 5) suggested, for the first time, that HB-EGF may represent one of the growth factors stimulating retinal cell responses in proliferative retinopathies.

Together with RPE and inflammatory cells, retinal glial cells represent a major cellular constituent of fibroproliferative membranes [4,5]. The expression of HB-EGF immunoreactivity by glial cells in these membranes, and the expression of HB-EGF mRNA by cells of the human Müller cell line, suggests that these cells may constitute one of the sources of this factor. Conspicuously, not all of the glial cells in the membranes expressed HB-EGF immunoreactivity (Figure 2A), and GFAP expressing fibrillary structures showed only partial costaining with HB-EGF immunoreactivity (Figure 2B). Whether the variable expression of HB-EGF immunoreactivity reflects different proliferative states of individual glial cells remains a subject for further investigation. It is likely that not only glial cells but also other types of retinal cells may express HB-EGF, and that a part of the HB-EGF immunoreactivity in the membranes represents deposits of HB-EGF protein released into the extracellular matrix.

ERK-1 and ERK-2 are the major MAPKs involved in cell proliferation. However, depending on the cell system and the stimulus used, other types of MAPKs may also contribute to the mediation of proliferative signals. We found that in human Müller cells the p38 MAPK is apparently not involved in the mediation of the HB-EGF evoked proliferation (Figure 3) in accordance with recently described observations in primary cultures of guinea pig Müller cells [15]. Likewise, the involvement of the PI3K in the proliferative signaling of human Müller cells (Figure 3) is similar to observations in guinea pig cells [15]. In contrast, the proliferation of MIO-M1 cells evoked by bFGF has been recently described to be dependent on activation of both ERK-1, ERK-2, and p38 MAPK [54].

Together with bFGF, VEGF is a major angiogenic factor in the retina [55], and Müller cells have been shown to release VEGF in response to hypoxic stimuli [48-50]. Here we show that HB-EGF stimulated the release of VEGF by Müller cells, via activation of ERK-1, ERK-2, and PI3K, whereas activation of p38 MAPK was not involved in the HB-EGF effect on VEGF secretion. The enhancement of VEGF secretion may be a selective effect of HB-EGF since it has been shown that HB-EGF does not stimulate the secretion of another growth factor (HGF) from MIO-M1 cells [12]. The involvement of MAPKs in the regulation of VEGF secretion is dependent on the types of cells and growth factors investigated. In MIO-M1 cells, VEGF secretion evoked by bFGF or by HGF is dependent on activation of the PI3K, but independent on activation of ERK-1 and ERK-2 [12,54]. In RPE cells, the VEGF secretion evoked by TGF-β is mediated by activation of ERK-1, ERK-2, and p38 MAPK [56]. We suggest that HB-EGF is one of the factors released into the retina during proliferative vitreoretinal disorders and which evoke the secretion of VEGF by glial cells.

In summary, the present results indicate that the neural retina of PVR patients undergoes a significant alteration in gene expression, with elevated mRNA levels of genes involved in cell proliferation, cell signaling, cell motility, and extracellular matrix remodeling. HB-EGF is a likely candidate factor promoting the uncontrolled proliferation and migration of retinal cells characteristic in PVR. Since the shedding of HB-EGF can be induced by activation of various types of receptor tyrosine kinases and G protein coupled receptors [14], HB-EGF may play a central role in mediating a variety of pathological events such as glial cell proliferation and migration, neovascularization, and leakage of ocular blood vessels.

Acknowledgements

References

1. Limb GA, Little BC, Meager A, Ogilvie JA, Wolstencroft RA, Franks WA, Chignell AH, Dumonde DC. Cytokines in proliferative vitreoretinopathy. Eye 1991; 5:686-93.

2. Wiedemann P. Growth factors in retinal diseases: proliferative vitreoretinopathy, proliferative diabetic retinopathy, and retinal degeneration. Surv Ophthalmol 1992; 36:373-84.

3. Charteris DG. Proliferative vitreoretinopathy: pathobiology, surgical management, and adjunctive treatment. Br J Ophthalmol 1995; 79:953-60.

4. Laqua H, Machemer R. Glial cell proliferation in retinal detachment (massive periretinal proliferation). Am J Ophthalmol 1975; 80:602-18.

5. Van Horn DL, Aaberg TM, Machemer R, Fenzl R. Glial cell proliferation in human retinal detachment with massive periretinal proliferation. Am J Ophthalmol 1977; 84:383-93.

6. Fisher SK, Erickson PA, Lewis GP, Anderson DH. Intraretinal proliferation induced by retinal detachment. Invest Ophthalmol Vis Sci 1991; 32:1739-48.

7. Geller SF, Lewis GP, Fisher SK. FGFR1, signaling, and AP-1 expression after retinal detachment: reactive Muller and RPE cells. Invest Ophthalmol Vis Sci 2001; 42:1363-9.

8. Bringmann A, Pannicke T, Moll V, Milenkovic I, Faude F, Enzmann V, Wolf S, Reichenbach A. Upregulation of P2X(7) receptor currents in Muller glial cells during proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 2001; 42:860-7.

9. Francke M, Weick M, Pannicke T, Uckermann O, Grosche J, Goczalik I, Milenkovic I, Uhlmann S, Faude F, Wiedemann P, Reichenbach A, Bringmann A. Upregulation of extracellular ATP-induced Muller cell responses in a dispase model of proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 2002; 43:870-81.

10. He PM, He S, Garner JA, Ryan SJ, Hinton DR. Retinal pigment epithelial cells secrete and respond to hepatocyte growth factor. Biochem Biophys Res Commun 1998; 249:253-7.

11. Lashkari K, Rahimi N, Kazlauskas A. Hepatocyte growth factor receptor in human RPE cells: implications in proliferative vitreoretinopathy. Invest Ophthalmol Vis Sci 1999; 40:149-56.

12. Hollborn M, Krausse C, Iandiev I, Yafai Y, Tenckhoff S, Bigl M, Schnurrbusch UE, Limb GA, Reichenbach A, Kohen L, Wolf S, Wiedemann P, Bringmann A. Glial cell expression of hepatocyte growth factor in vitreoretinal proliferative disease. Lab Invest 2004; 84:963-72.

13. Daub H, Weiss FU, Wallasch C, Ullrich A. Role of transactivation of the EGF receptor in signalling by G-protein-coupled receptors. Nature 1996; 379:557-60.

14. Prenzel N, Zwick E, Daub H, Leserer M, Abraham R, Wallasch C, Ullrich A. EGF receptor transactivation by G-protein-coupled receptors requires metalloproteinase cleavage of proHB-EGF. Nature 1999; 402:884-8.

15. Milenkovic I, Weick M, Wiedemann P, Reichenbach A, Bringmann A. P2Y receptor-mediated stimulation of Muller glial cell DNA synthesis: dependence on EGF and PDGF receptor transactivation. Invest Ophthalmol Vis Sci 2003; 44:1211-20.

16. Raab G, Klagsbrun M. Heparin-binding EGF-like growth factor. Biochim Biophys Acta 1997; 1333:F179-99.

17. Mishima K, Higashiyama S, Nagashima Y, Miyagi Y, Tamura A, Kawahara N, Taniguchi N, Asai A, Kuchino Y, Kirino T. Regional distribution of heparin-binding epidermal growth factor-like growth factor mRNA and protein in adult rat forebrain. Neurosci Lett 1996; 213:153-6.

18. Nakagawa T, Sasahara M, Hayase Y, Haneda M, Yasuda H, Kikkawa R, Higashiyama S, Hazama F. Neuronal and glial expression of heparin-binding EGF-like growth factor in central nervous system of prenatal and early-postnatal rat. Brain Res Dev Brain Res 1998; 108:263-72.

19. Tanaka N, Sasahara M, Ohno M, Higashiyama S, Hayase Y, Shimada M. Heparin-binding epidermal growth factor-like growth factor mRNA expression in neonatal rat brain with hypoxic/ischemic injury. Brain Res 1999; 827:130-8.

20. Opanashuk LA, Mark RJ, Porter J, Damm D, Mattson MP, Seroogy KB. Heparin-binding epidermal growth factor-like growth factor in hippocampus: modulation of expression by seizures and anti-excitotoxic action. J Neurosci 1999; 19:133-46.

21. Faber-Elman A, Solomon A, Abraham JA, Marikovsky M, Schwartz M. Involvement of wound-associated factors in rat brain astrocyte migratory response to axonal injury: in vitro simulation. J Clin Invest 1996; 97:162-71.

22. Kornblum HI, Zurcher SD, Werb Z, Derynck R, Seroogy KB. Multiple trophic actions of heparin-binding epidermal growth factor (HB-EGF) in the central nervous system. Eur J Neurosci 1999; 11:3236-46.

23. Xu KP, Ding Y, Ling J, Dong Z, Yu FS. Wound-induced HB-EGF ectodomain shedding and EGFR activation in corneal epithelial cells. Invest Ophthalmol Vis Sci 2004; 45:813-20.

24. Limb GA, Salt TE, Munro PM, Moss SE, Khaw PT. In vitro characterization of a spontaneously immortalized human Muller cell line (MIO-M1). Invest Ophthalmol Vis Sci 2002; 43:864-9.

25. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29:e45.

26. Yoshimura N, Kikuchi T, Kuroiwa S, Gaun S. Differential temporal and spatial expression of immediate early genes in retinal neurons after ischemia-reperfusion injury. Invest Ophthalmol Vis Sci 2003; 44:2211-20.

27. Kuroiwa S, Katai N, Shibuki H, Kurokawa T, Umihira J, Nikaido T, Kametani K, Yoshimura N. Expression of cell cycle-related genes in dying cells in retinal ischemic injury. Invest Ophthalmol Vis Sci 1998; 39:610-7.

28. Hall A. Rho GTPases and the actin cytoskeleton. Science 1998; 279:509-14.

29. Grisanti S, Heimann K, Wiedemann P. Origin of fibronectin in epiretinal membranes of proliferative vitreoretinopathy and proliferative diabetic retinopathy. Br J Ophthalmol 1993; 77:238-42.

30. Smith-Thomas L, Richardson P, Parsons MA, Rennie IG, Benson M, MacNeil S. Additive effects of extra cellular matrix proteins and platelet derived mitogens on human retinal pigment epithelial cell proliferation and contraction. Curr Eye Res 1996; 15:739-48.

31. Huh GS, Boulanger LM, Du H, Riquelme PA, Brotz TM, Shatz CJ. Functional requirement for class I MHC in CNS development and plasticity. Science 2000; 290:2155-9.

32. Corriveau RA, Huh GS, Shatz CJ. Regulation of class I MHC gene expression in the developing and mature CNS by neural activity. Neuron 1998; 21:505-20.

33. Amaratunga A, Abraham CR, Edwards RB, Sandell JH, Schreiber BM, Fine RE. Apolipoprotein E is synthesized in the retina by Muller glial cells, secreted into the vitreous, and rapidly transported into the optic nerve by retinal ganglion cells. J Biol Chem 1996; 271:5628-32.

34. Schneeberger SA, Iwahashi CK, Hjelmeland LM, Davis PA, Morse LS. Apolipoprotein E in the subretinal fluid of rhegmatogenous and exudative retinal detachments. Retina 1997; 17:38-43.

35. Benedit P, Paciucci R, Thomson TM, Valeri M, Nadal M, Caceres C, de Torres I, Estivill X, Lozano JJ, Morote J, Reventos J. PTOV1, a novel protein overexpressed in prostate cancer containing a new class of protein homology blocks. Oncogene 2001; 20:1455-64.

36. Santamaria A, Fernandez PL, Farre X, Benedit P, Reventos J, Morote J, Paciucci R, Thomson TM. PTOV-1, a novel protein overexpressed in prostate cancer, shuttles between the cytoplasm and the nucleus and promotes entry into the S phase of the cell division cycle. Am J Pathol 2003; 162:897-905.

37. Hoeflich A, Reisinger R, Lahm H, Kiess W, Blum WF, Kolb HJ, Weber MM, Wolf E. Insulin-like growth factor-binding protein 2 in tumorigenesis: protector or promoter? Cancer Res 2001; 61:8601-10.

38. Sueoka N, Lee HY, Wiehle S, Cristiano RJ, Fang B, Ji L, Roth JA, Hong WK, Cohen P, Kurie JM. Insulin-like growth factor binding protein-6 activates programmed cell death in non-small cell lung cancer cells. Oncogene 2000; 19:4432-6.

39. Schmid C, Keller C, Gosteli-Peter M, Zapf J. Mitogenic and antiapoptotic effects of insulin-like growth factor binding protein-6 in the human osteoblastic osteosarcoma cell line Saos-2/B-10. Biochem Biophys Res Commun 1999; 263:786-9.

40. Mathura JR Jr, Jafari N, Chang JT, Hackett SF, Wahlin KJ, Della NG, Okamoto N, Zack DJ, Campochiaro PA. Bone morphogenetic proteins-2 and -4: negative growth regulators in adult retinal pigmented epithelium. Invest Ophthalmol Vis Sci 2000; 41:592-600.

41. Trousse F, Esteve P, Bovolenta P. Bmp4 mediates apoptotic cell death in the developing chick eye. J Neurosci 2001; 21:1292-301.

42. Liu J, Wilson S, Reh T. BMP receptor 1b is required for axon guidance and cell survival in the developing retina. Dev Biol 2003; 256:34-48.

43. Sakaguchi H, Miyagi M, Darrow RM, Crabb JS, Hollyfield JG, Organisciak DT, Crabb JW. Intense light exposure changes the crystallin content in retina. Exp Eye Res 2003; 76:131-3. Erratum in: Exp Eye Res 2003; 77:121-2.

44. Vickers JC, Schumer RA, Podos SM, Wang RF, Riederer BM, Morrison JH. Differential vulnerability of neurochemically identified subpopulations of retinal neurons in a monkey model of glaucoma. Brain Res 1995; 680:23-35.

45. Levitzki A, Gazit A. Tyrosine kinase inhibition: an approach to drug development. Science 1995; 267:1782-8.

46. Kovalenko M, Gazit A, Bohmer A, Rorsman C, Ronnstrand L, Heldin CH, Waltenberger J, Bohmer FD, Levitzki A. Selective platelet-derived growth factor receptor kinase blockers reverse sis-transformation. Cancer Res 1994; 54:6106-14.

47. Pournaras CJ. Retinal oxygen distribution. Its role in the physiopathology of vasoproliferative microangiopathies. Retina 1995; 15:332-47.

48. Hata Y, Nakagawa K, Ishibashi T, Inomata H, Ueno H, Sueishi K. Hypoxia-induced expression of vascular endothelial growth factor by retinal glial cells promotes in vitro angiogenesis. Virchows Arch 1995; 426:479-86.

49. Behzadian MA, Wang XL, Shabrawey M, Caldwell RB. Effects of hypoxia on glial cell expression of angiogenesis-regulating factors VEGF and TGF-beta. Glia 1998; 24:216-25.

50. Eichler W, Kuhrt H, Hoffmann S, Wiedemann P, Reichenbach A. VEGF release by retinal glia depends on both oxygen and glucose supply. Neuroreport 2000; 11:3533-7.

51. Marikovsky M, Breuing K, Liu PY, Eriksson E, Higashiyama S, Farber P, Abraham J, Klagsbrun M. Appearance of heparin-binding EGF-like growth factor in wound fluid as a response to injury. Proc Natl Acad Sci U S A 1993; 90:3889-93.

52. Lanzrein M, Garred O, Olsnes S, Sandvig K. Diphtheria toxin endocytosis and membrane translocation are dependent on the intact membrane-anchored receptor (HB-EGF precursor): studies on the cell-associated receptor cleaved by a metalloprotease in phorbol-ester-treated cells. Biochem J 1995; 310:285-9.

53. Abramovitch R, Neeman M, Reich R, Stein I, Keshet E, Abraham J, Solomon A, Marikovsky M. Intercellular communication between vascular smooth muscle and endothelial cells mediated by heparin-binding epidermal growth factor-like growth factor and vascular endothelial growth factor. FEBS Lett 1998; 425:441-7.

54. Hollborn M, Jahn K, Limb GA, Kohen L, Wiedemann P, Bringmann A. Characterization of the basic fibroblast growth factor-evoked proliferation of the human Muller cell line, MIO-M1. Graefes Arch Clin Exp Ophthalmol 2004; 242:414-22.

55. Aiello LP, Pierce EA, Foley ED, Takagi H, Chen H, Riddle L, Ferrara N, King GL, Smith LE. Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A 1995; 92:10457-61.

56. Nagineni CN, Samuel W, Nagineni S, Pardhasaradhi K, Wiggert B, Detrick B, Hooks JJ. Transforming growth factor-beta induces expression of vascular endothelial growth factor in human retinal pigment epithelial cells: involvement of mitogen-activated protein kinases. J Cell Physiol 2003; 197:453-62.