Molecular Vision 2002; 8:483-493 <>
Received 7 November 2001 | Accepted 16 September 2002 | Published 20 December 2002

MAP kinase and β-catenin signaling in HGF induced RPE migration

Gregory I. Liou,1 Suraporn Matragoon,1 Sara Samuel,1 M. Ali Behzadian,2 Nai-Tse Tsai,2 Xiaolin Gu,2 Penny Roon,2 D. Margaret Hunt,3 Richard C. Hunt,3 Ruth B. Caldwell,2 Dennis M. Marcus1

Departments of 1Ophthalmology, and 2Cellular Biology & Anatomy, Medical College of Georgia, Augusta, GA, USA; 3Department of Microbiology, University of South Carolina Medical School, Columbia, SC, USA

Correspondence to: Gregory I. Liou, Ph.D., Department of Ophthalmology, Medical College of Georgia, Augusta, GA, 30912; Phone: (706) 721-4599; FAX: (706) 721-7913; email:


Purpose: Hepatocyte growth factor (HGF) has been implicated in retinal pigment epithelial (RPE) cell proliferation and migration that occurs in proliferative retinal diseases such as proliferative vitreoretinopathy (PVR). The aim of this study is to investigate HGF induced signaling pathways that lead to RPE cell migration.

Methods: Localization of β-catenin was determined by immunofluorescence. HGF induced migration of ARPE-19 cells was studied using a quantitative migration assay after wounding in the presence of a DNA polymerase inhibitor, and in the presence or absence of a mitogen activated protein kinase (MAP kinase) kinase inhibitor. C-jun expression was determined by semi-quantitative RT-PCR and by Northern blot analysis. P42/p44 MAP kinase activity was determined by western blot and by an immunoprecipitation kinase assay. Tyrosine phosphorylation of the HGF receptor (HGFR or c-met) and β-catenin was determined by immunoprecipitation and western blot analysis. Transactivation activity of β-catenin was determined by luciferase reporter gene analysis.

Results: β-catenin and E-cadherin were co-localized on the basal surface of the RPE in vivo. Diffusion of the cell surface-localized β-catenin occurs in migratory cells in vitro in the presence of HGF. HGF induced a MAP kinase dependent ARPE-19 cell migration, which is accompanied with a transient increase of c-jun expression and concomitant increases of MAP kinase activity, tyrosine phosphorylation of HGFR and β-catenin, increased cytosolic levels of β-catenin, and transactivation activity of β-catenin. Tyrosine phosphorylation of HGFR and β-catenin occurs in the primary or passaged RPE cultures or proliferative ARPE-19 cells, but not freshly isolated RPE or differentiated ARPE-19 cells.

Conclusions: This study defines the signal transduction pathways activated by HGF in RPE cells, leading to an increase in the MAP kinase activity and free pool of β-catenin, and changes in gene expression. These findings are consistent with the hypothesis that both β-catenin and MAP kinases are components of the HGF induced RPE migration that occurs in proliferative retinal diseases.


The morphologically and functionally polarized retinal pigment epithelium (RPE) has numerous functions including maintaining adjacent retinal photoreceptor cells and forming the outer blood-retinal barrier [1]. Differentiated RPE cells in the normal adult eye, however, retain the ability to migrate and proliferate in response to a wide range of pathological conditions [2], such as proliferative vitreoretinopathy (PVR; reviewed in [3,4]) and exudative age related macular degeneration [5]. Understanding the molecular mechanisms of RPE migration and proliferation is critical for development of therapeutic approaches for retinal diseases associated with proliferative RPE.

Recent studies have suggested that several autocrine or paracrine loops involving cytokines, growth factors, and their receptors may be critical as pathological insults to RPE cells. These include platelet derived growth factor (PDGF), vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), tumor necrosis factor (TNF)-α, interleukin (IL)-1 and 6, interferon-γ, transforming growth factor (TGF)-β, insulin-like growth factor (IGF), and hepatocyte growth factor (HGF) [6-9]. HGF, also known as scatter factor, may be particularly important due to its potent activity as a mitogen, motogen, and a pro-angiogenic factor that is involved in wound healing in many cell types (reviewed in [10]). Previous studies have suggested the role of HGF and its receptor, HGFR or c-met, in the migration and proliferation of cultured RPE cells or during epithelial-mesenchymal transition in PVR [8,11,12]. The HGF induced signaling pathways that lead to the epithelial-mesenchymal transition of RPE cells, however, have not been identified.

Integrity of the epithelium is largely regulated by the tumor suppressor protein, adenomatous polyposis coli (APC) and associated proteins (reviewed in [13]). APC plays an important role in epithelial cell proliferation and migration [14]. APC functions by interacting with and down regulating β-catenin, a bi-functional molecule that either mediates cell adhesion under normal conditions or promotes cell proliferation and migration under pathological conditions. In the absence of pathological insults, cell adhesion is maintained by the cytoskeleton/cadherin/catenin complex. Any uncomplexed, cytosolic β-catenin is removed via the glycogen synthase kinase 3b (GSK3b)/APC/axin complex mediated serine phosphorylation and ubiquitin dependent protein degradation (reviewed in [15]). The activity of the GSK3b/APC/axin complex is negatively regulated by the signaling pathways of many growth factors including wingless/int (Wnt), epidermal growth factor (EGF) and HGF (reviewed in [15,16]). In case of abnormal activities of these growth factors or mutations in the APC gene, β-catenin is tyrosine phosphorylated [17,18], dissociated from the cytoskeleton/cadherin/catenin complex, and accumulated in the cytosol. The cytosolic β-catenin then associates with members of the high mobility group [19] family of transcription factors, namely lymphoid enhancer factor-1 (LEF-1), T-cell factor 3 (Tcf-3), and Tcf-4 [20,21]. The β-catenin-Lef/Tcf complex then migrates to the nucleus where it activates the transcription of growth and migration promoting target genes [15,16]. To determine the role of APC in proliferative RPE cells, we have recently demonstrated RPE proliferation in a mouse model with a mutated APC gene [22,23]. This model recapitulates conditions seen in humans with APC mutation, familial polyposis and RPE hamartomatous fundus lesions [24]. We have also demonstrated that distinctive alternative APC transcripts exist in the neural retina and RPE cells and that the down regulation of a neural specific APC isoform in RPE cells may account for the ability of these cells to proliferate and migrate in pathologic conditions (unpublished data). Growth factor induced migration of RPE cells also involves the p42/p44 MAP kinase cascade [25]. P42/p44 MAP kinases are serine/threonine protein kinases which are rapidly activated as a downstream event in response to a number of external stimuli [26-28].

In the present study, we investigated the HGF induced signaling in migratory RPE cells. We demonstrated that both the p42/p44 MAP kinases and β-catenin signaling pathways are induced by HGF and are involved in activation of a migration promoting gene in RPE cells.


Reagents and antibodies

Mature, double chain recombinant human HGF was obtained from R & D System (Minneapolis, MN). The rabbit polyclonal antibodies for rat extracellular signal regulated kinase 1 (ERK 1; p44, also reactive with ERK 2 (p42)), human c-met receptor (HGFR), and mouse monoclonal antibodies for human p-ERK, pan-cytokeratin, Xenopus GSK3b, phosphotyrosine, as well as Protein A/G Plus Agarose used in immunocytochemistry, immunoprecipitation or western blot analysis were all purchased from Santa Cruz Biotechnologies (Santa Cruz, CA). The mouse monoclonal antibodies for human E-cadherin and mouse β-catenin were from Becton-Dickenson Transduction Laboratories (Los Angeles, CA), and for rabbit kidney Na+/K+ ATPase was from Upstate Biotechnology (Lake Placid, NY). The normal mouse IgG was from Calbiochem (San Diego, CA). The Oregon Green labeled goat anti-rabbit and goat anti-mouse antibodies were purchased from Molecular Probes (Eugene, Or). p42/p44 MAP Kinase Assay Kit was purchased from Cell Signaling/New England Biolabs (Beverly, MA). Peroxidase-labeled anti-rat or anti-mouse IgG and Enhanced Chemiluminascence (ECL) reagent was obtained from Amersham (Piscataway, NJ). Dulbecco's phosphate buffered saline (PBS), MAP kinase kinase (MEK) inhibitor PD98059, and protease inhibitor cocktail were obtained from Sigma (St. Louis, MO). DNase, oligo-dT, RNase inhibitor, MMLV reverse transcriptase, D-MEM/F-12 with L-glutamine, Ham's F10 and other cell culture reagents including ITS were obtained from Life Technologies (Grand Island, NY). Nitrocellulose membranes were obtained from Schleicher & Schuell (Keene, NH). Cell transfection reagent, Cytofectamine, standard protein assay or detergent compatible protein assay kits, and nonfat dry milk were obtained from Bio-Rad (Hercules, CA). RNA isolation reagent kit, RNeasy, and HotStart Taq DNA polymerase were obtained from Qiagen (Santa Clarita, CA). RNA isolation reagent kit, RNAwiz, and formaldehyde based Northern system were obtained from Ambion (Austin, TX).

Human Ocular tissues

Human ocular tissues were obtained from deceased donors and were distributed by the Cooperative Human Tissue Network (Philadelphia, PA). Donor ages ranged between 58 and 92 years. Donor eyes were kept at 4 °C and the average interval between death and retina dissection was 26 h. The research followed the tenets of the Declaration of Helsinki and was reviewed and approved by Human Assurance Committee of the Medical College of Georgia for using human post-mortem material.

Cell culture

ARPE-19 (ATCC, passage 22; [29]) cell line was maintained in D-MEM/F-12 with L-glutamine at 37 °C under 10% CO2 using a previously described method [29]. Primary cultured (non-passaged) and early passaged human RPE cells (passage 3 and 4; [30]) from human donors eyes obtained from the South Carolina Eye Bank, were cultured using the same method [29]. The purity of these RPE cells was evaluated by immunocytochemical staining of mouse monoclonal antibody for pan-cytokeratin. In this analysis, cells were found to be positive for cytokeratin. All media were supplemented with 10% fetal calf serum (unless otherwise mentioned), 0.1 mg/ml streptomycin, and 100 U/ml penicillin.

Migration assays

For in vitro wound assays, ARPE-19 cells were plated at a density of 7x104 cells/cm2 in 1% fetal calf serum containing medium on laminin coated 60 mm plates for 4 weeks. The cells were incubated in serum free medium for 16 h and were then incubated for 2 h with the specific inhibitor of DNA polymerase a aphidicolin (5 mM), as indicated. The confluent monolayer was scratched with a pipette tip to create a narrow, cell free area, HGF (50 ng/ml) was added, and wound closure was documented by photography. For quantitative cell migration analysis, the confluent monolayer on 60 mm plates with grid was serum starved for 16 h, and incubated with a aphidicolin (5 mM) for 2 h. Some cells were also incubated with PD98059 (30 mM in 0.1% DMSO). Cells were then scratched with a cell scraper to create a cell free area. HGF (50 ng/ml) were added and wound closure was documented by photography and analyzed using Metamorph Imaging System with Image Processing and Analysis software (Universal Imaging Corp., West Chester, PA).

Animals and ocular tissue processing for immunofluorescence

New Zealand White rabbits weighing between 4 and 5 pounds were obtained from Myrtle's Rabbit Farm (Thompson Station, TN), maintained on a 12 h light: 12 h dark lighting cycle and were fed the rabbit chow diet. Care and use of the animals adhered to the principles set forth in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

After overdose with sodium pentobarbital, whole rabbit eyes were enucleated and anterior chambers including the lenses were removed. The remaining eyecups were used for immunohistochemistry. The whole eyecups with the vitreous were frozen in OCT for at least 24 h and sectioned at 10 μm.


Immunofluorescence analyses were used to localize β-catenin in intact eyes from rabbits and in wounded and HGF (50 ng/ml) treated ARPE-19 cells on chamber slides. The 10 μm thick cryosections of rabbit eyes and cultured cells were fixed with ice cold acetone and methanol, respectively, and blocked with 4% normal goat serum. Samples were incubated with one of the monoclonal antibodies against β-catenin, E-cadherin and Na+/K+ ATPase. Tissue sections were incubated for 3 h at room temperature at a dilution of 1:100 followed by an overnight incubation at 4 °C; cells were incubated with the primary antibody for 3 h at room temperature at a dilution of 1:2000. Incubation with normal mouse IgG served as negative control. After rinsing, samples were incubated overnight at 4 °C with an Oregon Green conjugated goat anti-mouse IgG at a dilution of 1:500. Cryosections and cells were optically sectioned (z series) using a Bio-Rad MRC-600 Laser Scanning Confocal Imaging System.

MAP kinase activity assay

MAP kinase activity was measured using the p42/p44 MAP Kinase Assay Kit as described by the manufacturer. ARPE-19 cells seeded at a density of 2x104 cells/cm2 were cultured 2 days until 80-90% confluent. Cells were then starved in the serum free media for 24 h. Cells were then treated with 50 ng/ml HGF. Some cells were treated with MAPK kinase inhibitor PD98059 (30 mM in 0.1% DMSO) for 1 h prior to HGF treatment. Cells were lysed and a rabbit polyclonal antibody specific to phosphorylated MAP (Tyr204) was used to immunoprecipitate active MAPK. The immunoprecipitate was incubated with 1 μg of ELK 1 fusion protein for 5 min at 37 °C in the presence of 100 μM ATP and the provided kinase buffer. Phosphorylation at Ser383, a major MAPK phosphorylation site on ELK 1, was measured by western blot analysis using a phospho-specific ELK 1 (Ser383) antibody followed by chemiluminescent detection using Lumi GLO reagent provided in the kit.

Immunoprecipitation and western blot analysis

Cells were plated at a density of 1-2x104 cells/cm2 in 100 mm plates and cultured for 2-3 days until 80-90% confluent. The cells were then incubated in serum free medium for 18 h or longer when mentioned. When indicated, cells were treated with 50 ng/ml HGF, washed twice with ice cold PBS containing 0.5 mM phenylmethylsulfonyl fluoride (PMSF). For immunoprecipitation or western blot analysis, cells were lysed in ice cold lysis buffer containing 10 mM sodium phosphate, pH 7.0, 0.15 M NaCl, 1% NP-40, 1 mM Na3VO4, 50 mM NaF, and 10 μl/ml of protease inhibiter cocktail containing AEBSF, bestatin, aprotinin, E-64, pepstatin-A, and leupeptin. The lysates were centrifuged for 15 min at 14,000 rpm in a microcentrifuge. Protein content of the resultant supernatants was measured by Bio-Rad detergent compatible protein assay. The supernatants were then incubated for 2 h or overnight at 4 °C with 2 μg antibody that recognizes phosphotyrosine, human c-met, GSK3b, or β-catenin. Immune complexes were collected on Protein A/G Plus Agarose and washed 4x with the lysis buffer. Immunoprecipitates or cell lysates were resolved on 9 or 10% sodium dodecyl sulfate-polyacrylamide gel under reducing conditions and electroblotted to nitrocellulose membrane. The membranes were blocked for 1 h at room temperature in 5% nonfat dry milk in 1 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.05% Tween 20 (TTBS), and probed with indicated mouse monoclonal antibodies (1:250 in 0.5% nonfat dry milk in TTBS) for 1-2 h at room temperature. The membranes were then washed 3 times with TTBS, incubated with horseradish peroxidase labeled secondary rat anti-mouse IgG (1:3000) in 0.5% nonfat dry milk in TTBS, and washed 3 times in TTBS. The membranes were then developed with ECL reagent.

For western blot analysis of cytosolic β-catenin, which is according to the method by Giarre et al. [31], cells were washed twice in 10 mM Tris-HCl, pH 7.4, 140 mM NaCl and 2 mM CaCl2. After addition of 1 ml ice cold buffer containing 10 mM Tris-HCl, pH 7.4, 140 mM NaCl, 2 mM dithiothreitol (DTT), 5 mM EDTA and 10 μl/ml of protease inhibitor cocktail, the cells were collectred and homogenized in a Dounce homogenizer. The homogenates were centrifuged for 5 min at 2,000x g and the supernatants then centrifuged at 100,000x g for 60 min at 4 °C. Protein content in the supernatant, or cytosolic, fractions was measured by Bio-Rad protein assay. Cell cytosol resolution on gel and immunoblot were as above described. Films were subjected to a densitometric scanning (E-C Densitometer, St. Petersburg, FL) and analyzed with an HP integrator (Wilmington, DE).

RNA extraction, RT-PCR and northern blot analysis

RNA was extracted from cells using total RNA isolation reagent kit, RNAwiz, which is based on the use of detergent, chaotropic salts and phenolic extraction to denature the RNase and solubilize the total RNA. In some applications, RNA was also extracted by RNeasy, which is based on the selective binding properties of silica gel based membrane with the microspin technique.

RT-PCR was performed as follows: samples of 1 μg total RNA were treated with 1 unit of DNAse for 15 min at room temperature. The RNA samples were reverse transcribed in a total volume of 20 μl, containing 0.5 mM oligo-dT, 0.5 mM dNTPs, 10 mM DTT, 20 units RNase inhibitor, 10 units MMLV reverse transcriptase, 50 mM Tris-HCl, pH 8.4, 75 mM KCl and 3 mM MgCl2. Initially, total RNA was denatured at 70 °C for 10 min and immediately chilled on ice. First strand cDNAs were obtained after 90 min at 37 °C incubation. Following inactivation at 70 °C for 10 min, 1 ml of 2x diluted reverse transcribed mixture was used as a template and PCR was performed in a 50 μl reaction mixture containing 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 1.5-2.5 mM MgCl2, and 0.001% gelatin in the presence of 0.2 mM of each dNTP, 0.5 mM of each sense and anti-sense primers and 1 unit of Hot Start Taq DNA polymerase. Amplification was performed as follows: 95 °C for 15 min, and 27-30 cycles of 94 °C, 1 min, 55 °C, 1 min, and 72 °C, 2 min. The same cDNAs were used as templates for parallel PCR reactions performed in the presence of human c-jun primers 5'-ATGGAGTCCCAGGAGCGGATCAA-3' (position 1999-2021, Genbank Accession J04111) and 5'-GTTTGCAACTGCTGCGTTAG-3' (position 2249-2230), or human glyceraldehyde-3-phosphate dehydrogenase(G3PDH, an internal standard) primers 5'-TGCACCACCAACTGCTT-3' (position 514-530; [26]) and 5'-TACTCCTTGGAGGCCAT-3' (position 1067-1051). The RT-PCR products were analyzed by electrophoresis on 1.6% agarose gels and photographed. Quantitation of ethidium bromide stained RT-PCR products was performed using a densitometer (E-C Apparatus Corp., St. Petersburg, FL) and an integrator (HP3395, Hewlett Parkard, Wilmington, DE). Integrated volumes of each band, summing the intensities, was compared to G3PDH RT-PCR products, which were electrophresed in parallel.

For Northern blot analysis, 8 μg of total RNA was resolved on 1% formaldehyde-agarose gel and transferred to nitrocellulose membranes, baked and hybridized with RNA probes according to a formaldehyde based system from Ambion as described by the manufacturer. The antisense RNA probes for c-jun and G3PDH were transcribed and labeled with 32P-UTP (800 Ci/mm, Amersham). Films were subjected to a densitometric scanning and analysis as described above.

Luciferase reporter assays

ARPE-19 cells (1x106) were plated in a 100 mm plate and incubated overnight in the complete media without antibiotics. The cells were transfected with 5 μg each of TOPFLASH or FOPFLASH reporter plasmids [32], using 20 μl of Cytofectamine or lipofectamine following the manufacturer's protocol. On the following day, the cells were replated with 3 mM EDTA at 1x105 into 6x35 mm plates and incubated for 8 h in the complete media without antibiotics. The cells were washed and starved in serum free medium (or medium containing 0.5% serum for human RPE culture) overnight for 40 h. Cultures were incubated with HGF (37.5 ng/ml) for the indicated times. Reporter assays were performed using the luciferase reporter system (Promega) on a TD 20/20 luminometer (Turner Designs).

Statistical analysis

Calculations were performed with the aid of a computer equipped with appropriate software (Microstat, Advanced Graphics Software Inc., Sunnyvale, CA) designed for filling and processing data as well as for statistical analysis. The results of investigated parameters, which were expressed as mean plus or minus SEM, were compared with the control using student's t test. Differences between treatment time points and control were determined by ANOVA followed by post hoc tests (Scheffe). Significance is defined as p<0.05.


HGF promotes migration of ARPE-19 cells

The stationary RPE cells in the normal adult eye retain the ability to migrate and proliferate under pathological conditions. To understand the mechanism of RPE activation, we used the cultured RPE cell line ARPE-19 [29] as a model for the RPE disease process [33]. This cell line spontaneously arose from RPE cells isolated from a human donor and has a normal karyotype. This cell line retains certain features of RPE, which include the epithelial appearance and pigmentation and the ability to form polarized epithelial monolayers when cultured post-confluently on laminin coated porous filter supports [29] or chamber slides [34]. We first determined the effect of HGF on ARPE-19 cell migration by an in vitro wound assay. Confluent cultures were wounded and treated with HGF (50 ng/ml) for 12 h. Although HGF was a modest mitogen for RPE cells [12], to rule out proliferation as the cause of wound closure during the 12 h period, we used aphidicolin to inhibit DNA synthesis. As shown in Figure 1, at 12 h in the presence of HGF, migrated cells have almost covered the cell free area. In contrast, in control cultures at 12 h, only minimal cell migration was observed.

HGF and cell wounding promote re-distribution of β-catenin in cultured RPE cells

Integrity of the epithelium is largely controlled by β-catenin, which, under non-proliferative conditions, is primarily associated with the cytoskeleton/cadherin/catenin complex. To demonstrate the role of the cadherin/catenin complex in the normal RPE, we first determined the subcellular localization of β-catenin and E-cadherin in the rabbit retina, and then determined the effect of HGF and cell migration on the distribution of β-catenin in RPE cells. Figure 2A shows a hematoxylin and eosin stained, cryosectioned rabbit retina for comparison to non-stained sections subjected to immunohistochemistry. Figure 2B and Figure 2C show cryosections of rabbit retina labeled with antibodies directed against β-catenin and E-cadherin, respectively. To clearly define the RPE and the subcellular localization of these proteins, the nuclei were counter stained red with propidium iodide. Laser scanning confocal microscopy revealed an intense band of green fluorescence on the basal surface of the RPE for both proteins. For comparison, the antibody against the Na+/K+ ATPase (Figure 2D) was used in companion experiments, and the Na+/K+ ATPase was found to be localized to the apical surface as has been well established [35,36]. Sections incubated with normal mouse serum showed no positive labeling (Figure 2E).

To determine the effect of cell migration on the distribution of β-catenin in the RPE cells in culture, localization of β-catenin was first determined in the confluent ARPE-19 monolayer and cultured human RPE cells (passage number 3). The confluent cells were then starved overnight, wounded by scratching with a pipette tip, and localization of β-catenin was then determined in or away from the wound edge in the presence of 50 ng/ml HGF. Figure 2F and Figure 2J show that β-catenin was on the periphery of the confluent ARPE-19 or RPE cells. At 8 h after wounding in the presence of HGF away from the wound Figure 2G and Figure 2K, β-catenin was similarly located on the periphery of cells. However, more cells with diffusely distributed cytoplasmic β-catenin were observed as one approaches the wound edge Figure 2H and Figure 2L. Diffusely distributed β-catenin was observed in or very close to the wound edge Figure 2I and Figure 2M. In wounded cultures without HGF, only minimal cell migration was observed up to 12 h, and no β-catenin re-distribution was observed. This result suggests that the effect of HGF on β-catenin re-distribution is restricted to migratory cells, but not quiescent cells.

HGF induces a transient increase of C-jun expression in ARPE-19 cells

Migration of epithelial cells in response to growth factors is the result of modulation of many gene activities, some of which rely on the activation of AP-1 transcription factor complex [37]. To determine the effect of HGF on the activity of AP-1, we determined the expression of c-jun, which is one of the components of AP-1. Sub-confluent ARPE-19 cells were treated with HGF (50 ng/ml) for indicated time points. Using a semi-quantitative RT-PCR assay, which resulted in measurable signals within the linear range of amplification, we observed a rapid increase of c-jun expression that peaked at one hour after HGF treatment (Figure 3A,B). Similar results were observed in three independent Northern analyses (Figure 3C,D).

HGF induces p42/p44 MAP kinase activity in the migratory ARPE-19 cells

A primary signal transduction pathway activated by the HGFR, leading to motogenic and mitogenic responses, involves the Ras/Raf/MAP kinase cascade [38-40]. The activated MAP kinases phosphorylate transcription factors that induce expression of c-fos, c-jun and other early growth response genes [41-43]. Therefore, we analyzed the effect of HGF on p42/p44 MAP kinase activity in ARPE-19 cells. As shown in Figure 4, serum starved, sub-confluent ARPE-19 cells exhibited high MAP kinase activity, measured by the phosphorylated ELK-1 (Figure 4A) as well as the phosphorylation of MAP kinase (Figure 4B). However, incubation of these cells with HGF caused a time dependent increase in MAP kinase activity that peaked at 30-60 min and then declined. Preincubation of the cells with PD98059 (a p42/p44 MAP kinase kinase inhibitor) at 30 mM resulted in inhibition of MAP kinase activity. PD98059 at this concentration had no apparent toxic effect on ARPE-19 cells as determined by cell morphology and trypan blue exclusion (data not shown).

To determine whether the HGF induced migration of ARPE-19 cells involves the Ras/Raf/MAP kinase cascade, migration of differentiated ARPE-19 cells in the presence or absence of HGF and PD98059 was analyzed by a quantitative migration assay. Again, to rule out proliferation as the cause of wound closure, an aphidicolin was used to inhibit DNA synthesis, which did not interfere with migration. In contrast, inhibition of MAP kinase kinase by PD98059 completely abolished HGF induced cell migration during the first 22 h (Figure 5). HGF induced migration at 48 h was partially inhibited by PD98059, suggesting that the effect of the inhibitor was wearing off.

HGF induces tyrosine phosphorylation of HGFR and β-catenin in ARPE-19 cells

Previous studies have suggested that, in response to HGF or Wnt-1, β-catenin is tyrosine phosphorylated, dissociated from the cytoskeleton/cadherin/catenin complex, and serves to relay a downstream signal to other effectors [44]. In this process, β-catenin migrates to the nucleus where it stimulates Lef/Tcf dependent transcription from reporter genes [44]. It has also been shown that overexpression of the β-catenin gene in cells can activate several target genes of the β-catenin-Lef/Tcf complex [45]. These target genes include two genes that code for the components of the AP-1 transcription complex, c-jun and fra-1, and one gene, urokinase type plasminogen activator receptor, that is regulated by AP-1 [37].

To understand the role of β-catenin in RPE cells under pathological conditions, we analyzed HGF induced β-catenin signaling in ARPE-19 cells. Sub-confluent cells were treated with 50 ng/ml HGF for indicated times. Protein equivalent aliquots of cell extract made from each time point were immunoprecipitated with anti-HGFR or anti-β-catenin antibody and were then subjected to western blot analysis using anti-phosphotyrosine antibody. The result shown in Figure 6, which represents one of three independent experiments, demonstrates that tyrosine phosphorylation of both HGFR and β-catenin occurred within 15-30 min of HGF treatment. These results suggest that HGF not only induces tyrosine phosphorylation of its receptor in RPE cells as previous studies demonstrated [12], but also induces tyrosine phosphorylation of β-catenin in these cells.

HGF induces cytosolic accumulation of β-catenin

In many cell types, small but significant changes in uncomplexed, cytosolic β-catenin levels have been observed when cells were treated with EGF, HGF, or transformed with Wnt-1 [18,31,44]. Consequently, we studied the effect of HGF on the cytosolic pool of β-catenin in ARPE-19 cells. Sub-confluent cells were treated with HGF (50 ng/ml) for indicated time points. Cell lysates were extracted in a detergent free lysis buffer, which solubilizes only the cytosolic, but not the membrane associated, pool of β-catenin [31]. These lysates were subjected to Western blot analysis with anti-β-catenin antibody. GSK3ab proteins, whose levels are not altered by HGF treatment or Wnt-1 transformation [44], were used as the internal standard. The results of three experiments are shown in Figure 7. A small, but significant, β-catenin accumulation in the cytosol peaked at 1 h of HGF treatment.

HGF induces transcription activation of Lef/Tcf reporter constructs in ARPE-19 cells

Other studies have shown that increased cytosolic and nuclear levels of β-catenin can stimulate Lef/Tcf dependent transcription from reporter genes [20,21,32]. Therefore, we analyzed the effect of HGF on transcription from a Lef/Tcf dependent reporter gene construct [32]. ARPE-19 cells were transfected with a vector containing a multimerized Lef/Tcf binding site (TOPFLASH) upstream of a minimal thymidine kinase promoter that can direct expression of a luciferase gene. Cells were re-plated, treated with HGF for indicated times, and analyzed for luciferase activity. The results of 3-4 experiments are shown in Figure 8. The activity of the Lef/Tcf dependent promoter was minimal in untreated cells. Upon HGF treatment, the promoter activity increased rapidly to reach a peak within 60 min after which it declined. Transfection with the mutant Lef/Tcf (FOPFLASH) construct abrogated the HGF dependent increase in promoter activity.

HGFR and β-catenin signaling in ARPE-19 and cultured human RPE cells

As shown previously (Figure 2), HGF and cell wounding induced β-catenin re-localization occurs in both cultured human RPE cells and in ARPE-19 cells. This suggests that β-catenin signaling events are similar in these cells. To confirm this, we carried out the following experiment. It has been reported that the quiescent, differentiated RPE cells secret and respond to HGF when detached from retina and cultured in the serum [11]. To determine whether post-confluent, differentiated ARPE-19 cells behave similarly when induced to fast growing, we determined tyrosine phosphorylation of HGFR in freshly isolated or actively growing cultures of passaged or non-passaged RPE cells, and in post-confluent or actively growing ARPE-19 cells. We also determined whether β-catenin signaling is induced in these cells. Protein equivalent aliquots of cell extracts were immunoprecipitated with anti-HGFR or anti-β-catenin antibody, and immunoblotted with anti-phosphotyrosine antibody. Data of Figure 9 show that tyrosine phosphorylation of HGFR and β-catenin was only detected in the normal medium-cultured, passaged or non-passaged growing RPE or ARPE-19 cells, but not in freshly isolated, differentiated RPE or post-confluent, differentiated ARPE-19 cells.


Data presented here support a previously suggested role for HGF and HGFR in the migratory RPE involved in diseases such as PVR [8,11,12]. Importantly, these data are the first evidence for a β-catenin signaling pathway in multiple components of HGF induced migration in ARPE-19 cells and cultured human RPE cells.

HGF is a pleiotropic factor that functions as a mitogen, motogen, and morphogen for a variety of cells, particularly epithelial cells, bearing HGFR [46-48]. HGF is secreted as an inactive precursor form, and proteolytic activation of the precursor form in the extracellular milieu is a critical limiting step in the HGF induced signaling pathway [49]. A recent study has shown that RPE in epiretinal membranes expresses HGFR and that increased levels of HGF are detected in severe PVR [8]. Moreover, active HGF, measured by its scatter activity, was detected in the early phase of PVR [8] during which RPE transdifferentiation and migration occur [50,51]. These findings are consistent with the hypothesis that HGF may play a role in the RPE-mesenchymal transformation that typifies PVR [33].

As a primary signaling pathway activated by HGFR leading to mitogenic and motogenic responses involves the Ras/Raf/MAP kinase cascade in many cell types [38-40], we have shown HGF induced motogenic activity is associated with p42/p44 MAP kinase activity in ARPE-19 cells. Although high background MAP kinase activity was observed in the serum starved control cells, the effect of HGF on MAP kinase activity in these cells was specific as suggested by an almost complete inhibition by PD98059. MAP kinase pathways are highly conserved intracellular signal transduction pathways that communicate information between the plasma membrane and nucleus [41,42]. Interestingly, autocrine tyrosine phosphorylation of HGFR and increased phosphorylation of p42/p44 MAP kinases have been observed in the retina and RPE in early stages of PVR in a rabbit model [52]. It will be useful to determine whether the phosphorylated p42/p44 MAP kinases change their cellular localization from the cytoplasm to the nucleus in the retina and RPE during the early stages of PVR. It will also be useful to determine whether PD98059 inhibits MAP kinase activity in the PVR model, as it does in a rat model for retinopathy of prematurity (Dr. John S. Penn, personal communication).

Concomitant with p42/p44 MAP kinase phosphorylation and activation, we have shown HGF induced c-jun mRNA expression increases in ARPE-19 cells. This increase may be required for the HGF induced cell migration observed in the current study because members of the Fos and Jun families frequently become activated and up regulated in response to mitogenic and motogenic stimuli and often serve as early responding global regulators of gene expression [43].

Data presented here are also the first evidence for a transient effect of HGF on β-catenin pools and its transcriptional activation in RPE cells. Although these effects can be measured, they are transient and small because cytosolic diffusion of the cell surface localized β-catenin was not observed in the presence of HGF alone unless the differentiated RPE cells were wounded and became migratory. This suggests that the observed β-catenin re-distribution is the combined effect of HGF with other factors. Transcriptional activation of β-catenin may activate c-jun expression because the latter is one of the target genes of the β-catenin-Lef/Tcf complex [45].

Since HGF can stimulate RPE cell migration, these data offer a potential biochemical link, through β-catenin, between the activity of HGFR, APC and the migratory effects. In support of this conclusion, APC has been localized to areas of cytoplasmic processes that extend into actively migrating cell membranes [14]. As mentioned above, β-catenin is a bi-functional molecule that either mediates cell adhesion or promotes cell migration. These opposite functions are modulated by the activity of the GSK3/APC/axin complex. Under normal conditions, cell adhesion is maintained by the cytoskeleton/cadherin/catenin complex. Any uncomplexed β-catenin is removed via the GSK3/APC/axin complex mediated protein degradation. Mutations of the APC gene, or in the presence of abnormal activities of growth factors such as HGF, cell surface β-catenin becomes a transcription factor in the nucleus to promote cell migration (Figure 10). Therefore, APC, and by inference β-catenin, plays a role in modulating cell migration, which may help explain the molecular mechanism for the hamartomatous RPE proliferation found in mice [22,23] and humans [24] with APC mutations. The transient loss of junctional β-catenin observed in RPE cells in the current study as well as in other types of epithelial cells [18,44] may be due to the transient inhibition of the GSK3b activity by HGF treatment. HGF induced transient decrease in GSK3 activity and a parallel increase in the cytosolic pool of β-catenin have been reported in mouse mammary epithelial cells [44]. Further analysis of the GSK3b activity in the HGF treated RPE cells will be required to confirm this regulatory process.

Our in vitro study of the signaling pathways mediating migratory behavior in RPE cells in response to HGF suggests a potential for targeting HGF or HGFR to inhibit RPE migration in vivo. NK4, a specific antagonist to HGF, has been shown to competitively inhibit the specific binding of HGF to and phosphorylating its receptor and to inhibit the mitogenic, motogenic and morphogenic activities of HGF [53]. Application of NK4 in our rabbit model of PVR may inhibit HGF activation thereby reducing RPE migration and associated retinal damage.


This work was supported by the Combined Intramural Grants Program of Medical College of Georgia (GIL), an unrestricted departmental award from Research to Prevent Blindness Inc., New York, NY (GIL, AMB, and DMM), American Health Assistance Foundation, Rockville, MD (GIL), and grants EY04618 and EY11766 (RBC), EY10516 (RCH), EY12711 (DMH) from the National Institutes of Health. The authors thank Dr. Martin P. Playford for the TOPFLASH and FOPFLASH reporter constructs and Ms. Brenda Sheppard for her assistance in preparing the manuscript.


1. Zhao S, Rizzolo LJ, Barnstable CJ. Differentiation and transdifferentiation of the retinal pigment epithelium. Int Rev Cytol 1997; 171:225-66.

2. Grierson I, Hiscott P, Hogg P, Robey H, Mazure A, Larkin G. Development, repair and regeneration of the retinal pigment epithelium. Eye 1994; 8:255-62.

3. Ryan SJ. Traction retinal detachment. XLIX Edward Jackson Memorial Lecture. Am J Ophthalmol 1993; 115:1-20.

4. Hiscott P, Sheridan C, Magee RM, Grierson I. Matrix and the retinal pigment epithelium in proliferative retinal disease. Prog Retin Eye Res 1999; 18:167-90.

5. Grossniklaus HE, Green WR. Histopathologic and ultrastructural findings of surgically excised choroidal neovascularization. Submacular Surgery Trials Research Group. Arch Ophthal 1998; 116:745-9.

6. 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.

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

8. Briggs MC, Grierson I, Hiscott P, Hunt JA. Active scatter factor (HGF/SF) in proliferative vitreoretinal disease. Invest Ophthalmol Vis Sci 2000; 41:3085-94.

9. Mitamura Y, Takeuchi S, Matsuda A, Tagawa Y, Mizue Y, Nishihira J. Hepatocyte growth factor levels in the vitreous of patients with proliferative vitreoretinopathy. Am J Ophthalmol 2000; 129:678-80.

10. Steenfos HH. Growth factors and wound healing. Scand J Plast Reconstr Surg Hand Surg 1994; 28:95-105.

11. 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.

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

13. Polakis P. The adenomatous polyposis coli (APC) tumor suppressor. Biochim Biophys Acta 1997; 1332:F127-47.

14. Nathke IS, Adams CL, Polakis P, Sellin JH, Nelson WJ. The adenomatous polyposis coli tumor suppressor protein localizes to plasma membrane sites involved in active cell migration. J Cell Biol 1996; 134:165-79.

15. Dale TC. Signal transduction by the Wnt family of ligands. Biochem J 1998; 329:209-23.

16. Brown JD, Moon RT. Wnt signaling: why is everything so negative? Curr Opin Cell Biol 1998; 10:182-7.

17. Daniel JM, Reynolds AB. Tyrosine phosphorylation and cadherin/catenin function. Bioessays 1997; 19:883-91.

18. Muller T, Choidas A, Reichmann E, Ullrich A. Phosphorylation and free pool of beta-catenin are regulated by tyrosine kinases and tyrosine phosphatases during epithelial cell migration. J Biol Chem 1999; 274:10173-83.

19. Grosschedl R, Giese K, Pagel J. HMG domain proteins: architectural elements in the assembly of nucleoprotein structures. Trends Genet 1994; 10:94-100.

20. Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D, Grosschedl R, Birchmeier W. Functional interaction of beta-catenin with the transcription factor LEF-1. Nature 1996; 382:638-42.

21. Molenaar M, van de Wetering M, Oosterwegel M, Peterson-Maduro J, Godsave S, Korinek V, Roose J, Destree O, Clevers H. XTcf-3 transcription factor mediates beta-catenin-induced axis formation in Xenopus embryos. Cell 1996; 86:391-9.

22. Marcus DM, Rustgi AK, Defoe D, Brooks SE, McCormick RS, Thompson TP, Edelmann W, Kucherlapati R, Smith S. Retinal pigment epithelium abnormalities in mice with adenomatous polyposis coli gene disruption. Arch Ophthalmol 1997; 115:645-50.

23. Marcus DM, Rustgi AK, Defoe D, Kucherlapati R, Edelmann W, Hamasaki D, Liou GI, Smith SB. Ultrastructural and ERG findings in mice with adenomatous polyposis coli gene disruption. Mol Vis 2000; 6:169-77 <>.

24. Olschwang S, Tiret A, Laurent-Puig P, Muleris M, Parc R, Thomas G. Restriction of ocular fundus lesions to a specific subgroup of APC mutations in adenomatous polyposis coli patients. Cell 1993; 75:959-68.

25. Hinton DR, He S, Graf K, Yang D, Hsueh WA, Ryan SJ, Law RE. Mitogen-activated protein kinase activation mediates PDGF-directed migration of RPE cells. Exp Cell Res 1998; 239:11-5.

26. Graf K, Xi XP, Yang D, Fleck E, Hsueh WA, Law RE. Mitogen-activated protein kinase activation is involved in platelet-derived growth factor-directed migration by vascular smooth muscle cells. Hypertension 1997; 29:334-9.

27. Seger R, Krebs EG. The MAPK signaling cascade. FASEB J 1995; 9:726-35.

28. Zhu X, Assoian RK. Integrin-dependent activation of MAP kinase: a link to shape-dependent cell proliferation. Mol Biol Cell 1995; 6:273-82.

29. Dunn KC, Marmorstein AD, Bonilha VL, Rodriguez-Boulan E, Giordano F, Hjelmeland LM. Use of the ARPE-19 cell line as a model of RPE polarity: basolateral secretion of FGF5. Invest Ophthalmol Vis Sci 1998; 39:2744-9.

30. Hunt RC, Dewey A, Davis AA. Transferrin receptors on the surfaces of retinal pigment epithelial cells are associated with the cytoskeleton. J Cell Sci 1989; 92:655-66.

31. Giarre M, Semenov MV, Brown AM. Wnt signaling stabilizes the dual-function protein beta-catenin in diverse cell types. Ann N Y Acad Sci 1998; 857:43-55.

32. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW, Vogelstein B, Clevers H. Constitutive transcriptional activation by a beta-catenin-Tcf complex in APC-/- colon carcinoma. Science 1997; 275:1784-7.

33. Casaroli-Marano RP, Pagan R, Vilaro S. Epithelial-mesenchymal transition in proliferative vitreoretinopathy: intermediate filament protein expression in retinal pigment epithelial cells. Invest Ophthalmol Vis Sci 1999; 40:2062-72.

34. Chancy CD, Kekuda R, Huang W, Prasad PD, Kuhnel JM, Sirotnak FM, Roon P, Ganapathy V, Smith SB. Expression and differential polarization of the reduced-folate transporter-1 and the folate receptor alpha in mammalian retinal pigment epithelium. J Biol Chem 2000; 275:20676-84.

35. Okami T, Yamamoto A, Omori K, Takada T, Uyama M, Tashiro Y. Immunocytochemical localization of Na+,K(+)-ATPase in rat retinal pigment epithelial cells. J Histochem Cytochem 1990; 38:1267-75.

36. Gundersen D, Orlowski J, Rodriguez-Boulan E. Apical polarity of Na,K-ATPase in retinal pigment epithelium is linked to a reversal of the ankyrin-fodrin submembrane cytoskeleton. J Cell Biol 1991; 112:863-72.

37. Lengyel E, Wang H, Stepp E, Juarez J, Wang Y, Doe W, Pfarr CM, Boyd D. Requirement of an upstream AP-1 motif for the constitutive and phorbol ester-inducible expression of the urokinase-type plasminogen activator receptor gene. J Biol Chem 1996; 271:23176-84.

38. Hartmann G, Weidner KM, Schwarz H, Birchmeier W. The motility signal of scatter factor/hepatocyte growth factor mediated through the receptor tyrosine kinase met requires intracellular action of Ras. J Biol Chem 1994; 269:21936-9.

39. Ponzetto C, Bardelli A, Zhen Z, Maina F, dalla Zonca P, Giordano S, Graziani A, Panayotou G, Comoglio PM. A multifunctional docking site mediates signaling and transformation by the hepatocyte growth factor/scatter factor receptor family. Cell 1994; 77:261-71.

40. Zhu H, Naujokas MA, Fixman ED, Torossian K, Park M. Tyrosine 1356 in the carboxyl-terminal tail of the HGF/SF receptor is essential for the transduction of signals for cell motility and morphogenesis. J Biol Chem 1994; 269:29943-8.

41. Bernstein LR, Ferris DK, Colburn NH, Sobel ME. A family of mitogen-activated protein kinase-related proteins interacts in vivo with activator protein-1 transcription factor. J Biol Chem 1994; 269:9401-4.

42. Widmann C, Gibson S, Jarpe MB, Johnson GL. Mitogen-activated protein kinase: conservation of a three-kinase module from yeast to human. Physiol Rev 1999; 79:143-80.

43. Angel P, Karin M. The role of Jun, Fos and the AP-1 complex in cell-proliferation and transformation. Biochim Biophys Acta 1991; 1072:129-57.

44. Papkoff J, Aikawa M. WNT-1 and HGF regulate GSK3 beta activity and beta-catenin signaling in mammary epithelial cells. Biochem Biophys Res Commun 1998; 247:851-8.

45. Mann B, Gelos M, Siedow A, Hanski ML, Gratchev A, Ilyas M, Bodmer WF, Moyer MP, Riecken EO, Buhr HJ, Hanski C. Target genes of beta-catenin-T cell-factor/lymphoid-enhancer-factor signaling in human colorectal carcinomas. Proc Natl Acad Sci U S A 1999; 96:1603-8.

46. Zarnegar R, Michalopoulos GK. The many faces of hepatocyte growth factor: from hepatopoiesis to hematopoiesis. J Cell Biol 1995; 129:1177-80.

47. Matsumoto K, Nakamura T. Emerging multipotent aspects of hepatocyte growth factor. J Biochem (Tokyo) 1996; 119:591-600.

48. Matsumoto K, Nakamura T. Hepatocyte growth factor (HGF) as a tissue organizer for organogenesis and regeneration. Biochem Biophys Res Commun 1997; 239:639-44.

49. Miyazawa K, Shimomura T, Naka D, Kitamura N. Proteolytic activation of hepatocyte growth factor in response to tissue injury. J Biol Chem 1994; 269:8966-70.

50. Machemer R, Laqua H. Pigment epithelium proliferation in retinal detachment (massive periretinal proliferation). Am J Ophthalmol 1975; 80:1-23.

51. Machemer R, Aaberg TM, Freeman HM, Irvine AR, Lean JS, Michels RM. An updated classification of retinal detachment with proliferative vitreoretinopathy. Am J Ophthalmol 1991; 112:159-65.

52. Liou GI, Pakalnis VA, Matragoon S, Samuel S, Behzadian MA, Baker J, Khalil IE, Roon P, Caldwell RB, Hunt RC, Marcus DM. HGF regulation of RPE proliferation in an IL-1beta/retinal hole-induced rabbit model of PVR. Mol Vis 8:494-501 <>.

53. Date K, Matsumoto K, Shimura H, Tanaka M, Nakamura T. HGF/NK4 is a specific antagonist for pleiotrophic actions of hepatocyte growth factor. FEBS Lett 1997; 420:1-6.

Liou, Mol Vis 2002; 8:483-493 <>
©2002 Molecular Vision <>
ISSN 1090-0535