Activation and Role of MAP Kinase-Dependent Pathways in Retinal Pigment Epithelial Cells: ERK and RPE Cell Proliferation

Abstract

purpose. Retinal pigment epithelial (RPE) cell proliferation plays a key role in the pathogenesis of ocular diseases involving the posterior segment. Mitogen-activated protein kinases (MAPKs) are involved in the control of cell proliferation. This study was conducted to investigate the involvement of the extracellular signal–regulated kinase (ERK) pathway, the major MAPK pathway implicated in cell growth during the induction of RPE cell proliferation.

methods. RPE cell proliferation was stimulated with 10% fetal calf serum (FCS). Activation of the Ras/Raf/MAP kinase-ERK kinase (MEK)/extracellular signal-regulated kinase (ERK) signaling pathway was detected by Western blot analysis and immunochemistry with specific anti-phosphosignaling protein antibodies. Pharmacologic and antisense (AS) oligonucleotide (ODN) strategies were used to analyze the ERK signaling involved in serum-induced cell proliferation.

results. FCS (10%) induced more vigorous RPE cell proliferation than did FGF2, VEGF, platelet-derived growth factor (PDGF), or epidermal growth factor (EGF), alone or in combination. Pharmacologic inhibition of Ras and Raf-1 reduced cell proliferation by 67% to 100% and by 62% to 79%, respectively, demonstrating that activation of the Ras/Raf-1 pathway was essential for FCS-induced RPE cell proliferation. MEK1/2, ERK2, and P90 ribosomal S6 kinase (P90^RSK), the kinases downstream from ERK2, were strongly activated during cell proliferation. Pharmacologic inhibition of MEK1/2 abolished activation of ERK2, but reduced cell proliferation by only 32%, showing that MEK/ERK participates in the signaling involved in RPE cell growth. Both inhibition of ERK2 activation, which reduced cyclin D1 production, and inhibition of cyclin D1 by AS ODN decreased cell proliferation, suggesting that RPE cell proliferation is mediated by cyclin D1 through ERK2.

conclusions. The requirement for Ras and the regulatory role of ERK2 in cyclin D1 production and in cell proliferation suggest that the Ras/Raf/MEK/ERK pathway plays a key role in the control of RPE cell proliferation. These data may have important implications for the development of more selective methods for the inhibition of retinal proliferation.

Mitogen-activated protein kinases (MAPKs) are ubiquitous enzymes involved in signal transduction. ¹ Their activity is essential in numerous cellular functions, including proliferation and programmed cell death. Extracellular signal-regulated kinases (ERKs) 1 and 2 are the most well-characterized MAPKs. ² One of the most important activators of ERK1/2 is the Ras, guanosine triphosphatase (GTPase). Once Ras has been activated, it interacts with and facilitates activation of the serine/threonine kinase Raf-1. Activated Raf-1 phosphorylates and activates the downstream kinase MAP kinase/ERK kinase (MEK)1/2, which in turn phosphorylates and activates ERK1/2. ERK1/2 regulates activity by phosphorylating cytosolic kinases, including cytosolic phospholipase A2 (cPLA2) ³ and protein kinase P90^RSK. ⁴ The location of ERK1/2 is a key determinant of the functions of these kinases. In unstimulated cells, ERK1/2 are largely cytoplasmic. On stimulation, ERK1/2 is translocated to the nucleus, where they activate transcription factors involved in the regulation of cell cycle proteins. 

In vivo, the retinal pigment epithelium (RPE) is a monolayer of highly polarized cells that mediate the flow of nutrients, metabolites, and ions between the choroid and the neural retina. RPE cells have a limited proliferation capacity, corresponding to the normal increase in retinal area associated with growth and age. After injury to the choroid-retina, RPE cells are exposed to a variety of growth factors and cytokines from the serum. RPE cell activation results in cell proliferation. RPE cells play a key role in various proliferative diseases of the retina. Proliferation of these cells is an essential step in the pathogenesis of retinal diseases that lead to vision loss, such as proliferative vitreoretinopathy (PVR). The involvement of individual growth factors in the stimulation of RPE cell proliferation has been thoroughly investigated, ^{5 6 7 8} but nothing is yet known about the signaling and molecular mechanisms underlying these processes in vivo. Some of the molecular mechanisms involved in RPE cell proliferation have been partially elucidated by studies in vitro, but little is known about the intracellular signaling that mediates RPE cell proliferation. ERK1/2 has been shown recently to be involved in the RPE cell proliferation stimulated by fibroblast growth factors (FGFs), but no upstream or downstream activators of the ERK pathway were analyzed. ⁹ In vitro, RPE cell proliferation is a synergistic process, with multiple growth factors acting in concert. ¹⁰ It has been suggested that the activation of several different signaling pathways by serum components is required for the development and progression of various RPE proliferative diseases in vivo. It would therefore be useful to determine precisely which signaling pathway mediates RPE cell proliferation, so that this process can be inhibited specifically in RPE-related diseases such as PVR. 

In this study, we investigated various aspects of ERK signaling during serum-stimulated RPE cell proliferation. The role of the Ras/Raf/MEK/ERK signaling pathway in RPE cell proliferation was determined, using both pharmacologic and antisense oligonucleotide strategies to inhibit specific kinases of the signaling cascade. 

Materials and Methods

RPE Culture and Treatment of Cells

Human RPE cells were isolated as previously described. ¹¹ Primary human RPE cell cultures were seeded on 0.1% gelatin-coated dishes and grown in F99 RPE growth medium, and subcultures were grown in F99 basal medium (medium 199/Ham’s F12; GibcoBRL, Grand Island, NY), supplemented with 10% FCS, 1 mM sodium pyruvate, 1 μg/mL insulin, 50 μg/mL gentamicin, and 2.5 μg/mL amphotericin B. 

The mitogenic effects of FGF2 (18-kDa form, kindly provided by Carlo Erba, Farmitalia, Milan, Italy), VEGF (165 amino acids form, generously provided by Jean Plouet, GDR CNRS), platelet-derived growth factor (PDGF; R&D Systems, Oxford, UK), and epidermal growth factor (EGF; BioWhittaker, Emerainville, France) were studied in RPE cell cultures containing 0.1% FCS. Before comparing the mitogenic efficiency of the individual growth factors with that of a combination of the four growth factors and FCS, we determined the concentration required for the maximum mitogenic effect for each factor separately by dose-effect experiments, using growth factor concentrations of 0.01 ng/mL to 20 ng/mL in 0.1% FCS culture medium. Cells were used to seed 12-well plates at a density of 10⁴ cells per well and were cultured in basal medium supplemented with 10% FCS for 3 days. Cells were then stimulated with 0.1% FCS, 10% FCS, and FGF2, PDGF, EGF, and VEGF and with a combination of the four growth factors in basal medium supplemented with 0.1% FCS. Three days later, the culture media were removed, and the cells stimulated again with the growth factors. The mitogenic effects of the various growth factors were scored on the day of the first stimulation and on days 3 and 6 of the culture period. All experiments were run in triplicate and were performed at least three times. The proliferation of RPE cells was assessed by counting the number of cells, as previously described. ¹² The number of cells was determined by two methods: trypan blue exclusion and 3(4,5-dimethylthiazol-yl) 2,5 diphenyltetrazolium bromide (MTT) staining. ¹³ Maximum growth was observed with 1 ng/mL FGF2 and PDGF and 5 ng/mL EGF and VEGF, with the higher concentrations having no additional effect. 

In some experiments, specific inhibitors of Ras (AFC, FTS, FPT inhibitor III, and mevastatin), Raf-1 (Raf-1 inhibitor I), MEK1 (PD98059), and MEK1/2 (U0126) (Calbiochem, Meudon, France) were added 12 hours before the induction of cell proliferation, on the first day of stimulation and on day 3 of the cell proliferation assay. Cells were stimulated with 10% FCS on days 1 and 3. Stock solutions of each inhibitor were made in dimethyl sulfoxide (DMSO) and diluted in DMEM, so that the final concentration of DMSO in test solutions did not exceed 0.1% (a concentration that has no effect on RPE cell proliferation). The effects of each of the signaling inhibitors were assessed. 

Western Blot Analysis

After serum stimulation, RPE cells were washed twice in PBS, lysed in ice-cold lysis buffer (50 mM Tris-HCl, [pH 7.5], 100 mM NaCl, 50 mM NaF, 5 mM EDTA, 40 mM β-glycerophosphate, 0.2 mM sodium orthovanadate, 1 μg/mL leupeptin, and 1 μM pepstatin) and centrifuged at 4°C for 10 minutes at 10,000g. Monoclonal antibody directed against β-actin was used as an internal standard for checking protein loading. Cell lysates were mixed with 3× Laemmli buffer and heated for 5 minutes at 95°C. They were then resolved by SDS-PAGE (12%–15% polyacrylamide gel), transferred to nitrocellulose filters by electroblot, and probed with polyclonal antibodies directed against Raf-1, MEK1, ERK1/2, P90^RSK, cAMP-response element (CRE)-binding protein (CREB; dilution 1:100; Santa Cruz Biotechnology, Heidelberg, Germany), and cyclin D1 (NeoMarkers, Fremont, CA), to determine the amounts of these kinases present during the proliferation period. A polyclonal antibody directed against phospho-ERK1/2 (thr¹⁹² and tyr¹⁹⁴; 1:5000; Promega, Madison, WI); polyclonal antibodies directed against phospho-Raf-1 (ser³³⁸, dilution 1:1000; Chemicon, Harrow, UK) and phospho-Raf-1 (tyr³⁴⁰ and tyr³⁴¹; 1:1000, Biosource, Nivelles, Belgium); and polyclonal antibodies directed against phospho-MEK1/2 (ser²¹⁷ and ser²²¹), phospho-P90^RSK (ser³⁸¹) and (thr⁶⁰ and ser³⁶⁴), phospho-ATF2 (thr^69/71), and phospho-CREB/ATF1 (ser¹³³, 1:2000; New England Biolabs/Cell Signaling Technology, Boston, MA) were used to analyze the activation of intracellular signaling during RPE cell proliferation. The primary antibodies were detected with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody. Enhanced chemiluminescence substrate was used to detect the secondary antibody according to the manufacturer’s instructions, and the membrane was placed against autoradiographic film (Hyperfilm ECL; Amersham, Orsay, France). 

Immunochemistry

The nuclear translocation of phosphorylated ERK1/2 was confirmed by immunostaining. RPE cells plated on glass coverslips were fixed by incubation with 4% paraformaldehyde (PAF) for 15 minutes at room temperature. Cells were washed with PBS, treated with ice-cold methanol for 20 minutes, washed twice with PBS, and incubated with anti-phosphorylated ERK1/2 antibody (1:1000) for 1 hour. Control experiments were performed with antibody against phosphorylated ERK1/2 preadsorbed with an excess of phosphorylated active ERK1/2 peptide (Sigma, Saint Quentin-Fallavier, France). Antigen–antibody complexes were detected with rhodamine isothiocyanate (RITC)–conjugated anti-rabbit antibody (1:100; Biosys, Compiegne, France). Cells were washed with PBS and mounted in glycerol/PBS (1:1). 

Oligonucleotides and Oligonucleotide Treatment of Cells

Unmodified sense and antisense phosphodiester oligonucleotide (ODNs) directed against cyclin D1 were designed from the published sequence of the human cyclin D1 gene, such that they were targeted to the translation site of the cyclin D1 cDNA. The sense ODN sequence was identical with the conserved cyclin D1 gene sequences conserved in human, mouse, and rat. ¹⁴ The antisense (AS) cyclin D1 ODN was 5′-GGA-GCT-GGT-GTT-CCA-TGG-3′ (referred to as AS cyclin D1), and the corresponding sense ODN was 5′-CCA-TGG-AAC-ACC-AGC-TCC-3′ (referred to as S cyclin D1). We used lipofectin (GibcoBRL, Cergy Pontoise, France), a cationic lipid, to deliver the ODNs, because this method elicits high levels of ODN uptake and stability in the intracellular compartment, but does not affect the final nuclear location of ODNs after endocytosis and release from the endocytic compartment. All ODNs were synthesized commercially (Life Technology, Paris, France) and were purified by HPLC. Lipofectin-ODN complexes were produced according to the manufacturer’s instructions. All results presented herein were obtained with 20 μM cyclin D1 and 7 μg/mL lipofectin. Cells were incubated for 3 days in the presence of 10% heat-treated calf serum. RPE cells were then treated with ODNs-lipofectin for 4 hours and washed twice with serum-free medium. Fresh complete culture medium with the appropriate concentration of ODN was then added. The cells were cultured for 7 days. For 3 days, beginning on day 4 of incubation, cells were treated with the appropriate concentration of ODN without lipofectin. 

Statistics

Two-tailed Student’s t-test (normal distributions with equal variances) and Mann-Whitney tests (nonparametric tests) were used for statistical analysis. 

Results

Maximal Human RPE Cell Proliferation Involving Multiple Growth Factors

The effects of various mitogens on RPE cells cultured in medium F99 plus 0.1% FCS are presented in Figure 1 . RPE cells did not proliferate when cultured in medium F99 plus 0.1% FCS, but they survived for 6 days, whereas they died in basal medium F99 alone. FGF2 is a potent mitogen for RPE cells, as previously described. ¹² After 3 days of FGF2 treatment in 0.1% DMEM, there were 186% the number of cells present on day 1 of culture (Fig. 1A) . By day 6, FGF2 had induced a 348% increase in the total number of RPE cells. PDGF was also a potent mitogen for RPE cells (Fig. 1A) , stimulating proliferation by 218% and 367% on days 3 and 6, respectively. EGF was less mitogenic for RPE cells than FGF2 and PDGF, inducing an increase in RPE cell number of only 137% and 228% on days 3 and 6, respectively (Fig. 1A) . VEGF was only weakly mitogenic for RPE cells, inducing an increase in RPE cell proliferation of only 120% and 203% on days 3 and 6, respectively (Fig. 1A) . Stimulation with a combination of FGF2, PDGF, VEGF, and EGF induced RPE cell proliferation more strongly than did individual growth factors, showing that these growth factors had a synergistic effect on RPE cell proliferation, but the maximal proliferative activity was observed with FCS (Fig. 1A) . FCS, at a concentration of 10%, induced a 353% increase in RPE cell proliferation by day 3 of the culture period and a 648% increase in cell proliferation by day 6. This suggests that FCS, at a concentration of only 10%, was able to promote the multiple intracellular signaling necessary for the maximal RPE cell proliferation. A higher concentration of FCS had no greater effect on cell proliferation (data not shown). 

Involvement of the Ras/Raf-1/MEK/ERK Signaling Pathway in RPE Proliferation

Ras is a key component of the signal transduction pathway that mediates the proliferation of many types of cells. We therefore investigated the role of Ras signaling in serum-stimulated RPE cells, using various inhibitors of Ras synthesis, posttranslational modification, and activation (Fig. 1B) . RPE cell treatment with 100 μM of the farnesyl transferase (FTP) competitor N-acetyl-S-farnesyl-l-cysteine (AFC) resulted in a 67% reduction in cell proliferation after 6 days of culture, as assessed by cell counts. FPT III inhibitor (10 μM), a potent inhibitor of Ras farnesyl transferase, completely inhibited serum-stimulated RPE cell proliferation. Mevastatin (5 μM), an inhibitor of the posttranslational prenylation of Ras, abolished the RPE cell proliferation stimulated by serum (Fig. 1B) . Finally, S-trans,trans-farnesylthiosalicylic acid (FTS) (2.5 μM), a Ras antagonist that dislodges Ras from its membrane-anchoring sites also abolished FCS-induced RPE cell proliferation. Thus, Ras plays a key role in transmitting signals for serum-stimulated RPE cell proliferation. 

The ERK pathway is one of the major components of the signal transduction mediated by Ras and leading to growth. This pathway consists of a linear cascade of the protein kinases Raf, MEK, and ERK. We analyzed the state of activation of Raf-1 in serum-stimulated RPE cells by Western blot analysis with an antibody that specifically recognizes active forms of Raf-1 (Fig. 2A) . Low basal levels of Raf-1 phosphorylation were detected. Serum stimulation increased levels of Raf-1 phosphorylation. Raf-1 phosphorylation peaked after 2 hours of stimulation in serum-stimulated RPE cell cultures (Fig. 2A) . To test whether Ras activates Raf-1, serum-stimulated RPE cells were treated with FPT III inhibitor to inhibit Ras. Inhibition of Ras blocked Raf-1 activation (Fig. 2A) , whereas the production of Raf-1 remained constant (Fig. 2A) . To confirm the role of Raf-1 in RPE cell proliferation, cultures were then treated with Raf-1 inhibitor 1 (1–30 μM), a potent and selective inhibitor of Raf-1 (Fig. 2B) . Raf-1 inhibition by 30 μM Raf-1 inhibitor 1 reduced serum-stimulated RPE cell proliferation by 62% to 79% during the 7-day culture period. Thus, Raf-1 was also involved in the signaling pathway of serum-induced RPE cell proliferation. We then analyzed the involvement of MEK1/2 and ERK2 in signaling in serum-stimulated RPE cell proliferation. We investigated the state of phosphorylation of ERK2 by Western blot analysis with an antibody that specifically recognizes the active form of ERK2. Basal levels of ERK2 phosphorylation were very low (Fig. 2C) . The stimulation of RPE cells by serum induced changes in the level of phosphorylation of ERK2 during the 24-hour period of culture (Fig 2C) . ERK2 activation peaked within the first 10 minutes after the addition of serum, remained high for the following 6 hours, and then returned to activation levels similar to those detected before serum stimulation. To test whether MEK1/2 activate ERK1/2, serum-stimulated RPE cells were treated with PD98059 (20 μM), a potent and selective inhibitor of MEK1/2. Inhibition of MEK1/2 blocked ERK1/2 activation (Fig. 2C) . ERK2 production was constant throughout the culture period. 

We then investigated the role of MEK activation in RPE proliferation. To determine whether activation of the MEK/ERK signaling pathway by serum was necessary for RPE cell proliferation, we treated serum-stimulated cells with various specific pharmacologic inhibitors of the MEK/ERK pathway. Treatment of RPE cells with the potent and highly selective inhibitor of MEK1, PD98059, at concentrations up to 10 μM, had a weak effect on serum-stimulated RPE cell proliferation (Fig. 3A) . At the higher concentration of 30 μM, PD98059 inhibited serum-stimulated cell proliferation by only 24%, suggesting that the signaling involved in RPE cell proliferation was mildly sensitive to MEK1 inhibition and partially mediated by MEK1. It has been shown recently that PD98059 directly inhibits cyclooxygenase (COX)-1 and -2 ¹⁵ and that the COX 1/2 enzymes are involved in cell growth. ¹⁶ We therefore thought that COX might be the target of PD98059. Specific COX-1 and -2 inhibition with SC-560 (50 μM) and NS-398 (50 μM), respectively, did not affect RPE cell proliferation (Fig. 3B) . These results rule out the involvement of COX 1/2 in serum-stimulated RPE cell proliferation. We therefore investigated whether MEK2 was specifically involved in mediating RPE cell proliferation by treating serum-stimulated cells with the highly specific and potent inhibitor of MEK1/2, U0126. RPE treatment with up to 30 μM U0126 resulted in a reduction in serum-stimulated RPE cell proliferation of only 32% (Fig. 3C) , suggesting that the partial inhibition of RPE cell proliferation by PD98059 was not due to the absence of an inhibitory effect of this compound on MEK2 or to a dose-limited effect of the inhibitor. These data strongly suggest that MEK/ERK are involved in mediating serum signaling for RPE cell proliferation. 

Nuclear translocation of phosphorylated ERK1/2 is the limiting step in the transcriptional activity of ERK1/2 and is necessary for the induction of nuclear transcription factors and cell proliferation. Thus, the absence of or a decrease in translocation to the nucleus of phosphorylated ERK1/2 in serum-stimulated RPE cells would limit the effects of ERK on cell proliferation. Immunostaining for phosphorylated ERK1/2 showed that the active form of ERK1/2 was present only in tiny quantities in the cytoplasm of unstimulated cells (Fig. 4A) . After 5 minutes of stimulation with serum, active ERK1/2 staining was detectable in the cytoplasm of some cells, whereas other cells presented immunostaining for phosphorylated ERK1/2 in the nucleus (Fig. 4B) . After 15 minutes, staining for ERK1/2 was present in cytoplasm of some cells, whereas most of the nuclei were strongly stained (Fig. 4C) . After 2 hours of culture, staining for phosphorylated ERK1/2 was no longer observed in the nuclei of most of the stimulated cells, and moderate staining for active ERK1/2 was detected in the cytoplasm (Fig. 4D) . No immunostaining was observed if cells were incubated with antibody against active phosphorylated forms of ERK1/2, preadsorbed with an excess of the phosphopeptide corresponding to the phosphorylated sequence of active ERK1/2 (Fig. 4E) and if the RITC-conjugated anti-rabbit antibody was omitted (Fig. 4F) . This suggests that the partial effect of MEK/ERK activation on RPE cell proliferation was not due to partial blockade of the nuclear translocation of phosphorylated active ERK1/2 in stimulated cells. 

Analysis of the Downstream Effectors and Transduction Pathways of RPE Cell Proliferation Regulated by the ERK Pathway

ERK1/2 have been reported to activate the nonnuclear serine/threonine kinase, P90 ribosomal S6 kinase (P90^RSK). ¹⁷ Moreover, CREB protein is directly phosphorylated by P90^RSK. ¹⁴ We therefore investigated whether P90^RSK and the CREB protein were phosphorylated during RPE cell proliferation (Fig. 5A 5B) . Phosphorylation of P90^RSK was barely detectable in the basal state. Stimulation of RPE cell proliferation resulted in a strong increase in P90^RSK activation after 10 minutes of culture (Fig. 5A) . P90^RSK phosphorylation continued to be detectable for the following 2 hours and rapidly decreased thereafter. CREB protein phosphorylation was also barely detectable in the basal state, and stimulation of RPE cell proliferation resulted in an increase in CREB protein phosphorylation after 10 minutes (Fig. 5B) . CREB protein activation was constant over the first 30 minutes of stimulation and was undetectable thereafter. The kinetics of ATF-1 phosphorylation were similar to those of CREB protein phosphorylation (Fig. 5B) . Inhibition of MEK1/2 with U0126 (30 μM) greatly reduced P90^RSK phosphorylation, but did not affect the level of phosphorylation of CREB/ATF-1 (Figs. 5A 5B) . Although ATF-2 was phosphorylated after serum stimulation, MEK1/2 inhibition had no effect on the kinetics of ATF-2 phosphorylation (Fig. 5B) . These data suggest that the partial effect of the ERK1/2 pathway on RPE cell proliferation was not due to partial inhibition of the activation of P90^RSK within stimulated cells and that CREB/ATF-1, which may be involved in the activation of genes involved in cell proliferation, is not regulated by the ERK1/2 pathway. 

It has been demonstrated that levels of cyclin D1, which regulates cell cycle progression, may also be controlled by the ERK1/2 pathway, ^{18 19} whereas other studies have reported that the cyclin D1 gene is a direct target of ATF-2, which binds as a CREB protein complex. ²⁰ We therefore investigated whether cyclin D1 production was dependent on activation of ERK during RPE cell proliferation. Analysis of cyclin D1 levels by Western blot showed that serum greatly stimulated the production of cyclin D1 after 6 hours of culture (Fig. 6A) . RPE cell treatment with U0126 (30 μM) dramatically affected cyclin D1 production (Fig. 6A) , showing that the ERK1/2 pathway played an important role in regulation of the production of cyclin D1 during RPE cell proliferation. We further investigated the involvement of cyclin D1 in the proliferation of serum-stimulated RPE cells by downregulating cyclin D1 production by an antisense oligonucleotide strategy. Depletion of Cyclin D1 in FCS-stimulated cells by treatment with AS cyclin D1 ODNs decreased cell proliferation by 26% after 6 days of culture (Fig. 6B) , an inhibitory effect similar to that observed with MEK1/2 inhibition (compare Figs. 6B and 3C ). S cyclin D1 ODN treatment had no significant effect on the proliferation of FCS-stimulated RPE cells, when compared with control untreated cells and AS cyclin D1 ODN-treated cells (Fig. 6B) . These data suggest that ERK1/2 control RPE cell proliferation by regulating cyclin D1 production. 

The substrates of ERK1/2 also include the 85-kDa cPLA₂, which has been shown to be critical for cell proliferation in various cell types. ²¹ We therefore analyzed the effects of inhibition of cPLA₂ on serum-stimulated RPE cell proliferation. AACOCF₃, a trimethyl ketone analogue of arachidonic acid, recently shown to inhibit cPLA₂ dose dependently, inhibited RPE cell proliferation (Fig. 6C) , suggesting that cPLA₂ is also an important component of RPE cell proliferation. AACOCF₃ (10 μM) unexpectedly reduced cell proliferation by 72%, whereas the complete inhibition of ERK1/2 activation only reduced RPE cell proliferation by approximately 30% (compare Figs. 3A 3C and Fig. 6C ), suggesting that ERK1/2 did not directly control the activity of cPLA₂ in cell proliferation. In contrast, inhibition of the calcium-independent PLA₂ by 20 μM haloenol lactone suicide substrate (HELSS) did not significantly affect serum-induced cellular proliferation (Fig. 6C) . Higher concentrations of HELSS had no effect on FCS-stimulated RPE cell proliferation (data not shown). Thus, although cPLA₂ is a major component of the signaling pathway of FCS-induced cell proliferation, ERK1/2 did not control the PLA₂/COX pathway for cell proliferation in RPE cell cultures. 

Discussion

Regulation of the Ras/Raf/MEK/ERK Pathway during RPE Cell Proliferation

The purpose of this study was to determine whether MAP kinase signaling pathways are involved in the proliferation of RPE cells stimulated by serum. We demonstrated that the Ras/Raf/MEK/ERK signaling pathway was activated during RPE cell proliferation. A proposed scheme for activation of the ERK1/2 signaling pathway for RPE cell proliferation is presented in Figure 7 . In this pathway, the Ras/Raf module is essential for FCS-induced RPE cell proliferation, whereas the MEK/ERK module accounts for only one third of the signaling involved in RPE cell proliferation. Because the nuclear translocation of phosphorylated ERK is the limiting step in the transcriptional activity of ERK and ERK-mediated cell proliferation, we investigated the nuclear translocation of ERK. Immunostaining for phosphorylated ERK1/2 showed that ERK was translocated to the nucleus during serum-induced RPE cell activation. Thus, the partial effect of MEK/ERK activation on RPE cell proliferation was not due to a partial blockade of nuclear translocation of the phosphorylated active ERK1/2 in stimulated cells. Another possibility is that a signaling pathway may regulate the Ras/Raf/MEK/ERK cascade at the level of Raf-mediated activation of ERK. It has been shown that the ability of Ras to activate the downstream Raf/ERK cascade is impaired in cells treated with agents that increase cAMP concentration, leading to a decrease in ERK1/2 activation. ²² However, the exact point at which the cascade is inhibited by cAMP/protein kinase A (PKA) is still a matter for debate. Many studies have reported that the target of PKA is in most cases located downstream from Ras activation, resulting in a reduced ability of Ras to interact with Raf. ²³ However, it has also been demonstrated that cAMP-mediated growth inhibition does not involve MAPK inhibition, although MAPK activation is delayed. ²⁴ An alternative mechanism of inhibition of the MAPK pathway by PKA involves the inhibition of kinases that can activate Raf, such as certain isoforms of protein kinase C (PKC) or Src. The precise mechanism of cAMP/PKA crosstalk with the MAPK/ERK cascade remains to be identified. Surprisingly, it has recently been shown that cAMP-activated PKA stimulates cell proliferation by inhibiting ERK1/2, favoring the activation of PI3 K signaling. ^{25 26} Thus, the various data obtained concerning the effects of cAMP/PKA on cell proliferation are difficult to reconcile, and the precise role of the cAMP/PKA pathway in the control of ERK1/2-mediated cell proliferation remains unclear. Nevertheless, it would be of interest to investigate the role of the cAMP/PKA and PI3 kinase pathways in the regulation of ERK activation during RPE cell proliferation. 

We showed that transient activation of ERK mediates signals for cell proliferation in serum-stimulated RPE cells. It has recently been demonstrated that the sustained activation of ERK1/2 results in the transmission of antiapoptotic signals in serum-depleted RPE cell cultures. ^{9 14} These data suggest that the kinetics of ERK1/2 activation may control cell fate in RPE cell cultures, confirming previous data showing that the duration of ERK activation determines the cell’s fate in the PC12 cell line. ²⁶ It has recently been shown that ERK1/2 is phosphorylated in RPE cells within 15 minutes and remains so for several days after retinal detachment (RD), suggesting that ERK signaling may be involved in changes in gene expression, morphology, and function in RPE cells after RD. ²⁷ Thus, identification of the signaling pathways responsible for the prolonged or transient activation of ERK is of great importance in RPE cells. 

Multiple Crosstalk at and Downstream from ERK for Serum-Induced RPE Cell Proliferation

PLA₂ is a cytoplasmic protein that is regulated by ERK1/2. It is an interesting protein, because its products are substrates for the secondary generation of downstream signals, which may directly activate cell proliferation. Studies of the role of cPLA₂ in the signaling of proliferating RPE cells are of particular relevance, because both ERK1/2 and P38 kinase, another member of the MAPK family, contribute to the activation of cPLA ² . ¹⁵ We found that cPLA₂, and not the calcium-independent PLA₂, played a key role in RPE cell proliferation. However, we found no direct relationship between ERK1/2 and cPLA₂ because the complete inhibition of ERK2 reduced cell proliferation by 30%, whereas complete inhibition of cPLA₂ reduced cell growth by 70%. This suggests that another kinase controls activation of cPLA₂ and that the ERK1/2 pathway differs from the PLA₂ pathway for RPE cell proliferation. It has recently been demonstrated that P38 kinase and ERK1/2 cooperate to achieve the full activation of PLA₂. ²⁸ Activation of cPLA₂ is also involved in MAPK-independent pathways, including PKCs, suggesting that PKCs may be upstream regulators of PLA₂ and downstream effectors of PLA₂ for cell proliferation. ²⁰ Thus, it would be of interest to study the effect of P38 kinase and the various forms of PKC on PLA₂ phosphorylation and cell proliferation in RPE cell cultures. P38 kinase production in RPE cells has not been investigated, but PKCs have been shown to be widely produced in RPE cells. ^{29 30} PKC does not seem to be involved in RPE cell apoptosis, ³¹ but conflicting data have been obtained concerning the role of PKC in RPE cell proliferation. Acceleration of the onset of cell proliferation is dependent on conventional PKC, whereas long-term proliferation of RPE cells is not dependent on conventional PKC. ³² However, in a recent study with hypericin, the findings suggested that serum-induced RPE cell proliferation is PKC dependent. ³³ However, hypericin has also been reported to inhibit the activity of ERK, suggesting that hypericin may block serum-induced RPE cell proliferation through inhibition of ERK1/2. Moreover, the situation is more complicated than it appears, because conventional and atypical PKC isoforms have been demonstrated to activate the Ras-independent ERK1/2 pathway. ³⁴ It would be of interest to analyze the roles of the various PKC isoforms known to interact with the ERK1/2 pathway in serum-induced RPE cell proliferation. 

In conclusion, the Ras/Raf-1/MEK/ERK signaling cascade is involved in and tightly regulated during serum-induced RPE cell proliferation. Identification of the complete kinase signaling network, its substrates, and potential regulatory loops may help in the development of selective methods and strategies for the treatment of proliferative diseases in which RPE cells are involved. 

 Supported by Grants 9697 and 4474 from the Association for Research on Cancer (FM), Grant 75/01-RS/63 from The League against Cancer (FM), and the Ministry of Education and Research (GL).

 Submitted for publication March 1, 2002; revised April 26, 2002; accepted May 16, 2002.

 Commercial relationships policy: N.

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

 

Figure 1.

 

Comparison of the mitogenic activity of various mitogens on the proliferation of RPE cells and effects of inhibition of Ras on RPE cell proliferation. (A) Cells were stimulated with FCS, FGF2 (1 ng/mL), PDGF (1 ng/mL), EGF (5 ng/mL), or VEGF (5 ng/mL) and with a combination of the four growth factors in basal medium supplemented with 0.1% FCS (day 0). The mitogenic effects of the various growth factors were scored at the indicated time of the culture period. (B) The proliferation of RPE cells treated with AFC, the FPT III inhibitor, mevastatin (Meva), or FTS and then stimulated with FCS, was studied on days 1, 2, 3, and 6. Data are the mean ± SD of results in three separate experiments. Differences between means were analyzed by the Mann-Whitney test: *P < 0.05, **P < 0.01, ***P < 0.005.

Figure 2.

 

Analysis of the activation of Raf-1 and ERK2 during RPE cell proliferation and effects of Raf-1 inhibition on RPE cell proliferation. (A) RPE cells were cultured for 3 days and then stimulated with FCS. Cells were lysed at the indicated times, and proteins were reduced and subjected to SDS-PAGE and Western blot analysis with anti-active phosphorylated Raf-1 and anti-Raf-1 antibodies. (B) RPE cells were treated with the indicated amount of Raf-1 inhibitor 1, and cell proliferation was analyzed at the indicated time. (C) RPE cells were cultured in medium F99 with 10% FCS for 3 days and then stimulated with FCS in the presence and absence of the MEK inhibitor PD98059. Cells were lysed at the times indicated, and proteins were reduced and subjected to SDS-PAGE and Western blot analysis with anti-active phosphorylated ERK1/2 and anti-ERK2 antibodies. Similar results were obtained in three independent experiments.

Figure 3.

 

Effects of inhibition of MEK1, COX-1 and -2, and MEK1/2 on cell proliferation in FCS-stimulated RPE cells. RPE cells were cultured in DMEM with FCS for 3 days. RPE cells were then treated with the MEK1 inhibitor PD98059 (A), the COX 1 inhibitor SC-560, the COX-2 inhibitor NS-398 (B), and the MEK1/2 inhibitor U0126 (C) at the indicated concentrations, and cell proliferation was analyzed. Similar results were obtained in three independent experiments.

Figure 4.

 

Analysis of ERK1/2 nuclear translocation. Immunostaining for the active phosphorylated forms of ERK1/2 was performed after overnight incubation in serum-free culture medium (A) and 5 minutes (B, E, and F), 15 minutes (C), and 2 hours (D) after FCS stimulation of RPE cells. The specificity of anti-phosphorylated ERK1/2 immunoreactivity was checked by incubating an excess of the phosphopeptide corresponding to the phosphorylated sequence of active ERK1/2 with antibody against active phosphorylated forms of ERK2 (E). Omission of the RITC-conjugated anti-rabbit antibody (F). Similar results were obtained in three independent experiments.

Figure 5.

 

Effects of the inhibition of MEK1/2 activities on P90^RSK, CREB/ATF-1, and ATF-2, activation in FCS-stimulated RPE cell cultures. Cultures of FCS-stimulated cells were treated with the MEK1/2 inhibitor U0126. Phosphorylation of P90^RSK (A) and CREB/ATF1 and ATF2 (B) was analyzed by Western blot analysis, with anti-active P90^RSK and anti-P90^RSK antibodies, anti-active CREB/ATF1 and anti-CREB/ATF1 antibodies, and anti-active ATF-2 antibody. Similar results were obtained in three independent experiments.

Figure 6.

 

Effects of the inhibition of ERK1/2 activity on cyclin D1 production and of cyclin D1 depletion and PLA₂ inhibition on cell proliferation. Cultures of FCS-stimulated cells were treated with the MEK1/2 inhibitor U0126, and cyclin D1 production was analyzed by Western blot analysis (A). The role of cyclin D1 in cell proliferation was analyzed with an antisense ODN strategy, using AS ODNs against cyclin D1 (B). The effects of PLA₂ on cell proliferation were analyzed by a pharmacologic approach, with specific inhibitors of cytosolic (AACOCF₃) and calcium-independent (HELSS) PLA₂ activities (C). Similar results were obtained in three independent experiments. Data are the mean ± SD; differences between means were analyzed by the Mann-Whitney test: *P < 0.05, **P < 0.01, ***P < 0.005.

Figure 7.

 

Schematic representation of the activation of the Ras/Raf-1/MEK/ERK signaling pathway for cell proliferation. Serum engages two major parallel pathways involving ERK1/2 and cPLA₂. Once Ras has been activated, it activates Raf-1. Then, Raf-1 activates MEK1/2, which in turn activates ERK1/2. ERK1/2 stimulates P90^RSK activity and cyclin D1 production. ERK1/2-induced cyclin D1 participates in the stimulation of RPE cell proliferation. COX-2 does not regulate RPE cell proliferation through ERK1/2 activation. Although, P90^RSK activity is controlled by ERK1/2, it does not regulate activation of CREB, ATF1, or ATF2. Although the activity of cPLA₂ is not controlled by ERK1/2, it is an essential participant in RPE cell proliferation signaling.