Molecular Vision 2006; 12:821-831 <http://www.molvis.org/molvis/v12/a93/> Received 27 March 2006 | Accepted 21 July 2006 | Published 26 July 2006

Download Reprint

The positive feedback role of arachidonic acid in the platelet-derived growth factor-induced signaling in lens epithelial cells

Wei Zhang,^1,2 Yin Wang,³ Chao-Wei Chen,³ Kuiyi Xing,² S. Vivekanandan,² Marjorie F. Lou^2-4
 
(The first two authors contributed equally to this publication)
 
 

¹Beijing Institute of Ophthalmology, Capital University of Medical Sciences affiliated Beijing Tongren Ophthalmology Center, Beijing, China; ²Department of Veterinary and Biomedical Sciences and ³Department of Biochemistry, University of Nebraska-Lincoln, Lincoln, NE; ⁴Department of Ophthalmology, University of Nebraska Medical Center, Omaha, NE

Correspondence to: Marjorie F. Lou, 134 Veterinary Basic Science Building, Lincoln, NE, 68583-0905; Phone: (402) 472-0307; FAX: (402) 472-9690; email: mlou1@unl.edu

Abstract

Purpose: Platelet-derived growth factor (PDGF)-stimulated cell proliferation has been associated with reactive oxygen species (ROS)-mediated redox signaling. This study examined the role of arachidonic acid (AA) in PDGF-stimulated ROS generation in human lens epithelial B3 cells (HLE B3).

Methods: PDGF (1 ng/ml)-stimulated ROS generation was examined using dichlorofluorescein (DCFH)-activated fluorescence by laser confocal microscopy while AA (30-150 μM)-stimulated superoxide anion production was measured using lucigenin-amplified chemiluminescence in serum-starved HLE B3 cells. PDGF-stimulated AA release was quantified by cells prelabeled with ³H-AA with and without the presence of cytosolic phospholipase A₂ (cPLA₂) inhibitor (AACOCF³) and mitogen-activated protein (MAP) kinases (MEK) inhibitor (U0126). Western blot analysis was used to characterize the activated MAP kinase components in cell lysates or protein kinase C (PKC) translocation in isolated cytosolic and membrane fractions. Specific inhibitors to various enzymes were used in the study, including GF109203X for pan protein kinase C (PKC), AACOCF3 for cytosolic phospholipase A2 (cPLA₂), U0126 for MEK, and DPI for NADPH oxidase. Inhibitors for AA metabolism were also used to examine the role of AA in PDGF-stimulated ROS generation, including CDC and NDGA for pan lipoxygenase, AA861 for 5-lipoxygenase, indomethacin for cycloxygenase, and ketoconazole for cytochrome p450.

Results: We found that PDGF-stimulated ROS was eradicated by inhibitors to MEK, cPLA₂, 5-lipoxygenase, NADPH oxidase, or PKC. PDGF-stimulated AA release depended on both active cPLA₂ and ERK1/2. Exogenous AA showed a concentration-dependent ROS generation via NADPH oxidase activation that was insensitive to MEK inhibitor, but sensitive to PKC inhibitor, and could be attenuated by superoxide dismutase (SOD), mannitol, or DPI. This effect of AA was specific as other long chain fatty acids (leinoleic acid, stearic acid), or AA derivatives (eicosa-11Z, 14Z, 17Z-trienoic acid (20:3) and eicosa-11Z, 14Z-dienoic acid (20:2)) were ineffective. Inhibitor to lipoxygenase, in particular the 5-isoform, but not cycloxygenase or cytochrome p450, could diminish AA-stimulated luminescence generation. Western blot analysis showed that AA-treated cells transiently activated ERK1/2 and JNK, but not p38, in a time- and dose-dependent manner that was similar to that of PDGF. Finally, PDGF-stimulated PKC translocation depended on AA release while AA-stimulated PKC translocation was eradicated by lipoxygenase inhibition.

Conclusions: We conclude that PDGF signaling in HLE B3 cells is mediated by AA and its lipoxygenase metabolites, which provide a positive feedback loop for PDGF action, as AA and its metabolites can mobilize PKC and other factors needed for NADPH oxidase assembly and activation for ROS generation to facilitate cell proliferation. We further propose the role of AA in PDGF signaling.

Introduction

Reactive oxygen species (ROS), including superoxide anion (O₂^-), hydroxyl radical and hydrogen peroxide (H₂O₂), are known to be toxic and harmful molecules, that can lead to various degenerative diseases [1], including cataract [2-4]. However, low quantity of ROS has been shown lately to be beneficial to the cell and appears to mediate the physiological functions of growth factors and cytokines via redox signaling [5-8]. It has been found that ligand stimulation of nonphagocytic cells produces intracellular ROS [9-11]. For instance, platelet-derived browth factor (PDGF) stimulation of vascular smooth muscle cells displays transient ROS increase that is coincided with protein tyrosine phosphorylation [12]. Similar results were reported in other cell types [10,13,14], including lens epithelial cells [8,15]. It has been suggested that the membrane-bound NADPH oxidase may be the primary source for ROS used for cell signaling [16]. However, the mechanism(s) for the generation of endogenous ROS is not fully understood.

NADPH oxidase is known to associate with host-defense in phagocytic cells [17], but it is also found to be quite active in nonphagocytic cells [13,18]. One of the stimulants for NADPH oxidase is arachidonic acid (AA), which can be released from membrane phospholipids via cytosolic phospholipase A2 (cPLA₂), in response to a variety of stimuli, including activated ERK1/2 [13,19,20]. AA may activate NADPH oxidase by inducing conformational change in one of the subunits [21] or by recruiting Rac, a small GTPase, from cytosol to the membrane, and jointly activating the enzyme [22]. AA-activated NADPH oxidase has been demonstrated both in a cell-free system and in intact cells [13]. Studies also showed that AA could activate the signaling component of c-Jun NH₂-terminal kinase (JNK) to phospho-JNK, paralleled to superoxide anion generation [13,23,24].

The presence of NADPH oxidase and its components in lens epithelial cells has been recently reported by Rao et al. [15]. Our laboratory has demonstrated that PDGF can stimulate generation of ROS, which correlates with activated ERK1/2 and JNK cascades of MAP kinase, and cell proliferation in human lens epithelial cells [8]. However, the mechanisms of how the action of growth factor (such as PDGF) links to AA, how AA is being released, and how AA activates NADPH oxidase to produce ROS are not well understood. In this study, we examined the upstream factors that control AA release, and the downstream targets of AA that can lead to NADPH oxidase stimulation. We also investigated which of the metabolites of AA plays a major role in the signaling process. Our findings strongly indicate that the mitogenic function of PDGF is mediated through AA release, and that the released AA provides a positive feedback loop to PDGF by facilitating PKC and other key factors for NADPH oxidase stimulation and ROS generation. The generated ROS amplify activated ERK1/2 and JNK MAP Kinase to facilitate cell proliferation. A possible mechanism of PDGF-induced cell signaling is proposed.

Methods

Materials

Minimum Essential Medium Eagle (MEM), TC-199 organ culture medium, penicillin-streptomycin solution, and fetal bovine serum (FBS) were supplied by Invitrogen Corporation (Carlsbad, CA). HEPES, trypsin, AA, stearic acid (SA), linoleic acid (LA), nodihydroguaiaretic (NDGA), superoxide dismutase (SOD), mannitol, lucigenin, α-glycerphosphate, chymostatin, 3,4-DCI, E-64, leupeptin, pepstatin A, phenylmethylsulfonyl fluoride (PMSF), Na₃VO₄, and AA-carboxyl[¹⁴C] were obtained from Sigma-Aldrich (St. Louis, MO). [³H]-AA was obtained from Amersham Biosciences Inc. (Piscataway, NJ). AA-861 (5-lipoxygenase inhibitor), cinnamyl-3,4-dihydroxy-cyanocinnamate (CDC), nordihydroguaiaretic acid or NDGA (both lipoxygenase inhibitors), flavoprotein inhibitor diphenylene iodonium (DPI; an NADPH oxidase inhibitor), indomethacin (cycloxygenase inhibitor), ketoconazole (cytochrome p450 inhibitor), GF109203X (pan protein kinase C inhibitor), eicosa-(11Z,14Z)-dienoic acid (EDA), eicosa-(11Z,1Zz,17Z)-trienoic acid (ETA) were obtained from Biomol International, LP (Plymouth Meeting, PA). U0126 (MEK1/2 inhibitor or 1, 4-diamino-2, 3-dicyano-1, 4-bis [2-aminophenylthio] butadiene) was obtained from Cell Signaling Technology Inc. (Beverly, MA). M-Per, the mammalian protein extraction reagent, and protein assay reagent kit were purchased from Pierce Biotechnology, Inc. (Rockford, IL). Phospho-p42/44 MAPK (ERK1/2), phospho-JNK, phospho-p38 and phospho-MEK, pan P-PKC monoclonal antibodies were purchased from Cell Signaling Technology Inc. (Beverly, MA). Monoclonal anti-glyceraldehyde-3-phosphodehydrogenase (G3PD) antibody was obtained from Research Diagnostics Inc. (Flanders, NJ). Horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence (ECL) system components, anti-pan PKC rabbit polyclonal antibody were from Santa Cruz Biotechnology (Santa Cruz, CA). Hybond^TM-ECL^TM nitrocellulose membrane was from Amersham Biosciences Inc. (Piscataway, NJ). All other chemicals were of reagent grade.

Cell culture

Human lens epithelial cells (HLE) B3 were a gift from Dr. Usha Andley of Washington University (St. Louis, MO). Rabbit lens epithelial cells N/N 1003A were a gift from Dr. John Reddan of Oakland University (Rochester, MI). The cells were grown to confluence in MEM supplemented with 20% FBS and 5 μg/ml gentamicin, trypsinized, and gradually serum-starved by culturing in 2% FBS medium overnight. The cells were then washed with fresh MEM and incubated for 30 min in serum-free medium before used, following the methods described in Chen, et al. [8].

Quantitative image analysis of intracellular reactive oxygen species in live cells by confocal microscopy

HLE B3 cells (1.6 million) were gradually deprived of serum as described in the previous paragraph. The cells (no serum for 30 min) were then loaded with the fluorescent dye DCFDA (50 μM) for 5 min in the dark in a CO₂ incubator before stimulating with PDGF, and the fluorescence during the next 60 min was captured by confocal laser scanning microscope, following the method of Chen et al. [8]. In some experiments, cells were preloaded for 30 min with inhibitor either to cPLA₂ (AACOCF₃), pan PKC (GF 109203X), 5-lipoxygenase (AA861), MEK1/2, or NADPH oxidase (DPI) before PDGF stimulation. Cells without growth factor were used as controls. Data from five random images (90-100 cells per frame) within a given time were pooled and averaged to obtain the mean fluorescence intensities.

Stimulation of arachidonic acid release from fresh lens epithelial cells or cells in culture

A New Zealand white rabbit (2.5 kg) purchased from Harlen Sprague Dawley (Madison, WI) was sacrificed according to the Regulations of Association of Research in Vision and Ophthalmology Animal Users in Research. Lenses were removed from the eye globes surgically, and two epithelial layers were pooled and prelabeled with 0.5 μCi [¹⁴C]-AA in 1 ml TC 199 medium and incubated for 48 h in a CO₂ incubator. The layers were rinsed with PBS and placed in 1.0 ml unlabeled TC 199 medium for 30 min before addition of stimulant. Either EGF, IGF, or PDGF at 5 μg each was added to the medium and incubated with the prelabeled epithelial layers for up to 60 min. The culture medium (50 μl) was collected at 2, 5, 10, 20, 30, and 60 min and counted for radioactivity by a beta counter. Two unstimulated epithelial layers were pooled and used as the control.

HLE cells were also used for the studies of PDGF-induced AA release. Cells (1.3 million) were deprived of serum for 24 h as described in the previous paragraph. During the final 15 h, cells were incubated with 0.5 μCi/ml [³H]-AA. After incubation, unincorporated labeled AA was removed by washing thoroughly with MEM. These cells were then preloaded with AACOCF₃ (a specific inhibitor for cPLA₂) or U0126 (specific inhibitor for MEK1/2) in serum-free medium for 30 min before exposed to 50 ng/ml PDGF. Aliquots of media (100 μl) were collected at 0 (control), 2.5, 5, 10, and 20 min and counted for labeled AA release.

Treatment of cells with arachidonic acid

In dose-dependent studies, the serum-starved cells (1.6 million) described above were treated with 30, 60, 90, 120, or 150 μM arachidonic acid for 10 min. Cells treated with the same volume of ethanol (3% final concentration) were used as control. Serum starved cells were also used for time-dependent study and treated with 60 μM AA for 0, 2.5, 5, 10, 20, or 30 min.

Effect of arachidonic acid and its derivatives on superoxide anion production in intact lens epithelial cells

To examine AA-stimulated production of superoxide anion, serum-starved confluent cells (0.2 million) were washed twice with PBS, trypsinized in 37 °C for 5 min, and counted by a hemacytometer. Cell pellet was obtained by centrifugation at 200x g for 5 min at 4 °C, resuspended in HEPES-buffered saline (HBS) containing 130 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 1 mM CaCl₂, 35 mM phosphoric acid, and 20 mM HEPES (pH 7.4). The generation of superoxide was quantified using lucigenin-amplyfied chemiluminescence (LUCL) in the suspension of intact human lens epithelial (HLE B3) cells or rabbit lens epithelial cells (N/N 1003A). LUCL analysis is specific for superoxide anion since it can be induced effectively by superoxide anion but only weakly by H₂O₂ [25]. Measurement of LUCL was carried out by following a modified version of the method of Cui and Douglas [13]. Briefly, 2 million cells in suspension was mixed with lucigenin (0.5 mM final concentration), and the LUCL readings were recorded with a luminometer (LumiStar BMG) immediately upon addition of stimulant. The photoemission in terms of relative light units (RLU) was recorded automatically every 20 s for 10 min. Cells mixed with 3% ethanol (solvent for AA and other fatty acids) were used as the baseline. The stimulants tested included AA (C20: 4), AA derivatives of eicosa-(11Z, 14Z)-dienoic acid (C20: 2) and eicosa-(11Z, 14Z, 17Z)-trienoic acid (C20: 3), stearic acid (C18: 0) and linoleic acid (C18: 2).

The effect of antioxidants, and inhibitors to the arachidonic acid metabolic pathways on arachidonic acid-induced generation of luminescence in human lens epithelial cells

HLE B3 were trypsinized and resuspended in HEPES-buffered saline (HBS) as already described. Specific inhibitors to AA metabolic enzymes including cytochrome p450 inhibitor, ketoconazole (20 μM), cyclooxygenase inhibitor, indomethacin (20 μM), pan lipoxygenase inhibitor, CDC (20 μM), and 5-lipoxygenase inhibitor, AA861 (2 μM), were each incubated with cells for 30 min in HBS before AA stimulation and LUCL measurement. Some cells were pretreated with SOD or mannitol before AA stimulation. Cells mixed with 3% ethanol were used for baseline.

Platelet-derived growth factor or arachidonic acid-stimulated translocation of protein kinase C

Cells (2 million) were serum starved for 24 h, treated with either PDGF (20 ng/ml) or AA (60 μM) for 30 min, lysed by sonication in lysis buffer, containing HEPES (50 mM, pH 7.4), EDTA (1 mM), EGTA (1 mM), dithiothreitol (1 mM), and PMSF (0.2 mM). The lysate was then centrifuged (1,000x g) to remove cell debris, and the supernatant was saved for isolation of membrane and cytosolic fractions following the method of Lin et al. [26]. In brief, the supernatant from low speed centrifugation was centrifuged again at 100,000x g for 20 min at 4 °C. The supernatant (cytosol) was saved, and the pellet was resuspended in lysis buffer containing 1% Triton X-100 and centrifuged as before. The supernatant (membrane) fractions was saved and stored at -80 °C until immunoblot analysis was performed. In another experiment, the cells were pretreated with and without inhibitor either to 20 μM cPLA₂ (AACOCF₃) for 30 min or to 10 μM lipoxygenase (NDGA) for 60 min before stimulation by PDGF (20 ng/ml) or AA (60 μM), respectively. In all experiments, membrane and cytosolic fractions from unstimulated cells were used as controls.

Western blot analysis

A time-dependent (0, 2.5, 5, 10, 20, 30, or 60 min) or concentration-dependent (0, 30, 60, 90, 120, or 150 μM) study on AA-stimulated activation of MEK, ERK1/2, JNK, and p38 was carried out. Cells (1.6 million) were serum-starved, stimulated by AA, washed twice with ice-cold PBS and lysed in 500 μl lysis buffer (M-Per mammalian protein extraction reagent, containing 50 mM Tris-HCl, pH 7.4; 100 mM NaCl, 50 mM NaF, 5 mM EDTA, 40 mM β-glycerphosphate, 6 μg/ml chymostatin, 10 μM 3,4-DCI, 10 μM E-64, 1 μg/ml leupeptin, 1 μM pepstatin A, 10 μM PMSF, 200 μM Na₃VO₄). The soluble proteins (20 μg/well) were separated by 10% SDS-PAGE, and the gel was electro-blotted against Hybond^TM-ECL^TM nitrocellulose membrane, followed by probing with antiphosphorylated MEK, ERK1/2, JNK, p38, and G3PD antibodies, following the method described in Chen et al. [8]. The immunoblots were labeled with horseradish-labeled anti-mouse IgG and detected via enhancement with chemiluminescence reagent (ECL). Activation of PKC by PDGF was measured following the same procedure as described above.

Protein determination and statistical analysis

Protein concentration was determined by microanalysis of the BCA method [27]. The statistical analyses were done using Student's t-test. For all tests, a p<0.05 was considered significant.

Results

Confocal microscopic studies of the factors controlling platelet-derived growth factor-induced reactive oxygen species generation in human lens epithelial cells

It is known that cPLA₂ is responsible for AA release from the phospholipids in the membrane, thus inhibition of this enzyme may provide a clue as to whether PDGF-stimulated ROS generation is associated with AA or other factors. We pre-incubated HLE cells with a cPLA₂ specific inhibitor (AACOCF₃) before PDGF activation. Figure 1 shows that inhibition of cPLA₂ completely eradicates PDGF-stimulated ROS generation from HLEB3 cells. U0126 (MEK inhibitor), which inhibits ERK phosphorylation or stimulation also inhibits PDGF effect. Similarly, inhibiting 5-lipoxygenase by AA861 (a specific inhibitor to 5-lipoxygenase) also abolishes PDGF-induced ROS production significantly. Figure 1 also shows that PDGF-stimulated ROS generation is successfully eliminated when inhibitor to pan-PKC (GF109203X), or MEK (U0126), or NADPH oxidase (DPI) is preloaded into the cultured cells. These results indicate that PDGF-stimulated ROS generation depends on active enzymes of cPLA₂, 5-lipoxygenase, PKC, NADPH oxidase and the ERK MAPK signaling cascade.

Platelet-derived growth factor-stimulated arachidonic acid release in epithelial cells from cultured intact rabbit lens and in cultured human lens epithelial cells

In order to examine if AA plays any role in growth factor signaling in cells, we first exposed HLE to several growth factors and examined if these growth factors could stimulate AA release from cell membrane as a source for the endogenous AA pool. We used [¹⁴C]-AA to label two pooled epithelial layers from rabbit lens and cultured under organ culture conditions. The pre-labeled layers were then exposed to EGF, PDGF, or IGF at a physiological concentration (5 μg/ml). As shown in Figure 2A, each of the growth factors stimulated AA release into the culture medium as early as 5 min and reached plateau within 20 min. EGF has a stronger effect than PDGF and IGF but all are more efficient than the growth factor-free control.

HLE B3 cells (serum free) were also tested for AA release using PDGF (50 ng/ml) as a stimulant. As shown in Figure 2B, the control cells showed a gradual and fast release of AA with time, even in the absence of PDGF stimulation. The amount of labeled AA released into the medium from the unstimulated control cells was 2 fold at 2.5 min, 2.5 fold at 5 min, and 3 fold after 20 min in comparison to the zero time. However, in the presence of PDGF, the AA release was much more extensive throughout the experiment, in which AA release was 2.5 fold, 3.5 fold, and 4 fold at 2.5, 5, and 20 min, respectively, over the zero time. AA release in the unstimulated or stimulated cells was equally, and extensively inhibited in cells pretreated with cPLA₂ inhibitor, AACOCF₃, or a MEK inhibitor, U0126 (Figure 2B). Both inhibitors suppressed AA release to below the normal control level. This indicates that growth factors, including PDGF can stimulate AA release, and the stimulation is dependent on the presence of an active cPLA₂, an enzyme specifically cleaves AA from membrane phospholipids. ERK pathway apparently controls AA release, likely by regulating the bioavailability of activated cPLA₂.

Arachidonic acid stimulated superoxide anion generation in lens epithelial cells

The role of AA in stimulating NADPH oxidase was tested using intact HLE B3 cell suspension (0.2 million) in the presence of 30-150 μM AA. Production of superoxide occurred quickly within 30 s of AA addition, reached highest level by 240-300 s, and dissipated after 600 s (Figure 3A). The amount of superoxide anions, expressed as relative light unit (RLU) of luminescence, was proportional to the concentration of AA used. Figure 3A represents a typical pattern of superoxide anion production in the intact cell suspension when AA of 30, 60, 90, 120, and 150 μM were used. More ROS was produced when a higher number of cells was present (1.0 million; data not shown). AA-induced ROS generation was not species specific as a rabbit cell line (N/N 1003A) showed similar results (data not shown).

Specificity of arachidonic acid in generating superoxide anions

To prove that AA-induced ROS production is AA-specific, other long chain fatty acids and AA derivatives were used for comparison. As shown in Figure 3B, linoleic acid (18:2) at 120 μM and two AA derivatives, eicosa-11Z, 14Z, 17Z-trienoic acid (20:3) and eicosa-11Z, 14Z-dienoic acid (20:2), each at 90 μM showed a very weak effect while stearic acid (18:0) at 90 μM was ineffective in comparison to the strong luminescence generated by AA at the same concentration. This result indicates that AA-induced superoxide anion production is a specific physiological phenomenon.

Inhibition of superoxide anion generation by antioxidants

Since LUCL is mostly induced by superoxide anion and only weakly by H₂O₂ [25], we used superoxide dismutase (Mn SOD) preloaded cells (for 30 min) to test the presence of superoxide. Mannitol, a free radical scavenger, was also used for comparison after cells were preincubated with it for 30 min. As shown in Figure 3C, SOD at 0.4 mg/ml can suppress 50% superoxide production while mannitol at 20 mM can inhibit nearly 30% superoxide release.

To confirm that superoxide anion production in HLE B3 cells is generated from NADPH oxidase, we used a flavoprotein inhibitor DPI, which is known to inhibit NADPH oxidase. As shown in Figure 3C, about 50% of the AA-stimulated luminescence was produced in cells pretreated DPI (10 μM) in comparison with cells without DPI pretreatment. This inhibition was more efficient in rabbit lens epithelial cells under the same experimental conditions (data not shown).

Involvement of a specific metabolic pathway of arachidonic acid in stimulating superoxide anion production

AA is known to metabolize into various eicosanoids in several specific pathways, including cytochrome P-450 (produces 15-HETE), cyclooxygenase (produces prostaglandins), 15-lipoxygenase (produces 15-HPETE), 12-lipoxygenase (produces 12-HPETE) and 5-lipoxygenase (produces 5(S)-HPETE). We used inhibitors for each of these pathways to evaluate which of the AA metabolic pathway plays an important role. As shown in Figure 3D, AA-induced superoxide anion generation was not contributed by the metabolites either from cycloxygenase or cytochrome p450, as neither indomethacin (or Indo, 20 μM) nor ketoconazole (or Keto, 20 μM) was effective in preventing AA-stimulated superoxide production. Cinnamyl-3, 4-dihydroxy-cyanocinnamate (CDC), which is a specific inhibitor for 12-lipoxygenase at low concentration (IC₅₀ of 0.063 μM) but also inhibits 5- and 15-lipoxygenase at higher concentration (above 4 μM), extensively inhibited superoxide production when a high concentration of 20 μM was used (Figure 3D). Inhibition by CDC was concentration-dependent. At the peak time (240 s) of ROS generation, CDC at 2, 8, and 20 μM inhibited 5, 30, and 60% of the superoxide anion production, respectively (data not shown), thus it is likely that the inhibition was targeted at 5 and 15 lipoxygenases. However, AA-861 (a specific inhibitor of 5-lipoxygenase) completely blocked the luminescence production at 2 μM (Figure 3D). The inhibitory effect of AA-861 was observed at a concentration as low as 0.5 μM (data not shown). Therefore, it is likely that metabolite(s) from 5-lipoxygenase pathway is most essential for the activation of NADPH oxidase and the production of superoxide anion.

In N/N 1003A cells, CDC showed stronger inhibition on superoxide generation than AA-861 (data not shown). This is likely due to the fact that N/N 1003A cell has 12-lipoxygenase while HLE B3 cell has no 12-lipoxygenase (Zelenka P., personal communication, May, 2003), but may have both 5- and 15-lipoxygenase enzymes.

Involvement of protein kinase C and MAPK pathway in arachidonic acid stimulated superoxide anion production

The potential role of PKC in activating NADPH oxidase was explored by preloading the cells with a pan-PKC inhibitor, GF109203X (1-10 μM), for 30 min before subjecting the cells to AA stimulation. As shown in Figure 3E, GF109203X extensively inhibited the production of luminescence, and the inhibition was proportional to the amount of inhibitor used. Luminescence generation was eliminated nearly 70% in the presence of 10 μM GF109203X. In contrast, inhibitor to MEK, U0126, showed no effect (Figure 3E), contrast to the result shown in Figure 1 when PDGF was used as the stimulant. This suggests that PKC is downstream from AA and is involved in NADPH oxidase activation in the lens epithelial cells. Although activated ERK pathway is essential for PDGF-stimulated AA released for ROS generation (see Figure 1), however P-ERK is not essential when exogenous AA is supplied to the cells under this experimental condition.

Activation and translocation of protein kinase C

PDGF (1 ng/ml) stimulated transient activation of PKC in HLE B3 cells (Figure 4A) during the 60 min experimental period. The activation began about 10 min, reached maximum at 15 min, and subsided at 30 min. PDGF (20 ng/ml) also stimulated translocation of PKC in cells pre-exposed to PDGF for 30 min. As shown in Figure 4B, PKC is mainly present in the cytosolic fraction with negligible amount in the membrane fraction of unstimulated cells. After 30 min of PDGF exposure, a heavy protein band reactive to anti-PKC antibody can be seen in the membrane fraction. PDGF-stimulated PKC translocation was abolished when cPLA₂ was inactivated by AACOCF₃ in cells preloaded with this inhibitor for 60 min (Figure 4B). This indicates that PDGF-stimulated PKC translocation depends on the active cPLA₂ enzyme to release AA for mobilizing PKC from cytosol to the membrane.

Similar to PDGF, exogenous AA directly stimulated PKC translocation from cytosol to the membrane. As shown in Figure 4C, AA at 60 μM mobilizes a considerable amount of cytosolic PKC to the membrane after only 30 min of pre-AA treatment in comparison with the unstimulated control. However, the translocation is inhibited in cells pretreated with a pan lipoxygenase specific inhibitor, NDGA. The inhibitor not only prevented PKC translocation to occur in AA-treated cells, but also abolished any residual membrane PKC in the AA untreated cells. These results suggest that PDGF-stimulated translocation of PKC is dependent on the intracellular release of AA in HLE B3 cells, and likely the metabolites of AA in the lipoxygenase pathway play an important role in mobilizing PKC.

Activation of MAPK by exogenous arachidonic acid

To prove that AA stimulated free radical generation can subsequently influence the MAPK pathways, we conducted time- and concentration-dependent studies of AA in HLE B3 cells and examined the activation of signaling components in ERK and JNK pathways. The possible activation of the p38 stress signaling pathway was also examined for comparison. Western blot analysis of the cell lysate showed that AA stimulated ERK and JNK pathways but not the p38 pathway (Figure 5). AA (60 μM) stimulated activation of ERK was quick, appeared at 2.5 min, and reached optimum at 5 min. The transient activation subsided to the basal level by 20 min (Figure 5A). Activation of ERK was concentration-dependent (30-150 μM) as shown in Figure 5B. Similar transient activation of MEK, the upstream kinase of ERK, was observed under the same experimental conditions (Figure 5C).

AA also stimulated transient JNK activation in a concentration-dependent (Figure 5D) and time-dependent (Figure 5E) manner. The stimulation was extensive even at low concentration of 30 μM (Figure 5D). When AA at 60 μM was used, the stimulation started at 1 min, reached maximum at 5-10 min and returned to basal level by 30 min (Figure 5E). The stress-associated p38 was not affected under the same experimental conditions (Figure 5F). Equal intensity of G3PD shown in western blot of the same transblot indicated that the same amount of proteins was applied to the gels (Figure 5G). These data suggest that AA-stimulated ROS generation affects the downstream Raf-MEK-ERK cascade and the stress-associated JNK pathway for further physiological functions.

Discussion

We have demonstrated recently [8] that the mitogenic action of PDGF was closely associated with the production of ROS in human lens epithelial cells. The current study confirmed this finding and provided evidence that PDGF-induced ROS generation depended on AA release from the cells. Besides PDGF, we found that EGF and IGF also stimulated AA release in the intact rabbit lens. This supports the findings of Rao et al. [15] that all three growth factors could stimulate ROS generation in the lens epithelial cells. The growth factor-associated AA release is not species-specific as it was observed in both human and rabbit lens epithelial cells.

The results of the AA-stimulated generation of superoxide anions shown in HLE B3 cells were quite similar to that of kidney epithelial cells [13]. This stimulation was concentration-dependent, and specific to AA as other long chain fatty acids, either saturated or unsaturated, could not. Based on the inhibitory studies, it is likely that AA metabolites generated from 5-lipoxygenase pathway play a major role. That PKC inhibitor could eradicate both the AA-stimulated (Figure 3E), and PDGF-stimulated (Figure 1) ROS generation in lens epithelial cells, indicated that PKC was essential for NADPH oxidase activation, and that PKC acted downstream of AA and upstream of NADPH oxidase. Indeed, we observed in this study that PDGF not only stimulated PKC activation, but also its translocation. Furthermore, PKC translocation depended on the presence of an intracellular AA pool, either released endogenously via PDGF stimulation, or exogenously via AA enrichment (Figure 4C,D).

This study also demonstrated that the superoxide anions produced via AA stimulation was associated with NADPH oxidase as DPI effectively inhibited superoxide production. The nature of superoxide anion was confirmed by the inhibitory effect of SOD and mannitol. Recently, Rao et al. [15] demonstrated the presence of all components of NADPH oxidase in lens epithelial cells, similar to other cell types [15], but little is known how NADPH oxidase is stimulated and controlled.

In phagocytic cells, it is known that phosphorylation-induced conformational change of the p47^phox subunit of NADPH oxidase can activate the enzyme [28]. PKC has been demonstrated to play a role in the phosphorylation and translocation of p47^phox in several cell types [29,30]. Thus, it is reasonable to assume that PKC may play a similar role in the lens epithelial cells. However, it is not clear how AA or its metabolites from lipoxygenase pathway can mobilize PKC to the membrane, and also whether the membranes of nuclear and endothelial reticulum besides the plasma membrane are involved. Recent findings in rabbit lens epithelial cells [31] suggest that PKC can be specifically translocated to the plasma membrane by metabolites in the 12-lipoxygenase pathway. On the other hand, Sharma et al. [32] showed that it was the metabolites of 15-lipoxygenase pathway that mobilized PKCγ to the membrane of corneal epithelial cells during EGF stimulation. Based on our results in this study, we speculate that HLE B3 cell, which have no 12-lipoxygenase (Zelenka P., personal communication, May, 2003), may mobilize PKC to the membrane via metabolites of 5- or 15-lipoxygenase or both. However considering the extreme sensitivity of AA-stimulated superoxide anion production to 5-lipoxygenase inhibitor (Figure 3D), the metabolites of 5-lipoxygenase may play a major role.

The response of cells to PDGF stimulation appears to be modulated by the activation and translocation of cPLA₂, and the subsequent release of intracellular AA from the membrane lipids. Several factors, including ERK1/2 and calcium, are known to activate and translocate cPLA₂ in other cell types [33-36]. AA release could be seen even in unstimulated cells (Figure 2B). However, that AA release from both stimulated and unstimulated cells was completely eradicated in the presence of cPLA₂ or MEK inhibitor, indicating the ERK 1/2 and cPLA₂ are essential in AA release in HLE cells. ERK1/2 is likely involved in cPLA₂ function in HLE B3 cells as when ERK phosphorylation was prevented by MEK specific inhibitor (U0126), PDGF-stimulated AA release and ROS generation were both attenuated. However, U0126, was unable to eliminate exogenous AA stimulated-ROS generation, indicating that an activated ERK1/2 is required for cPLA₂ function and the downstream AA release, but not needed once the AA pool is enriched through exogenous source. Therefore, we propose that AA acts as a positive feedback to PDGF signaling, in which the low level of signals induced by activated PDGF receptor, such as activated ERK1/2, and calcium release are only enough to activate cPLA₂ for AA release, but not enough to exert their effect for the activation of membrane-bound NADPH oxidase. The released AA in turn recruits PKC and other factors, such as Rac, needed for NADPH oxidase assembly and activation. The mechanisms for activation of cPLA₂ and NADPH oxidase require further studies.

Our data also show that the ERK1/2 and JNK MAP kinase, but not p38, were activated through AA stimulation, similar to the PDGF effect reported earlier [8]. Therefore, these MAPK ERK1/2 and JNK signals downstream from AA can be directed for cellular function, including proliferation. It is also possible that AA-amplified P-ERK1/2 may facilitate further activation of cPLA₂ and AA release, providing another round of activated NADPH oxidase for ROS production. PDGF and AA stimulated activation of JNK appears to parallel the activation of ERK1/2, similar to the report of Cui and Douglas [13], except in this study we have demonstrated the in vivo JNK activation that is concentration- and time-dependent. The function of JNK activation is likely also associated with cell proliferation, as activated JNK is known to be bifunctional, both in cell proliferation and in cell apoptosis. What role JNK plays in PDGF stimulation in lens epithelial cells is not clear, and is worthy of further examination.

In conclusion, the role of AA in mediating PDGF signaling can be summarized based on the following evidence in this study. First, the PDGF-stimulated ROS production was eradicated by a cPLA₂ inhibitor, indicating that AA presence is required. Second, P-ERK1/2 inhibition led to complete elimination of PDGF-stimulated ROS generation but not AA-induced ROS formation, indicating that P-ERK1/2 is upstream from AA to facilitate AA release and the downstream AA-induced NADPH oxidase activation. Third, inactivation of NADPH oxidase by DPI not only inhibited PDGF-stimulated ROS generation but also abolished AA-stimulated superoxide anions in cells, indicating that NADPH oxidase is the main contributor to ROS. We therefore propose the signaling cascade of PDGF in the lens epithelial cells as depicted in Figure 6: PDGF first binds and stimulates PDGF receptor, which dimerizes and autophosphorylates to initiate a series of protein phosphorylation and dephosphorylation events downstream to activate ERK1/2. P-ERK1/2 in turn activates cPLA₂ and translocates cPLA₂ to the cell membrane to allow AA release from the membrane phospholipids. The metabolites of AA, either 5- or/and 15-lipoxygenase pathway, act on PKC and translocate it to NADPH oxidase complex, which in turn is activated to produce ROS. The ROS generated can stimulate ERK1/2 and JNK, which are translocated to the nucleus to facilitate targeted gene expressions for cell proliferation.

Acknowledgements

We thank Peggy Zelenka of the National Eye Institute for her advice and suggestions. We also appreciate Dr. You Zhou of the University of Nebraska for his assistance in the confocal microscopy studies. This research is supported by NIH grant RO1 EY10595 (MFL). Portions of this research were presented at the annual meeting for the Association for Research in Vision and Ophthalmology, Fort Lauderdale, FL, May, 2003. This research was in partial fulfillment of doctoral dissertations for YW and CWC.

References

1. Finkel T, Holbrook NJ. Oxidants, oxidative stress and the biology of ageing. Nature 2000; 408:239-47.

2. Augusteyn RC. Protein modification in cataract: possible oxidative mechanisms. In: Duncan G, editor. Mechanisms of cataract formation in the human lens. New York: Academic Press; 1981. p. 72-115.

3. Spector A. Oxidative stress-induced cataract: mechanism of action. FASEB J 1995; 9:1173-82.

4. Andley U, Liang J and Lou MF. Biochemical mechanisms of age-related cataract. In: Albert DM, Jakobiec FA, Azar DT, Gragoudas ES, Power SM, Robinson NL, editors. Principles and practice of ophthalmology. 2nd ed. Vol. 2. Philadelphia: Saunders; 2000. p. 1428-49.

5. Burdon RH. Superoxide and hydrogen peroxide in relation to mammalian cell proliferation. Free Radic Biol Med 1995; 18:775-94.

6. Cotgreave IA, Gerdes RG. Recent trends in glutathione biochemistry--glutathione-protein interactions: a molecular link between oxidative stress and cell proliferation? Biochem Biophys Res Commun 1998; 242:1-9.

7. Sen CK, Packer L. Antioxidant and redox regulation of gene transcription. FASEB J 1996; 10:709-20.

8. Chen KC, Zhou Y, Xing K, Krysan K, Lou MF. Platelet derived growth factor (PDGF)-induced reactive oxygen species in the lens epithelial cells: the redox signaling. Exp Eye Res 2004; 78:1057-67.

9. Meier B, Radeke HH, Selle S, Younes M, Sies H, Resch K, Habermehl GG. Human fibroblasts release reactive oxygen species in response to interleukin-1 or tumour necrosis factor-alpha. Biochem J 1989; 263:539-45.

10. Lo YY, Cruz TF. Involvement of reactive oxygen species in cytokine and growth factor induction of c-fos expression in chondrocytes. J Biol Chem 1995; 270:11727-30.

11. Krieger-Brauer HI, Kather H. Human fat cells possess a plasma membrane-bound H2O2-generating system that is activated by insulin via a mechanism bypassing the receptor kinase. J Clin Invest 1992; 89:1006-13.

12. Sundaresan M, Yu ZX, Ferrans VJ, Irani K, Finkel T. Requirement for generation of H2O2 for platelet-derived growth factor signal transduction. Science 1995; 270:296-9.

13. Cui XL, Douglas JG. Arachidonic acid activates c-jun N-terminal kinase through NADPH oxidase in rabbit proximal tubular epithelial cells. Proc Natl Acad Sci U S A 1997; 94:3771-6.

14. Griendling KK, Minieri CA, Ollerenshaw JD, Alexander RW. Angiotensin II stimulates NADH and NADPH oxidase activity in cultured vascular smooth muscle cells. Circ Res 1994; 74:1141-8.

15. Rao PV, Maddala R, John F, Zigler JS Jr. Expression of nonphagocytic NADPH oxidase system in the ocular lens. Mol Vis 2004; 10:112-21 <http://www.molvis.org/molvis/v10/a15/>.

16. Finkel T. Oxygen radicals and signaling. Curr Opin Cell Biol 1998; 10:248-53.

17. Chanock SJ, el Benna J, Smith RM, Babior BM. The respiratory burst oxidase. J Biol Chem 1994; 269:24519-22.

18. Lo YY, Wong JM, Cruz TF. Reactive oxygen species mediate cytokine activation of c-Jun NH2-terminal kinases. J Biol Chem 1996; 271:15703-7.

19. Piomelli D. Arachidonic acid in cell signaling. Curr Opin Cell Biol 1993; 5:274-80.

20. Sumida C, Graber R, Nunez E. Role of fatty acids in signal transduction: modulators and messengers. Prostaglandins Leukot Essent Fatty Acids 1993; 48:117-22.

21. Shiose A, Sumimoto H. Arachidonic acid and phosphorylation synergistically induce a conformational change of p47phox to activate the phagocyte NADPH oxidase. J Biol Chem 2000; 275:13793-801.

22. Abo A, Webb MR, Grogan A, Segal AW. Activation of NADPH oxidase involves the dissociation of p21rac from its inhibitory GDP/GTP exchange protein (rhoGDI) followed by its translocation to the plasma membrane. Biochem J 1994; 298:585-91.

23. Hii CS, Ferrante A, Edwards YS, Huang ZH, Hartfield PJ, Rathjen DA, Poulos A, Murray AW. Activation of mitogen-activated protein kinase by arachidonic acid in rat liver epithelial WB cells by a protein kinase C-dependent mechanism. J Biol Chem 1995; 270:4201-4.

24. Alexander LD, Cui XL, Falck JR, Douglas JG. Arachidonic acid directly activates members of the mitogen-activated protein kinase superfamily in rabbit proximal tubule cells. Kidney Int 2001; 59:2039-53.

25. Li Y, Zhu H, Kuppusamy P, Roubaud V, Zweier JL, Trush MA. Validation of lucigenin (bis-N-methylacridinium) as a chemilumigenic probe for detecting superoxide anion radical production by enzymatic and cellular systems. J Biol Chem 1998; 273:2015-23.

26. Lin D, Zhou J, Zelenka PS, Takemoto DJ. Protein kinase Cgamma regulation of gap junction activity through caveolin-1-containing lipid rafts. Invest Ophthalmol Vis Sci 2003; 44:5259-68.

27. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC. Measurement of protein using bicinchoninic acid. Anal Biochem 1985; 150:76-85. Erratum in: Anal Biochem 1987; 163:279.

28. Groemping Y, Lapouge K, Smerdon SJ, Rittinger K. Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 2003; 113:343-55.

29. Bey EA, Xu B, Bhattacharjee A, Oldfield CM, Zhao X, Li Q, Subbulakshmi V, Feldman GM, Wientjes FB, Cathcart MK. Protein kinase C delta is required for p47phox phosphorylation and translocation in activated human monocytes. J Immunol 2004; 173:5730-8.

30. Kitada M, Koya D, Sugimoto T, Isono M, Araki S, Kashiwagi A, Haneda M. Translocation of glomerular p47phox and p67phox by protein kinase C-beta activation is required for oxidative stress in diabetic nephropathy. Diabetes 2003; 52:2603-14.

31. Zhou J, Fariss RN, Zelenka PS. Synergy of epidermal growth factor and 12(S)-hydroxyeicosatetraenoate on protein kinase C activation in lens epithelial cells. J Biol Chem 2003; 278:5388-98.

32. Sharma GD, Ottino P, Bazan NG, Bazan HE. Epidermal and hepatocyte growth factors, but not keratinocyte growth factor, modulate protein kinase Calpha translocation to the plasma membrane through 15(S)-hydroxyeicosatetraenoic acid synthesis. J Biol Chem 2005; 280:7917-24.

33. Lin LL, Wartmann M, Lin AY, Knopf JL, Seth A, Davis RJ. cPLA2 is phosphorylated and activated by MAP kinase. Cell 1993; 72:269-78.

34. van Rossum GS, Klooster R, van den Bosch H, Verkleij AJ, Boonstra J. Phosphorylation of p42/44(MAPK) by various signal transduction pathways activates cytosolic phospholipase A(2) to variable degrees. J Biol Chem 2001; 276:28976-83.

35. Xu J, Weng YI, Simonyi A, Krugh BW, Liao Z, Weisman GA, Sun GY, Simoni A. Role of PKC and MAPK in cytosolic PLA2 phosphorylation and arachadonic acid release in primary murine astrocytes. J Neurochem 2002; 83:259-70. Erratum in: J Neurochem 2002; 83:1239.

36. Hayama M, Inoue R, Akiba S, Sato T. ERK and p38 MAP kinase are involved in arachidonic acid release induced by H(2)O(2) and PDGF in mesangial cells. Am J Physiol Renal Physiol 2002; 282:F485-91.