Role of PKC and MAPK in cytosolic PLA2 phosphorylation and arachadonic acid release in primary murine astrocytes

Authors


Address correspondence and reprint requests to Grace Y. Sun, M121 Medical Sciences Building, Biochemistry Department, University of Missouri-Columbia, Columbia, MO, USA. E-mail: sung@health.missouri.edu

Abstract

Although Group IV cytosolic phospholipase A2 (cPLA2) in astrocytes has been implicated in a number of neurodegenerative diseases, mechanisms leading to its activation and release of arachidonic acid (AA) have not been clearly elucidated. In primary murine astrocytes, phorbol myristate acetate (PMA) and ATP stimulated phosphorylation of ERK1/2 and cPLA2 as well as evoked AA release. However, complete inhibition of phospho-ERK by U0126, an inhibitor of mitogen-activated protein kinase kinase (MEK), did not completely inhibit PMA-stimulated cPLA2 and AA release. Epidermal growth factor (EGF) also stimulated phosphorylation of ERK1/2 and cPLA2[largely through a protein kinase C (PKC)-independent pathway], but EGF did not evoke AA release. These results suggest that phosphorylation of cPLA2 due to phospho-ERK is not sufficient to evoke AA release. However, complete inhibition of ATP-induced cPLA2 phosphorylation and AA release was observed when astrocytes were treated with GF109203x, a general PKC inhibitor, together with U0126, indicating the important role for both PKC and ERK in mediating the ATP-induced AA response. There is evidence that PMA and ATP stimulated AA release through different PKC isoforms in astrocytes. In agreement with the sensitivity of PMA-induced responses to PKC down-regulation, prolonged treatment with PMA resulted in down-regulation of PKCα and ε in these cells. Furthermore, PMA but not ATP stimulated rapid translocation of PKCα from cytosol to membranes. Together, our results provided evidence for an important role of PKC in mediating cPLA2 phosphorylation and AA release in astrocytes through both ERK1/2-dependent and ERK1/2-independent pathways.

Abbreviations used
AA

arachidonic acid

bFGF

basic fibroblast growth factor

BSA

bovine serum albumin

cPLA2

cytosolic phospholipase A2

EGF

epidermal growth factor

ERK

extracellular signal-regulated kinase

HELSS

haloenol lactone suicide substrate

MAFP

methyl arachidonyl fluorophosphonate

MAPK

mitogen-activated protein kinase

MEK

mitogen-activated protein kinase kinase

PKC

protein kinase C

PMA

phorbol myristate acetate.

Stimulation of arachadonic acid (AA) release and subsequent generation of eicosanoid metabolites have been shown to play an important role in the pathogenesis of a number of neurodegenerative diseases (Farooqui et al. 1997a, 1997b; Montine et al. 1999). Among several biochemical pathways involved in AA metabolism, e.g. the diacylglycerol lipase pathway, many studies have focused on PLA2, enzymes that are ubiquitously expressed in mammalian cells (Murakami et al. 1997; Six and Dennis 2000). The Group IV cPLA2 is of special interest because this enzyme is regulated by phosphorylation and a Ca2+-dependent translocation mechanism (Schievella et al. 1995; Leslie 1997; Murakami et al. 1997; Qiu et al. 1998; Gijon et al. 1999). This form of PLA2 is present in astrocytes (Stephenson et al. 1994), and increased expression has been associated with a number of neurodegenerative diseases, including Alzheimer's disease (Clemens et al. 1996; Stephenson et al. 1996). Despite its role in neurodegeneration, the intracellular signaling pathways regulating cPLA2 in glial cells have not been examined in detail.

The presence of a consensus phosphorylation site for mitogen-activated protein kinase (MAPK) at Ser505 of cPLA2 has led to studies demonstrating the role of MAPKs, such as the ERKs and p38 MAPK, in the phosphorylation and activation of this enzyme (Lin et al. 1993b; Qiu and Leslie 1994; Kramer et al. 1996; Börsch-Haubold et al. 1999; Husain and Abdel-Latif 1999; Geijsen et al. 2000; Gijon et al. 2000; Hefner et al. 2000). However, several studies also implicated a role for protein kinase C (PKC) in the regulation of cPLA2 (Nemenoff et al. 1993; Qiu and Leslie 1994; Xing and Insel 1996; Xing et al. 1997; Husain and Abdel-Latif 1998). Although phorbol myristate acetate (PMA) has been shown to stimulate AA release in astrocytes (Chen and Chen 1998), the precise relationship between PKC and the MAPK pathways in activating cPLA2 is still largely unknown.

A number of studies have demonstrated the presence of G protein-coupled P2Y purinergic (nucleotide) receptors in astrocytes (Kastritsis et al. 1992; King et al. 1996; Centemeri et al. 1997). Stimulation of astrocytes by P2Y receptor agonists has been linked to AA release and prostanoid synthesis (Pearce et al. 1989; Bruner and Murphy 1990, 1993; Seregi et al. 1992; Stella et al. 1997; Chen and Chen 1998). Recent studies further demonstrated that, besides coupling to the poly-phosphoinositide pathway (Kastritsis et al. 1992; Salter and Hicks 1995), activation of P2Y2 receptors by ATP/UTP also led to stimulation of the MAPK cascade (Neary et al. 1998, 1999; Soltoff 1998; Weisman et al. 1998, 1999). In this study, we investigated the signaling pathways stimulated by PMA and ATP on cPLA2 phosphorylation and AA release in murine primary astrocytes. Our results demonstrated an important role of PKC and MAPK in regulating phosphorylation of cPLA2 and AA release in these cells.

Materials and methods

Materials

Dulbecco's modified Eagle's medium (DMEM), α-minimum essential medium (α-MEM), Ham's F12 nutrient mixture, 100 × N2 supplement, trypsin–EDTA, recombinant human basic fibroblast growth factor (rhbFGF), and recombinant human epidermal growth factor (rhEGF) were purchased from Gibco-BRL (Gaithersburg, MD, USA). ATP and UTP were purchased from Pharmacia Biotech (Piscataway, NJ, USA). Phorbol-12-myristate-13-acetate (PMA), GF109203x, Gö6976, methyl arachidonyl fluorophosphonate (MAFP), 12-epi-scalaradial, and haloenol lactone suicide substrate (HELSS) were purchased from Calbiochem (La Jolla, CA, USA). U0126 was obtained from Promega (Madison, MI, USA). Bovine serum albumin (BSA), 4-α-phorbol 12-myristate 13-acetate (4-α-PMA), sodium pyruvate, β-nicotinamide adenine dinucleotide reduced form (β-NADH), BIOMAX MR films, and all other chemicals were purchased from Sigma (St Louis, MO, USA). [14C]Arachidonic acid (50 mCi/mmol) was purchased from NEN Life Science Products (Boston, MA, USA). Rabbit polyclonal anti-human cPLA2 antibody, rabbit polyclonal anti-rat ERK1/2 antibody, horseradish peroxidase (HRP)-linked goat anti-rabbit IgG and HRP-linked goat anti-mouse IgG were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Mouse monoclonal anti-human phospho-ERK1/2 antibody was obtained from New England Biolabs (Beverly, MA, USA). PKC Sampler Kit including mouse monoclonal antibodies against PKC α, β, γ, δ, ɛ, η, τ, ι, λ, ζisoforms was purchased from BD Transduction Laboratories (San Diego, CA, USA). Mouse monoclonal anti-PKC and β-actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). DC protein assay kit, and pure nitrocellulose membrane (0.45 µm) were obtained from Bio-Rad Laboratories (Hercules, CA, USA). LumiGLO chemiluminescent substrate was obtained from Kirkegaard & Perry Laboratories (Gaithersburg, MD, USA).

Primary mouse astrocyte cultures

Adult male and female C57BL/6 J mice were obtained from Harlan (Indianapolis, IN, USA) and a breeding colony was maintained at the Laboratory Animal Facilities of the University of Missouri-Columbia Health Sciences Center. Cortical astrocytes were prepared from 1- to 3-day post-natal pups, using methods described by McCarthy and de Vellis (1980) with slight modifications. Briefly, cerebral cortices were dissected and meninges were removed. The tissues were minced and then suspended in 0.05% (w/v) trypsin and 0.53 mm EDTA for 10 min at 37°C. The suspension was passed through a 20-gauge needle five times. Cells were filtered through 85 µm nylon mesh, sedimented by centrifugation, suspended in 10% (v/v) fetal bovine serum (FBS) in DMEM containing 100 IU/mL penicillin and 100 µg/mL streptomycin, and transferred to culture flasks. Medium was changed after 24 h and twice weekly thereafter. When cells became confluent (around 1 week), the flasks were shaken at 225 rpm on an orbital shaker (Fisher Scientific, Pittsburgh, PA, USA) for 4 h at room temperature (22°C) to remove microglial cells. After shaking, cells were rinsed three times with phosphate-buffered saline (PBS), suspended in trypsin-containing solution as above, and subcultured at 3 × 105 cells/35-mm dish. These cell cultures contained over 95% astrocytes, as determined by immunostaining for glial fibrillary acidic protein. Experiments were performed with cells at approximately 90% confluence on 35-mm dishes. Prior to each experiment, cells were routinely examined under the microscope (Nikon DIAPHOT 300) for cell morphology and contamination. In experiments to test cytotoxic effects of inhibitors, the release of lactate dehydrogenase (LDH) into the culture medium was assayed according to a procedure described by Xue et al. (1999).

Measurement of AA release

The protocol for measurement of AA release from [1−14C]-AA-labeled cells was essentially described by Xue et al. (1999). Briefly, primary astrocytes cultured in DMEM containing 0.5% (w/v) bovine serum albumin (BSA) were incubated with 0.1 µCi of [14C]-AA (50 Ci/mol) per 35-mm dish for 4 h. During this time period, greater than 90% of the labeled AA was incorporated into cell membrane phospholipids (Xue et al. 1999). After incubation, unincorporated labeled AA was removed by washing twice with DMEM containing 0.5% BSA. Cells were then equilibrated in DMEM containing 0.5% BSA at room temperature for 1 h and inhibitors were added after 30 min. Cells were stimulated with PMA, ATP, UTP, or growth factors at 37°C for 30 min. After stimulation, the culture media were transferred to Eppendorf tubes, centrifuged at 12 000 g for 10 min and supernatant was taken for determination of radioactivity using a Beckman LS5800 liquid scintillation counter. Radioactivity in the cells was measured after adding methanol (1 mL) to the dish and then scraping the cells into scintillation vials. Controls were dishes incubated with vehicle without inhibitors or stimulating agents. Also, radioactivity of cells and medium without 30-min incubation with stimulating agents were obtained and regarded as basal radioactivity. In each experiment, basal radioactivity in the medium was subtracted to obtain net release of labeled AA during the 30-min incubation with or without agonists. Total incorporation of radioactivity was defined as the sum of radioactivity in the medium and the cells, and the amount of labeled AA released into the medium was expressed as a percentage of the total.

Western blot analyses

Astrocytes were incubated, as described in the figure legends, in serum-free medium containing 1 : 1 (v/v) α-MEM/Ham's F12 nutrient mixture with 1 × N2 supplement. Cell lysates prepared in sample buffer heated to 100°C, as above, were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) under reducing conditions, using methods described previously (Xue et al. 1999). For immunoblotting of cPLA2, 40 µg of sample protein were separated in a 20-cm 10% polyacrylamide gel at 40 mA/gel for 6 h at room temperature using a PROTEIN II xi electrophoresis cell (Bio-Rad). For immunoblotting of ERK1/2, 20 µg of sample protein were resolved in a 10% polyacrylamide mini-gel at 125 V for 1.5 h at room temperature using a Mini-PROTEIN II electrophoresis cell (Bio-Rad). After electrophoresis, proteins were electro-blotted onto nitrocellulose membrane. The membrane was incubated for 1 h at room temperature in TBS-T solution [10 mm Tris–HCl (pH 7.6), 150 mm NaCl and 0.1% (v/v) Tween-20] containing 5% (w/v) non-fat dry milk (NFDM). The membrane was subsequently incubated for 2 h at room temperature with a 1 : 500 dilution of mouse anti-human phospho-ERK1/2 antibody, a 1 : 500 dilution of rabbit anti-rat ERK1/2 antibody, or a 1 : 250 dilution of rabbit anti-human cPLA2 antibody in TBS-T solution containing 5% NFDM. After three washes with TBS-T, the membrane was incubated for 1 h at room temperature with a 1 : 2000 dilution of HRP-linked goat anti-mouse IgG or HRP-linked goat anti-rabbit IgG in TBS-T containing 5% NFDM. The protein bands were visualized on autoradiographic film using the Lumi GLO chemiluminescent substrate. For cPLA2, the electrophoretic mobility of phosphorylated cPLA2 was slightly retarded as compared to the mobility of the non-phosphorylated cPLA2. The images in the autoradiographs were determined using a Lazer Densitometer Ultro Scan XL (Pharmacia LKB Biotechnology, Sweden).

PKC down-regulation and translocation of PKC isoforms

To examine PKC isoforms expressed in murine astrocytes, cells were lysed in sample buffer and proteins in lysates were applied to 10% SDS–polyacrylamide gels cast with Prep/2D combs in a Mini-PROTEAN 3 Cell (Bio-Rad). After transfer of proteins, the nitrocellulose membranes were mounted to the Mini-PROTEAN II Multiscreen Apparatus (Bio-Rad) and incubated with primary antibodies against specific PKC isoforms, including PKC α, β, γ, δ, ɛ, η, τ, ι, λ and ζ isoforms with concentrations suggested by the manufacture. The remaining blotting procedures were similar as described above in the western blot analyses section.

To examine the translocation of PKC isoforms, we isolated the membrane and the cytosol fractions of the cell lysates using previously described procedures with slight modifications (Qiu et al. 1998). Astrocytes were scraped and sonicated for 10 s in ice-cold buffer containing 10 mm HEPES (pH 7.4), 1 mm EDTA, 1 mm EGTA, 0.34 m sucrose, 10% glycerol, 10 mg/mL leupeptin, 10 mg/mL aprotinin, 1 mm benzamidine and 1 mm phenylmethylsulfonyl fluoride. Membrane and cytosol were then separated by centrifugation at 100 000 g at 4°C for 1 h. Cell membranes were lysed with the same buffer supplemented with 1% Triton X-100. Proteins in each fraction were analyzed by western blot as described above.

Statistics

Data were analyzed by one-way anova followed by Bonfferoni's multiple comparison. All analyses were performed using the GraphPad Prism program (V3.0, GraphPad Software Inc., San Diego, CA, USA).

Results

ATP/UTP or PMA stimulated AA release from primary murine astrocytes: effects of PLA2 inhibitors

Stimulation of [14C]-AA-labeled primary murine astrocytes with ATP or UTP resulted in a dose-dependent increase in the release of labeled AA into the culture medium (Fig. 1a). AA release was also stimulated by PMA, reaching a plateau around 50 nm (Fig. 1b). In subsequent studies, stimulations were carried out using 100 µm ATP or 100 nm PMA. Treatment of astrocytes with 4-α-PMA, an inactive analog of PMA, failed to stimulate AA release (Fig. 1b, inset).

Figure 1.

ATP-, UTP- and PMA-stimulated AA release from primary murine astrocytes. Astrocytes labeled with [14C]-AA for 4 h at 37°C were washed and treated with different doses of (a) ATP (●), UTP (○), or (b) PMA for 30 min in a medium containing DMEM with 0.5% (w/v) BSA. Release of [14C]-AA was measured as described under Materials and methods and results are expressed as a percentage of the total radioactivity incorporated into the cells. In most experiments, [14C]-AA incorporated into the cells ranged around 7–10 × 105 dpm and the amount of radioactivity in the culture medium prior to stimulation with agonists ranged around 3–5% of total. Similar results with 100 µm ATP or UTP were obtained from two other experiments (data not shown). Inset: Results to indicate that 4-α-PMA is an inactive isomer of PMA.

In order to identify the subtypes of PLA2 involved in PMA- and ATP-stimulated AA release, astrocytes were pre-treated with inhibitors for cPLA2, sPLA2 or iPLA2. As shown in Fig. 2, methyl arachidonoyl fluorophosphonate (MAFP) 20 µm, an irreversible inhibitor of both cPLA2 and iPLA2 (Balsinde and Dennis 1996; Lio et al. 1996), inhibited PMA- or ATP-stimulated AA release by more than 75%. Increasing the concentration of MAFP to 50 µm did not further inhibit PMA- or ATP-stimulated AA release (data not shown). We also tested scalaradial (2 µm), an inhibitor of sPLA2 (Barnette et al. 1994), and HELSS (20 µm), an inhibitor of iPLA2 (Lehman et al. 1993; Balsinde and Dennis 1996; Yang et al. 1999). As shown in Fig. 2, neither scalaradial nor HELSS caused significant inhibition of PMA- or ATP-induced AA release in these cells.

Figure 2.

Effects of PLA2 inhibitors on PMA- or ATP-stimulated AA release in primary astrocytes. [14C]-AA-labeled astrocytes were pre-treated with vehicle (control), MAFP (20 µm), scalaradial (2 µm) or HELSS (20 µm) for 30 min at 37°C prior to stimulation with PMA (100 nm) or ATP (100 µm) for 30 min at 37°C. Net release of labeled AA due to PMA or ATP stimulation was determined as described under Materials and methods and the effects of inhibitors were expressed as a percentage of the net stimulated release without inhibitors (control). Data are expressed as the mean ± SEM of results from three independent experiments with each condition performed in triplicate. *p < 0.001, compared to control.

PMA-, ATP-, and EGF-activated ERK1/ERK2 MAPK cascade

The ability of PMA and ATP to stimulate phosphorylation of ERK1/2 in primary astrocytes was investigated using immunodetection of phospho- and total ERK1/2 proteins. Primary murine astrocytes showed both ERK1 and ERK2 (Fig. 3). However, upon treatment with PMA (100 nm), ATP (100 µm) or EGF (10 ng/mL), more ERK2 became phosphorylated than ERK1 (Fig. 3). Pre-treatment of astrocytes with U0126 (20 µm) completely abolished ERK1/2 phosphorylation induced by all three agents (Fig. 3). On the other hand, pre-treatment of astrocytes with GF109203x (10 µm), a general PKC inhibitor (Wilkinson et al. 1993), completely inhibited phospho-ERK2 stimulated by PMA but only partially inhibited the phospho-ERK2 stimulated by ATP or EGF (Fig. 3b). SB203580 (10 µm), an inhibitor of p38 MAPK, did not significantly alter phospho-ERK1/2 caused by PMA, ATP or EGF (Fig. 3b).

Figure 3.

Effects of kinase inhibitors on ERK1/2 phosphorylation stimulated by PMA, ATP, and EGF. Astrocytes were pre-treated with vehicle (control), U0126 (20 µm), SB203580 (10 µm) or GF109203x (10 µm) for 30 min at 37°C followed by incubation with vehicle, PMA (100 nm), ATP (100 µm) or EGF (10 ng/mL) for 20 min at 37°C. Cell lysates were prepared in hot sample buffer, and western blot analyses of phospho-ERK1/2 and total ERK1/2 were performed as described under Materials and methods. (a) The autoradiographs are representative of results from two to three independent experiments. (b) Densitometry of ERK2 bands from the autoradiograph in (a) expressed in arbitrary units (a.u).

Growth factors did not stimulate AA release in astrocytes

Using [14C]AA-labeled astrocytes, we examined whether growth factors, which are known to stimulate ERK1/2 phosphorylation, could induce AA release. As shown in Fig. 4(a), neither EGF nor bFGF could stimulate AA release. Furthermore, pre-treatment of astrocytes with EGF (20 min) prior to stimulation with PMA or ATP did not further enhance the levels of AA release (Fig. 4b). These results indicate that, despite stimulation of phospho-ERK1/2 by growth factors, these agents could not elicit AA release.

Figure 4.

Growth factors do not stimulate AA release in primary murine astrocytes. (a) [14C]-AA-labeled astrocytes were stimulated with vehicle, PMA (100 nm), bFGF (10 ng/mL), or EGF (10 ng/mL) for 30 min at 37°C. (b) [14C]-AA-labeled astrocytes were pre-treated with vehicle or EGF (10 ng/mL) for 20 min at 37°C and then incubated with vehicle, PMA (100 nm) or ATP (100 µm) for 30 min at 37°C. Release of [14C]-AA was expressed as the –fold increase over the response of untreated (control) cells (fold of control). Data are expressed as the mean ± SEM of results from three independent experiments performed in triplicate. *p < 0.001, compared to control.

PKC and MAPK inhibitors on ATP- and PMA-stimulated AA release

Prolonged exposure of cells to PMA has been shown to cause down-regulation of some PKC isoforms (Webb et al. 2000). In this study, astrocytes were pre-treated with PMA (100 nm) for 6, 12 or 24 h, labeled with [14C]-AA, and stimulated with PMA or ATP. Prolonged PMA treatment completely inhibited the ability of astrocytes to respond to PMA is mediating AA release and this occurred as early as 6 h (Fig. 5a). On the other hand, similar treatment conditions only slightly altered ATP-stimulated AA release (Fig. 5a and Table 1). Treatment of astrocytes (for 24 h) with 4-α-PMA, the inactive analog of PMA, did not alter the ability of PMA or ATP to stimulate AA release (Fig. 5b). Astrocytes treated with PMA became highly differentiated and more actively incorporated radioactive AA as compared to the non-treated controls (data not shown). Nevertheless, the increased incorporation of [14C] AA into phospholipids did not influence the AA release response.

Figure 5.

Effects of PKC down-regulation on AA release. (a) Astrocytes were incubated with or without 100 nm PMA for 6, 12, 18 or 24 h. Four hours prior to the end of PMA treatment, [14C]-AA was added to each dish and cells were labeled for 4 h. After this time, cells were washed and then stimulated with vehicle (◊, control), PMA (●, 100 nm) or ATP (□, 100 µm) for 30 min at 37°C. (b) Astrocytes were treated with or without 4-α-PMA for 20 h and followed by labeling with [14C]-AA for 4 h. Cells were washed and subsequently stimulated with PMA, ATP and 4-α-PMA for 30 min at 37°C. Release of [14C]-AA was expressed as percentage of total radioactivity after subtracting basal. Data are the mean ± SD of one experiment performed in triplicate. Similar results with PMA treatment for 24 h were observed in five other experiments (Table 1).

Table 1.  Effects of PKC and MAPK inhibitors on PMA- or ATP-stimulated AA release in primary astrocytes
Pre-treatmentPMAATP
MeanSEMnMeanSEMn
  1. Astrocytes were labeled with [14C]-AA for 4 h at 37°C, washed and pretreated with vehicle (control) or inhibitors for 30 min at room temperature, prior to stimulation with vehicle (control), PMA (100 nm) or ATP (100 µm) for 30 min at 37°C. To down-regulate PKC, astrocytes were treated with PMA (100 nm) or vehicle (control) for 20 h at 37°C. Cells were labeled with [14C]-AA for 4 h, washed and then stimulated with PMA and ATP as described in text. Net stimulated AA release was determined as described under ‘Materials and Methods’ and effects of inhibitors were expressed as a percentage of the net stimulated AA release without inhibitors (control). Data are expressed as the mean ±S.E.M. of results from the indicated number (N) of independent experiments with each condition performed in triplicate. aDenotes significant difference (p < 0.05) comparing treatment groups with controls based on one-way anova followed by the Bonfferoni post-test.

Control100.0100.0
PMA (100 nm, 24 h) 27.1a1.33 86.0a2.53
GF109203x (10 μm, 30 min)  1.6a4.27 38.9a4.17
Gö6976 (10 μm, 30 min) 27.6a4.73 94.64.13
U0126 (20 μm, 30 min) 50.7a5.68 59.3a6.05
SB203580 (10 μm, 30 min) 70.7a3.93 86.03.23

GF109203x was used to test for the involvement of PKC in AA release stimulated by PMA and ATP. As shown in Fig. 6(a), GF109203x at 1 µm completely inhibited PMA-stimulated AA release, whereas inhibition of ATP-stimulated AA release occurred at much higher concentrations of GF109203x (Fig. 6a). Gö6976 (10 µm), an inhibitor for the Ca2+-dependent PKC isoforms (Martiny-Baron et al. 1993), inhibited PMA-stimulated AA release by 72% but did not significantly alter ATP-stimulated AA release (Table 1). Under these treatment conditions, assay of LDH release revealed no apparent cytotoxicity (data not shown).

Figure 6.

Effects of PKC and MAPK inhibitors on AA release. [14C]-AA-labeled astrocytes were pre-treated for 30 min without (control) or with indicated doses of (a) GF109203x or (b) U0126 and then stimulated with vehicle, PMA (100 nm) or ATP (100 µm) for 30 min at 37°C. Net stimulated release of labeled AA was determined as described under Materials and methods, and the effects of inhibitors were expressed as a percentage of the net stimulated release without inhibitors (control). Data are expressed as the mean ± SEM from three independent experiments performed in triplicate. *p < 0.01, **p < 0.001, compared to control.

The MEK inhibitor, U0126, was used to examine effects of ERK1/2 on PMA- and ATP-stimulated AA release. As shown in Fig. 6(b), U0126 (1–20 µm) partially inhibited PMA- or ATP-stimulated AA release and the inhibition was not dose-dependent. This partial inhibitory effect of U0126 was observed in at least five other experiments (Table 1). Under these treatment conditions, assay of LDH release revealed no apparent cytotoxicity due to treatment with 20 µm of U0126 (data not shown). In the initial phase of our studies, we also tested the effects of PD098059, another MEK inhibitor. Our results indicated that, despite the ability for PD098059 to inhibit phospho-ERK, this inhibitor altered the AA release process resulting in high basal activity. Because recent studies with U0126 indicated that this inhibitor could inhibit MEK1 and -2 with potency 100-fold higher than PD098059 (Favata et al. 1998), U0126 was subsequently used in our studies. Treating astrocytes with SB203580 (10 µm), an inhibitor of p38 MAPK, resulted only in small inhibitions of PMA- and ATP-stimulated AA release (Table 1). These results demonstrated that although both PMA and ATP can stimulate phosphorylation of ERK1/2, complete inhibition of phospho-ERK1/2 did not abolish AA release stimulated by either agent.

PMA-, ATP- and EGF-stimulated cPLA2 phosphorylation: response to kinase inhibitors

Exposure of astrocytes to PMA (100 nm) or ATP (100 µm) for 20 min resulted in a complete mobility shift of the non-phosphorylated form of cPLA2 (lower band) to the phosphorylated form (upper band; Fig. 7a). Pre-treatment of astrocytes with U0126 reduced PMA-, ATP-, and EGF-stimulated cPLA2 phosphorylation, with greatest effect on EGF. Pre-treatment of astrocytes with GF109203x (10 µm) completely inhibited PMA-induced cPLA2 phosphorylation, but only partially inhibited ATP- or EGF-stimulated cPLA2 phosphorylation (Fig. 7a). Unlike U0126, the p38 MAPK inhibitor, SB203580 (10 µm), had little effect on cPLA2 phosphorylation stimulated by PMA, ATP or EGF (Fig. 7a). On the other hand, treatment of astrocytes with a mixture of U0126 (20 µm) and GF109203x (10 µm) completely inhibited cPLA2 phosphorylation as well as AA release stimulated by PMA and ATP (Figs 7b and c). Based on LDH data, the inhibitory effects of the kinase inhibitors were not due to cytotoxicity of the compounds (data not shown).

Figure 7.

Effects of PKC and MAPK inhibitors on phosphorylation of cPLA2. Astrocytes were pretreated with vehicle (control), U0126 (20 µm), SB203580 (10 µm) or GF109203x (10 µm) for 30 min at 37°C. Then, the cells were incubated with vehicle (basal), PMA (100 nm), ATP (100 µm) or EGF (10 ng/mL) for 20 min at 37°C. Cell lysates were prepared in sample buffer and western blot analysis of cPLA2 was performed, as described under Materials and methods. (a) The western blots are representative of results from three independent experiments. (b) Astrocytes were pretreated for 30 min at 37°C with vehicle (control), U0126 (20 µm) and/or GF109203x (10 µm) and then incubated with vehicle (basal), PMA (100 nm) or ATP (100 µm) for 20 min at 37°C. Cell lysates were prepared in sample buffer and western blot analysis of cPLA2 was performed, as described under Materials and methods. The western blots are representative of results from two independent experiments. (c) [14C]-AA-labeled astrocytes were pre-treated without (control) or with U0126 (20 µm), GF109203x (10 µm) or U0126 + GF109203x as decribed in Fig. 6(a). The cells were then incubated with vehicle (basal), PMA (100 nm) or ATP (100 µm) for 30 min at 37°C. Net release of labeled AA due to PMA or ATP stimulation was determined as described under Materials and methods and the effects of inhibitors were expressed as a percentage of the net stimulated release without inhibitors (control). Data are expressed as the mean ± SEM of results from three independent experiments performed in triplicate. p < 0.05 comparing treatment groups with controls.

PKC isoforms in murine astrocytes

The results showing involvement of PKC have prompted us to further investigate the isoform(s) of PKC responsible for PMA- and ATP-stimulated AA release. Despite the presence of many PKC isoforms in the rat brain, primary murine astrocytes contained mainly PKCα, ε, ι and λ(Fig. 8a). Prolonged exposure of astrocytes to PMA resulted in down-regulation of PKCα and ε. PKCε levels declined rapidly and nearly disappeared 6 h after PMA treatment (Fig. 8b). No change was observed in PKCι and PKCλ upon prolonged PMA treatment (Fig. 8b).

Figure 8.

PKC isoforms in primary murine astrocytes and down-regulation by PMA. (a) Rat brain tissue sample (positive control) or whole-cell lysates prepared from untreated primary murine astrocytes were subject to SDS–PAGE and blotted with the Multiscreen Apparatus for PKC isoforms as described under Materials and methods. (b) Primary murine astrocytes were treated with PMA (100 nm) for 0, 3, 6, 12, 18 or 24 h and cell lysates were prepared in sample buffer. Western blot for PKCα, ε, ι and λ were performed as described under Materials and methods. Data are representative of results from three independent experiments.

Western blot analysis of cytosol and membrane fractions seperated by centrifugation indicated the presence of a large portion of PKCα in the cytosol fraction. While PKCε was present mainly in the membrane fraction, PKCι and λ were distributed in both cytosol and membrane fractions (Fig. 9a). We further tested the ability of PMA and ATP to cause the translocation of PKC isoforms. As shown in Fig. 9(b), PMA but not ATP treatment resulted in a rapid translocation of PKCα from cytosol to membrane. No apparent change in translocation was found with PKCε, ι and λ.

Figure 9.

Translocation of PKC isoforms upon stimulation with PMA or ATP. (a) Untreated primary murine astrocytes were lysed in non-denaturing buffer and the membrane and cytosol fractions were prepared as described under Materials and methods. Western blot was performed with the Multiscreen Apparatus for the indicated PKC isoforms. (b) Primary murine astrocytes were treated without or with PMA (100 nm) or ATP (100 µm) for indicated time periods and cell lysates were prepared in non-denaturing buffer. The membrane (M) and cytosol (C) fractions were prepared as above, and western blot for PKCα, ε, ι, λ and actin was performed. Data represent similar results from three independent experiments.

Discussion

This study provided new information about the role of PKC and ERK MAPK in the regulation of Group IV cPLA2 in response to PMA and ATP in primary murine astrocytes. Results from this study provided evidence that: (i) both PMA, an activator of conventional and novel PKC isoforms, and ATP/UTP, agonists for the G protein-coupled P2Y2 receptor, stimulated cPLA2 phosphorylation and AA release through both ERK-dependent and ERK-independent pathways; (ii) stimulation of AA release by PMA and ATP likely involves different PKC isoforms; and (iii) phosphorylation of cPLA2 due to stimulation of the ERK pathway alone is not sufficient to evoke AA release. The scheme depicted in Fig. 10 summarizes the pathways involved in these stimulations.

Figure 10.

A scheme depicting the role of PKC and ERK in phosphorylation of cPLA2 and AA release in response to PMA, ATP/UTP and growth factors in murine astrocytes.

Our previous study with immortalized rat astrocytes (DITNC) demonstrated rapid incorporation of [14C]-AA into membrane phospholipids and agents such as ATP, PMA, or A23187 could evoke the release of labeled AA from the cells (Xue et al. 1999). In the present study, these same agents also stimulated AA release from primary astrocytes isolated from 1- to 3-day-old mouse brain. Because C57BL/6J mice are known to have a natural mutation causing a frame shift disruption in the Group IIA sPLA2 gene (Kennedy et al. 1995), AA release due to this type of sPLA2 can be eliminated. The inhibition of AA release by MAFP, an inhibitor of cPLA2 and iPLA2 (Balsinde and Dennis 1996; Lio et al. 1996), and the lack of inhibition of AA release by HELSS, inhibitor for iPLA2, strongly suggest a major involvement of cPLA2 in ATP- and PMA-stimulated AA release in these cells (Fig. 2). The lack of effect by scalaradial, an inhibitor for sPLA2, further indicated that sPLA2 did not contribute significantly to the AA release due to stimulation by PMA and ATP.

Results from this study provided evidence that both PKC and ERK1/2 can independently phosphorylate cPLA2 in astrocytes. This is in agreement with the notion that multiple phosphorylation sites are present in cPLA2 (de Carvalho et al. 1996; Gijon et al. 1999) and that multiple protein kinases may regulate its activity (Börsch-Haubold et al. 1998; Geijsen et al. 2000; Gijon et al. 2000). Phosphorylation of cPLA2 at Ser505 by ERK1/2 and/or p38 MAPK has been shown to cause a shift in the electrophoretic mobility of cPLA2 (Lin et al. 1993; Kramer et al. 1996; Gijon et al. 2000) as well as an increase in enzyme activity (Lin et al. 1993b). Although the role of PKC in regulating cPLA2 activity has been demonstrated in a number of cell types (Qiu and Leslie 1994; Xing et al. 1997; Husain and Abdel-Latif 1998; Chen et al. 1999), the possibility for PKC to directly phosphorylate cPLA2 has only been reported in a cell-free system (Nemenoff et al. 1993). In agreement with the ability for PMA to activate PKC species, GF109203x, an inhibitor known to preferentially inhibit the conventional and novel PKC isoforms, completely inhibited the PMA-induced ERK and cPLA2 phosphorylation and AA release (Fig. 6a). On the other hand, PMA-stimulated cPLA2 phosphorylation and AA was only partially inhibited by U0126 (1–20 µm) at concentrations that completely inhibited phospho-ERK (Fig. 6b). These results lead to the conclusion that despite that PKC activation led to an increase in phospho-ERK, complete inhibition of the PKC-dependent ERK pathway was not sufficient to block all the AA release stimulation by PMA. Because EGF is known to stimulate the ERK pathway, which is largely PKC-independent, our observations that EGF stimulated cPLA2 phosphorylation but not AA release (Fig. 4) further support the notion that stimulation of cPLA2 phosphorylation by the ERK pathway is not sufficient to elicit AA release. In studies with macrophages and neutrophils, factors such as colony-stimulating factor (CSF-1) could also stimulate the MAPK pathway leading to phosphorylation of cPLA2. However CSF-1 could not elicit AA release (Nahas et al. 1996; Qiu et al. 1998; Geijsen et al. 2000). Similarly, lipopolysaccharide (LPS) could convert cPLA2 to its phosphorylated form in P388D1 macrophages, but this stimulation was not sufficient to cause AA release (Balboa et al. 2000). Together, these results suggest that agents that stimulate cPLA2 phosphorylation by the MAPK pathway may not result in cPLA2 activation and AA release.

This study also provided several lines of evidence suggesting the involvement of different PKC isoforms in the pathways for PMA- and ATP-stimulated AA release:

  • 1Prolonged exposure of murine astrocytes to PMA (100 nm), an agent known to cause down-regulation of the conventional and novel PKC isoforms (Webb et al. 2000), resulted in a time-dependent down-regulation of PKCα and ε isoforms in these cells. Down-regulation of PKCε correlated well with the rapid decline in PMA- but not ATP-stimulated AA release (Fig. 5a and Table 1).
  • 2Treatment with PMA (but not with ATP) caused a rapid translocation of PKCα from cytosol to membranes (Fig. 9). Because murine astrocytes contain mainly PKCα, ε, ι, and λ, these observations help to rule out the involvement of PKCα and ε in the ATP-stimulated AA release pathway (Fig. 5 and Table 1). Study with keratinocytes had implicated the involvement of atypical PKCλ and ι isoforms in the pathway for cPLA2 activation by cytokines (Anthonsen et al. 2001). Our results also suggest the involvement of the atypical PKC isoforms in the ATP-stimulated cPLA2 phosphorylation and AA release in murine astrocytes. It is worth mentioning that PKCδ, an isoform implicated in a number of tyrosine kinase signaling pathways (Neary et al. 1999; Konishi et al. 2001), is not found in murine astrocytes (Fig. 8a). In fact, differences in PKC isoforms may explain the differences in cPLA2 activation between our observations with murine astrocytes and those with rat astrocytes (Chen and Chen 1998). Study with endothelial cells had implicated the involvement of PKCε for activation of cPLA2 by purinergic receptor agonists in endothelial cells (Chen et al. 1999). In astrocytes, however, ATP-stimulated AA release was not altered by prolonged exposure to PMA, which down-regulated the PKCε.
  • 3Our results also show differences in response to GF109203x between PMA- and ATP-stimulated AA release in astrocytes (Fig. 6a and Table 1). Although PMA-stimulated cPLA2 phosphorylation and AA release was completely abolished by low concentrations (1 µm) of GF109203x, much higher concentrations were required to inhibit ATP-mediated AA release (Fig. 6 and Table 1). The differences in sensitivity are likely due to the fact that GF109203x prefers the conventional and novel PKC isoforms than other isoforms.

Studies with murine astrocytes as well as rat astrocytes have demonstrated the ability of ATP/UTP to stimulate AA release through activation of P2Y nucleotide receptors (Bruner and Murphy 1990, 1993; Chen and Chen 1998). The equal potencies of ATP and UTP (Fig. 1a) are characteristic of the agonist potency profile of the P2Y2 receptor subtype (Weisman et al. 1999). P2Y2 receptors belong to the G protein receptor family, which stimulate polyphosphoinositide hydrolysis and the release of second messengers for calcium mobilization and for activation of PKC (Lin et al. 1993a; Weisman et al. 1999). However, recent studies further demonstrated the ability of P2Y2 receptor to confer other functions, including activation of the MAPK cascade by both PKC-dependent and PKC-independent pathways (Soltoff 1998; Weisman et al. 1998, 1999; Neary et al. 1999; Erb et al. 2001). Studies with rat astrocytes have indicated the involvement of Ca2+-independent PKC on the ATP-stimulated ERK pathway (Neary et al. 1999). Our observations that ATP-stimulated AA release was largely insensitive to Gö6976, a PKC inhibitor for the Ca2+-dependent PKC isoforms, further support the involvement of a Ca2+-independent PKC/ERK pathway for activation of cPLA2 in astrocytes (Table 1). These results clearly demonstrated the complex signaling pathways associated with this receptor and its role in modulating diverse cellular events in astrocytes.

The presence of a C-2 domain in the cPLA2 molecule implicated an important role for Ca2+ in translocation of cPLA2 from cytosol to membranes (Leslie 1997). In the present study with astrocytes, chelating extracellular calcium by EGTA completed abrogated AA release stimulated by PMA and ATP (data not shown). Because PMA is known not to cause an increase in intracellular Ca2+ (Qiu et al. 1998), the amount of Ca2+ present intracellularly is apparently sufficient to cause cPLA2 translocation. Our data clearly show the ability of PMA to cause a rapid translocation of PKCα isoform (Fig. 9). Despite that Ca2+-mobilizing P2 receptors are present widely in cells in the CNS (Pearce et al. 1989; Weisman et al. 1999) as well as cells in the peripheral systems, e.g. macrophages (Qiu et al. 1998) and endothelial cells (Chen et al. 1999), the role of intracellular Ca2+ mobilization on stimulation and activation of cPLA2 remains to be an important subject for further investigation.

Taken together, results from this study provided new information on the role of PKC and ERK1/2 in mediating phosphorylation of cPLA2 and AA release in response to PMA and ATP in primary murine astrocytes (Fig. 10). In addition, our study revealed that the signaling pathways leading to activation of cPLA2 might differ depending on the type of agonists used to stimulate the cells and the PKC isoforms involved. As stimulation of cPLA2 and subsequent conversion of AA to prostaglandins have been related to many neurodegenerative diseases (Clemens et al. 1996; Stephenson et al. 1996; Farooqui et al. 1997a, 1997b), understanding these signaling pathways that regulate activity of the AA cascade may have important clinical implications.

Acknowledgements

This work was supported by AA06661, ES10535 and AG18357 from NIH, the Missouri Alzheimer's Disease and Related Disorders Program, and the University of Missouri-Columbia Food for the 21st Century Program.

Ancillary