GLIA 51:199–208 (2005)

Signal Transduction Pathways Regulating Cyclooxygenase-2 in Lipopolysaccharide-Activated Primary Rat Microglia 1 € RAVI SHANKAR AKUNDI,1 EDUARDO CANDELARIO-JALIL,1 SANDRA HESS,1 MICHAEL HULL, 1 2 1* KLAUS LIEB, PETER J. GEBICKE-HAERTER, AND BERND L. FIEBICH 1 Department of Psychiatry, University of Freiburg Medical School, Freiburg, Germany 2 Department of Psychopharmacology, Central Institute for Mental Health, Mannheim, Germany

KEY WORDS brain inflammation; lipid second messengers; neuroimmunology; protein kinases; signal transduction

ABSTRACT Microglia are the major cell type involved in neuroinflammatory events in brain diseases such as encephalitis, stroke, and neurodegenerative disorders, and contribute significantly to the release of prostaglandins (PGs) during neuronal insults. In this report, we studied the immediateearly intracellular signalling pathways in microglia, following bacterial lipopolysaccharide (LPS) stimulation, leading to the synthesis and release of PGE2. Here we show that LPS induces cyclooxygenase (COX) 2 by activating sphingomyelinases leading to the release of ceramides, which in turn, activate the p38 mitogen-activated protein kinases (MAPK), but not the p42/44 MAPK. We further show that exogenously added ceramide analogue (C2-ceramide) also induce PGE2 synthesis through a p38 MAPK-dependent pathway. This potential nature of ceramides in activating microglia suggests that endogenously produced ceramides during neuronal apoptosis in ischemia or neurodegenerative diseases could also contribute to the amplification of neuroinflammatory events. In contrast to protein kinase C (PKC) and phosphocholine-specific phospholipase C (PCPLC), which transcriptionally regulate LPS-induced COX-2 synthesis, inhibition of phospholipase A2 (PLA2) has no effect on COX-2 transcription, although it inhibits the release of PGE2. Transcriptional regulation of LPS-induced COX-2 by PKC is further proved by the ability of the PKC inhibitor, G€o 6976, to inhibit LPS-induced 8-isoprostane synthesis, but not affecting LPS-induced COX-2 activity. Our data with 8-isoprostane also indicates that COX-2 plays a major role in ROS production in LPS-activated microglia. This detailed view of the intracellular signaling pathway in microglial activation and COX-2 expression opens a new therapeutic window in the search for new and more effective central anti-inflammatory agents. C 2005 V

Wiley-Liss, Inc.

INTRODUCTION Microglia are believed to play a major role in inflammatory and immune reactions in the brain, including neurodegenerative disorders such as Alzheimer’s disease (AD) (Streit et al., 2004). Microglial cells are derived from bone marrow precursors that invade the brain at an early stage of embryonic development and reside C 2005 V

Wiley-Liss, Inc.

there as tissue-specific macrophages. Activated microglia release a variety of pro-inflammatory mediators such as cytokines, prostanoids, and complement components, and potentially neurotoxic substances such as reactive oxygen species (ROS), nitric oxide (NO) and excitatory amino acids. Although prostanoids serve multiple roles in neuronal signaling, regulation of sleep/wake cycle and body temperature, gastric and renal maintenance, and certain hormonal control; their intracerebral levels are elevated dramatically during neuronal insults and neuropathological conditions (Seibert and Masferrer, 1994). In the brain, prostanoids can be synthesized by both isoforms of the enzyme, cyclooxygenase (COX): the constitutive COX-1 and the inducible COX-2 (Smith et al., 2000). COX-2, but not COX-1, has been shown to be induced through molecules such as adenosine (Fiebich et al., 1996), inflammatory cytokines (Fiebich et al., 2000b; O’Banion et al., 1996), bacterial lipopolysaccharide (LPS), and mitogens (Minghetti and Levi, 1995; Bauer et al., 1997), and during insults such as hypoxia, ischemia and in various neurodegenerative disorders (Akiyama et al., 2000; Candelario-Jalil et al., 2004). COX-2 has been shown to be instrumental in Parkinson’s disease (Teismann et al., 2003), as well as in AD, where both COX-1 and COX-2 mRNAs were found to be elevated in the brain (Kitamura et al., 1999), with a concomitant increase in prostaglandin metabolites in the cerebrospinal fluid (Montine et al., 1999). Prostaglandin E2 (PGE2), the major enzymatic product of COX-2, is able to induce interleukin-6 (IL-6) in astrocytes (Fiebich et al., 1997), neuronal apoptosis in some conditions (Takadera et al., 2004), increase intracellular levels of ROS (Montine et al., 2002), and spread COX-2 synthesis from activated to resting microglial cells or between PGE2-producing neurons and microglia (Minghetti et al., 1997). PGE2 mediates its effects via four different recep-

Grant sponsors: Alexander von Humboldt Foundation (Bonn, Germany); Deutsche Forschungsgemeinschaft *Correspondence to: Bernd L. Fiebich, Department of Psychiatry, Neurochemisches Labor I, University of Freiburg Medical School, Hauptstrasse 5, 79104 Freiburg, Germany. E-mail: bernd.fi[email protected] Received 24 November 2004; Accepted 12 January 2006 DOI 10.1002/glia.20198 Published online 30 March 2005 in Wiley InterScience (www.interscience. wiley.com).

200

AKUNDI ET AL.

tors named EP1 through EP4 (Narumiya et al., 1999). Both EP1 and EP2 receptors are found on cultured microglia, although EP3 was also detected in activated microglia in vivo (Slawik et al., 2004). Microglia-driven chronic inflammatory responses are thus considered a major factor in the progression of neurodegenerative disorders. Studies on LPS-activated primary rat microglia have been well documented (Minghetti and Levi, 1995; Bauer et al., 1997). LPS, in association with the LPS binding protein, mediates its effects through the CD14 receptor, a glycosylphosphatidylinositol-linked membrane protein that is present on the microglial cells (Becher et al., 1996). The LPS–CD14 complex, together with other adaptor proteins, binds to the toll-like receptor 4 (TLR4), an evolutionarily conserved cell-surface receptor related to the Drosophila gene toll, that is present on microglia, but not on astrocytes, oligodendrocytes, or cortical neurons (Lehnardt et al., 2003). TLR4 interacts with the adaptor protein, myeloid differentiation factor 88 (MyD88), that activates interleukin-1 receptor type I-associated protein kinases (IRAKs), which, in turn, phosphorylate the tumor necrosis factor (TNF) receptorassociated factor 6 (TRAF6) (Fitzgerald et al., 2001; Akira, 2003). Several kinases have been implicated downstream of TRAF6, including nuclear factor jB (NFjB)-inducing kinase (NIK), mitogen-activated protein kinase (MAPK), kinase kinase 1 (MEKK1), and transforming growth factor-b-activated kinase 1 (TAK1) (Akira, 2003). MEKK1 is strongly selective for the c-Jun N-terminal kinase (JNK), while TAK1 activates both JNK and p38 MAPK (Kyriakis and Avruch, 2001). NIK, as well as TAK1, activates IjB kinase (IKK), whose activation, in turn, leads to the phosphorylation and degradation of the inhibitory jB (IjB) subunit and consequent release of NF-jB (Karin and Ben Neriah, 2000; Rhee and Hwang, 2000). MyD88-independent pathways have also been identified, that lead to the activation of JNK and NF-jB, although involving different set of adaptor proteins (Kawai et al., 2001). With the exception of ERK5/big MAPK1 (BMK1) (Lee et al., 1995), all the remaining three members of the MAPK family, consisting of p38 MAPK, extracellular signal-regulated kinases (ERK1/2, or p44/42 MAPK), and JNK have been shown to be activated following LPS stimAbbreviations

AACOCF3 AD COX CREB ERK JNK LPS MAPK NF-jB PC-PLC PGE2 PKC PLA2 ROS SMase TLR4

arachidonyl trifluoromethylketone Alzheimer’s disease cyclooxygenase cyclic AMP responsive element binding protein extracellular signal-regulated kinases c-Jun N-terminal kinase lipopolysaccharide mitogen-activated protein kinase nuclear factor jB phosphocholine phospholipase C prostaglandin E2 protein kinase C phospholipase A2 reactive oxygen species sphingomyelinase toll-like receptor 4.

ulation (Hambleton et al., 1996). All MAPKs have also been implicated in neuroinflammatory events. However, the very early signaling events occurring in LPSinduced microglial activation are not completely understood. In this report, we show that within 1 min of LPS stimulation, sphingomyelinases are activated in microglia, followed by phosphorylation of p38 and p42/44 MAP kinases. Furthermore, we demonstrate that ceramides, the products of sphingomyelinase activity, can themselves lead to the release of PGE2 through a p38 MAPK-dependent pathway. MATERIALS AND METHODS Reagents LPS from Salmonella typhimurium (Sigma, Taufkirchen, Germany) was resuspended in sterile phosphate buffered saline (PBS; Cell Concepts, Umkirch, Germany) as 5 mg/ml stock, and was used at a final concentration of 100 ng/ml in culture. N-acetyl-D-sphingosine (C2 ceramide; Sigma), SC-58125 (Cayman Chem, Ann Arbor, MI) and the kinase inhibitors, SB203580, PD98059, G€ o 6976 and GF109203X (all from Calbiochem, Bad Soden, Germany) were solubilized in dimethylsulfoxide (DMSO), tricyclodecan-9-yl-xanthogenate (D609, Sigma) in water, and arachidonyl trifluoromethylketone (AACOCF3, Sigma) in ethanol, while arachidonic acid (Sigma) was reconstituted in methanol. Solvent concentration in the culture media was maintained at <0.1%. All agents, used at the given concentrations, do not affect the viability of the cells, as observed through a luminescent kit (Promega, Madison, WI) that measures metabolic ATP levels (data not shown). The cDNA probe encoding for rat COX-2 gene has been cloned previously (Bauer et al., 1997). The cDNA probe for b-actin was kindly provided by G. Finkenzeller (Institute for Tumor Biology, Freiburg, Germany). Antibodies against COX-1 and p38 MAPK were from Santa Cruz Biotechnology. Goat polyclonal antibody against rat COX-2 was obtained from two sources: Santa Cruz Biotechnology (Heidelberg, Germany) or Cayman Chemicals (Ann Arbor, MI). Rabbit polyclonal antibodies against phospho-specific p38 MAPK (Thr 180/Tyr 182) were from Cell Signaling Tech. (supplied by New England Biolabs, Frankfurt, Germany), while the antibody against phospho-specific p42/ 44 MAPK (Thr 183/Tyr 185 of the p42 enzyme) was from Promega (Mannheim, Germany). Cell Culture Primary astroglial cell cultures were established from cerebral cortices of 1-day neonatal Sprague-Dawley rats as previously described (Keller et al., 1985). Briefly, forebrains were minced and gently dissociated by repeated pipetting in PBS and filtered through a 70-mm cell strainer (Falcon). Cells were collected by centrifugation (1,000 rpm, 10 min), resuspended in Dulbecco’s modified Eagle’s medium (MEM) containing 10% fetal calf serum (FCS; Sigma, Taufkirchen, Germany) and antibiotics

REGULATION OF COX-2 IN MICROGLIA

(40 U/ml penicillin and 40 mg/ml streptomycin, both from PAA Laboratories, Linz, Austria), and cultured on 10-cm cell culture dishes (5 3 105 cells/plate; Falcon) in 5% CO2 at 37°C. Medium was prepared taking extreme care to avoid LPS contamination (Gebicke-Haerter et al., 1989). Floating microglia were harvested every week (between 2–7 weeks) and re-seeded into a 75-cm2 culture flask (for Northern and Western blots and sphingomyelinase assays) or 24-well plates (for 8-isoprostane/PGE2 estimation) to give pure microglial cultures. The following day, cultures were washed to remove non-adherent cells, and fresh medium was added. The purity of the microglial culture was >98% as previously determined by immunofluorescence and cytochemical analysis (Gebicke-Haerter et al., 1989).

201

44 MAPK) polyacrylamide gel under reducing conditions. Separated proteins were then transferred onto a polyvinylidene fluoride (PVDF) membrane (Millipore, Bedford, MA) by semi-dry blotting and blocked overnight at 4°C using Rotiblock (Roth, Karlsruhe, Germany). Primary antibodies used were goat anti-COX-1 (M-20), goat anti-COX-2 (M-19), or goat anti-p38 MAPK, diluted 1:500, or rabbit anti-phospho-p38 MAPK or rabbit antiphospho-p42/44 MAPK, diluted 1:1,000, in Tris-buffered saline (TBS) containing 0.1% Tween 20 (Merck, Darmstadt, Germany) and 1% BSA for 2 h. Secondary antibody was diluted 1:100,000 (for goat) or 1:25,000 (for rabbit) in 1% BSA in TBS-T and incubated for 1 h at room temperature. Following extensive washing, proteins were detected using enhanced chemiluminescence (ECL) reagents from Amersham (Freiburg, Germany).

RNA Extraction and Northern Blot Analysis Total RNA was extracted by using an RNA extraction kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. Total RNA (10 mg per lane) was separated on an agarose-formaldehyde gel electrophoresis, followed by blotting onto a positively charged Nylon membrane (Pharmacia, Freiburg) and cross-linked by incubation at 120°C for 30 min. The membrane was then prehybridized in 50% formamide, 0.25 M phosphate buffer, 0.25 M NaCl, 1 mM EDTA, 100 mg/ml denatured salmon sperm DNA and 7% sodium dodecyl sulfate (SDS) at 43°C for 2 h. cDNA probes were labeled with 50 mCi [32P] dCTP (Amersham, Braunschweig, Germany) using a random priming kit from Stratagene (Heidelberg, Germany). Unincorporated nucleotides were removed by using a nucleotide removal kit from Qiagen (Hilden, Germany). The radiolabeled probe was denatured and added to a fresh prehybridization buffer for an overnight incubation at 43°C. Membranes were washed three times in 2 3 standard saline citrate (SSC) containing 0.1% SDS (53°C, 20 min) and exposed to Kodak XAR films at 280°C. For rehybridization, probes were removed by heating the membrane at 95°C in distilled water.

Western Blotting Microglial cultures plated in flasks were pre-treated with various inhibitors 30 min before LPS stimulation. For the determination of p38 and p42/44 MAPK, medium was maintained serum-free. At the end of stimulation, cells were washed in PBS and lysed in SDS lysis buffer composed of 42 mM Tris-HCl, pH 6.8, 1.3% (w/v) SDS, 6.5% glycerol, and 100 mM orthovanadate (Laemmli 1970). Protein content was measured using the bicinchoninic acid method (Pierce, distributed by KFC Chemikalien, M€ unchen, Germany), using bovine serum albumin (BSA) as the standard. Dithiothreitol (DTT; final concentration, 10 mM) and bromophenol blue were added to the samples and heated at 95°C, 5 min, just before electrophoresis. 60 mg of the sample was loaded on a 7.5% (for COX-1/2) or 12% (for p38 and p42/

Determination of 8-Isoprostane and PGE2 Release Microglial cells pre-treated with various inhibitors for 30 min were treated with LPS or ceramide for 24 h. Supernatants were harvested and the levels of 8-isoprostane (Cayman Chemicals) and PGE2 (Assay Design, distributed by Biotrend, K€ oln, Germany) were measured by an enzyme immunoassay according to the manufacturer’s instructions. The standards were used in the range of 3.9–500 pg/ml (detection limit of 5 pg/ml) for 8-isoprostane, and 39–2,500 pg/ml for PGE2 (sensitivity of 36.2 pg/ml). In assays measuring COX-2 activity, microglial cells were pre-incubated with 100 ng/ml LPS for 18 h, washed and re-incubated in a serum-free media with or without the inhibitor (G€ o 6976, 10 nM) for 15 min; 30 mM arachidonic acid was then added, and the cells were further incubated for 15 min, at the end of which media supernatants were collected for the determination of PGE2 as described above. Determination of Sphingomyelinase Activity For the assay of acidic and neutral sphingomyelinases (SMase), the Amplex Red SMase kit (Molecular Probes, distributed by MoBiTec, G€ ottingen, Germany) was employed. The assay involves a four-step enzymatic procedure performed in a 96-well plate (Falcon) in a reaction buffer composed of 0.1 M Tris-HCl, 10 mM MgCl2, and added components, as described, in a total volume of 200 ml (37°C, 30 min). The reaction is initiated by the SMase present in the sample, which hydrolyzes added sphingomyelin (0.5 mM in 0.2% Triton X-100) into ceramide and phosphorylcholine. The action of alkaline phosphatase (8 U/ml) on phosphorylcholine yields choline, which is oxidized by choline oxidase (0.2 U/ml) into betaine and hydrogen peroxide. Finally, hydrogen peroxide, in the presence of horseradish peroxidase (HRP; 2 U/ml), reacts with Amplex Red (10-acetyl-3,7-dihydroxyphenoxazine, 0.1 mM) in a 1:1 stoichiometry to generate the highly fluorescent product, resorufin. Resorufin was

202

AKUNDI ET AL.

measured with a fluorescence microplate reader (BMG Lab Technologies, Offenburg, Germany) using excitation wavelength at 530 nm and emission at 590 nm. For the analysis of acidic SMase, a two-step assay was employed in which the initial SMase reaction was performed in 50 mM sodium acetate, pH 5.0, (37°C, 60 min), followed by a rise in pH to 7.5 by the addition of 0.1 M Tris-HCl, pH 8.0, to permit detection with the Amplex Red reagent. Samples for the above SMase reaction were obtained from microglial cells cultured in 75-cm2 flasks. Following LPS (100 ng/ml) stimulation for 0, 1, or 5 min, the cells were lysed by three cycles of freeze-thaw in a buffer containing 25 mM Tris-HCl, pH 7.4, 2 mM EDTA, 5 mM DTT, 1 mM phenylmethylsulfonyl fluoride and a protease inhibitor cocktail (Sigma). The cells were homogenized by repeatedly forcing through an 18-gauge needle. The homogenates were spun at 800g for 10 min (4°C) to remove nuclei and cell debris. Clear supernatants were collected, and
25 mg protein was used for the SMase reaction as described above.

Statistical Analysis

RESULTS Involvement of p38 and p42/44 MAPK in COX-2 Expression

Fig. 1. Mitogen-activated protein kinases are involved in COX-2 expression. Rat primary microglial cells were treated with 100 ng/ml LPS for 24 h with a 30-min pre-stimulation in the presence of SB203190 (10 mM) or PD98059 (10 mM). The cells were harvested to collect mRNA for Northern (A) or Western (B) blots for COX-2, with b-actin and COX-1 used as internal controls for the experiment. Culture supernatants were collected for the measurement of PGE2 released (C). Data are expressed in mean 6 SD and are representative of at least three independent measurements from three different primary cultures (§P < 0.001, w.r.t. control; ***P < 0.001 w.r.t (with respect to) LPS). To further confirm the role of p38 and p42/44 MAPKs, microglial cells were treated with LPS for different intervals of time (D) or in the presence of various inhibitors for 30 min (E), followed by a Western blot for the detection of p38 and p42/44 MAPKs in their phosphorylated state. SB203190, PD98059 and GF109203X were used at a concentration of 10 mM, G€ o 6976 at 1 mM and D609 at 50 mg/ml and were used 30 min before the addition of LPS (100 ng/ml).

Several reports have previously shown that LPS strongly activates microglia and induces COX-2 gene leading to the synthesis of PGE2 (Bauer et al., 1997; Fiebich et al., 2003). We were interested in investigating the early signal transduction events in highly purified microglia obtained from neonatal rat cortices. The MAPKs are ubiquitous enzymes and have been implicated in LPS-induced COX-2 expression in other celltypes. Northern blot analysis revealed that COX-2 mRNA was undetectable under basal conditions, but showed an intense increase with LPS treatment (Fig. 1A). This increase was strongly suppressed by the MAPK inhibitors, SB203580 and PD98059, that specifically inhibit the p38 and p42/44 (ERK2/1) MAPKs, respectively (Fig. 1A). This inhibition was also reflected at the protein level. LPS strongly induces COX-2, but not COX-1, and both SB203580 and PD98059 inhibited the LPS-induced COX-2 (Fig. 1B). The MAPK inhibitors also caused a suppression of PGE2 release (Fig. 1C). In the resting microglia, 225 6 120 pg/ml PGE2 was detected. Following LPS stimulation, a 20-fold induction in PGE2 levels was observed, with
4771 6 2,700 pg/ml PGE2 detectable in the medium supernatants. Both the

MAPK inhibitors significantly inhibited PGE2 release, with the p38 MAPK inhibitor reducing the PGE2 levels to 1409.8 6 1140 pg/ml and the p42/44 MAPK inhibitor to 568 6 492 pg/ml PGE2 (Fig. 1C). To confirm the involvement of MAPK in COX-2 regulation, microglial cells were treated with LPS (100 ng/ ml) for different intervals of time. Stimulation with LPS resulted in the phosphorylation of both p38 and p42/44 MAPK within 15 min, a peak at 30 min and sustaining up to 120 min after stimulation (Fig. 1D). Total p38 MAPK or p42 MAPK protein levels were not affected by LPS (data not shown). The MAPK inhibitors inhibited the phosphorylation of their respective kinases. SB203580 (1 mM), but not PD98059, specifically inhibited the phosphorylation of p38 MAPK (Fig. 1E). In contrast, other kinase inhibitors, such as GF109203X or G€ o 6976 (PKC inhibitors) and D609 (phosphocholine-phospholipase C, PC-PLC, inhibitor) failed to inhibit LPS-induced p38 MAPK phosphorylation (Fig. 1E). Similarly, PD98059 (1 mM), but not SB203580, specifically inhibited the phosphorylation of p42/44 MAPK (Fig. 1E). Similar to p38 MAPK, the

Raw data were converted into percentage values of LPS-control so as to minimize culture-culture and age variation. Data are represented as mean 6 SD and analyzed using Student’s t-test (two-groups) or one–way analysis of variance (ANOVA) with post hoc Student– Newman–Keuls test (multiple comparison). Differences were considered significant when P < 0.05.

REGULATION OF COX-2 IN MICROGLIA

Fig. 2. Role of phospholipase and protein kinase C in LPS-induced PGE2 synthesis. Microglial cells were treated with AACOCF3, D609, GF109203X, or G€ o 6976, at the indicated concentrations for 30 min, followed by LPS (100 ng/ml) for 24 h (§P < 0.01 and §§P < 0.001, w.r.t. control). The culture supernatants were tested for PGE2, and the cells were harvested for immunoblotting against rat COX-1 and COX-2. While AACOCF3 inhibited LPS-induced PGE2 dose-dependently (A), it had no effect on COX-2 synthesis (B). In contrast, D609 inhibited LPSinduced COX-2 (C) and PGE2 release (D). Similarly, the PKC inhibitors, GF109203X and G€o 6976, inhibited LPS-induced COX-2 (E) and PGE2 (F), at 10 mM and 10 nM, respectively (*P < 0.05, **P < 0.01, ***P < 0.001, w.r.t. LPS).

phosphorylation of p42/44 MAPK was not affected by GF109203X, G€o 6976, or D609.

Phospholipases and PKC are Crucial for PGE2 Release The cytosolic phospholipase A2 (PLA2) inhibitor, AACOCF3, strongly inhibited LPS-induced PGE2 synthesis (Fig. 2A). AACOCF3 inhibits LPS-induced PGE2 release from a concentration of 10 nM onward, with below basal levels beyond 1 mM (P < 0.001). At 10 mM, there was a reduction in basal PGE2 levels as well. However, AACOCF3 showed no effect on COX-1 protein or LPS-induced COX-2 synthesis (Fig. 2B). In contrast, the PC-PLC inhibitor, D609, strongly inhibited COX-2 synthesis (Fig. 2C) as well as the release of PGE2 (P < 0.05; Fig. 2D). Although the PKC inhibitors do not inhibit the phosphorylation of p38 or p42/44 MAPK, both GF109203X and G€o 6976 suppress LPS-induced COX-2 synthesis (Fig. 2E). GF109203X inhibited LPS-induced COX-2 protein at 10 mM, while, in contrast, G€ o 6976 was highly potent and inhibited LPS-induced COX-2 protein at a low dose of 10 nM. Corresponding observations were also made at the PGE2 level, with GF109203X inhibiting LPS-induced PGE2 by 40% at 10 mM (P < 0.02), while G€ o 6976 inhibited LPS-induced PGE2 by more than 50% at 10 nM (P 5 0.002; Fig. 2F).

203

Fig. 3. LPS induces the formation of 8-isoprostanes. A: For the measurement of free radicals formed in activated microglia, cells were treated with LPS (100 ng/ml) in the presence of various inhibitors (AACOCF3, 1 mM; G€ o 6976, 10 nM; SC58125, 100 nM) for 24 h. Media supernatants were collected for the measurement of 8-epi-PGF2a (§P < 0.01, w.r.t. control; *P < 0.05, **P < 0.01, w.r.t. LPS). B: To identify whether G€ o 6976 directly inhibited COX-2 enzyme activity, activity assay was performed as described under Materials and Methods. Culture supernatants were then measured for the release of PGE2.

G€ o 6976 Blocks LPS-Induced 8-Isoprostane Formation To study the role of reactive oxygen species (ROS) in microglial activation, we measured the levels of 8-isoprostanes, an index of free radical generation in vivo (Morrow et al., 1990), in activated microglia. As shown in Figure 3A, LPS induced the formation of 8-isoprostanes, when measured 24 h after stimulation. From a basal level of 33.4 6 0.5 pg/ml, LPS increases the levels of 8-epi-PGF2a up to 67.4 6 9 pg/ml (P < 0.01). The PLA2 inhibitor, AACOCF3, moderately repressed the induced levels of 8-isoprostane to 48 6 4 pg/ml (P < 0.05), while the PKC inhibitor, G€ o 6976, potently reduced the levels of 8-isoprostanes to nearbasal levels of 39 6 6 pg/ml (P < 0.01), similar to the COX-2-specific inhibitor SC-58125 (24 6 8 pg/ml, P < 0.01). To identify if G€ o 6976-mediated inhibition of 8-isoprostanes was a direct effect on COX-2 activity, we measured pre-activated microglial cells (100 ng/ml LPS, 18 h), with or without G€ o 6976 (10 nM, 15 min), in the presence of exogenously added arachidonic acid (30 mM; Fig. 3B). A basal level of 297 6 148 pg/ml increased to 896 6 355 pg/ml PGE2 (P < 0.001) when incubated with arachidonic acid, and further to 1614 6 370 pg/ml PGE2 in pre-activated cells (P 5 0.002). However, G€ o 6976 was unable to significantly inhibit PGE2 in pre-activated cells treated with arachidonic acid (Fig. 3B).

204

AKUNDI ET AL.

Fig. 4. Role of sphingomyelinase in LPS-induced PGE2 release. A: Cells treated with LPS (100 ng/ml) for short intervals were lysed and the activity of sphingomyelinases measured as described in Materials and Methods (**P < 0.01). B: To determine whether ceramide could induce PGE2 release, cells were treated with 100 mM C2-ceramide for 24 h in the presence of SB203190 (10 mM) or PD98059 (10 mM). Culture supernatants were used for the measurement of PGE2 (P < 0.01, § w.r.t. control, and, ** w.r.t. LPS). Cells were also treated with different concentrations of C2-ceramide (C) or 100 mM C2-ceramide, in the presence of MAPK inhibitors (10 mM, D), or 100 ng/ml LPS (E) for 30 min, and harvested for immunoblotting against phosphorylated and nonphosphorylated forms of p38 MAPK or phosphorylated p42/44 MAPK.

Role of Sphingomyelinases in LPS-Induced PGE2 Release In addition to MAPK, PKC, and PLA2, we were interested in identifying other enzymes regulating microglial PGE2 release. In this respect, the role of sphingomyelinase (SMase) was explored. Indeed, LPS induced the activation of both forms of SMase; with the acidic SMase being more potently activated in comparison to neutral SMase, when measured using a fluorimetric kit (Fig. 4A). Activity of SMase in the samples is measured by an increase in fluorescence caused by the release of phosphorylcholine, which in turn, through a series of enzymatic steps as described in Materials and Methods, generates the highly fluorescent resorufin product. Stimulation of microglial cells with 100 ng/ml LPS increased the fluorescent signal from a basal level of 14,512 6 302 relative fluorescent units (RFU) to 18,458 6 2,140 RFU in 5 min (P < 0.01; Fig. 4A). Under acidic conditions of the assay, for the measurement of the activity of acidic SMase, there was an elevation of fluorescence from a basal value of 15,469 6 1,120 RFU to 18504 6 1876 RFU within 1 min and to 24,560 6 7,477 RFU in 5 min

(P 5 0.01). Similar acute, but short, changes in ceramide concentration, occurring within 1–2 min, have also been described for tumor necrosis factor-a (TNF-a) and interleukin-1 signaling (Mathias et al., 1993). SMases hydrolyze membrane sphingomyelins in order to yield ceramide (N-acylsphingosine) and phosphorylcholine (Barnholz et al., 1966). We were interested in knowing if exogenously added ceramides could also activate microglia leading to PGE2 synthesis. The activity of intracellular ceramides can be demonstrated using a cell-permeable analogue of ceramide, C2-ceramide (N-acetylsphingosine; Hannun and Bell, 1989; Lee et al., 1996). C2-ceramide, when used at a concentration of 100 mM, caused a moderate, but significant, increase in PGE2 released by the cells; from a basal level of 228 6 102 pg/ml PGE2 to 564 6 272 pg/ml (by 2.5-fold, P < 0.01; Fig. 4B). Furthermore, this release was inhibited by the MAPK inhibitors, SB203580 and PD98059, suggesting the involvement of p38 and p42/44 MAPK in ceramide-mediated PGE2 synthesis. To identify further the role of MAPK signaling by ceramides, Western blots recognizing the phosphorylated forms of p38 and p42/44 MAPK were performed. C2-ceramide, at doses as low as 1 mM, induced the phosphorylation of p38 MAPK, and was very potent at 10 mM (Fig. 4C). In addition, ceramide-induced phosphorylation of p38 MAPK was inhibited by SB203580, but not by PD98059 (Fig. 4D). However, in contrast to p38 MAPK, ceramide alone, could not induce the phosphorylation of p42/44 MAPK (Fig. 4E), although PD98059, inhibited ceramide-induced PGE2 synthesis (Fig. 4B).

DISCUSSION Much of the knowledge of microglial functions is based largely on in vitro studies of primary microglial cultures exposed to activating stimuli such as pro-inflammatory cytokines, bacterial products and toxins (Bauer et al., 1997; Minghetti and Levi, 1995; Hoozemans et al., 2001). In this study, we show that both LPS and ceramide induce PGE2 synthesis through the MAPK pathway (Fig. 5). LPS induces the activation of SMase, which in turn hydrolyzes membrane sphingomyelin into ceramides and phosphorylcholine (Barnholz et al., 1966). Ceramides, in turn, activate p38 MAPK, which is essential for the maintenance of COX-2 mRNA stability and expression. The p42/44 MAPK and the PC-PLC-PKC pathways, however, act in parallel, but are essential for PGE2 synthesis. Furthermore, this study also shows that, at the substrate level, the production of PGE2 after COX-2 protein synthesis depends on the provision of the PGE2 substrate, arachidonic acid, whose synthesis depends on PLA2. Although under in vivo situations, other factors might modulate PGE2 synthesis, the above results allows us to construct a complex scheme of the various signal transduction pathways mediating LPSinduced PGE2 release in primary rat microglial cells (Fig. 5).

REGULATION OF COX-2 IN MICROGLIA

Fig. 5. Hypothetical schematic of LPS-induced PGE2 release. LPS, in association with several binding and adaptor proteins, interacts with TLR4 receptor, resulting in the activation of Raf-MEK-ERK pathway, PC-PLC-PKC pathway and sphingomyelinase (SMase) that releases ceramides and phosphorylcholine (P-Chol) from membrane sphingomyelin (SM). Ceramides can activate p38 MAPK directly, an early response, or is metabolized into sphingosine-1-phosphate (S1P) and/or ceramide-1phosphate (Cer-1-P) generated through the actions of ceramidase (CERase)-sphingosine kinase (SK) and ceramide kinase (CK), respectively, which mediate their effects on p42/44 MAPK or PLA2 in a late response. While p42/44 MAPK and PKC are involved in COX-2 expression, through CREB and NF-jB pathways; p38 MAPK is involved in providing mRNA stability and COX-2 synthesis. Formation of PGE2 is, however, dependent on the activity of PLA2 that releases arachidonic acid (AA) from membrane-bound phospholipids. The site of action of various inhibitors used in this study is included in gray for reference.

Identifying the downstream signaling pathways following LPS stimulation in microglia is of fundamental importance in the understanding of the molecular mechanisms mediating bacterial encephalitis during Gram negative bacterial infections (Parrillo, 1993; Kim, 2003). In addition to the direct role of LPS in inducing PGE2 synthesis in microglia, the release of ceramides during neuronal apoptosis (Hannun, 1996) could also lead to moderate increases in PGE2 synthesis, as shown in this report. Neuronal apoptosis has been shown to occur in a variety of brain injuries such as bacterial meningitis (Kim, 2003), stroke (Charriaut-Marlangue et al., 1996), and neurodegenerative disorders (Braun et al., 1999; Behl, 2000; Tatton et al., 2003). Several studies have shown an increased level of ceramides in the cerebrospinal fluid of AD patients (Cutler et al., 2004), in human immunodeficiency virus-infected brain (Haughey et al., 2004), as well as in lipid metabolism disorders (Gornati et al., 2002; Soreghan et al., 2003). Ceramides can also be released in response to a variety of stress stimuli, including radiation (Buckman et al., 1998) and anti-cancer drugs (Bieberich et al., 2002). Further, the activation of cytokine receptors (Ballou et al., 1992), CD95 (Fas) receptor (Grullich et al., 2000), and the low-affinity neurotrophin receptor (Dobrowsky et al., 1994) also lead to the release of ceramides, which in turn can potentiate the inflammatory reaction (Ballou et al., 1992), or lead to apoptosis (Yu et al., 2000). In addition to providing innate immunity against bacterial pathogens, a role of CD14 receptors in mediating inflammation in AD has been suggested (Fassbender et al., 2004). Ceramides show a strong structural similarity to a region of lipid A in LPS (Joseph et al., 1994).

205

It has been shown that in monocytes, ceramides, like LPS, can bind to the CD14 receptor (Pfeiffer et al., 2001). In contrast, the LPS-CD14 cluster, together with other binding and adaptor proteins, mediates its intracellular effects through TLR4 (Beutler and Rietschel, 2003), or can also be internalized, and activate downstream ceramide targets, such as ceramide-activated protein kinase, without activating SMase (Joseph et al., 1994). However, unlike LPS, that shows a strong induction of p42/44 MAPK (Fig. 1D), ceramides do not induce p42/44 MAPK in microglia (Fig. 4E). This is in contrast to human mammary epithelial cells where ceramides induced the activation of p42/44 MAPK, JNK as well as p38 MAPK leading to the expression of COX-2 gene (Subbaramaiah et al., 1998). Differential signaling mediated by CD14 receptors has been cited in other works, such as in macrophages, where downstream pathways, such as between pro-inflammatory or phagocytotic pathways mediated by CD14, are dependent, respectively, on the interaction with ‘‘non-self ’’ and ‘‘self ’’ components (Devitt et al., 1998). The p42/44 MAPK are activated by MAPK kinase (MEK) 1 and 2 which, in turn, are activated by Raf proteins that are regulated by all four G-protein subfamilies and receptor protein tyrosine kinases (Sugden and Clerk, 1997). LPS is capable of activating p42/44 MAPK through the ras/raf-1/MEK pathway in macrophages (Geppert et al., 1994). PD98059 inhibits the phosphorylation of p42/44 MAPK (Fig. 2B) by an upstream inhibition of MEK1 (Dudley et al., 1995), and consequently, inhibits the release of PGE2 as shown in this study (Fig. 1C). However, phosphorylation of p42/44 MAPK is not suppressed by inhibitors of PKC or PC-PLC (Fig. 1E), suggesting that the p42/44 MAPK activation is independent of the PC-PLC-PKC pathway (Fig. 5). Phosphorylation of p38 MAPK is mediated through ceramides following activation of SMase as shown in Figure 4. The p38 MAPK plays a critical role in the transcriptional regulation of several pro-inflammatory genes, including COX-2 (Clark et al., 2003). SB203580-mediated inhibition of p38 MAPK decreases COX-2 mRNA and protein levels (Figs. 1A and 4D), suggesting that p38 is essential in the maintenance of COX-2 mRNA stability and efficient expression, similar to previous other observations in human monocytes (Dean et al. 1999) and neuroblastoma cells (Fiebich et al., 2000b). Exploring the role of phospholipases in PGE2 synthesis, we found that while PC-PLC transcriptionally regulated COX-2 expression, cytosolic PLA2 is involved only in the provision of arachidonic acid (Fig. 2). The PC-PLC specific inhibitor, D609, inhibited LPS-induced COX-2 expression, whereas the PLA2 inhibitor, AACOCF3, had no effect on COX-2 synthesis, although both inhibited PGE2 release. Thus, cytosolic PLA2 is important in the generation of arachidonic acid, which forms the substrate for COX-2 activity. Further, cytosolic PLA2 also activates secretory PLA2, which shows a functional coupling with COX-2, but not COX-1, in activated macrophages (Balsinde et al., 1998). Activation of PC-PLC results in the release of diacylglycerol, a strong activator

206

AKUNDI ET AL.

of PKC, which is essential for COX-2 synthesis and PGE2 release (Fig. 2E,F). However, since inhibition of PC-PLC with D609 had no effect on the activation of p38 or p42/44 MAPK, the PC-PLC-PKC pathway is induced in parallel to the MAPK pathways (Fig. 5). The regulation of PGE2 synthesis by ROS can occur through enhanced arachidonic acid release (Martinez et al., 2000) or COX-2 expression due to increased IjB degradation (Fiebich et al., 2000b). Although several sources of intracellular ROS exist, NADPH oxidasemediated ROS has been reported to be responsible for the regulation of PGE2 in microglia (Wang et al., 2004). 8-Isoprostanes are formed in response to free radical attack on arachidonic acid present on membrane phospholipids, and whose release depends on the activation of PLA2 (Morrow et al., 1990). Our results with 8-isoprostane suggest that ROS generation in LPS-induced microglia is COX-dependent (Fig. 3A). LPS strongly induced 8-isoprostane formation, which was significantly reduced by inhibition of PLA2, the enzyme responsible for the release of membrane-bound 8-isoprostanes. Similarly, G€o 6976, which showed a very potent inhibition of COX-2 protein (Fig. 2E,F), strongly inhibited 8-isoprostane formation. Since G€ o 6976 showed no significant inhibition of COX-2 enzyme activity (Fig. 3B), suppression of 8-isoprostanes must be due to an inhibition of COX-2 synthesis. The COX-2 specific inhibitor SC-58125 significantly inhibited LPS-induced ROS as well. These results suggest that 8-isoprostane generation in microglia are COX-2-dependent, similar to other reports in human monocytes (Pratico and FitzGerald, 1996). The degree of inhibition by various inhibitors used in our studies support the schematic for COX-2 expression in microglia, as shown in Figure 5, at least under in vitro conditions. Although the p38, p42/44 MAPKs and PC-PLC inhibitors independently inhibit the synthesis of PGE2, suggesting their importance in microglial activation, the absence of cross-inhibition suggests the parallel nature of these pathways. However, unlike SB203580, PD98059 shows a more potent inhibition of LPS-induced PGE2 release, supporting the view that p38 MAPK is mostly involved in posttranscriptional regulation, while p42/44 MAPK is required for COX-2 expression. The COX-2 promoter contains binding sites for several transcription factors including NF-jB and cyclic AMP responsive element binding protein (CREB), and thus, both PKC and Ras/Raf-mediated p42/44 MAPK signaling can lead to COX-2 gene transcription (Chun and Surh, 2004). In macrophages, p42/44 MAPKmediated phosphorylation of CREB is sufficient for COX-2 transcription without the requirement of NF-jB (Wadleigh et al., 2000; Eliopoulos et al., 2002). This also supports the weaker induction of PGE2 by C2-ceramide in comparison with LPS (2.5-fold versus 20-fold induction), where the activation of p42/44 MAPK by LPS, but not by ceramide, increases COX-2 expression significantly. In the case of C2-ceramide, other pathways seem to govern COX-2 expression. SMases hydrolyze membrane sphingomyelin into ceramides and phosphorylcholine (Fig. 5). Phosphorylcholine

has been shown to enhance cellular proliferation through the activation of phosphotidylinositol 3-kinase (Chung et al., 1997) and, in association with sphingosine-1-phosphate, induces p42/44 MAPK in NIH 3T3 cells (Kiss and Mukherjee, 1997). Sphingosine-1-phosphate is formed by the degradation of ceramide into sphingosine (through the action of ceramidase), followed by its phosphorylation (through sphingosine kinase), and have been implicated in several vital biological processes (Spiegel and Milstien, 2002). Sphingosine-1-phosphate has also been shown to induce COX-2 and PGE2 synthesis directly (Davaille et al., 2000; Pettus et al., 2003b). In contrast, ceramides can also be phosphorylated through the action of ceramide kinase (Bajjalieh et al., 1989). Ceramide-1-phosphate has been reported to be a potent modulator of arachidonic acid release and necessary for calcium-dependent translocation of cytosolic PLA2 in response to cytokines and calcium ionophore (Pettus et al., 2003a). Although anti-inflammatory drugs, and particularly the nonsteroidal anti-inflammatory agents (NSAIDs) such as indomethacin, showed promising results in the treatment of neurodegenerative disorders, including clinical trials, further tests with specific COX inhibitors are still under way to overcome the several side effects observed with NSAIDs (Hull et al., 2000). We have previously demonstrated an inhibition of PGE2 release through caffeine and paracetamol, alone or in combination, with acetylsalicylic acid (Fiebich et al., 2000a) and a synergistic inhibition of PGE2 release by ascorbic acid and acetylsalicylic acid in microglia (Fiebich et al., 2003). Tocopherol and its metabolites (Grammas et al., 2004), estrogen (Vegeto et al., 2001), lipocortin-1 (Minghetti et al., 1999), and several other anti-oxidants and antiinflammatory substances have been tested for the inhibition of microglial PGE2 synthesis. An alternative approach would be an interference in the intracellular signaling pathways leading to the activation of COX-2 gene. This has been shown to be effective in various cell culture and animal models, such as inhibition of PKC in the inflammatory pathways (Lahn et al., 2003; Nezhat et al., 2004). However, the ubiquitous nature of various kinases limits their therapeutic potential. In contrast, the sphingomyelin pathway has shown potential results; for instance, inhibition of de novo synthesis of ceramides prevented amyloid-induced death of hippocampal neurons (Cutler et al., 2004); acidic SMase knock-out mice which failed to show CD95-dependent apoptosis (Kirschnek et al., 2000); or novel acidic/neutral SMase inhibitors which can prevent neuronal death in cerebral ischemia (Yu et al., 2000; Soeda et al., 2004). Targeted SMase inhibition could, thus, represent a novel approach to suppress COX-2 activation and PGE2 release in microglia.

ACKNOWLEDGMENTS The authors thank Silke Hunzelmann, Brigitte G€ unter, and Ulrike G€ otzinger-Berger for skillful technical assistance. E.C.J. was supported by the Alexander

REGULATION OF COX-2 IN MICROGLIA

von Humboldt Foundation (Bonn, Germany). This work was supported, in part, by grant Fi 683/1-1 from the Deutsche Forschungsgemeinschaft.

REFERENCES Akira S. 2003. Toll-like receptor signaling. J Biol Chem 278:38105– 38108. Akiyama H, Barger S, Barnum S, Bradt B, Bauer J, Cole GM, Cooper NR, Eikelenboom P, Emmerling M, Fiebich BL, Finch CE, Frautschy S, Griffin WS, Hampel H, Hull M, Landreth G, Lue L, Mrak R, Mackenzie IR, McGeer PL, O’Banion MK, Pachter J, Pasinetti G, Plata-Salaman C, Rogers J, Rydel R, Shen Y, Streit W, Strohmeyer R, Tooyoma I, Van Muiswinkel FL, Veerhuis R, Walker D, Webster S, Wegrzyniak B, Wenk G, Wyss-Coray T. 2000. Inflammation and Alzheimer’s disease. Neurobiol Aging 21:383–421. Bajjalieh SM, Martin TF, Floor E. 1989. Synaptic vesicle ceramide kinase. A calcium-stimulated lipid kinase that co-purifies with brain synaptic vesicles. J Biol Chem 264:14354–14360. Ballou LR, Chao CP, Holness MA, Barker SC, Raghow R. 1992. Interleukin-1-mediated PGE2 production and sphingomyelin metabolism. Evidence for the regulation of cyclooxygenase gene expression by sphingosine and ceramide. J Biol Chem 267:20044–20050. Balsinde J, Balboa MA, Dennis EA. 1998. Functional coupling between secretory phospholipase A2 and cyclooxygenase-2 and its regulation by cytosolic group IV phospholipase A2. Proc Natl Acad Sci USA 95:7951–7956. Barnholz Y, Roitman A, Gatt S. 1966. Enzymatic hydrolysis of sphingolipids. II. Hydrolysis of sphingomyelin by an enzyme from rat brain. J Biol Chem 241:3731–3737. Bauer MK, Lieb K, Schulze-Osthoff K, Berger M, Gebicke-Haerter PJ, Bauer J, Fiebich BL. 1997. Expression and regulation of cyclooxygenase-2 in rat microglia. Eur J Biochem 243:726–731. Becher B, Fedorowicz V, Antel JP. 1996. Regulation of CD14 expression on human adult central nervous system-derived microglia. J Neurosci Res 45:375–381. Behl C. 2000. Apoptosis and Alzheimer’s disease. J Neural Transm 107:1325–1344. Beutler B, Rietschel ET. 2003. Innate immune sensing and its roots: the story of endotoxin. Nat Rev Immunol 3:169–176. Bieberich E, Hu B, Silva J, MacKinnon S, Yu RK, Fillmore H, Broaddus WC, Ottenbrite RM. 2002. Synthesis and characterization of novel ceramide analogs for induction of apoptosis in human cancer cells. Cancer Lett 181:55–64. Braun JS, Novak R, Herzog KH, Bodner SM, Cleveland JL, Tuomanen EI. 1999. Neuroprotection by a caspase inhibitor in acute bacterial meningitis. Nat Med 5:298–302. Buckman SY, Gresham A, Hale P, Hruza G, Anast J, Masferrer J, Pentland AP. 1998. COX-2 expression is induced by UVB exposure in human skin: implications for the development of skin cancer. Carcinogenesis 19:723–729. Candelario-Jalil E, Gonzalez-Falcon A, Garcia-Cabrera M, Leon OS, Fiebich BL. 2004. Wide therapeutic time window for nimesulide neuroprotection in a model of transient focal cerebral ischemia in the rat. Brain Res 1007:98–108. Charriaut-Marlangue C, Margaill I, Represa A, Popovici T, Plotkine M, Ben Ari Y. 1996. Apoptosis and necrosis after reversible focal ischemia: an in situ DNA fragmentation analysis. J Cereb Blood Flow Metab 16:186–194. Chun KS, Surh YJ. 2004. Signal transduction pathways regulating cyclooxygenase-2 expression: potential molecular targets for chemoprevention. Biochem Pharmacol 68:1089–1100. Chung T, Crilly KS, Anderson WH, Mukherjee JJ, Kiss Z. 1997. ATPdependent choline phosphate-induced mitogenesis in fibroblasts involves activation of pp70 S6 kinase and phosphatidylinositol 30 -kinase through an extracellular site. Synergistic mitogenic effects of choline phosphate and sphingosine 1-phosphate. J Biol Chem 272:3064–3072. Clark AR, Dean JL, Saklatvala J. 2003. Post-transcriptional regulation of gene expression by mitogen-activated protein kinase p38. FEBS Lett 546:37–44. Cutler RG, Kelly J, Storie K, Pedersen WA, Tammara A, Hatanpaa K, Troncoso JC, Mattson MP. 2004. Involvement of oxidative stressinduced abnormalities in ceramide and cholesterol metabolism in brain aging and Alzheimer’s disease. Proc Natl Acad Sci USA 101:2070–2075. Davaille J, Gallois C, Habib A, Li L, Mallat A, Tao J, Levade T, Lotersztajn S. 2000. Antiproliferative properties of sphingosine 1-phosphate in human hepatic myofibroblasts. A cyclooxygenase-2 mediated pathway. J Biol Chem 275:34628–34633.

207

Dean JL, Brook M, Clark AR, Saklatvala J. 1999. p38 mitogen-activated protein kinase regulates cyclooxygenase-2 mRNA stability and transcription in lipopolysaccharide-treated human monocytes. J Biol Chem 274:264–269. Devitt A, Moffatt OD, Raykundalia C, Capra JD, Simmons DL, Gregory CD. 1998. Human CD14 mediates recognition and phagocytosis of apoptotic cells. Nature 392:505–509. Dobrowsky RT, Werner MH, Castellino AM, Chao MV, Hannun YA. 1994. Activation of the sphingomyelin cycle through the low-affinity neurotrophin receptor. Science 265:1596–1599. Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR. 1995. A synthetic inhibitor of the mitogen-activated protein kinase cascade. Proc Natl Acad Sci USA 92:7686–7689. Eliopoulos AG, Dumitru CD, Wang CC, Cho J, Tsichlis PN. 2002. Induction of COX-2 by LPS in macrophages is regulated by Tpl2dependent CREB activation signals. EMBO J 21:4831–4840. Fassbender K, Walter S, Kuhl S, Landmann R, Ishii K, Bertsch T, Stalder AK, Muehlhauser F, Liu Y, Ulmer AJ, Rivest S, Lentschat A, Gulbins E, Jucker M, Staufenbiel M, Brechtel K, Walter J, Multhaup G, Penke B, Adachi Y, Hartmann T, Beyreuther K. 2004. The LPS receptor (CD14) links innate immunity with Alzheimer’s disease. FASEB J 18:203–205. Fiebich BL, Biber K, Lieb K, van Calker D, Berger M, Bauer J, Gebicke-Haerter PJ. 1996. Cyclooxygenase-2 expression in rat microglia is induced by adenosine A2a-receptors. Glia 18:152–160. Fiebich BL, Hull M, Lieb K, Gyufko K, Berger M, Bauer J. 1997. Prostaglandin E2 induces interleukin-6 synthesis in human astrocytoma cells. J Neurochem 68:704–709. Fiebich BL, Lieb K, Hull M, Aicher B, van Ryn J, Pairet M, Engelhardt G. 2000a. Effects of caffeine and paracetamol alone or in combination with acetylsalicylic acid on prostaglandin E2 synthesis in rat microglial cells. Neuropharmacology 39:2205–2213. Fiebich BL, Mueksch B, Boehringer M, Hull M. 2000b. Interleukin-1beta induces cyclooxygenase-2 and prostaglandin E2 synthesis in human neuroblastoma cells: involvement of p38 mitogen-activated protein kinase and nuclear factor-kappaB. J Neurochem 75:2020–2028. Fiebich BL, Lieb K, Kammerer N, Hull M. 2003. Synergistic inhibitory effect of ascorbic acid and acetylsalicylic acid on prostaglandin E2 release in primary rat microglia. J Neurochem 86:173–178. Fitzgerald KA, Palsson-McDermott EM, Bowie AG, Jefferies CA, Mansell AS, Brady G, Brint E, Dunne A, Gray P, Harte MT, McMurray D, Smith DE, Sims JE, Bird TA, O’Neill LA. 2001. Mal (MyD88-adapter-like) is required for Toll-like receptor-4 signal transduction. Nature 413:78–83. Gebicke-Haerter PJ, Bauer J, Schobert A, Northoff H. 1989. Lipopolysaccharide-free conditions in primary astrocyte cultures allow growth and isolation of microglial cells. J Neurosci 9:183–194. Geppert TD, Whitehurst CE, Thompson P, Beutler B. 1994. Lipopolysaccharide signals activation of tumor necrosis factor biosynthesis through the ras/raf-1/MEK/MAPK pathway. Mol Med 1:93–103. Gornati R, Berra B, Montorfano G, Martini C, Ciana G, Ferrari P, Romano M, Bembi B. 2002. Glycolipid analysis of different tissues and cerebrospinal fluid in type II Gaucher disease. J Inher Metab Dis 25:47–55. Grammas P, Hamdheydari L, Benaksas EJ, Mou S, Pye QN, Wechter WJ, Floyd RA, Stewart C, Hensley K. 2004. Anti-inflammatory effects of tocopherol metabolites. Biochem Biophys Res Commun 319:1047– 1052. Grullich C, Sullards MC, Fuks Z, Merrill AH Jr, Kolesnick R. 2000. CD95(Fas/APO-1) signals ceramide generation independent of the effector stage of apoptosis. J Biol Chem 275:8650–8656. Hambleton J, Weinstein SL, Lem L, DeFranco AL. 1996. Activation of c-Jun N-terminal kinase in bacterial lipopolysaccharide-stimulated macrophages. Proc Natl Acad Sci USA 93:2774–2778. Hannun YA. 1996. Functions of ceramide in coordinating cellular responses to stress. Science 274:1855–1859. Hannun YA, Bell RM. 1989. Functions of sphingolipids and sphingolipid breakdown products in cellular regulation. Science 243:500–507. Haughey NJ, Cutler RG, Tamara A, McArthur JC, Vargas DL, Pardo CA, Turchan J, Nath A, Mattson MP. 2004. Perturbation of sphingolipid metabolism and ceramide production in HIV-dementia. Ann Neurol 55:257–267. Hoozemans JJ, Rozemuller AJ, Janssen I, De Groot CJ, Veerhuis R, Eikelenboom P. 2001. Cyclooxygenase expression in microglia and neurons in Alzheimer’s disease and control brain. Acta Neuropathol (Berl) 101:2–8. Hull M, Lieb K, Fiebich BL. 2000. Anti-inflammatory drugs: a hope for Alzheimer’s disease? Expert Opin Invest Drugs 9:671–683. Joseph CK, Wright SD, Bornmann WG, Randolph JT, Kumar ER, Bittman R, Liu J, Kolesnick RN. 1994. Bacterial lipopolysaccharide has structural similarity to ceramide and stimulates ceramide-activated protein kinase in myeloid cells. J Biol Chem 269:17606–17610.

208

AKUNDI ET AL.

Karin M, Ben Neriah Y. 2000. Phosphorylation meets ubiquitination: the control of NF-[kappa]B activity. Annu Rev Immunol 18:621–663. Kawai T, Takeuchi O, Fujita T, Inoue J, Muhlradt PF, Sato S, Hoshino K, Akira S. 2001. Lipopolysaccharide stimulates the MyD88-independent pathway and results in activation of IFN-regulatory factor 3 and the expression of a subset of lipopolysaccharide-inducible genes. J Immunol 167:5887–5894. Keller M, Jackisch R, Seregi A, Hertting G. 1985. Comparison of prostanoid forming capacity of neuronal and astroglia cells in primary culture. Neurochem Int 7:655–665. Kim KS. 2003. Pathogenesis of bacterial meningitis: from bacteraemia to neuronal injury. Nat Rev Neurosci 4:376–385. Kirschnek S, Paris F, Weller M, Grassme H, Ferlinz K, Riehle A, Fuks Z, Kolesnick R, Gulbins E. 2000. CD95-mediated apoptosis in vivo involves acid sphingomyelinase. J Biol Chem 275:27316–27323. Kiss Z, Mukherjee JJ. 1997. Phosphocholine and sphingosine-1-phosphate synergistically stimulate DNA synthesis by a MAP kinasedependent mechanism. FEBS Lett 412:197–200. Kitamura Y, Shimohama S, Koike H, Kakimura J, Matsuoka Y, Nomura Y, Gebicke-Haerter PJ, Taniguchi T. 1999. Increased expression of cyclooxygenases and peroxisome proliferator-activated receptor-gamma in Alzheimer’s disease brains. Biochem Biophys Res Commun 254:582–586. Kyriakis JM, Avruch J. 2001. Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation. Physiol Rev 81:807–869. Laemmli UK. 1970. Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature 227:680–685. Lahn M, Sundell K, Moore S. 2003. Targeting protein kinase C-alpha (PKC-alpha) in cancer with the phosphorothioate antisense oligonucleotide aprinocarsen. Ann NY Acad Sci 1002:263–270. Lee JD, Ulevitch RJ, Han J. 1995. Primary structure of BMK1: a new mammalian map kinase. Biochem Biophys Res Commun 213: 715–724. Lee TC, Ou MC, Shinozaki K, Malone B, Snyder F. 1996. Biosynthesis of N-acetylsphingosine by platelet-activating factor: sphingosine CoAindependent transacetylase in HL-60 cels. J Biol Chem 271:209–217. Lehnardt S, Massillon L, Follett P, Jensen FE, Ratan R, Rosenberg PA, Volpe JJ, Vartanian T. 2003. Activation of innate immunity in the CNS triggers neurodegeneration through a Toll-like receptor 4-dependent pathway. Proc Natl Acad Sci USA 100:8514–8519. Martinez J, Sanchez T, Moreno JJ. 2000. Regulation of prostaglandin E2 production by the superoxide radical and nitric oxide in mouse peritoneal macrophages. Free Radic Res 32:303–311. Mathias S, Younes A, Kan CC, Orlow I, Joseph C, Kolesnick RN. 1993. Activation of the sphingomyelin signaling pathway in intact EL4 cells and in a cell-free system by IL-1 beta. Science 259:519–522. Minghetti L, Levi G. 1995. Induction of prostanoid biosynthesis by bacterial lipopolysaccharide and isoproterenol in rat microglial cultures. J Neurochem 65:2690–2698. Minghetti L, Polazzi E, Nicolini A, Creminon C, Levi G. 1997. Up-regulation of cyclooxygenase-2 expression in cultured microglia by prostaglandin E2, cyclic AMP and non-steroidal anti-inflammatory drugs. Eur J Neurosci 9:934–940. Minghetti L, Nicolini A, Polazzi E, Greco A, Perretti M, Parente L, Levi G. 1999. Down-regulation of microglial cyclo-oxygenase-2 and inducible nitric oxide synthase expression by lipocortin 1. Br J Pharmacol 126:1307–1314. Montine TJ, Sidell KR, Crews BC, Markesbery WR, Marnett LJ, Roberts LJ, Morrow JD. 1999. Elevated CSF prostaglandin E2 levels in patients with probable AD. Neurology 53:1495–1498. Montine TJ, Milatovic D, Gupta RC, Valyi-Nagy T, Morrow JD, Breyer RM. 2002. Neuronal oxidative damage from activated innate immunity is EP2 receptor-dependent. J Neurochem 83:463–470. Morrow JD, Hill KE, Burk RF, Nammour TM, Badr KF, Roberts LJ. 1990. A series of prostaglandin F2-like compounds are produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed mechanism. Proc Natl Acad Sci USA 87:9383–9387. Narumiya S, Sugimoto Y, Ushikubi F. 1999. Prostanoid receptors: structures, properties, and functions. Physiol Rev 79:1193–1226. Nezhat F, Wadler S, Muggia F, Mandeli J, Goldberg G, Rahaman J, Runowicz C, Murgo AJ, Gardner GJ. 2004. Phase II trial of the combination of bryostatin-1 and cisplatin in advanced or recurrent carcinoma of the cervix: a New York Gynecologic Oncology Group study. Gynecol Oncol 93:144–148. O’Banion MK, Miller JC, Chang JW, Kaplan MD, Coleman PD. 1996. Interleukin-1b induces prostaglandin G/H synthase-2 (cyclooxygenase-2) in primary murine astrocyte cultures. J Neurochem 66: 2532–2540.

Parrillo JE. 1993. Pathogenetic mechanisms of septic shock. N Engl J Med 328:1471–1477. Pettus BJ, Bielawska A, Spiegel S, Roddy P, Hannun YA, Chalfant CE. 2003a. Ceramide kinase mediates cytokine- and calcium ionophoreinduced arachidonic acid release. J Biol Chem 278:38206–38213. Pettus BJ, Bielawski J, Porcelli AM, Reames DL, Johnson KR, Morrow J, Chalfant CE, Obeid LM, Hannun YA. 2003b. The sphingosine kinase 1/sphingosine-1-phosphate pathway mediates COX-2 induction and PGE2 production in response to TNF-alpha. FASEB J 17: 1411–1421. Pfeiffer A, Bottcher A, Orso E, Kapinsky M, Nagy P, Bodnar A, Spreitzer I, Liebisch G, Drobnik W, Gempel K, Horn M, Holmer S, Hartung T, Multhoff G, Schutz G, Schindler H, Ulmer AJ, Heine H, Stelter F, Schutt C, Rothe G, Szollosi J, Damjanovich S, Schmitz G. 2001. Lipopolysaccharide and ceramide docking to CD14 provokes ligand-specific receptor clustering in rafts. Eur J Immunol 31:3153–3164. Pratico D, FitzGerald GA. 1996. Generation of 8-epiprostaglandin F2alpha by human monocytes. Discriminate production by reactive oxygen species and prostaglandin endoperoxide synthase-2. J Biol Chem 271:8919–8924. Rhee SH, Hwang D. 2000. Murine TOLL-like receptor 4 confers lipopolysaccharide responsiveness as determined by activation of NF kappa B and expression of the inducible cyclooxygenase. J Biol Chem 275:34035–34040. Seibert K, Masferrer JL. 1994. Role of inducible cyclooxygenase (COX-2) in inflammation. Receptor 4:17–23. Slawik H, Volk B, Fiebich B, Hull M. 2004. Microglial expression of prostaglandin EP3 receptor in excitotoxic lesions in the rat striatum. Neurochem Int 45:653–660. Smith WL, DeWitt DL, Garavito RM. 2000. Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem 69:145–182. Soeda S, Tsuji Y, Ochiai T, Mishima K, Iwasaki K, Fujiwara M, Yokomatsu T, Murano T, Shibuya S, Shimeno H. 2004. Inhibition of sphingomyelinase activity helps to prevent neuron death caused by ischemic stress. Neurochem Int 45:619–626. Soreghan B, Thomas SN, Yang AJ. 2003. Aberrant sphingomyelin/ceramide metabolic-induced neuronal endosomal/lysosomal dysfunction: potential pathological consequences in age-related neurodegeneration. Adv Drug Deliv Rev 55:1515–1524. Spiegel S, Milstien S. 2002. Sphingosine 1-phosphate, a key cell signaling molecule. J Biol Chem 277:25851–25854. Streit WJ, Mrak RE, Griffin WS. 2004. Microglia and neuroinflammation: a pathological perspective. J Neuroinflammation 1:14. Subbaramaiah K, Chung WJ, Dannenberg AJ. 1998. Ceramide regulates the transcription of cyclooxygenase-2. Evidence for involvement of extracellular signal-regulated kinase/c-Jun N-terminal kinase and p38 mitogen-activated protein kinase pathways. J Biol Chem 273:32943–32949. Sugden PH, Clerk A. 1997. Regulation of the ERK subgroup of MAP kinase cascades through G protein-coupled receptors. Cell Signal 9:337–351. Takadera T, Shiraishi Y, Ohyashiki T. 2004. Prostaglandin E2 induced caspase-dependent apoptosis possibly through activation of EP2 receptors in cultured hippocampal neurons. Neurochem Int 45: 713–719. Tatton WG, Chalmers-Redman R, Brown D, Tatton N. 2003. Apoptosis in Parkinson’s disease: signals for neuronal degradation. Ann Neurol 53(Suppl 3):S61–S70. Teismann P, Tieu K, Choi DK, Wu DC, Naini A, Hunot S, Vila M, Jackson-Lewis V, Przedborski S. 2003. Cyclooxygenase-2 is instrumental in Parkinson’s disease neurodegeneration. Proc Natl Acad Sci USA 100:5473–5478. Vegeto E, Bonincontro C, Pollio G, Sala A, Viappiani S, Nardi F, Brusadelli A, Viviani B, Ciana P, Maggi A. 2001. Estrogen prevents the lipopolysaccharide-induced inflammatory response in microglia. J Neurosci 21:1809–1818. Wadleigh DJ, Reddy ST, Kopp E, Ghosh S, Herschman HR. 2000. Transcriptional activation of the cyclooxygenase-2 gene in endotoxin-treated RAW 264.7 macrophages. J Biol Chem 275:6259–6266. Wang T, Qin L, Liu B, Liu Y, Wilson B, Eling TE, Langenbach R, Taniura S, Hong JS. 2004. Role of reactive oxygen species in LPSinduced production of prostaglandin E2 in microglia. J Neurochem 88:939–947. Yu ZF, Nikolova-Karakashian M, Zhou D, Cheng G, Schuchman EH, Mattson MP. 2000. Pivotal role for acidic sphingomyelinase in cerebral ischemia-induced ceramide and cytokine production, and neuronal apoptosis. J Mol Neurosci 15:85–97.