Address correspondence and reprint requests to Bernd L. Fiebich, PhD, Department of Psychiatry, University of Freiburg Medical School, Hauptstr. 5, D-79104 Freiburg, Germany. E-mail: firstname.lastname@example.org
Both interleukin-1β (IL-1β) and prostaglandins (PGs) are important mediators of physiological and pathophysiological processes in the brain. PGE2 exerts its effects by binding to four different types of PGE2 receptors named EP1–EP4. EP3 has found to be expressed in neurons, whereas expression of EP3 in glial cells has not been reported in the brain yet. Here we describe IL-1β-induced EP3 receptor expression in human astrocytoma cells, primary astrocytes of rat and human origin and in rat brain. Using western blot, we found a marked up-regulation of EP3 receptor synthesis in human and rat primary glial cells. Intracerebroventricular administration of IL-1β stimulated EP3 receptor synthesis in rat hippocampus. The analysis of involved signal transduction pathways by pathway-specific inhibitors revealed an essential role of protein kinase C and nuclear factor-κB in astrocytic IL-1β-induced EP3 synthesis. Our data suggest that PGE2 signaling in the brain may be altered after IL-1β release due to up-regulation of EP3 receptors. This might play an important role in acute and chronic conditions such as cerebral ischemia, traumatic brain injury, HIV-encephalitis, Alzheimer's disease and prion diseases in which a marked up-regulation of IL-1β is followed by a prolonged increase of PGE2 levels in the brain.
An increased expression of cyclooxygenase (COX), the rate-limiting enzyme of PGE2 synthesis and an elevation of PGE2 levels have been observed in chronic progressive neurodegenerative disorders such as Alzheimer's disease or prion's disease (Minghetti et al. 2000; Hoozemans et al. 2001) and during acute conditions leading to neuronal loss such as cerebral ischemia (Iadecola et al. 1999) or traumatic brain injury (Kunz et al. 2002). Non-steroidal anti-inflammatory drugs (NSAIDs), which inhibit COX and consecutive PGE2 release, have shown promising results as possible therapeutic tools in brain ischemia (Iadecola and Alexander 2001; Candelario-Jalil et al. 2004) and models of Alzheimer's disease (Lim et al. 2000; Hull et al. 2002), pointing towards a key role of prostaglandins as mediators of neuroinflammatory reactions.
PGE2 exerts its diverse effects by binding to four different EP receptor subtypes, named EP1–EP4, resulting in the activation of different intracellular signal transduction pathways. EP2 and EP4 receptors couple to Gs protein leading to the elevation of cAMP, whereas stimulation of EP1 receptors leads to increases in intracellular calcium through a Gq independent mechanism. Moreover, the EP3 receptor exists in multiple splice variants generated by alternative splicing of the C-terminal tail. In humans, at least eight isoforms have been identified (Kotani et al. 1997). Although most of them have been shown to couple to Gi proteins decreasing cAMP levels, individual splice variants seem to be capable of IP3 generation and cAMP stimulation (Hata and Breyer 2004). In contrast to EP1, EP2 and EP4 receptors, which have been found in restricted neuronal cell groups only, EP3 receptors show a widespread constitutive expression in neurons throughout the brain (Zhang and Rivest 1999; Nakamura et al. 2000; Candelario-Jalil et al. 2005).
While numerous studies have focused on the expression and regulation of COX in neuroinflammatory diseases, it is only recently that the role of individual EP receptors in mediating neurotoxic or neuroprotective properties of PGE2 and their differential expression under proinflammatory stimuli is gaining more attention.
In this study, we focused on the regulation of the EP3 receptor in astrocytes. Although the expression of EP3 receptors in glial cells in vivo has not been reported yet, detection of EP3 mRNA in cultured glial cells suggests that EP3 receptors may be induced in this cell type under pathophysiological conditions (Kitanaka et al. 1996). One of the key mediators in neuroinflammatory reactions seems to be interleukin-1 (IL-1) (Basu et al. 2004). IL-1 can be synthesized by a wide variety of cells in the CNS and IL-1 receptors have been found on neurons and glial cells. IL-1β is expressed constitutively only at low levels in the healthy brain, but it is rapidly induced during neuroinflammatory reactions to ischemia or traumatic brain injury, and a chronic overexpression of IL-1 has been observed in progressive degenerative brain disorders (Rothwell and Luheshi 2000). We were interested if IL-1β affects the synthesis and expression of EP3 receptors in glial cells, therefore we investigated the influence of the cytokine on primary human and rat glial cells and elucidated the signal transduction mechanisms leading to EP3 synthesis in the human astrocytoma cell line U373 MG. In addition, we studied the effects of intracerebroventricular IL-1β injections on EP3 receptor synthesis in the rat brain.
Materials and methods
IL-1β was obtained from Roche (Mannheim, Germany) for cell culture experiments and from R & D Systems (Minneapolis, MN, USA) for in vivo experiments. Inhibitors of p38 mitogen-activated protein kinase (MAPK) (SB202190), p42/44 MAPK (PD98059) and c-jun N-terminal kinase (JNK) (SP600125), inhibitors of protein kinase C (PKC) (GF109203X, Gö6976, Gö6983, Ro-31-8425, rottlerin) and the proteasome inhibitor MG-132 were obtained from Calbiochem (Bad Soden, Germany). The sesquiterpene lactone parthenolide was obtained from Sigma-Aldrich (Taufkirchen, Germany). SC-514 was kindly provided by Dr Kishore (Department of Arthritis and Inflammation Pharmacoloy, Pharmacia Corp. St. Louis, MO, USA). Stock solutions (10 mm) were prepared in dimethylsulfoxide (PKC inhibitors, MAPK inhibitors, MG-132, parthenolide) and stored at −20°C. Further dilutions were carried out in dimethylsulfoxide or distilled water. Effects of dimethylsulfoxide alone in the used range of concentrations were not detectable.
Cell Titer-Glo™ cell viability assays (Promega, Mannheim, Germany) were used to determine cytotoxic effects of the various inhibitors. Assays were performed according to the manufacturer's instructions.
The human astrocytoma cell line U373 MG was obtained from the American Type Culture Collection (Rockville, MD, USA) and cultured in 75-cm2 culture flasks (Greiner, Frickenhausen, Germany). Cells were grown (at 37°C; 5% CO2) as monolayer cultures in modified Eagle's medium (Earle's MEM; PAA, Coelbe, Germany) containing 5% fetal calf serum (PAN, Aidenbach, Germany), l-glutamine, antibiotics, vitamins, essential amino acids and pyruvate (Invitrogen, Karlsruhe, Germany). For stimulation experiments, cells were trypsinized and seeded into six-well plates (2 × 105 cells per well) for protein analysis or in 10-cm diameter dishes (1 × 106 cells per dish) for RNA extraction. Cultures were grown for 3–4 days. Medium was changed 2 h prior to cell treatment.
Primary human astrocyte cultures were established and characterized as described previously (De Groot et al. 1997). Tissue was obtained from the cerebral white matter or prefrontal cortex with a relatively short postmortem delay (the Netherlands Brain Bank, Amsterdam, the Netherlands; coordinator Dr R. Ravid). Tissue specimens were collected in Dulbecco's modified Eagle's medium (DMEM)/nutrient mixture HAM's F-10 (HAM-F10) (Gibco Life Technologies, Breda, the Netherlands) (1 : 1) containing 50 µg/mL gentamycin (Sigma). Tissue devoid of meninges and blood vessels was minced and subsequently digested for 20 min at 37°C in Hank's balanced salt solution (HBSS) containing 0.025 mg/mL trypsin (Sigma) and 0.1 mg/mL bovine pancreatic DNase I (Roche). After digestion, cell suspensions were washed with culture medium [DMEM/HAMF-10; 1 : 1, supplemented with 2 mm l-glutamin (Gibco), 10% fetal bovine serum (Integro, Zaandam, the Netherlands), 100 U/mL penicillin (Sigma) and 50 µg/mL streptomycin (Sigma)] and transferred to poly-l-lysine (15 µg/mL, Sigma) coated 80-cm2 culture flasks (Greiner, Alphen a/d Rijn, the Netherlands), whereafter cells were grown (at 37°C; 5% CO2) as monolayer cultures in culture medium. More than 95% of the human astrocyte cultures from several passages expressed the astrocyte-specific markers glial fibrillary acidic protein (GFAP) (Dako, Glostrup, Denmark), glutamine synthetase (Affinity, Nottingham, UK) and vimentin (Roche). No staining for LeuM5 (CD11c) and LeuM3 (CD14) (both obtained from Becton Dickinson, Erembodegem, Aalts, Belgium) was seen, indicating that contamination with macrophage/microglia was negligible. Three primary astrocyte cultures (passage 3–5) established from three different postmortem brains were trypsinized and transferred to six-well plates (Falcon) at 8 × 104 cells/well or to 10-cm diameter dishes (4 × 105 cells/dish). Cells were grown for 1 week. 14 h prior to stimulation experiments medium was replaced by culture medium with a reduced serum concentration (5%).
Primary mixed astroglial cell cultures were established from midbrain and cortex of 1-day neonatal Sprague-Dawley rats, which were gently dissociated by repeated pipetting in Dulbecco's phosphate-buffered saline (Cell Concepts, Umkirch, Germany) and filtered through a 70 µm cell strainer (Falcon). Cells were collected by centrifugation (200 g, 10 min), resuspended in DMEM (30 mL/midbrain; 60 mL/cortex) containing 10% fetal calf serum (Sigma) and antibiotics, and cultured on 10-cm diameter cell culture dishes (Falcon) or in 6-well plates in 5.5% CO2 at 37°C. Medium was prepared taking extreme care to avoid lipopolysaccharide contamination (Gebicke-Haerter et al. 1989). Cultures were grown for 4 weeks. Medium was changed on the second day, afterwards once a week and 14 h before treatment.
Primary rat microglia cell cultures were prepared as described elsewhere (Gebicke-Haerter et al. 1989). In brief, microglia were harvested from 2 to 8-weeks-old mixed glial cultures derived from the forebrain of 1-day postnatal Sprague Dawley pups, by carefully collecting the medium. Cells were then re-seeded into 75-cm2 culture flasks in fresh medium to give pure microglial cultures.
Oligodendrocyte precursor cells (OPCs) enriched cultures were prepared according to a modified protocol based on the method of McCarthy and de Vellis (McCarthy and de Vellis 1980; Othman et al. 2003). In brief, rat mixed glial cultures were prepared as described above and grown in 75-cm2 culture flasks until cultures reached confluence. To remove microglia from the mixed glial cultures, flasks were shaken on a rotary shaker for 60 min at 200 r.p.m. Medium was renewed and cultures were further shaken for 16 h at 250 r.p.m. to isolate OPCs. Floating cells were transferred to 10-cm dishes. After 60 min, the loosely attached OPCs were detached by gentle shaking, leaving behind firmly attached microglia and astrocytes. OPCs were then reseeded on culture dishes. Eight hours after reseeding, immunocytochemistry and stimulation experiments were performed. About 55% of the cells expressed the OPCs-specific marker A2B5 (Chemicon, Chandler's Ford, Hampshire, UK), about 40% of the cells were positive for the microglia marker ED-1 (Serotec, Düsseldorf, Germany) and less than 5% of the cells in the OPCs-enriched cultures expressed the astrocyte-specific marker GFAP (DAKO, Hamburg, Germany). All experiments were carried out according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.
Primary human monocytes (CD4/CD8 positive mononuclear cells) were prepared from buffy coats from healthy blood donors and were isolated by using the Ficoll-gradient isolation method. Cells were then seeded in 10-cm diameter dishes (107 cells/mL) (Falcon) and were allowed to attach for 1 h. After two washing steps, fresh medium (low-endotoxin RPMI 1640) (Gibco, Paisley, Scottland) containing 1% human serum was added and the cells were incubated at 37°C, 5% CO2 for 2 h before cell lysis.
RNA extraction and reverse transcription–polymerase chain reaction analysis
Total RNA was extracted using the guanidine isothiocyanate method according to Chomczynski and Sacchi (1987). For RT–PCR, 2 µg of total RNA was reverse transcribed using M-MLV reverse transcriptase (Promega) and random hexamers (Promega). PCR was carried out using Taq polymerase (Promega), dNTP master mix (Invitek, Berlin, Germany). Primers were designed using PrimerSelect Software from DNA Star Inc. (Madison, WI, USA) (see Table 1). For EP3 transcript variants, we used the nomenclature used in the NCBI database; EP3 splice variants are numbered as published by Kotani et al. (1997). Primer sequences for the human EP1 receptor and EP3 transcript variants 3, 5 and 7 have been described previously (Schlotzer-Schrehardt et al. 2002; Wing et al. 2003).
Table 1. Primer sequences and annealing temperatures of the respective primers used to amplify each EP receptor and the EP3 receptor splice variants
EP receptor primer
EP3 variant 3
EP3 variant 5
EP3 variant 7
EP3 variant 4,9,10
Equal equilibration was determined using S12 primers (human: forward, 5′-GGAGGTGTAATGGACGTTA-3′; reverse, 5′-CTGAGACTCCTTGCCATAG-3′, 56°C, 25 cycles, product length: 332 bp, rat: forward, 5′-ACGTCAACACTGCTCTACA-3′; reverse, 5′-CTTTGCCATAGTCCTTAAC-3′, 56°C, 25 cycles product length: 312 bp) or primers for β-actin (human: forward, 5′-ATCTGGCACCACACCTTCTACAATGAGCTGCG-3′; reverse, 5′-CGTCATACTCCTGCTTGCTGATCCACATCTGC-3′, 60°C, 25 cycles, product length 838 bp; rat: forward, 5′-ATGGATGACGATATCGCT-3′; reverse, 5′-ATGAGGTAGTCTGTCAGGT-3′, 48°C, 35 cycles, 569 bp). PCR products were separated electrophoretically on a 2% agarose gel. Potential contamination by genomic DNA was controlled by omitting reverse transcriptase and using S12 or β-actin primers in the subsequent PCR amplification.
Western blot analysis
For stimulation experiments, cells were exposed to IL-1β in the presence or absence of the specific inhibitors for the indicated periods of time. Cells were washed with ice-cold phosphate-buffered saline (Cell Concepts). Cell lysis was conducted using sodium dodecyl sulfate-sample buffer (42 mm Tris-HCL pH 6.8, 1.3% sodium dodecyl sulfate buffer, 100 µm orthovanadate, 6.5% Glycerine) (Laemmli 1970) and samples were homogenized by repeated pipetting followed by an incubation step at 95°C for 5 min. Protein content was determined using the bicinchoninic acid method (BCA protein assay kit, Pierce, distributed by KFC Chemikalien, München, Germany). For EP3 protein analysis, 65 µg of total cell protein from U373 MG derived samples, 55 µg cell protein from rat mixed astroglial cultures and 35 µg from human primary astrocyte cultures were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis on a 7.5% gel under reducing conditions. Proteins were transferred onto a polyvinylidene fluoride membrane (Millipore, Bedford, MA, USA) by semidry blotting. The membrane was blocked overnight at 4°C using Rotiblock (Roth, Karlsruhe, Germany). Rabbit anti-EP3 receptor polyclonal antibody (Cayman, Ann Arbor, MI, USA) was diluted (1 : 750) in Tris-buffered saline–Tween containing 1% bovine serum albumin, rabbit anti-IκBα (C-21) was used in a 1 : 600 dilution (Santa Cruz Biotechnology, Heidelberg, Germany) and rabbit anti-actin (20–33, Sigma-Aldrich, Taufkirchen, Germany) was diluted 1 : 500. Membranes were incubated for 2 h at room temperature with the primary antibody, and for 1 h with the secondary antibody horseradish peroxidase-linked anti-rabbit IgG (Amersham-Biosciences, Freiburg, Germany). Subsequent detection was performed using the ECL western blotting system (Amersham-Biosciences) according to the manufacturer's instructions. All western blot experiments were carried out at least three times.
Quantification of the western blots was performed using ScanPack 3.0 software (Biometra, Göttingen, Germany). Relative immunoreactivity was evaluated as band intensity in percentage of an unspecific control band within the same blot. Equal protein loading and transfer were assessed by subjection of each sample to an actin western blot. For descriptive purposes, means and SD were calculated and Student's t-test was performed to assess significance. A value of p < 0.05 was considered to be statistically significant.
In vivo experiments
All experiments were carried out according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Ethical Committee of the National Laboratory Center, University of Kuopio, Finland. Nine male Wistar rats weighing 250–350 g were used. Animals were housed at a standard temperature (22 ± 1°C) in a light-controlled environment (lights on from 07:00 h to 21:00 h) with access to food and water ad libitum. The rats were anaesthetized (3 mL/kg equithesin, i.p.) and immobilized in a stereotaxic frame. A midline skin incision and a craniectomy of 2 mm diameter were performed at −0.9 mm anterioposterior and 1.4 mm lateral to the midline. Coordinates for the injection were −0.92 mm anterioposterior, 1.6 mm mediolateral and 3.3 mm dorsoventral to the bregma and from dura according to the atlas of Paxinos and Watson. Then 20 ng of IL1-β (R & D Systems) diluted in 10 µL volume or the vehicle solution (phosphate-buffered saline) only was infused into the right ventricle at 1 µL/min speed with a nanolitre infusion pump (TSE Systems, Bad Homburg, Germany). The cannula was left in place for another 5 min before being withdrawn. The exposed cortex was covered with repair material, the wound closed and disinfected. Animals were allowed to recover from anesthesia, returned to the cages and allowed free access to food and water. At 24 h postinjection, animals were decapitated; the brains were removed and dissected on ice. The right hippocampus was collected, frozen and stored at −80°C until further processing. Hippocampi were homogenized using a Branson Sonifier 250. Proteins were isolated using Trizol LS Reagent (Invitrogen) according to the manufacturer's instructions and resuspended in sodium dodecyl sulfate-sample buffer. Protein contents were determined using BCA-Protein Assay Kit (Pierce) and 40 µg of the protein extracts/lane were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis on a 10% gel under reducing conditions. Western blotting was performed as described above. For comparative analysis, the membranes were stripped and reprobed. Membranes were covered with elution buffer (62.5 mm Tris-HCL, pH 6.7, 2% sodium dodecyl sulfate, 100 mm 2-mercaptoethanol) and incubated under gentle shaking at 70°C for 60 min. Membranes were washed in water and Tris-buffered saline–Tween and reprobing was performed using a GFAP antibody (Dako, Hamburg, Germany) diluted 1 : 10 000.
Expression of prostaglandin E receptors in glial cells
We investigated the expression of the four known EP receptors, EP1–EP4, in glial cells of human and rat origin. Human primary monocytes (Fig. 1a) were used as control, which showed the known expression of EP2, EP3 and EP4 but not of EP1 receptors (Iwasaki et al. 2003). U373 MG astrocytoma cells did not express EP1 receptor mRNA, but expressed EP2 and EP4 receptor mRNA as detected by RT–PCR, whereas EP3 receptor mRNA was visible to a lower extent (Fig. 1a). In primary human astrocytes we were able to demonstrate that EP receptor mRNA expression was similar to the expression pattern of EP receptor mRNA in U373 MG astrocytoma cells (Fig. 1b). Both, EP2 and EP4 receptor mRNAs were expressed in untreated cultured primary human astrocytes. Although EP3 mRNA expression was detectable in untreated cultures from donor A (left side), it appeared to be absent in cultures from donor B (right side). In parallel to U373 MG astrocytoma cells, primary human astrocytes did not express EP1 receptor mRNA. We further studied the expression of EP receptor mRNA in primary rat mixed cortical and midbrain glial cultures and microglia (Fig. 1c). We found that rat mixed glial cultures either derived from cortical or midbrain origin expressed mRNA for all four EP receptors (Fig. 1c, upper panel). Rat microglia cultures expressed EP1 and EP2 receptors (Fig. 1c, second panel), as has been previously reported (Caggiano and Kraig 1999). In contrast to the prior study, we also found an expression of EP4 receptor mRNA in rat microglia.
Interleukin-1β induces EP3 receptor synthesis in U373 astrocytoma cells and primary astrocytes
We were interested if IL-1β is able to regulate the expression of EP3 receptors in glial cells. We found that IL-1β dose-dependently induced EP3 immunoreactivity (Fig. 2a) in U373 MG astrocytoma cells. Maximal EP3 levels were obtained with 50 U/mL of IL-1β, and this dose was then used in most other experiments. A time kinetic experiment revealed that EP3 synthesis reached maximal levels after 24 h. EP3 immunoreactivity was visible for at least 72 h after IL-1β treatment (Fig. 2b). Densitometrical analysis demonstrated an eight-fold increase of EP3 protein synthesis (Fig. 2c) in U373 MG astrocytoma cells after stimulation with 50 U/mL IL-1β for 24 h (n = 6, p < 0.0001). We next were interested if IL-1β-induced EP3 protein synthesis is paralleled by expression of EP3 mRNA. Using qualitative RT–PCR with specific primers for the human EP3 receptor, we found that IL-1β induced EP3 mRNA expression in U373 MG astrocytoma cells (Fig. 2d). EP3 mRNA expression was detectable after 4 h of IL-1β treatment, lasted for 48 h and was no longer detectable after 72 h of IL-1β treatment (Fig. 2d). By using specific primers for different EP3 receptor transcript variants as described by Wing and colleagues (Wing et al. 2003), we identified the splice variants of the EP3 receptor expressed in U373 astrocytoma cells (Fig. 2e). We found mRNA transcript variants 3 and 5 encoding for the receptor isoforms EP3-VI and EP3-II, respectively. Furthermore, mRNA for transcript variants 4, 9 and 10, which all encode for the same EP3 receptor isoform, namely isoform EP3-I, was detected in the astrocytic cell line U373 MG. Transcript variant 7, resulting in isoform EP3-III, was not expressed in U373 MG astrocytoma cells, although a PCR product for transcript variant 7 was detected in the neuroblastoma cell line SK-N-SH used as a positive control (Fig. 2e). In contrast to the splice variants 3, 4, 9 and 10, splice variant 5 was not induced by IL-1β (Fig. 2e).
In parallel to our findings in U373 astrocytoma cells, IL-1β also induced EP3 protein synthesis in primary human astrocytes (Fig. 2f). We investigated astrocyte cultures obtained from postmortem tissues of three different donors. An example of IL-1β-induced EP3 immunoreactivity in primary human astrocytes from one donor (donor A) is shown in Fig. 2(f). Four hours after stimulation with 50 U/mL IL-1β, an EP3-immunoreactive band was visible at an apparent molecular weight of 72 kDa (lane 2), which corresponds to the molecular weight of EP3-immunoreactivity in U373 MG astrocytoma cells (Fig. 2a) and rat kidney (data not shown). This molecular weight has previously been shown to be a major immunoreactive band for EP3 receptors in the rat kidney (Nakamura et al. 1999). The densitograph (Fig. 2g) obtained from the western blot of the three brain-derived human astrocyte cultures revealed that IL-1β induced an approximately three-fold increase in EP3 immunoreactivity in human primary astrocytes (n = 3, ***p < 0.001).
Interleukin-1β induces EP3 receptor mRNA and protein synthesis in rat mixed glial cultures
Stimulation with IL-1β led to a time-dependent increase in EP3 receptor protein synthesis in rat mixed glial cell (Fig. 3a). In line with the results obtained at the protein level, EP3 receptor mRNA expression was detected by qualitative RT–PCR in rat glial cultures after treatment with IL-1β (Fig. 3b). There were no differences between primary rat mixed glial cultures derived from cortical or midbrain tissues concerning the action of IL-1β on the expression of the EP3 receptor (data not shown). IL-1β did not induce EP3 expression in primary rat microglia (Fig. 3c). EP3 receptor mRNA was detected in OPCs (oligodendrocyte precursor cells) enriched cultures (approx. 55% 2A5B positive OPC's, 40% ED-1 positive microglia) under basal conditions and treatment with IL-1β had no influence on the expression of EP3 receptor mRNA in these cultures (Fig. 3d).
Protein kinase C and nuclear factor-κB are key signal transducers of interleukin-1β-induced EP3 synthesis
We were further interested in the signal transduction pathways involved in IL-1β-induced EP3 synthesis in astrocytes. IL-1β is known to induce various signaling pathways in astrocytes including PKC and the MAP kinases p38 MAPK and p42/44 MAPK (Moynagh et al. 1993; Zhang et al. 1996; Molina-Holgado et al. 2000). We studied several inhibitors of those kinases in the presence of IL-1β in U373 MG cells. Whereas the MAPK inhibitors SB202190 (p38 MAPK), PD98059 (p42/44 MAPK) and SP600125 (JNK) had no effect on IL-1β-induced EP3 synthesis up to concentrations of 10 µm (data not shown), various PKC inhibitors prevented IL-1β-mediated EP3 protein synthesis. We used several unselective inhibitors of PKC activity: GF109203X, all PKC isoforms; Ro-31-8425, all conventional PKCs and PKC ε (Way et al. 2000); Gö6983, all conventional and novel PKCs (Gschwendt et al. 1996). We also used two relatively selective PKC inhibitors: rottlerin, preferential for PKC δ (Gschwendt et al. 1994) and Gö6976, a PKC α/βΙ preferential inhibitor (Martiny-Baron et al. 1993).
As shown in Figs 4(a) and (b), GF109203X and Gö6983 strongly inhibited IL-1β-induced EP3 synthesis in U373 MG astrocytoma cells at a dose of 10 µm. In addition, the unselective PKC inhibitor Ro-31-8425 proved to be a potent inhibitor of IL-1β−induced EP3 receptor synthesis (Fig. 4c), whereas the PKC alpha/beta inhibitor Gö6976 had only a slight inhibitory effect at a dose of 10 µm (Fig. 4a). Rottlerin dose-dependently reduced IL-1β-induced EP3 protein synthesis with a marked reduction at a concentration of 1 and 10 µm (Fig. 4d). Statistical analysis of densitographic scans of independent western blots revealed a marked reduction of EP3 immunoreactivity by inhibitors of PKC reaching 70%, 64%, 84% with 10 µm of GF109203X, Gö6983 and Ro-31-8425, respectively. The inhibitor of conventional PKC isoforms, Gö6976, reduced IL-1β-induced EP3 receptor immunoreactivity by 45% at 10 µm only. Rottlerin showed a nearly complete inhibition of IL-1β-induced EP3-immunoreactivity at 1 µm. Cell viability assays based on the quantitation of ATP as an indicator of the number of metabolic active cells were performed to exclude that the observed effects of the various PKC inhibitors on EP3 protein synthesis were only secondary to toxic effects of the substances (data not shown).
IL-1β is known to induce nuclear factor-κB (NF-κB) activation in a variety of cell types (O'Neill and Greene 1998). Using three inhibitors of the NF-κB pathway, we found that parthenolide, which has been shown to interact on different levels with the NF-κB complex and thereby inhibits NF-κB activation (Hehner et al. 1998, 1999), the IκB kinase-2 (IKK-2) inhibitor SC-514 (Kishore et al. 2003) and the proteasome inhibitor MG-123 (Palombella et al. 1994) were able to block IL-1β-induced EP3 protein synthesis, suggesting a possible involvement of the NF-κB pathway in IL-1β-mediated EP3 synthesis (Fig. 5). Parthenolide dose-dependently inhibited IL1-β-induced EP3 immunoreactivity (Fig. 5a). Total inhibition of EP3 receptor synthesis by parthenolide was achieved at a concentration of 10 µg/mL (Fig. 5a, lane 6). This concentration of parthenolide was sufficient to prevent IL-1β-induced IκBα degradation in U373 MG in our experiments (Fig. 5b). Using statistical analysis of densitographic scans, we demonstrated that parthenolide causes a highly significant reduction of IL-1β-induced EP3 immunoreactivity (p < 0.0001) (Fig. 5c). The compound SC-514 has previously been described to be a potent IKK-2 inhibitor (Kishore et al. 2003). Here we show that SC-514 had an inhibitory effect on IL-1β-induced EP3 synthesis in a concentration of 50 µm (Fig. 5d, lane 5). Total inhibition was achieved by using 100 µm (Fig. 5d, lane 6). The densitograph of three different western blot experiments revealed a significant inhibition of IL-1β-induced EP3 immunoreactivity at 100 µm of SC-514 (Fig. 5e). The proteasome inhibitor MG-132 also completely inhibited IL-1β mediated EP3 receptor induction when used at concentrations of 1 µm or 5 µm (Fig. 5f, lanes 5 and 6). We found that these concentrations were sufficient to block IL1-β-induced IκBα degradation (Fig. 5g). Densitometrical and statistical analysis of western blot experiments with MG-132 showed a marked reduction of IL-1β-induced EP3 immunoreactivity, which was reduced to basal levels (Figs 5h, n = 3, p < 0.005). Whereas cell viability assays revealed mild cytotoxic effects of MG-132 at the doses used in these experiments, exposure to SC-514 or parthenolide did not seem to affect the viability of the cells (data not shown).
Interleukin-1β increases EP3 protein levels in vivo
In extension of our in vitro investigations, we investigated the effect of IL-1β on EP3 receptor protein levels in the CNS in vivo. Therefore, we injected 20 ng (the equivalent to 1000 U) of IL-1β or vehicle solution alone into the right lateral ventricle of adult Wistar rats. After a period of 24 h, during which the animals were allowed to recover, the rats were decapitated and brains were removed and dissected. Protein isolated from homogenized tissue of the right hippocampus was used for western blot analysis (Fig. 6a). For comparative analysis, membranes were reprobed using a GFAP antibody. After normalization using GFAP as an internal standard, we observed an approximately 20-fold increase in hippocampal EP3 receptor protein levels by the injection of the vehicle solution alone, which might be caused by the induction of local cytokine synthesis due to the canula lesion (Fig. 6a, lanes 2 and 5, Fig. 6b). Injection of IL-1β significantly further increased EP3 immunoreactivity in the rat hippocampus by two-fold when compared to the group that had experienced the canula lesion and vehicle injection only (Fig. 6a, lanes 3 and 4, Fig. 6b).
In this study, we show for the first time that IL-1β induces EP3 receptor mRNA and protein in primary human astrocytes, primary rat glial cells and human U373 MG astrocytoma cells. U373 MG cells have been shown previously to react comparable to primary astrocytes when stimulated with IL-1β (Lieb et al. 1996a,b) or PGE2 (Fiebich et al. 1997, 2001).
In all investigated astrocytic cell cultures, we found a basal expression of EP2 and EP4 mRNA. Whereas primary human astrocytes and human U373 MG astrocytoma cells did not show any basal expression of EP1 receptor mRNA, EP1 receptor mRNA was detectable in primary mixed glial cultures from the rat. This difference in basal expression of EP1 receptors between cultured human and rat glial cells might either be due to species-specific differences in EP1 expression in astrocytes or due to the presence of microglia in these mixed glial cultures, which express EP1 (Caggiano and Kraig 1999).
In all studied astroglial cell cultures, IL-1β led to a time and concentration-dependent increase in EP3 immunoreactivity. In western blot experiments, a major EP3 immunoreactive band was detected at an apparent molecular weight of approximately 72 kDa in U373 MG cells, rat glial cells and human astrocyte cultures, similar to findings in rat kidney and COS-7 cells (Nakamura et al. 1999). EP3 immunoreactivity in the rat brain was detectable at an apparent molecular weight of 60 kDa, similar to previous studies that also showed additional bands at lower molecular weights (Nakamura et al. 1999). The observed differences between the apparent molecular weights detected in sodium dodecyl sulfate–gel electrophoresis and the predicted molecular mass of the human (40.5–47.3 kDa) or rat (39.5–39.9 kDa) EP3 receptor isoforms may be due to post-translational addition of carbohydrate moieties. Differences between glial cells derived from different brain regions in response to IL-1β have been described; however, we did not find significant differences between cortical and midbrain glial cultures concerning the responsiveness to IL-1β with regard to EP3 synthesis. However, a selective response in smaller specialized brain regions would have escaped the detection in our broader sampling of subcortical tissue.
As mixed rat glial cultures contain mainly astrocytes and also oligodendrocyte precursor cells (OPCs) and microglia, we investigated whether IL-1β influences the expression of EP3 receptors in theses cell types. In pure microglial cultures, IL-1β did not induce the expression of EP3 receptor mRNA. Direct effects of IL-1β on cultured microglia remain questionable as IL-1β failed to activate intracellular signal transduction pathways in murine microglia cells in a recent study (Pinteaux et al. 2002). However, microglial EP3 expression is inducible in vivo by excitotoxins, indicating that EP3 receptors are regulated in this cell type (Slawik et al. 2004).
In contrast to microglia, OPCs have been reported to express the IL-1 receptor type I (Vela et al. 2002). In our hands, OPC-enriched glial cultures showed a basal expression of EP3 without any influence of an additional stimulation with IL-1β. However, refined methods to obtain pure OPCs are necessary to study the regulation of EP receptors in depth in this cell type.
Alternative splicing of the C-terminal tail generates multiple splice variants of the EP3 receptor. Although originally the EP3 receptor has been described to couple to Gi proteins, some data suggest that activation of individual EP3 isoforms can also lead to an increase of cAMP or intracellular calcium (Hata and Breyer 2004). We focused in this study on the expression of the isoforms EP3-I, EP3-II and EP3-III, which inhibit cAMP formation and increase intracellular calcium levels (An et al. 1994), and on the isoform EP3-VI, which mediates both, increases and decreases of cAMP concentrations (Kotani et al. 2000). In unstimulated U373 MG astrocytoma cells, we detected mRNA encoding for the isoforms EP3-I, EP3-II and EP-VI, but not for EP3-III. Interestingly, transcript variant 5 was not induced by IL-1, whereas expression of variants 3, 4, 9 and 10 was up-regulated by IL-1β, suggesting that EP3 splice variants are differentially regulated by IL-1β. However, the physiological significance of the differences in EP3 receptor isoform signaling needs to be elucidated in more detailed studies using overexpression or knock-down of individual EP3 isoforms.
Effects of IL-1β on cultured glial cells are dependent on the IL-1 receptor type I, which has been recently demonstrated using glial cells derived from IL-1 receptor type I knock-out animals (Parker et al. 2002). We analysed putative intracellular second messenger systems involved in IL-1β-induced EP3 expression. IL-1β has been shown to activate NF-κB, PKC, JNK, p38 MAPK and p42/44 MAPK in glial and U373 MG human astrocytoma cells (Lieb et al. 1996b; Molina-Holgado et al. 2000; Parker et al. 2002). Using specific inhibitors, we did not find evidence for a critical involvement of JNK, p38 MAPK or p42/44 MAPK. We used various PKC inhibitors to assess the potential role of PKC isoenzymes in signal transduction of IL-1β-induced EP3 receptor synthesis. The bisindolylmaleimides, GF109203X and Gö6983, which have been described to inhibit most isoforms of PKC (Toullec et al. 1991; Gschwendt et al. 1996), effectively inhibited IL-1β-induced EP3 receptor synthesis. This was further confirmed by the strong effect of Ro-31-8425, an unselective inhibitor of conventional and novel PKC isoforms (Way et al. 2000).
Gö6976, an indolocarbazole that preferentially inhibits conventional PKC isoforms (α, β, γ) but not PKC δ and ε (Martiny-Baron et al. 1993; Gschwendt et al. 1996), only slightly affected IL-1β-induced EP3 receptor synthesis. Although we can not rule out completely an involvement of conventional PKC isoforms, IL-1β-induced EP3 receptor expression seems to be mediated by novel PKC isoforms such as δ and ε, which are expressed in U373 MG cells (Shih et al. 1999) The potent inhibition of EP3 receptor expression by the PKC δ inhibitor rottlerin might point to PKC δ as the major PKC isoform involved in this process. However, rottlerin has been shown lately to interfere also with other signal transduction pathways (Way et al. 2000). In U373 MG cells, PKC ε has been already implicated in gene regulation (Lieb et al. 2003, 2005) and may also play a role in IL-1β-induced EP3 synthesis. However further PKC isoforms (PKC η, θ, τ, λ) or a cross-talk between various isoforms of PKC cannot be ruled out (Kim et al. 1997; Shih et al. 1999).
The promoter region of the human EP3 receptor contains an NF-κB binding site, no AP-1 binding sites and seven copies of the AP-2 consensus sequences (Kotani et al. 1997). A NCBI databank search revealed a supercontig containing the rat EP3 promoter which includes three NF-κB sites (−160, −182 and −1799 relative to the ATG start codon), an AP-2 consensus sequence (−99) but no AP-1 consensus sequences. IκB degradation and activation of NF-κB has already been shown in U373 MG cells after stimulation with IL-1β (Fiebich et al. 1999). We demonstrated here that three compounds that interfere with NF-κB activation, parthenolide (Hehner et al. 1998, 1999), SC-514 (Kishore et al. 2003) and MG-132 (Palombella et al. 1994), disrupt IL-1β-stimulated EP3 synthesis in U373 MG cells. Therefore, an activation of NF-κB seems to be necessary for the synthesis of EP3 besides the activition of PKC. An upstream involvement of PKC δ in activation of NF-κB has recently been demonstrated (Tojima et al. 2000; Wang et al. 2003).
We further studied whether IL-1β induces EP3 receptor expression in vivo, investigating rat hippocampus that expresses the IL-1 receptor type I (Lynch and Lynch 2002). We found a strong up-regulation of EP3 receptor protein in rat hippocampus after application of the vehicle, suggesting that this manipulation might have already stimulated endogenous IL-1β synthesis. However, additional application of IL-1β through the canula markedly enhanced cerebral EP3 protein production, suggesting that IL-1β is a potent inductor of EP3 not only in vitro but also in vivo. We used GFAP immunoreactivity to normalize EP3 receptor band intensity for densitometric scans and statistical analysis. Although an increase in GFAP expression after IL-1 application has been described before, we were not able to detect up-regulation of GFAP protein levels in the short time period of 24 h. A concomitant up-regulation of GFAP protein synthesis by IL-1β in our study would rather lead to an underestimation of the effect of IL-1β on EP3 protein synthesis. We were not able to identify the cellular source of EP3 expression after IL-1β injection. Although the in vitro experiments are suggestive that astrocytes are the cellular source of EP3 protein in IL-1β-injected brains, the hypothesis has to be confirmed in further experiments using sensitive histological techniques.
Studies investigating EP receptor regulation in models of neuroinflammation are rare. A few groups have investigated cerebral EP receptor expression in response to systemic inflammation. Zhang and Rivest showed that the peripheral application of lipopolysaccharide or IL-1β results in the expression of the EP2 and EP4 receptor subtypes in circumscribed neuronal populations (Zhang and Rivest 1999). EP3 receptor regulation has been studied 3 h after systemic application of IL-1β by in situ hybridization (Ek et al. 2000). De-novo expression of EP3 in glial cells or neurons was not reported in this study. However, the amount of circulating IL-1β that crosses the blood–brain barrier may be minimal.
Although neuronal EP3 receptors have been implied in the generation of fever (Ushikubi et al. 1998), the activation of the hypothalamo-pituitary axis (Matsuoka et al. 2003) and catecholaminergic transmitter release (Nakamura et al. 2001), the role of IL-1-induced glial EP3 receptor expression remains unclear. PGE2 has been shown to have both neuroprotective and neurotoxic properties (Akaike et al. 1994; Cazevieille et al. 1994; Prasad et al. 1998). Recently it has been shown that the EP2 receptor subtype, via a cAMP-dependent mechanism, mediates neuroprotective effects of PGE2 in models of excitotoxicity and cerebral ischemia (McCullough et al. 2004). Consequently, a change in the EP receptor expression pattern of neuronal or glial cells under neuroinflammatory conditions may crucially alter the function of PGE2 in brain pathology. A better understanding of the role of specific EP receptor subtypes and their differential expression under the influence of proinflammatory stimuli may help to develop novel therapeutic strategies that target specific EP receptor pathways in neuroinflammatory diseases.
The technical assistance of Ulrike Goetzinger-Berger, Doris Gmeiner and Franziska Klott is gratefully acknowledged. We would like to thank Dr Kishore (Pfizer) for supplying the compound SC-514. This work was supported in part by grant Fi 683/1-1 from the Deutsche Forschungsgemeinschaft and VivaCell Biotechnology GmbH, Denzlingen, Germany. ECJ was supported by a research fellowship from the Alexander von Humboldt Foundation (Bonn, Germany).