The p38 mitogen-activated protein kinase ... - Wiley Online Library

Immunology 2003 108 502±512

The p38 mitogen-activated protein kinase regulates interleukin-1b-induced IL-8 expression via an effect on the IL-8 promoter in intestinal epithelial cells KULJIT PARHAR,* ANDREW RAY,* URS STEINBRECHER,* COLLEEN NELSONy & BALJINDER SALH* DEPARTMENTS OF *Medicine and yProstate Cancer Research, The Jack Bell Research Centre, Vancouver, British Columbia, Canada

SUMMARY Several lines of evidence implicate the p38 mitogen-activated protein kinase (p38 MAPK) in the proin¯ammatory response to bacterial agents and cytokines. Equally, the transcription factor, nuclear factor (NF)-kB, is recognized to be a critical determinant of the in¯ammatory response in intestinal epithelial cells (IECs). However, the precise inter-relationship between the activation of p38 MAPK and activation of the transcription factor NF-kB in the intestinal epithelial cell (IEC) system, remains unknown. Here we show that interleukin (IL)-1b activates all three MAPKs in Caco-2 cells. The production of IL-8 and monocyte chemotactic protein 1 (MCP-1) was attenuated by 50% when these cells were preincubated with the p38 MAPK inhibitor, SB 203580. Further investigation of the NF-kB signalling system revealed that the inhibitory effect was independent of the phosphorylation and degradation of IkBa, the binding partner of NF-kB. This effect was also independent of the DNA binding of the p65 Rel A subunit, as well as transactivation, determined by an NF-kB luciferase construct, using both SB 203580 and dominant±negative p38 MAPK. Evaluation of IL-8 and MCP-1 RNA messages by reverse transcription±polymerase chain reaction (RT±PCR) revealed that the inhibitory effect of SB 203580 was associated with a reduction in this parameter. Using an IL-8±luciferase promoter construct, an effect of p38 upon its activation by both pharmacological and dominant±negative p38 construct co-transfection was demonstrated. It is concluded that p38 MAPK in¯uences the expression of chemokines in intestinal epithelial cells, through an effect upon the activation of the chemokine promoter, and does not directly involve the activation of the transcription factor NF-kB.

phenomenon remain unknown, it is probably dependent upon the co-ordination of different cell systems through elaborate signalling networks. Dysregulation of in¯ammation is seen in in¯ammatory bowel diseases (IBD) such as Crohn's disease (CD) and ulcerative colitis (UC), and is often accompanied by an increased expression of the proin¯ammatory cytokines interleukin (IL)-1b and tumour necrosis factor-a (TNF-a). It is also characterized by an increased in®ltration with immune cells such as monocytes, neutrophils and T lymphocytes.3,4 Although the underlying aetiology of these chronic disorders is unknown, knowledge of aberrant cytokine production has led to the development and effective use of immunoregulatory therapies, best exempli®ed by anti-TNF monoclonal antibodies (mAbs) for CD.5 Attraction and in®ltration of leucocytes into extravascular body compartments involves the expression of a subfamily of small cytokines, called chemokines. These molecules are classi®ed into two major groups (C±C and C±X±C) on the basis of the arrangement of the two N-terminal cysteine residues. IECs have been shown to express these molecules in response to

INTRODUCTION Intestinal epithelial cells (IECs) are an integral component of the enteric immune system. They have the capacity to express antigens to T cells, and they produce cytokines and chemokines in response to bacterial invasion. Through poorly understood mechanisms, they are involved in mounting an immune response against harmful pathogens, whilst suppressing a response against the normal enteric ¯ora.1,2 As a result, there is a basal level of physiological in¯ammation within the lamina propria of the intestine. This in¯ammatory response is highly regulated, and although the precise mechanisms underlying this

Received 22 July 2002; revised 4 December 2002; accepted 23 December 2002. Correspondence: Dr B. Salh, Division of Gastroenterology, University of British Columbia, 100±2647 Willow Street, Vancouver, British Columbia, Canada V5Z 3P1. E-mail: [email protected]

502

# 2003 Blackwell Publishing Ltd

p38 regulation of chemokines in IECs cytokines.6 Increased expression levels of chemokines have been correlated with the pathophysiology of IBD. IL-8, a potent attractant for neutrophils, has been shown to be elevated in both UC and CD.7 Other chemokines dysregulated in IBD include monocyte chemotactic factor 1 (MCP-1), MCP-3 and epithelialcell-derived neutrophil-activating peptide 78 (ENA-78).8±11 Furthermore, the chemokine, regulated on activation, normal, T-cell expressed, and secreted (RANTES), has been shown to be involved in the progression of acute to chronic disease, in a rodent model of colitis.12 One of the key components of the in¯ammatory response is the multifunctional cytokine, IL-1b. Its signi®cance in IBD is documented by the demonstration of its increased expression in affected tissues13 and the attenuation of its damaging properties by its regulatory partner the IL-1 receptor (IL-1R) antagonist.14,15 After binding of IL-1 to its receptor, IL-1R1 and interleukin receptor associated kinases (IRAK)-1 and -2 are recruited via the adapter protein, MyD88 (reviewed in ref. 16). Following their phosphorylation, downstream signalling events are initiated through TNF receptor associated factor (TRAF)-6, which appears to be a prerequisite for the activation of both the nuclear factor (NF)-kB and c-Jun N-terminal protein kinase (JNK) signalling pathways.17 TRAF-2 has been demonstrated to play a role in this process in HT-29 and IEC-6 cells.18 NF-kB is an inducible heterodimeric transcription factor that recognizes the consensus motif 50 -GGGRNNYYCC-30 , and is composed most commonly of the Rel A (p65) and NF-kB1 (p50) subunits.19 Normally sequestered in the cytoplasm by NF-kB, IL1b stimulation results in the phosphorylation and subsequent degradation of IkBa.20 This allows NF-kB to translocate to the nucleus, where it can effect the transcription of many genes, including IL-6, IL-8, MCP-1, IL-1b and TNF-a.21 Notably, NFkB is activated in IBD22,23 and therapies that target it have been shown to attenuate in¯ammatory disease activity.24±26 Mitogen-activated protein kinases (MAPKs) are a group of protein kinases that are activated in response to a variety of extracellular stimuli and are responsible for a diverse range of biological processes, including response to stress, growthrelated stimuli, in¯ammation and differentiation.27,28 MAPKs are comprised of three subfamilies: JNKs; extracellular signalregulated kinases (ERKs); and p38 MAPK. Their precise roles in the regulation of in¯ammatory processes within the intestine are currently unknown. p38 MAPK is a stress-activated protein kinase, ®rst discovered as a kinase activated by hyperosmolarity and endotoxin29 and later shown to be activated in response to many physical stresses such as heat and ultraviolet (UV) irradiation, as well as cytokines such as IL-1b and TNF-a. It is regulated by phosphorylation via the MAPK kinases (MAPKK) MKK3 and MKK6.30,31 Several lines of evidence indicate an important role for this molecule in the in¯ammatory process. These include interferon-g (IFN-g) production by lymphocytes;32 inducible nitric oxide synthase (iNOS) production;33,34 and Cox-2, IL-6 and TNF production by cells of the monocytic series.35,36 Inhibition of this molecule by speci®c pharmacological antagonists has resulted in the attenuation of in¯ammation seen in a number of animal models.37±39 In this report we have examined the role that p38 MAPK plays in chemokine expression by IECs, in response to IL-1b. Furthermore, owing to the potential link between p38 MAPK # 2003 Blackwell Publishing Ltd, Immunology, 108, 502±512

503

and NF-kB,35,40 we addressed the possible involvement of the p38 MAPK in several aspects of regulation of the latter molecule. Our ®ndings indicate a role for p38 MAPK in the reduction of IL-8 expression, at the level of the chemokine promoter, and that p38 MAPK does not in¯uence NF-kB activation directly. We propose that p38 MAPK inhibitors may have a supplementary role in the treatment of IBD. MATERIALS AND METHODS Cell culture HT29 and Caco-2 cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA). The latter cell line has been well characterized to undergo differentiation into functional enterocytes, and was therefore chosen for detailed studies.41 Both cell lines were pretreated for 2 hr using PD 98059, curcumin, SB 203580 and DRB (Calbiochem, San Diego, CA) at the appropriate concentrations before stimulation with IL-1b (Calbiochem) for the indicated lengths of time. HT29 and Caco-2 human colonic epithelial cells were grown in M199 media containing 10% heat-inactivated fetal bovine serum (FBS) (Hyclone, Logan, UT), supplemented with penicillin and streptomycin (Life Technologies, Burlington, Ont., Canada). Transient transfections with p38 MAPK (wild-type and kinasedead) plasmids were carried out using Effectene. These plasmids were kindly donated by Dr J. Han (Scripps, La Jolla, CA). Nuclear extracts Caco-2 cells were seeded onto 60-mm plates and grown to con¯uence, then pretreated with the appropriate inhibitor for 2 hr before stimulation with IL-1b for 30 min. Cells were washed once with ice-cold phosphate-buffered saline (PBS) and scraped into 1 ml of PBS. They were then centrifuged at 4000 g and resuspended in 200 ml of Buffer A [10 mM HEPES, pH 7
9, 10 mM KCl, 0
1 mM EDTA, 0
2 mM EGTA, 1 mM dithiothreitol (DTT), 0
5 mM phenylmethylsulphonyl ¯uoride (PMSF)] for 15 min, before the addition of 13 ml of 10% Nonidet P-40 (NP-40). The cell suspension was then vortexed for 10 seconds before being centrifuged again at 4000 g for 30 seconds. The supernatant was removed and the nuclear pellet resuspended in 30 ml of Buffer C (20 mM HEPES, pH 7
9, 0
4 M NaCl, 1 mM EDTA, 1 mM EGTA, 25% glycerol, 1 mM DTT, 1 mM PMSF), rocked vigorously for 15 min, then centrifuged for 5 min at 4000 g. The supernatant was retained and the protein concentration determined by using the Bradford assay (Bio-Rad, Mississuagua, Ont., Canada). Samples were stored at 808 until use. Electromobility shift assay A synthetic kb oligonucleotide was cloned into the cloning vector, pBS (Stratagene, La Jolla, CA), using the EcoRI and HindIII sites, to create pBS±EMSAkb. To radiolabel the probe, it was excised from pBS±EMSAkb using EcoRI and HindIII, and labelled using [g-32P]dCTP (Amersham, Montreal, QueÂbec, Canada) and the Klenow fragment of DNA polymerase (New England Biolabs, Missisauga, Ont., Canada). The probe was then puri®ed by running on a 5% non-denaturing gel, cutting out the fragment, and incubating the gel slice in elution buffer [(0
6

504

K. Parhar et al.

M ammonium acetate, 1 mM EDTA, 0
1% sodium dodecyl sulphate (SDS)] overnight. Ten micrograms of nuclear extract was preincubated for 15 min in binding buffer (20 mM HEPES, pH 7
9, 100 mM KCl, 10% glycerol, 1 mM DTT) containing 1 mg of poly dIdC (Amersham), 20 000 counts per minute (c.p.m.) of probe was then added and the reaction mixture incubated at room temperature for 30 min before electrophoresis on a 5% non-denaturing polyacrylamide gel in 0
25  Tris 89 mM, boric acid 89 mM, EDTA 2 mM (TBE) at 200 V for 1
5 hr. The gel was subsequently dried for 45 min before phospho-imaging analysis was carried out using a Bio-Rad molecular imager FX (or alternatively exposed to ®lm overnight at 808 and then developed). For supershift or cold competitor reactions, the nuclear extract was preincubated with 1 mg of anti-p65 antibody (Calbiochem), or 100-fold excess of unlabelled probe with binding buffer and poly dI-dC for 30 min before adding the radiolabelled probe. The probes used are as follows: NF-kB: 50 -AATTCGGTTACAAGGGACTTTCCGCTGA-30 and 30 -GCCAATGTTCCCTGAAAGGCGACTTCGA50 ; and AP1: 50 -CGCTTGATGACTCAGCCGGAA-30 and 30 GCGAACTACTGAGTCGGCCTT-50 .

Western analysis Cells were washed once with ice-cold PBS, resuspended in homogenization buffer (20 mM MOPS, 50 mM b-glycerophosphate, 5 mM EGTA, 50 NaF, 1 mM DTT, 1 mM sodium vanadate, and 1 mM PMSF) for 30 min, sonicated for 15 seconds and centrifuged at 4000 g for 15 min. The protein concentration in the supernatant was determined by using the Bradford assay (Bio-Rad). Fifty micrograms of protein from each sample was resolved using 10% sodium dodecyl sulphate± polyacrylamide gel electrophoresis (SDS±PAGE) before transferring to nitrocellulose membranes (Bio-Rad). The blots were blocked in 5% skim milk in TBS-T (20 mM Tris±HCl, pH 7
4, 250 mM NaCl, 0
05% Tween-20) for 1 hr before probing for 2± 4 hr using the appropriate primary antibody. The blots were washed with TBS-T three times (10 min each wash), before incubation with the appropriate secondary antibody for 1
5 hr. Following three further washes in TBS-T, they were developed using the enhanced chemiluminescence detection system (ECL; Amersham). Phospho-p38, phospho-IkB, phospho-tyrosine, phospho-JNK and IkBa antibodies were all purchased from New England Biolabs. Antibodies to p38 protein were purchased from Stressgen Biotechnologies (Victoria, BC, Canada). MAPKAPK2 assay This was carried out as an in vitro assay using crude lysate. Brie¯y, Caco-2 cells were preincubated with SB 203580 for 2 hr, stimulated with IL-1b (2 ng/ml) for 30 min and then harvested in homogenization buffer (described under Western analysis above). Ten microlitres (equal to 10 mg of protein) of cell lysate was then used in a kinase assay, using heat shock protein 27 (hsp 27) (1 mg; Stressgen Biotechnology) as the substrate and 0
5 mg of ATP (250 mM ATP, 1 mCi [g-32P]labelled), for 20 min at 308. Samples were then boiled in 5 sample buffer and resolved on SDS±PAGE. After transfer to nitrocellulose membranes, the substrate bands were visualized using autoradiography. Ponceau staining was carried out to demonstrate equal loading.

Transient transfections Transient transfections were performed using the Effectene reagent (Qiagen, Toronto, Ont., Canada) as per the manufacturer's instructions. Cells were seeded and transfected upon reaching 90% con¯uency. For one 35-mm well, 0
2 mg of DNA was transfected using a DNA : Effectene ratio of 1 : 25. Transfections were carried out overnight before replacing with fresh medium for 24 hr prior to conducting experiments. Reporter gene assays An NF-kB-dependent reporter, containing four repeats of the kb consensus sequence from the IL-1b gene cloned into the BglII site of the pGL3-promoter vector (Promega, Madison, WI), creating 4 x kb-luc, was co-transfected with a LacZ plasmid (Kindly donated by William Jia, UBC) using Effectene (Qiagen, Toronto, Ont., Canada). Cells were pretreated for 2 hr with the appropriate inhibitor, then stimulated with IL-1b for 6 hr, before being harvested. Luciferase and b-galactosidase activities were measured according to the manufacturer's instructions (Promega). Light emission was measured using a luminometer, and the results were normalized using b-galactosidase. The transfection ef®ciency was normalized for, by assaying for b-galactosidase activity from a constitutively expressed LacZ plasmid that was co-transfected with the reporter construct. IL-8 reporter (kindly supplied by Dr Gary Wu, Philadelphia) assays were performed and corrected in a similar manner. ELISA assays for IL-8 and MCP-1 Cells were pretreated with the appropriate inhibitors and stimulated for 36 hr with IL-1b. The supernatants were sampled and the chemokine concentration was determined, in triplicate, by using enzyme-linked immunosorbent assay (ELISA) (BD Pharmingen, Missisaugua, Ont., Canada), as per the manufacturer's instructions. Isolation of RNA and reverse transcription±polymerase chain reaction (RT±PCR) Cells were pretreated for 2 hr with the appropriate inhibitor, and stimulated with IL-1b. RNA was isolated using the TRIZOL method (Life Technologies). The purity of the RNA was determined by running 1 mg of RNA on a 1% agarose gel for 1
5 hr at 75 V. One microgram of RNA was reverse transcribed, for 50 min at 378, in a reaction mix of 20 ml containing: 0
5 mg of oligo (dT)12 18 (Amersham), 1 ml of 10 mM dNTPs, 2 ml of 0
1 M DTT and 40 U of RNA-guard (Amersham, Montreal, Quebec) in 1 ®rst-strand buffer (Life Technologies) using 200 U of murine Moloney leukaemia virus (M-MLV) reverse transcriptase. Two microlitres of cDNA was used in each subsequent PCR reaction. For each 50-ml PCR reaction, 2 U of TAQ (PE Biosystems, Branchburg, NJ), 1  PCR Buffer (PE Biosystems), 10 pmol of each primer, 1 ml of 10 mM dNTPs, and 3 ml of 25 mM MgCl2 were used. The PCR temperatures were 948 for 45 seconds (denaturation), 568 for 45 seconds (annealing), and 728 for 1 min (extension). Ten-microlitre aliquots of the reaction mix were resolved on a 1
5% agarose gel containing ethidium bromide. Negative controls for cDNA synthesis were run without template, and also without RT. The linearity of PCR reactions was determined in the range between # 2003 Blackwell Publishing Ltd, Immunology, 108, 502±512

p38 regulation of chemokines in IECs 20 and 40 cycles. Densitometry was performed using Bio-Rad Quantity-One software. The sizes of the PCR products were 289 bp, 510 bp and 235 bp for IL-8, MCP-1 and actin, respectively. Primers The primers used are listed below:  IL-8 Forward: 50 -TCTGCAGCTCTGBTGTGAAGGTGCAGTT-30 ;  IL-8 Reverse: 50 -TTCCTTTGACCCACGTCTCCCAA-30 ;  MCP-1 Forward: 50 -TCTGTGCCTGCTGCTCATAGC-30 ;  MCP-1 Reverse: 50 -GGGTAGAACTGTGGTTCAAGAGG-30 ;  b-actin Forward 50 -CCAACCGCGAGAAGATGACC-30 ; and  b-actin Reverse 50 -GATCTTCATGAGGTAGTCAGT-30 . RESULTS Il-1b activates protein tyrosine phosphorylation and MAPKs in IECs We ®rst examined the effect of IL-1b stimulation on the temporal characteristics of the activation of the MAPK family members. The data indicates that all three MAPKs were activated within 10±15 min, with p38 MAPK exhibiting the earliest activation (Fig. 1a). Additionally it is apparent that the activation was sustained for at least 60 min. The lowest panel shows that proteins corresponding to MAPKs (with a molecular mass of 40±44 kDa) become tyrosine phosphorylated at the same times. To con®rm that the SB 203580 inhibited signalling through to the MAPKAPK2, in vitro phosphorylation reactions were performed using its substrate, hsp27. The data indicates that preincubation with the p38 inhibitor attenuated activation of this downstream target (Fig. 1b). SB 203580 attenuates IL-8 and MCP-1 production by IL-1b-stimulated IECs We next sought to examine the effect of the p38 MAPK inhibitor, SB 203580, on the production of an alpha-chemokine (IL-8) and a beta-chemokine (MCP-1) by IECs. This inhibitor acts by binding in the ATP-binding pocket of p38. Pretreatment of both Caco-2 cells and HT-29 with SB 203580 led to a signi®cant reduction in IL-8 production (Fig. 1c). A similar inhibitory effect of the p38 MAPK inhibitor was observed on MCP-1 production in Caco-2 cells (Fig. 1d). HT29 cells do not produce any MCP-1 in response to IL-1b alone, as previously reported.42 PD 98059 (an inhibitor of MEK via upstream activator-dependent phosphorylation) had no effect upon the production of either chemokine in the Caco-2 cell line (lane 4). The data indicate that p38 MAPK is activated in IECs in response to IL-1b and is involved in the production of important immunoregulatory chemokines. IjBa phosphorylation and degradation are independent of p38 MAPK NF-kB has been reported to be a major regulator of IL-8 and MCP-1 transcription.43,44 The phosphorylation and subsequent degradation of IkBa is an integral step in the activation of # 2003 Blackwell Publishing Ltd, Immunology, 108, 502±512

505

NF-kB. Previously, protein kinase casein kinase 2 (CK2) has been shown to phosphorylate IkBa at both the PEST region45±47 and serine 32/36 in vitro.48 Furthermore, it has been demonstrated that the stress-induced activation of p38 MAPK activates CK2.49 Therefore, it seemed logical to explore the possibility that p38 MAPK or CK2 were involved in the phosphorylation and degradation of IkBa. However, as the data indicates, pretreatment of cells with either SB 203580 or DRB, a speci®c inhibitor of protein kinase CK2 at the concentrations used,50 did not prevent IL-1b-induced phosphorylation and degradation of IkBa (Fig. 2a). NF-jB DNA binding and trans-activation are independent of p38 MAPK activation NF-kB is regulated by its ability to translocate to the nucleus and bind to its consensus sequence. Furthermore, it has been shown that activation of NF-kB is accompanied by the CK2induced phosphorylation of the p65 subunit,51 which has been mapped to the serine 529 site of Rel A.52 Therefore, we investigated the potential roles of p38 MAPK and CK2 on this aspect of NF-kB activation. Electromobility shift assays (EMSAs) were performed on nuclei isolated from Caco-2 cells. Pretreatment of cells with SB 203580 or DRB did not prevent nuclear translocation or the ability of NF-kB to bind to its consensus sequence (Fig. 2b). It can be seen that there was no effect upon the bound probe signal with either of the two inhibitors. In order to con®rm that in fact p65 was the major Rel family member binding to the NF-kB consensus sequence, a supershift assay was performed, by preincubating nuclear extracts with the anti-p65 antibody before addition of the probe. As seen, the Rel A band was almost entirely shifted up with the addition of the antibody (Fig. 2c). DNA binding alone does not confer transcriptional activation, as NF-kB requires multiple trans-activating phosphorylations in order to be functionally active once bound to DNA.53 To investigate the possibility that p38 MAPK may be phosphorylating and regulating the trans-activation of p65, we performed reporter assays with a synthetic NF-kB promoter fused to a luciferase gene. The data indicates that there is a threefold increase in trans-activation upon exposure of Caco-2 cells to IL1b (Fig. 3a). Pretreatment with SB 203580 did not attenuate the trans-activation of NF-kB. Curcumin was found to attenuate trans-activation, in accordance with a previous report.54 This data excludes an effect of p38 MAPK upon NF-kB activation in this cell system. In order to substantiate these ®ndings, we investigated the effects of transient transfections of the p38 MAPK constructs upon NF-kB trans-activation. The data indicated (Fig. 3b) that despite the small changes observed in the DNA-binding study (Fig. 3c), this is not translated into a reduction in the transactivation effect. We believe that this complements the data obtained with the SB 203580 inhibitor, and establishes that p38 MAPK is not involved at any step of NF-kB activation within the Caco-2 cell system. To con®rm that there was no involvement of p38 MAPK upon NF-kB DNA binding, we transiently transfected Caco-2 cells with both the wild-type and kinase-dead versions of p38 MAPK, and investigated their in¯uence upon DNA binding in

506

K. Parhar et al.

Figure 1. Interleukin (IL)-1b stimulation of intestinal epithelial cells results in activation of extracellular signal-regulated kinase (ERK), c-Jun N-terminal protein kinase (JNK) and p38 (a), and p38 mitogen-activated protein kinase (p38 MAPK) signalling can be inhibited using the speci®c pharmacological inhibitor SB 203580 (10 mM). (a) Caco-2 cells were stimulated with IL-1b (2 ng/ml) for the indicated times, before being harvested. Western immunoblotting was carried out as described in the Materials and methods. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; p-Tyr, tyrosine phosphorylation. (b) Caco-2 cells were preincubated with SB 203580 for 2 hr before being stimulated with IL-1b (2 ng/ml) for 30 min and harvested. Ten microlitres of cell lysate was then used in a kinase assay, with the heat shock protein 27 (hsp 27) as substrate. Samples were boiled in 5 sample buffer and resolved by sodium dodecyl sulphate±polyacrylamide gel electrophoresis (SDS±PAGE). After transfer to nitrocellulose membranes, the substrate bands were visualized using autoradiography. The panel on the left is an autoradiogram and the one on the right is a Ponceau stain of the same membrane, showing co-localization of the hsp 27 band. The results shown are representative of at least three independent experiments. (c) Cells were pretreated with SB 203580 or PD 98059 (25 mM) for 2 hr, stimulation with IL-1b (2 ng/ml) for 36 hr. Supernatants were collected and analysed for the presence of IL-8 (c) or monocyte chemotactic protein 1 (MCP-1) (d) by enzyme-linked immunosorbent assay (ELISA). (Lane 1, control; lane 2, IL-1b; lane 3, IL-1b with SB 203580; lane 4, IL-1b with PD 98059.) The results shown are representative of at least three independent experiments. The results are expressed as mean  standard error of the mean (SEM) (*P