Blockade of p38 Mitogen-activated Protein Kinase Pathway Inhibits Inducible Nitric-oxide Synthase Expression in Mouse Astrocytes*

Treatment of mouse astrocyte cultures with combined interleukin (IL)-1α and tumor necrosis factor (TNF)-α induced expression of inducible nitric-oxide synthase (iNOS), resulting in sustained release of large amounts of nitric oxide, whereas TNF-α and IL-1α individually were unable to induce iNOS expression in astrocytes. The role of MAPK cascades and of NF-κB activation in the early intracellular signal transduction involved in iNOS transcription in TNF-α/IL-1α-stimulated astrocytes was investigated. TNF-α and IL-1α activated all p42/44MAPK, p38MAPK, and p54JNK pathways as determined by immunoprecipitation kinase assays using specific antibodies and substrates. The p38MAPK pathway is specifically involved in TNF-α/IL-1α-induced iNOS expression, since iNOS protein and nitric oxide release in the presence of a specific inhibitor of p38MAPK, 4-(4-fluorophenyl)-2–2-(4-hydroxyphenyl)-5-(4-pyridyl)-imidazole (FHPI), were dramatically diminished. In contrast, PD98059, a specific inhibitor of MEK1 had no effect on iNOS expression. p38MAPKdid not couple NF-κB to iNOS transcription, but NF-κB had a clear role in iNOS transcription regulation. Northern blot analysis showed that the p38MAPK pathway controlled iNOS expression at the transcriptional level, since iNOS mRNA was reduced in the presence of FHPI in TNF-α/IL-1α-stimulated astrocytes. iNOS expression was investigated with TNF receptor (TNFR)-1- and TNFR-2-deficient mice. The TNF-α activity in TNF-α/IL-1α-stimulated astrocytes was exclusively mediated through TNFR-1, most likely because TNFR-2-mediated signals in astrocytes did not connect to the p38MAPK pathway. These data suggest that TNF-α/IL-1α-induced iNOS expression depends on a yet undetermined second pathway in addition to p38MAPK.

synthases (NOS), convert L-arginine to L-citrulline and NO. In the brain, three genes encode NO synthase isoforms, with significant differences in their regulation (6 -8). Neuronal and endothelial NOS are constitutively expressed in astrocytes and in subpopulations of neurons. Their activity is predominantly regulated through intracellular calcium/calmodulin signals in response to hormone or neurotransmitter stimulation. A third NOS, inducible NOS (iNOS), is expressed only in response to cell activation by cytokines and stress as well as by bacterial cell wall products in different cell types, including astrocytes and microglia cells (9,10). Cellular NO release after iNOS induction is therefore thought to be important in many pathological conditions such as infections (11), ischemia (12), and multiple sclerosis (13).
Little is known about the intracellular signaling pathways of iNOS induction. The murine iNOS promoter revealed the presence of 24 transcription factor binding sites, including NF-B and AP-1 sites (14 -16). Protein kinase C, and particularly protein kinase C-⑀, has been reported to induce iNOS (17). The role of the mitogen-activated protein kinase (MAPK) cascades, p42/44 MAPK (ERK1 and ERK2), p38 MAPK , and p54 JNK in the control of iNOS expression has not yet been clearly defined (18,19). p38 MAPK is activated by treatment of cells with lipopolysaccharide, cytokines, and stress (20,21). MAPKAP kinase-2 was first identified as a p38 MAPK substrate, which in turn phosphorylated HSP-27 (22,23). Subsequently, several transcription factors have been found to be activated by p38 MAPK . These transcription factors included ATF-2 (24); CHOP/GADD 153, a member of the C/EBP family expressed in stressed cells (25); Max, which is bound and phosphorylated by Mxi-2, a homologue to CSBP 1 and 2 (26), CREB, and ATF-1, reported to be under the control of MAPKAP kinase-2 and p38 MAPK (27,28); and the myocyte enhancer factor 2C (MEF2C), belonging to the MADS family (29). However, p38 MAPK also has a major role in the regulation of gene expression at the post-transcriptional level (30), probably by a mechanism depending on the AUUUA sequence motifs in the 3Ј-untranslated region of their transcripts (31). Important roles of p38 MAPK have been shown in the control of TNF and IL-1 expression by THP-1 cells (30), of IL-8 expression in monocytes (32), and of IL-6 expression in TNF-treated astrocytes (33,34). More recently, p38 MAPK has also been shown to play a role in the expression control of adhesion molecules in endothelial cells, including E-selectin (35) and VCAM-1 (36).
The present study investigates the role of p38 MAPK in the regulation of iNOS expression. The results show that a p38 MAPKdependent pathway transduces iNOS expression signals elicited by combined TNF-␣ and IL-1␣ stimulation in mouse primary astrocytes. p38 MAPK activation is necessary but not sufficient, since stimulation by TNF-␣ or IL-1␣ alone fully activated p38 MAPK but was unable to induce iNOS transcription. The p38 MAPK -mediated signal acts indirectly, since iNOS gene transcription is blocked by protein synthesis inhibition. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Isolation and Culture of Mouse Neonate Astrocytes-Astrocyte precursors were isolated from 4 -5-day-old mouse pups as described elsewhere (40). Briefly, cerebella were dissected and stripped of the meninges in a calcium-free salt solution, CSSϪ/ϩ, pH 7.3 (120 mM NaCl, 5.5 mM KCl, 30 mM Tris⅐Cl, pH 7.4, 15 mM glucose, 1.7 mM MgCl 2 ) on ice, and transferred to 3 ml of 1% trypsin in phosphate-buffered saline for 10 min at room temperature. Tissue was triturated 10 -20 times through the fire-polished tip of a Pasteur pipette. Cells were sedimented at 200 ϫ g and resuspended in serum-free medium (Dulbecco's modified Eagle's medium; 10 ng/ml epidermal growth factor, 10 g/ml insulin, 100 g/ml transferrin, 1 g/liter bovine serum albumin, 100 units/ml penicillin, 100 g/ml streptomycin). Cells were seeded in 75cm 2 flasks precoated with poly-D-lysine. To differentiate cells in mature astrocytes, the serum-free medium was exchanged with a medium containing 10% heat-inactivated fetal calf serum. After 2-3 days, cultures contained Ͼ98% astrocytes as determined by glial fibrillary acidic protein staining. For kinase studies, astrocytes were rendered quiescent by a 3-day incubation in Dulbecco's modified Eagle's medium containing 0.25% serum.
Northern Blotting-Astrocytes grown to confluence in 75-cm 2 flasks were treated for various times with cytokines or other reagents, and total cell RNA was isolated as described by Chomczynski and Sacchi (41). 10-g samples of total RNA were electrophoresed through 1% agarose gels, transferred to nylon membranes, and fixed by UV exposure. The filters were prehybridized in hybridization mix (1% bovine serum albumin, 1 mM EDTA, 1 M sodium phosphate, pH 7.2, 7% SDS) for 1 h and hybridized overnight at 65°C in hybridization mix containing DNA probes produced by polymerase chain reaction and labeled with [␣-32 P]dCTP. Polymerase chain reaction primers to prepare DNA probes were 5Ј-TGCCAGGGTCACAACTTTACAGG (forward iNOS primer), 5Ј-GGTCGATGTCACATGCAGCTTGTC (reverse iNOS primer), 5Ј-TGCTGTTCACAGTTGCCGGC (forward MCP-1 primer), and 5Ј-CGGGTCAACTTCACATTCAAAG (reverse MCP-1 primer). The human GAPDH probe was purchased from CLONTECH. Filter membranes were washed twice for 10 min at low stringency (5% SDS, 40 mM sodium phosphate, 1 mM EDTA, 65°C) and twice for 20 min at high stringency (1% SDS, 40 mM sodium phosphate, 1 mM EDTA, 65°C) and subjected to autoradiography.
Western Blotting-Aliquots of cell lysates containing 20 g of total protein in Laemmli buffer were separated by 10 or 12% SDS-PAGE and transferred to PVDF membranes. The filter blots were blocked with 5% nonfat milk in blocking buffer (Tris-buffered saline; 50 mM Tris⅐Cl, pH 7.5, 150 mM NaCl, 0.1% Tween 20) and incubated with specific antibody for 2 h and with peroxidase-conjugated secondary antibody for 1 h at ambient temperature. Specific bands were revealed using the ECL system. Rainbow markers, range 14.3-220 kDa (Amersham, Buckinghamshire, UK), were used as molecular mass standards.
Immune Complex Kinase Assays-Cell lysates samples equivalent to 50 g of total protein were precleared three times for 20 min at 4°C with 30 l of protein A-Sepharose beads and incubated with 1 g of anti-p38 MAPK , anti-p42/44 MAPK , and anti-p54 JNK antibodies for 1 h at 4°C under constant agitation. Immune complexes were allowed to bind to 13 l of protein A-Sepharose beads for 30 min. The beads were washed with lysis buffer (2 times) and 10 mM Hepes and 10 mM magnesium acetate (2 times). Complexes were incubated for 30 min at 30°C in 30 l of kinase assay buffer (20 mM Hepes, 25 mM MgCl 2 , 5 M ATP, 2 mM DTT, 0.1 mM Na 3 VO 4 ) containing 5 Ci of [␥-33 P]ATP and 1 g of GST-ATF-2, GST-c-Jun, or MBP. Reactions were terminated by the addition of loading buffer, and samples were boiled and separated by 12 or 15% SDS-PAGE. The gels were dried, and incorporated 33 P was quantitated using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Nitrite Assays-Astrocytes were seeded in 24-well plates at 3 ϫ 10 5 cells/well and stimulated with cytokines, and culture supernatants were collected at the indicated time points. The accumulation of NO 2 Ϫ , a stable end product of NO formation, was used as a relative measurement of NO production. 100 l of supernatant were incubated with 100 l of Griess reagent (1% sulfanilic acid, 0.1% naphtylethylenediamine dihydrochloride, and 2.5% phosphoric acid) for 10 min at room temperature, and the optical absorbance was monitored at 570 nm with a microtiter plate reader, using sodium nitrite as a standard.
Preparation of Nuclear Extracts-Nuclear extracts were prepared as described previously (42), with minor modifications. Untreated or cytokine-treated astrocytes were washed twice with phosphate-buffered saline, 500 l of hypotonic buffer A (10 mM Hepes, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.5% Nonidet P-40, 1 mM PMSF, 10 mM DTT) were added, and cells were allowed to swell on ice for 30 min. The lysates were centrifuged at 6,500 rpm for 5 min at 4°C, the supernatants were discarded, and the volume of the nuclear pellet was estimated. The nuclear pellets were resuspended in 3 volumes of high salt buffer C (20 mM Hepes, pH 7.9, 25% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 200 M EDTA, pH 7.5, 1 mM PMSF, 10 mM DTT), vortexed vigorously for 10 s, and incubated on ice for 1 h with intermittent mixing. Samples were centrifuged at 6,500 rpm for 5 min at 4°C, and 3 volumes of low salt buffer D (20 mM Hepes, pH 7.9, 20% (v/v) glycerol, 100 mM KCl, 0.2 mM EDTA, 1% Nonidet P-40, 1 mM PMSF, 10 mM DTT) were added to the supernatants representing the nuclear extracts. The protein content was measured with the BCA protein assay kit (Pierce) with bovine serum albumin as a standard. Nuclear extracts were stored at Ϫ80°C until use.
Gel Shift Assays-Electrophoretic mobility shift assays were performed by incubating 6 g of nuclear extracts with 2 g of poly(dI⅐dC) and 30 g of bovine serum albumin in 10 l of binding buffer (100 mM Tris, pH 7.5, 500 mM NaCl, 10 mM DTT, 10 mM EDTA, pH 7.5, 50% glycerol) for 20 min at room temperature. A double-stranded oligonucleotide with two tandem B elements from the human immunodeficiency virus long terminal repeat (5Ј-ATCAGGGACTTTCCGCTGGG-GACTTTCCG-3Ј) was used as a probe to detect specific NF-B binding activity. Oligonucleotides were end-labeled with [␥-32 P]ATP using polynucleotide kinase; 60,000 cpm of 32 P-labeled probe were added to the nuclear extracts for 20 min at room temperature. Samples were electrophoresed using native 4% polyacrylamide gels.

p38 MAPK Mediates iNOS Induction by Combined TNF-␣ and IL-1␣ Treatment of Primary Astrocytes-
The expression of iNOS protein in cytokine-stimulated mouse astrocyte cultures was first examined in cytoplasmic extracts by immunoblotting using a specific polyclonal anti-iNOS antibody. Fig. 1A shows the time dependence of iNOS expression in astrocyte stimulated by combined TNF-␣/IL-1␣ treatment. iNOS protein was first detected at a very low level after 9 h, and high levels were reached after 24 and 48 h. To correlate the time dependence of iNOS expression and NO production, astrocyte cultures were treated with TNF-␣/IL-1␣, the cell culture supernatants were harvested, and nitrite concentrations were determined by a colorimetric assay. An accumulation of nitrites in the supernatants consistent with the observed iNOS protein expression time profile was found (Fig. 1B).
To investigate whether a p38 MAPK -dependent pathway mediates the TNF-␣/IL-1␣-stimulated iNOS induction, a specific inhibitor, FHPI, was used to block p38 MAPK activity (30). As shown in Fig. 1C, preincubation of astrocytes with 10 M FHPI reduced the TNF-␣/IL-1␣-induced iNOS protein expression to nondetectable levels at the 24-h time point. As a consequence, the stimulated NO release determined from the nitrite concentrations in the culture supernatants after 48 h was decreased by more than 80% to background levels (Fig. 1D). In contrast, iNOS expression 24 h after TNF-␣/IL-1␣ stimulation was only partially affected by the pretreatment with PD98059, a specific inhibitor of the MEK1-dependent pathway (Fig. 1C). As a consequence, the NO release 48 h after TNF-␣/IL-1␣ stimulation was merely reduced by 15% in the presence of 20 M of PD98059, a treatment that fully inactivates the MEK1 pathway (43). The FHPI dose dependence of the inhibition of iNOS expression was determined in astrocytes that had been preincubated for 45 min with various concentrations of p38 MAPK inhibitor before TNF-␣/IL-1␣ stimulation (Fig. 2, A and B). iNOS expression was inhibited in a regular dose dependence, with a half-inhibitory concentration, IC 50 , of about 0.5 M, congruent with the IC 50 of FHPI determined in assays using recombinant p38 MAPK (data not shown). The NO production in astrocyte cultures in the presence of increasing concentrations of FHPI agreed well with the iNOS protein expression (Fig.   2B). The inhibition of NO release by PD98059 treatment saturated at about 40 M at a level of about 10% inhibition.
p38 MAPK Regulates iNOS Expression at the Transcriptional Level-The expression of iNOS had been previously shown to be regulated at the transcriptional level (16), but the dependence on p38 MAPK activation and the presence of AUUUA motifs in the 3Ј-untranslated region sequence of the iNOS mRNA suggested that a regulation at the translational level might also occur. To further investigate the molecular mechanisms of the regulation of iNOS gene expression by the p38 MAPK pathway, iNOS mRNA levels were studied by Northern blot analysis in TNF-␣/IL-1␣-stimulated astrocytes in the presence or the absence of p38 MAPK inhibitor. Combined TNF-␣ and IL-1␣ treatment resulted in a transient increase of the 4.4-kilobase pair iNOS transcript (Fig. 3A). The transcript was first detected at 3 h, reached a maximum at 6 h, and had substantially declined after 24 h. Pretreatment of the cells with 10 M FHPI resulted in a 3-4-fold reduction in iNOS transcript level after 6 h, and only faint bands were detected at the 3-, 12-, and 24-h time points. The fact that residual iNOS mRNA levels are detected agrees well with incomplete inhibition of iNOS protein expression in the presence of 10 M FHPI, the lowest inhibitor concentration causing a substantial reduction in iNOS protein expression (Fig. 2A). These findings support a dominant transcriptional control of iNOS by p38 MAPK .
To test for a selective activity of FHPI on p38 MAPK -mediated iNOS regulation, the Northern blots were reprobed for another inducible gene, the MCP-1 gene. The induction of TNF-␣/IL-1␣-stimulated MCP-1 transcription was found independent of FHPI treatment of the cells and therefore must be controlled by signal pathways independent of p38 MAPK activity (Fig. 3A). Interestingly, the production of MCP-1 protein by the astrocytes was inhibited by FHPI treatment in a dose-dependent fashion, with an IC 50 of 0.5 M, despite the fact that the MCP-1 mRNA levels were unaffected, suggesting that p38 MAPK exerts iNOS mRNA Induction Requires New Protein Synthesis-To investigate whether the induction of iNOS mRNA transcription by TNF-␣/IL-1␣ treatment depended on new protein synthesis, the astrocytes were cultured in the presence of cycloheximide (1 or 10 g/ml) added 1 h before the start of the cytokine treatment. The cells were harvested after 6 h of cytokine stimulation, at a time when in control cultures iNOS transcription was maximal. Cycloheximide at 1 and 10 g/ml fully prevented induction of iNOS transcription as detected by Northern blot analysis (Fig. 4), whereas transcription of MCP-1 and GAPDH was not affected by the cycloheximide treatment.
iNOS Induction Requires Combined TNF-␣ and IL-1␣ Stimulation-To dissect the role of the TNF-␣ and IL-1␣ in the iNOS expression, astrocyte cultures were stimulated with TNF-␣, IL-1␣, or combined TNF-␣/IL-1␣ for 24 h. Cytoplasmic extracts were prepared and separated by 10% SDS-PAGE, and PVDF membrane filter blots were probed for iNOS protein by immunoblotting. iNOS expression was observed after combined TNF-␣/IL-1␣ treatment but not after treatment with TNF-␣ or IL-1␣ or after osmotic shock by sorbitol used as control (Fig. 5A). Northern analysis showed that the astrocytes responded to TNF-␣ and to IL-1␣ stimulation with induced transcription of the MCP-1 gene, but iNOS transcription required the combined activity of both cytokines (Fig. 5B). As a consequence, combined TNF-␣/IL-1␣ stimulation is required for accumulating NO release after 48 h (Fig. 5C); preliminary studies had shown that the differences among TNF-␣, IL-1␣, combined TNF-␣/IL-1␣, and sorbitol stimulation are not due to different kinetic profiles of NO release from 12 to 72 h. Interestingly, sorbitol was not able to substitute for either TNF-␣ or IL-1␣ to induce the release of NO studied by nitrite determination.
TNF-␣ and IL-1␣ Alone Fully Activate p38 MAPK -MAPKs are activated by upstream dual specificity kinases, which phosphorylate the threonine and tyrosine residues in the TEY, TGY, or TPY motifs of p42/44 MAPK , p38 MAPK , or p54 JNK , respectively. Various cytokines, including TNF-␣ and IL-1␣ have been shown to activate these different pathways. To dissect the role of the different MAPK pathways in the TNF-␣-, IL-1␣-, and combined TNF-␣/IL-1␣-induced iNOS expression, immunoprecipitation kinase assays were performed. Astrocytes were stimulated with TNF-␣, IL-1␣, combined TNF-␣/IL-1␣, and sorbitol for various times. p38 MAPK , p42/44 MAPK , and p54 JNK were immunoprecipitated sequentially from the cytoplasmic extracts with specific antibodies, and immune complex kinase assays were performed using [␥-33 P]ATP and GST-ATF-2, MBP, or GST-c-Jun as specific substrates, respectively. The reactants were separated by SDS-PAGE, and the radioactivity incorporated in the respective bands was determined (Fig. 6). p38 MAPK in TNF-␣-stimulated astrocytes reached a maximal activity as early as 5 min and declined 1 h after the start of TNF-␣ treatment (Fig. 6). The p42/44 MAPK and p54 JNK activity peaks were delayed when compared with p38 MAPK , reaching maxima about 15 min after TNF-␣. IL-1␣ treatment also induced a transient activation of p38 MAPK , p42/44 MAPK , and p54 JNK with kinetics similar to those observed with TNF-␣ stimulation (Fig.  6). The induction of iNOS expression in the astrocytes required combined TNF-␣ and IL-1␣ stimulation (Fig. 1), yet both TNF-␣ and IL-1␣ independently were able to strongly activate p38 MAPK . Furthermore, the treatment with both cytokines combined did not activate p38 MAPK in a quantitatively or qualitatively significantly different manner from either TNF-␣ or IL-1␣ alone. Similarly, the level and time dependence of the activation of p42/44 MAPK and p54 JNK were not affected by the presence of combined TNF-␣/IL-1␣ (Fig. 6).
p38 MAPK and p54 JNK are also activated by stress factors such as osmotic shock. Sorbitol in hyperosmolar concentrations was therefore used to investigate the kinase responses to osmotic stress. Preliminary studies had shown a regular dose response of p38 MAPK and p54 JNK activation by sorbitol, first saturating at 300 mM. The further studies therefore were carried out at 300 mM sorbitol, which elicited a transient activation of p38 MAPK and p54 JNK (Fig. 6). Surprisingly, sorbitol was also able to activate the p42/44 MAPK pathway generally thought to be under the control of mitogenic or growth factor stimuli.
The above findings were in full agreement with the results of immunoblot analyses of p38 MAPK and p42/44 MAPK activation. Cytoplasmic extracts of astrocytes were prepared; samples  4. iNOS transcription in TNF-␣/IL-1␣-stimulated astrocytes is sensitive to protein synthesis inhibition. Northern blot analysis of astrocytes exposed to excipient (lanes 1 and 2) or cycloheximide at 1 g/ml (lane 3) or 10 g/ml (lanes 4 and 5) for 1 h before cytokine treatment. Cells were either left untreated (lanes 1 and 5) or stimulated with combined TNF-␣/IL -1␣ (lanes 2, 3, and 4) for 6 h. 10 g of total RNA was analyzed using sequentially iNOS, MCP-1, and GAPDH DNA probes.
with calibrated amounts of total protein were separated by SDS-PAGE and transferred to PVDF filter membranes; and the phosphorylation state of p38 MAPK and p42/44 MAPK was analyzed by Western blotting using specific anti-phospho-Tyr-182 p38 MAPK or anti-phospho-Tyr-204 p42/44 MAPK antibodies that selectively target the fully phosphorylated p38 MAPK and p42/44 MAPK enzymes. As shown in Fig. 7, TNF-␣, IL-1␣, and combined TNF-␣/IL-1␣ induced a transient increase in p38 MAPK and p42/44 MAPK Tyr phosphorylation that was clearly visible as early as 5 and 15 min, respectively. The signals were significantly reduced after 60 min and were practically at background level after 360 min.
The p38 MAPK Pathway Does Not Regulate NF-B Activity-Several transcription factors, such as NF-B, have been implicated in mediating cytokine-induced iNOS expression (14,16). NF-B indeed participates in the control of genes of the cellular response to stress and inflammatory mediators. Therefore, the possibility that p38 MAPK may couple the activation of NF-B to the response to cytokines and stress in astrocyte cultures was explored. The effects of TNF-␣, IL-1␣, combined TNF-␣/IL-1␣, and sorbitol treatments in the presence or absence of FHPI on the activation of NF-B were examined by electrophoretic mobility shift assays of astrocyte nuclear extracts using a doublestranded B-specific 32 P-labeled oligonucleotide probe. An NF-B-specific band was detected with all TNF-␣-, IL-1␣-, and combined TNF-␣/IL-1␣-treated cells that was absent in untreated or sorbitol-treated nuclear extracts (Fig. 8). Interestingly, the inhibition of p38 MAPK by 10 M FHPI did not interfere with the intensities of these bands in any of the stimulated cells, suggesting that p38 MAPK is not involved in the coupling of the activation and nuclear translocation of NF-B to TNF-␣ and IL-1␣ receptor activation.
p38 MAPK Mediates iNOS Expression treatment. These results were further confirmed by nitrite measurements conducted after 48 h (Fig. 9B). Large amounts of accumulating nitrites were detected in supernatants of TNF-␣/IL-1␣-stimulated wild type and tnfr-2°astrocytes, whereas nitrite concentrations with tnfr-1°and tnfr-1°/tnfr-2°astrocytes remained at background levels. Interestingly, the TNF-␣-stimulated activation of p38 MAPK , p42/44 MAPK , and p54 JNK in the astrocytes is also exclusively mediated by TNFR-1, in contrast to findings with primary fibroblasts isolated from these mice (39). DISCUSSION The necessity to stimulate cells with combined cytokines such as TNF-␣ plus IL-1␤ or IFN-␥ or with cytokines combined with lipopolysaccharide to induce iNOS expression has been reported in several cell types such as astrocytes (45), fibroblasts (46), and endothelial cells (47), although a single cytokine can induce iNOS expression in some other cell systems including hepatocytes (48), islet cells (49), and vascular smooth muscle cells (50). In the present astrocyte cultures, combined stimulation with IL-1␣ and TNF-␣ was required to drive iNOS transcription; p38 MAPK activation was necessary but not sufficient to transduce the signal, since either IL-1␣ or TNF-␣ alone fully activated p38 MAPK . Osmotic stress also fully activated p38 MAPK but could not substitute for either TNF-␣ or IL-1␣ in the TNF-␣/IL-1␣ combination. Furthermore, combinations of either IL-1␣ or TNF-␣ with IFN-␥ were ineffective in stimulating iNOS transcription, although IFN-␥ activity enhanced the iNOS response to TNF-␣/IL-1␣. 2 Combined TNF-␣/IL-1␣ stimulation of the astrocytes was not a general requirement, since both TNF-␣ and IL-1␣ individually were fully competent to stimulate MCP-1 transcription in the astrocyte cultures.
The effect of FHPI treatment demonstrated that activation of the p38 MAPK pathway is necessary for iNOS induction. It might have been argued that the requirement for two cytokines to induce iNOS transcription reflects a synergy leading to stronger and more extended p38 MAPK activation. However, the extent and kinetics of p38 MAPK as well as p42/44 MAPK and p54 JNK kinase activation did not provide evidence for any such additive or enhancing effects by the TNF-␣ and IL-1␣ combination treatment. It must rather be proposed that in addition to the p38 MAPK pathway a TNF-␣-or IL-1␣-activated second pathway is required to generate a sufficient signal for iNOS induction.
NF-B has been reported to be involved in iNOS induction (17,51). It was strongly activated by TNF-␣ as well as IL-1␣ in the present astrocytes. NF-B and p38 MAPK lie on two distinct pathways as shown by the present finding that p38 MAPK inhibition did not interfere with activation of NF-B, confirming previous reports (52,53), and by the lack of evidence that NF-B is upstream of p38 MAPK . However, NF-B may represent a second independent pathway for iNOS induction. NF-B has been reported to be important for iNOS transcription, since the pretreatment of rat alveolar macrophages and glomerular mesangial cells with pyrrolidine dithiocarbamate, an inhibitor of NF-B activation, completely blocked iNOS transcription (54,55). Pyrrolidine dithiocarbamate treatment of the present astrocytes also suppressed iNOS transcription. 2 However, given the time scale and dependence on new protein synthesis of iNOS induction, a more general or toxic effect of pyrrolidine dithiocarbamate could not be ruled out. The responses of the p38 MAPK and NF-B systems are fast and could both induce independently one or two separate genes, yet to be defined, whose products may lead to induced iNOS transcription. The present results do not allow us to definitely resolve the nature of the second pathway required for iNOS induction, but they provide clear evidence for a parallel signaling cascade in addition to p38 MAPK to induce TNF-␣/IL-1␣-stimulated iNOS transcription.
A role of p54 JNK and p42/44 MAPK must also be considered in the control of iNOS expression, since both kinases were shown to be activated by TNF-␣ and IL-1␣ in slightly delayed but otherwise similar kinetics as p38 MAPK in the mouse astrocyte cultures. The lack of effect by PD98059 rendered the involvement of the p42/44 MAPK pathway in iNOS control unlikely. This view was further supported by a recent report where the inhibition of IFN-␥-activation of ERK1 and ERK2 by PD98059 or by Ras dominant negative expression did not affect iNOS FIG. 7. p38 MAPK and p42/44 MAPK phosphorylation by TNF-␣, IL-1␣, and combined TNF-␣/IL-1␣ treatment. Immunoblot analyses of cytoplasmic extracts of astrocytes left untreated or treated with TNF-␣ (100 ng/ml), IL-1␣ (100 ng/ml), or combined TNF-␣/IL-1␣ (100 ng/ml each) for indicated times. Cell lysates were separated by SDS-PAGE and transferred to PVDF membranes. Phosphorylation of p38 MAPK and p42/44 MAPK was analyzed using anti-phospho-Tyr-182 p38 MAPK (pp38) or anti-phospho-Tyr-204 p42/44 MAPK (ppERK1, -2) antibody. The same blots were reprobed for anti-p38 MAPK (p38) or anti-p42/44 MAPK (pERK1, -2) expression, indicating equal protein loading under all conditions. FIG. 8. p38 MAPK does not regulate activation of NF-B. Electrophoretic mobility shift assays of NF-B activation of astrocytes left unstimulated or stimulated for 15 min with TNF-␣ (100 ng/ml), IL-1␣ (100 ng/ml), combined TNF-␣/IL-1␣ (100 ng/ml each), or sorbitol (300 mM) in the absence or presence of FHPI (10 M) are shown. Nuclear extracts (10 g of total protein) were incubated with a specific 32 Plabeled oligonucleotide probe containing B elements. The specificity of the band shift was shown in a parallel gel using a point-mutated B probe.
induction in C6 glioma cells (56). In contrast, inhibition of Ha-Ras farnesylation, which blocks Ha-Ras processing, correlated with inhibition of iNOS induction in vascular smooth muscle cells, suggesting a role of the Ras/ERK pathway in the control of iNOS induction by IL-1␤ (57). Furthermore, while the present data show that p38 MAPK activity is required for the transcriptional induction of iNOS, it has been demonstrated in serum-starved mesangial cells that the inhibition of p38 MAPK promoted IL-1␤-induced iNOS expression and subsequent NO production (58). The most likely explanation for these seemingly inconsistent results is that the complex regulation of iNOS expression is tissue-specific. One further example of the intriguing cell type specificity of signal pathway connectivities is provided by the coupling of TNF receptors to the MAPK pathways. Studies of primary fibroblast cultures of the same mice from which the astrocyte cultures had been isolated had shown that p38 MAPK , p42/44 MAPK , and p54 JNK are all fully activated by TNF-␣ treatment of wild type, tnfr-1°, and tnfr-2°f ibroblasts (39), 2 demonstrating that both TNFR-1 and TNFR-2 in these fibroblasts couple to all three MAPK pathways, whereas only TNFR-1-mediated signals acceded to the three MAPK pathways in the astrocytes. Previous studies in tnfr-1°h epatic cells (59), peritoneal macrophages (60), and in in vivo models (61) had reported TNFR-1 activity to be necessary to trigger NO release. With the present tnfr-2°astrocytes, TNF-␣-activated pathways, when combined with IL-1␣ signals, were also sufficient to induce iNOS expression and produced comparable levels of iNOS protein and nitrites as wild type astrocytes, demonstrating the sufficient role of TNF-1. In contrast to the astrocytes, IL-1␣ stimulation alone suffices to induce iNOS expression in the fibroblasts of the same mouse lines, 2 further demonstrating the complex and tissue-specific regulation of iNOS expression.