Role of the Cyclic AMP-Protein Kinase A Pathway in Lipopolysaccharide-induced Nitric Oxide Synthase Expression in RAW 264.7 Macrophages

The signaling pathway for lipopolysaccharide (LPS)-induced nitric oxide (NO) release in RAW 264.7 macrophages involves the protein kinase C and p38 activation pathways (Chen, C. C., Wang, J. K., and Lin, S. B. (1998) J. Immunol. 161, 6206–6214; Chen, C. C., and Wang, J. K. (1999) Mol. Pharmacol. 55, 481–488). In this study, the role of the cAMP-dependent protein kinase A (PKA) pathway was investigated. The PKA inhibitors, KT-5720 and H8, reduced LPS-induced NO release and inducible nitric oxide synthase (iNOS) expression. The direct PKA activator, Bt2cAMP, caused concentration-dependent NO release and iNOS expression, as confirmed by immunofluorescence studies. The intracellular cAMP concentration did not increase until after 6 h of LPS treatment. Two cAMP-elevating agents, forskolin and cholera toxin, potentiated the LPS-induced NO release and iNOS expression. Stimulation of cells with LPS or Bt2cAMP for periods of 10 min to 24 h caused nuclear factor-κB (NF-κB) activation in the nuclei, as shown by detection of NF-κB-specific DNA-protein binding. The PKA inhibitor, H8, inhibited the NF-κB activation induced by 6- or 12-h treatment with LPS but not that induced after 1, 3, or 24 h. The cyclooxygenase-2 (COX-2) inhibitors, NS-398 and indomethacin, attenuated LPS-induced NO release, iNOS expression, and NF-κB DNA-protein complex formation. LPS induced COX-2 expression in a time-dependent manner, and prostaglandin E2production was induced in parallel. These results suggest that 6 h of treatment with LPS increases intracellular cAMP levels via COX-2 induction and prostaglandin E2 production, resulting in PKA activation, NF-κB activation, iNOS expression, and NO production.

functions mediated by activated macrophages, including antimicrobial and tumoricidal activity, implicated in the pathogenesis of tissue damage associated with acute and chronic inflammation (2,3). Macrophages generate NO from the guanidino moiety of L-arginine via a reaction catalyzed by the inducible form of nitric oxide synthase (iNOS) (4). iNOS has been identified in a wide variety of cell types including macrophages, mesangial cells, vascular smooth muscle cells, keratinocytes, chondrocytes, osteoclasts, and hepatocytes and can be induced by many immune stimuli (1,5). Changes in NO formation in iNOS-expressing cells usually correlate with similar changes in iNOS mRNA levels, indicating that a major part of iNOS regulation occurs at the transcription level. The promoter region of the iNOS gene contains several binding sites for transcriptional factors, such as nuclear factor-B (NF-B) and activator protein-1, as well as for various members of the CCAAT/enhancer-binding protein, activating transcription factor/cAMPresponse element-binding protein, and Stat families of transcriptional factors (6). Of these, the proteins of the NF-B family appear to be essential for the enhanced iNOS gene expression seen in macrophages exposed to the active component of endotoxin, lipopolysaccharide (LPS) (7). In unstimulated cells, NF-B is retained in the cytoplasm by binding to IB but is released by signal induction and translocates to the nucleus, activating the responsive gene (8). In macrophages, iNOS induction by LPS requires initiation of gene expression and de novo protein synthesis over a period of several hours (9).
The intracellular signaling pathways by which LPS causes iNOS expression in macrophages involve a series of events resulting in the transmission of the signal from the plasma membrane through the cytoplasm to the nucleus, where iNOS gene expression is up-regulated. Previous studies have shown that LPS first binds to LPS-binding protein and then to membrane CD14 and that it also activates phosphatidylinositolphospholipase C and phosphatidylcholine-PLC by tyrosine phosphorylation, thus causing PKC activation (10). Tyrosine phosphorylation also causes p38 activation (11). These phosphorylation processes result in stimulation of NF-B DNAprotein binding and the initiation of iNOS expression and NO release (10,11). An increase in intracellular cAMP levels is an important intracellular signaling mechanism involved in the regulation of gene expression. Certain in vitro studies have shown that an increase in cAMP levels causes iNOS induction (12)(13)(14), whereas in other studies, increased cAMP levels caused a reduction in iNOS (15,16). In the present study, we explored the intracellular signaling pathway for the LPS-induced increase in cAMP levels and its involvement in LPSstimulated NO production in RAW 264.7 macrophages. The results show that, after 6 h of treatment, LPS can increase cAMP levels by induction of cyclooxygenase-2 (COX-2) and formation of prostaglandin E 2 (PGE 2 ), resulting in the activation of PKA and NF-B, iNOS expression, and NO production. The PKA activation pathway explored in this study had a delayed onset (6 h), whereas the previously reported PKC and p38 activation pathways have rapid onsets (10 min) (10,11).
Cell Culture-RAW 264.7 cells, a murine macrophage cell line, were obtained from the ATCC and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, and 100 g/ml streptomycin in 12-well plates (nitrite assay, iNOS and COX-2 expression, and PGE 2 production), on 24-mm glass coverslips in 35-mm dishes (immunofluorescence staining), in 6-well plates (cAMP assay), or in 10-cm dishes (NF-B gel shift assay).
Determination of NO Concentration-NO production in culture supernatants was assessed by measuring nitrite, its stable degradation product, using Griess reagent as described previously (10). The Dulbecco's modified Eagle's medium was changed to phenol red-free medium before the cells were stimulated for 24 h with 1 g/ml LPS or 100 M Bt 2 cAMP. After stimulation, the supernatants were centrifuged and mixed with an equal volume of Griess reagent and then incubated for 10 min at room temperature before measuring the absorbance at 550 nm in a microplate reader. NaNO 2 was used as a standard. In pretreatment experiments, the cells were incubated for 30 min with KT-5720 or H8 (PKA inhibitors), with actinomycin C or cycloheximide (transcriptional or translational inhibitors, respectively), or with NS-398 or indomethacin (COX-2 inhibitors) before the addition of LPS or Bt 2 cAMP.
Immunofluorescence Staining-RAW cells grown on coverslips were treated for 24 h with LPS or Bt 2 cAMP in growth medium and then rapidly washed with phosphate-buffered saline and fixed at room temperature for 10 min with 2% paraformaldehyde. After washing with phosphate-buffered saline, the cells were blocked for 15 min with 1% bovine serum albumin in TTBS (50 mM Tris-HCl, pH 7.5, 0.15 M NaCl, 0.05% Tween 20) containing 0.1% Triton X-100, incubated with anti-iNOS antibody (1:100) for 1 h, washed extensively, and stained for 30 min with anti-rabbit IgG-fluorescein (1:2,000). After additional washes, the coverslips were mounted on glass slides using mounting medium (2% N-propyl gallate in 60% glycerol, 0.1 M phosphate-buffered saline, pH 8.0). Optical sections of the immunostained cells were observed and photographed using a Zeiss Axiovert inverted microscope equipped with the photo MicroGraph digitized integration system (MGDS).
Preparation of Cell Extracts and Western Blot Analysis of iNOS and COX-2-Following treatment with LPS or Bt 2 cAMP, with or without pretreatment with various inhibitors, the cells were harvested and collected. Cell lysates were prepared and subjected to SDS-polyacrylamide gel electrophoresis using 7.5% (iNOS) or 10% (COX-2) running gels as described previously (10). The proteins were transferred to nitrocellulose, and the membrane was incubated successively with 0.1% milk in TTBS at room temperature for 1 h, with rabbit antibody specific for iNOS or COX-2 for 1 h, and with horseradish peroxidase-labeled anti-rabbit antibody for 30 min. After each incubation, the membrane was washed extensively with TTBS. The immunoreactive band was detected using ECL detection reagent and developed with Hyperfilm-ECL.
Determination of Intracellular cAMP Concentrations-After cells were treated with LPS for 1, 3, 6, 12, or 24 h or with CTX for 24 h or with forskolin for 10 min, the reaction was terminated by aspiration of the growth medium and addition of 0.1 N HCl. The cells were scraped into Eppendorff tubes and the suspensions were centrifuged; the supernatants were then neutralized with 10 N NaOH and assayed for cAMP levels using an enzyme immunoassay kit from Amersham Pharmacia Biotech.

Preparation of Nuclear Extracts and the Electrophoretic Mobility
Shift Assay (EMSA)-Control cells or H8, NS-398, or indomethacinpretreated cells were treated with LPS or Bt 2 cAMP for various amounts of time, and then nuclear extracts were prepared as described previously (10). A double-stranded oligonucleotide probe containing NF-B binding sequences was purchased (5Ј-AGTTGAGGGGACTTTC-CCAGGGC-3Ј, Santa Cruz Biotechnology) and end-labeled with [␥-32 P]ATP using T4 polynucleotide kinase. The nuclear extracts (6 -10 g) were incubated at 30°C for 20 min with 1 ng of 32 P-labeled NF-B probe (40,000 -60,000 cpm) in 10 l of binding buffer containing 1 g of poly(dI⅐dC), 15 mM HEPES, pH 7.6, 80 mM NaCl, 1 mM EGTA, 1 mM dithiothreitol, and 10% glycerol as described previously (10). DNAnuclear protein complexes were separated from the DNA probe by electrophoresis on a native 6% polyacrylamide gel, and the gel was vacuum-dried and subjected to autoradiography using an intensifying screen at Ϫ80°C. The quantitative data were obtained using a computing densitometer and ImageQuant software (Molecular Dynamics). PGE 2 Production-After treatment of cells with LPS for 1, 3, 6, 12, or 24 h, PGE 2 levels in the culture media were measured using an enzyme immunoassay kit from Amersham Pharmacia Biotech.

Inhibitory Effect of PKA Inhibitors on LPS-induced NO Production and iNOS Expression and
Effect of Bt 2 cAMP-To determine whether PKA was involved in the LPS-induced NO production, the PKA inhibitors, KT-5720 and H8, were used. When cells were pretreated for 30 min with 1 or 3 M KT-5720 or with 30 or 50 M H8, LPS-induced NO production (Fig. 1A) and iNOS expression (Fig. 1, B and C) were inhibited in a dose-dependent manner. The respective levels of inhibition were 31 or 62% for 1 or 3 M KT-5720 and 35 or 65% for 30 or 50 M H8.
Because both LPS-induced NO production and iNOS expression were inhibited by KT-5720 or H8, indicating the involvement of the PKA pathway in the LPS effect, the direct PKA activator, Bt 2 cAMP, was used. Exposure of RAW cells to  Fig. 2A). In the following NO release experiment, the cells were treated with 100 M Bt 2 cAMP for 24 h. Under these conditions, either the transcriptional inhibitor, actinomycin D, or the translational inhibitor, cycloheximide, inhibited the Bt 2 cAMP-induced nitrite production and iNOS expression (data not shown). Bt 2 cAMP-induced iNOS expression was also demonstrated by immunofluorescence staining; as shown in Fig. 3, iNOS expression was not seen in the basal state ( Fig. 3B) but was induced in the cytoplasm after treatment with either LPS (Fig. 3D) or Bt 2 cAMP (Fig. 3F).
Because the PKA pathway had been shown to be involved in LPS-induced NO production and because Bt 2 cAMP stimulated NO production, intracellular cAMP levels were measured following LPS treatment. When cells were treated with 1 g/ml LPS for various times, cAMP levels increased slightly after 3 h (121% of basal), reached a maximum at 6 h (243% of basal), and then declined (161% of basal after 12 h) (Fig. 4A). Following treatment of cells with 1 g/ml CTX for 24 h or with 100 M forskolin for 10 min, cAMP levels increased to 292 and 202% of basal, respectively (Fig. 4B).
Effect of Cyclic AMP-elevating Agents on LPS-induced NO Production and iNOS Expression-Forskolin or CTX themselves had no effect on nitrite production but enhanced the LPS-stimulated increase in nitrite production and iNOS expression (Fig. 5, A-C). Ten or 30 M forskolin, which had no effect on cAMP levels in RAW cells (data not shown), also had no effect on LPS-induced NO production and iNOS expression, whereas 100 M forskolin, which increased cAMP levels 2-fold (Fig. 4B), also increased LPS-induced NO production and iNOS expression (Fig. 5, A and  B). CTX potentiated the LPS effect over the range of 10 -1,000 ng/ml (Fig. 5, A and C). A similar parallel enhancement of the LPS-stimulated increase in NO production and iNOS expression was seen using Bt 2 cAMP (Fig. 6).
Kinetics of NF-B-specific DNA-Protein Complex Formation in Nuclei Stimulated with LPS or Bt 2 cAMP and the Inhibitory Effect of H8-The time course of NF-B activation after treatment with 1 g/ml LPS or 100 M Bt 2 cAMP was studied. Nuclear extracts prepared from RAW cells were assayed for activated NF-B in an EMSA. As shown in Fig. 7A, NF-Bspecific DNA-protein complex formation increased after treatment with LPS for 1, 3, 6, 12, or 24 h. When cells were exposed to 100 M Bt 2 cAMP for 10 min, increased formation of the NF-B-specific DNA-protein complex was also seen (Fig. 7B), whereas after treatment with Bt 2 cAMP for 3 or 24 h, the intensity of these complexes decreased but was still stronger than in resting cells (Fig. 7B). The bands in the upper and lower complex were previously identified as the p65/p50 heterodimer and p50/p50 homodimer, respectively (10). After pretreatment of the cells for 30 min with 50 M H8, the activation of NF-B-specific DNA-protein complex formation induced following 1, 3, or 24 h of LPS treatment was not affected, whereas that induced following 6 or 12 h of LPS treatment was inhibited, the extent of inhibition being 46% and 26%, respectively (Fig. 8A). The activation of NF-B-specific DNA complex formation seen after 6 h of LPS treatment was inhibited by H8 in a dose-dependent manner (30, 50, and 75 M) (Fig. 8B).

Inhibitory Effect of COX-2 Inhibitors on LPS-induced NO Production, iNOS Expression, and NF-B DNA-Protein Complex Formation and Induction of COX-2 by LPS-
The fact that the cAMP-PKA pathway had been shown to be involved in LPS-induced NO production and iNOS expression, that LPS caused an increase in cAMP levels after 6 h of treatment, and that H8 inhibited LPS-induced NF-B-specific DNA-protein complex formation following 6 h of treatment indicated that the cAMP formation was a delayed response. To determine whether the increased cAMP levels were because of PG formation produced as a result of COX-2 expression, the COX-2 inhibitors, NS-398 and indomethacin, were used. As shown in Fig. 9A, LPS-induced NO production and iNOS expression were inhibited by 10 M NS-398 or indomethacin. The inhibition of NO production was 35 and 38%, respectively. NF-B DNA-protein complex formation induced after 6 h of treatment with LPS was also inhibited by 10 M NS-398 or indomethacin (Fig. 9B). LPS-elicited COX-2 expression was also examined. Exposure of RAW cells to 1 g/ml LPS resulted in a time-dependent COX-2 expression; no expression was seen after 1 h of treatment, but expression was observed at 3 h and continued to increase to 24 h (Fig. 10A). Fig. 10B shows the time-dependent production of PGE 2 in response to 1 g/ml LPS. The basal release of PGE 2 was 0.838 pg/g of total protein, whereas after treatment with LPS for 1, 3, 6, 12, or 24 h, this rose to 1.82, 9.23, 37.8, 71.2, and 112.3 pg/g protein, respectively.

DISCUSSION
The effects of cAMP on iNOS expression have been of increasing interest since the first report that cAMP-elevating agents induced iNOS in cultured vascular smooth muscle cells and that this induction was synergistic with that elicited by inflammatory cytokines (17). Similar effects have also been seen in renal mesangial cells (13,18) and brown adipocytes (19). Although cAMP alone does not induce iNOS in unstimulated cardiac myocytes, it augments iNOS induction in interleukin-1␤-stimulated cells (20). In 3T3 fibroblasts, different signaling pathways, including elevation of cAMP, lead to the induction of iNOS by NF-B mediation (21). In contrast, elevation of cellular cAMP levels has been shown to down-regulate iNOS in LPS-or cytokine-activated astrocytes, hepatocytes, or Kuffer cells (15,16,22). In RAW 264.7 macrophages, NF-B/Rel is positively regulated by the cAMP cascade, thus helping to initiate iNOS gene expression in response to LPS stimulation, and inhibition of adenylate cyclase attenuates LPS-induced activation of iNOS gene expression (12), indicating that, in inducing iNOS expression in these cells, LPS acts by increasing cAMP levels. In the present study, the PKA inhibitors, KT-5720 and H8, inhibited LPS-induced NO release and iNOS expression in a dose-dependent manner, indicating that the LPS effect is indeed related to the cAMP-PKA activation pathway. The cAMP analogue, Bt 2 cAMP, also increased NO release and iNOS expression; immunofluorescence staining also demonstrated iNOS expression in the cytoplasm. LPS caused a time-dependent increase in cAMP levels that was maximal with 6-h treatment and then declined. In J774 macrophages, the increase in cAMP levels occurs after 6 h of treatment with LPS (23), Cells were incubated at 37°C with 1 g/ml LPS for various time intervals; the medium was then removed and analyzed for PGE 2 production by enzyme-linked immunosorbent assay (B), while the cells were lysed and subjected to Western blotting using anti-COX-2 antibody as described under "Experimental Procedures" (A). B, the results are expressed as the mean Ϯ S.E. of one typical experiment performed in duplicate; similar results were obtained in three independent experiments. and PKA is involved in the LPS-induced activation of junB and NF-B (24). As previously reported (10,11), activation of NF-Bspecific DNA-protein complex formation was seen after 10-min to 24-h treatment with LPS, and a similar time course of activation of this complex was seen using Bt 2 cAMP (Fig. 7B). In contrast with the inhibition seen using PKC or p38 inhibitors (10,11), when cells were pretreated for 30 min with H8, the NF-Bspecific DNA-protein complex formation seen after 1 h of LPS treatment was unaffected (Fig. 8A). However, the complex formation seen after 6 h of LPS treatment was inhibited by H8, thus correlating with the maximal cAMP level seen after 6 h of treatment (Fig. 4A). Thus, in contrast with the PKC and p38 activation pathways, which are rapid (10 min) (10,11), the cAMP-PKA activation pathway is a delayed event in LPS-induced NF-B activation. cAMP may modulate NF-B activation and iNOS transcription via cAMP-dependent PKA-mediated phosphorylation of the cAMP response element-binding protein (25). In RAW cells, the rapid activation of NF-B by PKC and p38 pathways, together with the delayed activation of NF-B by the cAMP-PKA pathway, contributes to the LPS-induced iNOS expression and NO release.
Because the cAMP-PKA activation pathway is a much delayed event (6 h) in LPS-induced NF-B activation, the mechanism involved in LPS-induced increase in cAMP levels was further explored. Both NS-398 and indomethacin had an inhibitory effect on LPS-induced NO release, iNOS expression, and NF-B activation, indicating the involvement of COX-2 expression in LPS-stimulated NO release. When the effect of various periods of LPS treatment was studied, no COX-2 expression was seen in unstimulated cells or after 1 h of treatment, but COX-2 expression was seen after 3 h of treatment and continued to increase to 24 h. COX is a key enzyme in prostanoid synthesis, as it catalyzes the conversion of arachidonic acid to PGH2, which is then metabolized by one or more terminal synthases to a variety of active prostanoids (26). It possesses both fatty acid cyclooxygenase activity and PG hydroperoxidase activity (converting PGG2 to PGH2). COX-2 is a COX isoform that is induced in a number of cells by proinflammatory stimuli and is thought to contribute to the generation of pros-tanoids at sites of inflammation (27,28); it is considered to be responsible for high production of PGs (29). PGE 2 production following LPS treatment was also measured, and the increases after 1, 3, 6, 12, or 24 h of treatment were, respectively, 2-, 11-, 45-, 85-, and 134-fold of basal levels, paralleling the increase in COX-2 expression. PGE 2 acts via receptor-mediated generation of cAMP and activation of PKA (30). As seen in a study on the effects of interleukin-1␤ on human bronchial smooth muscle cells (31), in the present study, induction of PGE 2 synthesis precedes the increase in cAMP, and PGE 2 acts as an autocrine factor for adenylate cyclase activation. LPS-induced tumor cell killing in EC4 cells is also because of increased levels of cAMP, and this effect is inhibited by indomethacin (32). In peritoneal macrophages, LPS is reported to act via PGE 2 to increase cAMP levels (33).
In summary, in RAW 264.7 cells, LPS increases iNOS expression via a prostanoid-and cAMP-dependent pathway, and this is followed by PKA activation of NF-B. The increase in PGE 2 is because of COX-2 expression. This effect has a more delayed onset (6 h) compared with those involving the PKC and p38 activation pathways (10 min). A schematic representation of the signaling pathway for the LPS-induced NO release in RAW cells is shown in Fig. 11.