cAMP targeting of p38 MAP kinase inhibits thrombin-induced NF-κB activation and ICAM-1 expression in endothelial cells

Arshad Rahman, Khandaker N. Anwar, Mohd. Minhajuddin, Kaiser M. Bijli, Kamran Javaid, Andrea L. True, Asrar B. Malik

Abstract

We investigated the mechanisms by which elevated intracellular cAMP concentration inhibits the thrombin-induced ICAM-1 expression in endothelial cells. Exposure of human umbilical vein endothelial cells to forskolin or dibutyryl cAMP, which increase intracellular cAMP by separate mechanisms, inhibited the thrombin-induced ICAM-1 expression. This effect of cAMP was secondary to inhibition of NF-κB activity, the key regulator of thrombin-induced ICAM-1 expression in endothelial cells. The action of cAMP occurred downstream of IκBα degradation and was independent of NF-κB binding to the ICAM-1 promoter. We observed that cAMP interfered with thrombin-induced phosphorylation of NF-κB p65 (RelA) subunit, a crucial event promoting the activation of the DNA-bound NF-κB. Because p38 MAPK can induce transcriptional activity of RelA/p65 without altering the DNA binding function of NF-κB, we addressed the possibility that cAMP antagonizes thrombin-induced NF-κB activity and ICAM-1 expression by preventing the activation of p38 MAPK. We observed that treating cells with forskolin blocked the activation of p38 MAPK, and inhibition of p38 MAPK interfered with phosphorylation of RelA/p65 induced by thrombin. Our data demonstrate that increased intracellular cAMP concentration in endothelial cells prevents thrombin-induced ICAM-1 expression by inhibiting p38 MAPK activation, which in turn prevents phosphorylation of RelA/p65 and transcriptional activity of the bound NF-κB.

  • endothelial adhesivity
  • polymorphonuclear leukocytes
  • IκBα degradation
  • RelA/p65 phosphorylation
  • mitogen-activated protein kinase
  • nuclear factor-κB
  • intercellular adhesion molecule-1
  • cyclic adenosine 5′-monophosphate

the procoagulant serine protease thrombin, released during intravascular coagulation initiated by tissue injury or sepsis (11, 14), promotes the adhesion of polymorphonuclear leukocytes (PMN) to the endothelium by a mechanism involving the endothelial cell surface expression of intercellular adhesion molecule-1 (ICAM-1) (28, 29, 40). ICAM-1, a member of the immunoglobulin supergene family (38), serves as a counterreceptor for leukocyte β2-integrins CD11a/CD18 (lymphocyte function-associated antigen-1) and CD11b/CD18 (Mac-1) and mediates the arrest of PMN and other leukocytes and their migration across the vessel wall barrier (36). We showed that the transcription factor NF-κB p65 (RelA) is an essential regulator of thrombin-induced ICAM-1 gene transcription following thrombin-induced activation of the heterotrimeric G protein-coupled receptor protease-activated receptor-1 in endothelial cells (28, 30).

NF-κB, typically a heterodimer of 50- (p50) and 65-kDa (RelA) subunits, is sequestered in the cytoplasm bound to IκB proteins that mask the nuclear localization sequence of NF-κB (3). NF-κB activity is in part regulated at the level of IκBα degradation, which is accomplished through serine phosphorylation (Ser32 and Ser36) of IκBα (41) by IκBβ kinase (16, 45). Phosphorylation targets IκBα for ubiquitination and proteasome-mediated degradation (31). The liberated NF-κB is translocated to the nucleus and binds to NF-κB-responsive elements in genes. Studies have shown an additional regulatory pathway involving phosphorylation of RelA/p65 that stimulates the activation of NF-κB (2, 23). In contrast with IκBα phosphorylation, RelA/p65 phosphorylation can be mediated by several kinases depending on the stimulus and cell type (12, 33, 43, 44).

The p38 MAPKs are widely expressed serine/threonine kinases activated by a number of stimuli, including thrombin (21, 29, 34). Activation of p38 MAPK has been implicated in the mechanism of NF-κB activation (1, 29, 42). We have recently shown that inhibition of p38 MAPK prevented thrombin-induced transcriptional activity of NF-κB without altering its DNA binding function (29), indicating the involvement of p38 MAPK in promoting the transactivation of NF-κB (42). cAMP is an ubiquitous second messenger in cells. An important function of cAMP is to activate protein kinase A (39), which has been implicated in the regulation of numerous genes through phosphorylation and activation of the cAMP response element binding (CREB) protein (9). Studies show that cAMP regulates the activation of NF-κB in a cell-specific manner. In the promyelocytic cell line HL-60, elevated cAMP levels induced NF-κB activation (35), whereas exposure of endothelial cells to cAMP prevented NF-κB activation (24). Because cAMP modulation of NF-κB activity in endothelial cells may be important in controlling ICAM-1 expression during inflammation, in the present study we addressed the mechanism of this modulatory cAMP action. We provide evidence herein that cAMP inhibits thrombin-induced NF-κB activity and ICAM-1 expression in endothelial cells by blocking p38 MAPK activation, which in turn prevents the phosphorylation of RelA/p65 and activation of the DNA-bound NF-κB.

MATERIALS AND METHODS

Materials.

Human thrombin (activity of 3,170 NIH U/mg) was purchased from Enzyme Research Laboratories (South Bend, IN). Polyclonal antibodies against p38 MAPK, IκBα, and NF-κB p65 and a monoclonal antibody against ICAM-1 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies that detect p38 MAPK when activated by dual phosphorylation at Thr180 and Tyr182, RelA/p65 when activated by phosphorylation at Ser536, or IκBα when phosphorylated at Ser32 and Ser36 were obtained from Cell Signaling Technology (Beverly, MA). In addition, polyvinylidene difluoride (PVDF) membrane was from Millipore (Bradford, MA). Forskolin and dibutyryl cAMP (DBcAMP) were from Sigma Chemical (St. Louis, MO). SB-203580 was from Calbiochem-Novabiochem (La Jolla, CA). Protein assay kit and nitrocellulose membrane were from Bio-Rad Laboratories (Hercules, CA). Plasmid Maxi Kit was from Qiagen (Valencia, CA). All other materials were from Fisher Scientific (Pittsburgh, PA).

Cell culture.

Human umbilical vein endothelial cells (HUVEC; Clonetics, La Jolla, CA) were cultured as described (29) in gelatin-coated flasks using endothelial basal medium 2 (EBM2) with Bullet Kit additives (Clonetics). Confluent cells were incubated for 2–12 h in heat-inactivated 0.5–1% FBS containing EBM2 before thrombin challenge. All experiments, except where indicated, were made in cells under the eighth passage.

Flow cytometry analysis.

Flow cytometry analysis was performed as described (28). Briefly, HUVEC monolayers in six-well tissue culture dishes were pretreated with forskolin (20 μM) or DBcAMP (0.5 mM) for 30 min before stimulation with thrombin for 12–15 h. After the incubation period was completed, cells were washed twice with cold PBS, removed by careful trypsinization, and washed again with Ca2+/Mg2+-free PBS before being incubated with 20% horse serum for 30 min. After two washes, cells were incubated with a mouse MAb directed against human ICAM-1, BIRR0001 (kindly provided by Dr. Robert Rothlein, Boehringer Ingelheim, Ridgefield, CT) (32), in Ca2+/Mg2+-free PBS containing 3% horse serum for 30 min at 4°C. Cells were then washed twice with PBS/horse serum and incubated for 30 min at 4°C with a goat anti-mouse IgG FITC-conjugated secondary antibody. Cells were then fixed with 2% paraformaldehyde and analyzed by flow cytometry in a FACScan cytofluorometer (Becton Dickinson, Mountain View, CA), and the results were gated for mean fluorescence intensity above the fluorescence produced by the secondary antibody alone.

Cell lysis and immunoblotting.

After treatment, the cells were lysed in radioimmune precipitation buffer (50 mM Tris·HCl, pH 7.4, 150 mM NaCl, 0.25 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 5 mM sodium fluoride, 1 mM PMSF, and 1 μg/ml each of leupeptin, pepstatin A, and aprotinin). Cell lysates were analyzed by SDS-PAGE and transferred onto nitrocellulose (Bio-Rad Laboratories) or PVDF (Millipore) membranes, and the residual binding sites on the filters were blocked by incubating with 5% (wt/vol) nonfat dry milk in 10 mM Tris, pH 8.0, 150 mM NaCl, and 0.05% Tween 20 for 1 h at room temperature or overnight at 4°C. The membranes were subsequently incubated with the indicated antibodies and developed using an enhanced chemiluminesence method as described (29).

p38 MAPK assay.

Cells were serum-starved by overnight incubation in EBM2-1% FBS. The cells were subsequently challenged with thrombin (5 U/ml) for 5 min in the absence and presence of forskolin (20 μM) or DBcAMP (0.5 mM), which was added 30–60 min before thrombin treatment. The cells were then lysed with a phosphorylation lysis buffer (50 mM HEPES, 150 mM NaCl, 200 μM sodium orthovanadate, 10 mM sodium pyrophosphate, 100 mM sodium fluoride, 1 mM EDTA, 1.5 mM magnesium chloride, 10% glycerol, 0.5–1% Triton X-100, 1 mM PMSF, and 10 μg/ml aprotinin). Cell lysates were immunoprecipitated with an antibody against p38 MAPK using protein G-Sepharose (Amersham Pharmacia Biotech) as described (29). The immunocomplexes were washed three times with phosphorylation lysis buffer and two times with kinase buffer (25 mM HEPES, pH 7.4, 25 mM MgCl2, 25 mM β-glycerophosphate, 2 mM DTT, 0.1 mM sodium orthovanadate, and 20 μM ATP) and resuspended in 30 μl of kinase buffer containing 5 μg of activating transcription factor (ATF)-2 and 20–30 μCi of [γ-32P]ATP. The reaction was incubated for 15–30 min at room temperature and was terminated by the addition of SDS-sample buffer. Proteins were analyzed by SDS-PAGE, and the phosphorylated form of ATF-2 was detected by autoradiography.

Northern blot analysis.

Total RNA was isolated from HUVEC with RNeasy kit (Qiagen) according to the manufacturer's recommendations. Quantification and purity of RNA were assessed by A260:A280 absorption, and an aliquot of RNA (20 μg) from samples with a ratio >1.6 was fractionated using a 1% agarose formaldehyde gel. The RNA was transferred to Duralose-UV nitrocellulose membrane (Stratagene, La Jolla, CA) and covalently linked by ultraviolet irradiation using a Stratalinker UV crosslinker (Stratagene). Human ICAM-1 (0.96-kb SalI-to-PstI fragment) (38) and rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 1.1-kb PstI fragment) were labeled with [α-32P]dCTP using the random primer kit (Stratagene), and hybridization was carried out as described (28). Briefly, the blots were prehybridized for 30 min at 68°C in QuikHyb solution (Stratagene) and hybridized for 2 h at 68°C with random primed 32P-labeled probes. After hybridization, the blots were washed 2× for 30 min at room temperature in 2× SSC with 0.1% SDS followed by two washes for 15 min each at 60°C in 0.1× SSC with 0.1% SDS. Autoradiography was performed with an intensifying screen at −70°C for 12–24 h. The nitrocellulose membrane was soaked for stripping the probe with boiled water or 0.1× SSC with 0.1% SDS.

Reporter gene constructs, endothelial cell transfection, and luciferase assay.

The plasmid pNF-κB-LUC containing five copies of consensus NF-κB sequences linked to a minimal E1B promoter-luciferase gene was purchased from Stratagene. Transfections were performed using the DEAE-dextran method (22) with slight modifications (29). Briefly, 5 μg of DNA were mixed with 50 μg/ml of DEAE-dextran in serum-free EBM2, and the mixture was added onto cells that were 70–80% confluent. We used 0.125 μg of pTKRLUC plasmid (Promega, Madison, WI) containing Renilla luciferase gene driven by the constitutively active thymidine kinase promoter to normalize the transfection efficiencies. After 1 h, cells were incubated for 4 min with 10% DMSO in serum-free EBM2. The cells were then washed 2× with EBM2–10% FBS and grown to confluency. Cell extracts were prepared and assayed for firefly and Renilla luciferase activities using Promega Biotech Dual Luciferase Reporter Assay System. The data were expressed as a ratio of firefly and Renilla luciferase activity.

Cytoplasmic and nuclear extract preparation.

After treatment, cells were washed 2× with ice-cold Tris-buffered saline and resuspended in 400 μl of buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, and 0.5 mM PMSF). After 15 min, Nonidet P-40 was added to a final concentration of 0.6%. Samples were centrifuged to collect the supernatants containing cytosolic proteins for determining IκBα degradation by Western blot analysis. The pelleted nuclei were resuspended in 50 μl of buffer B (20 mM HEPES, pH 7.9, 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, and 1 mM PMSF). After 30 min at 4°C, lysates were centrifuged, and supernatants containing the nuclear proteins were transferred to new vials. Protein concentration of the extract was measured using a Bio-Rad protein determination kit.

EMSA.

EMSAs were performed as described (29). Briefly, 10 μg of nuclear extract were incubated with 1 μg of poly(dI-dC) in a binding buffer (10 mM Tris·HCl, pH 7.5, 50 mM NaCl, 0.5 mM DTT, 10% glycerol; 20 μl final volume) for 15 min at room temperature. Then, end-labeled, double-stranded oligonucleotides containing an NF-κB site (30,000 counts per minute each) were added, and the reaction mixtures were incubated for 15 min at room temperature. The DNA-protein complexes were resolved in 5% native PAGE in low ionic strength buffer (0.25× Tris-borate-EDTA). The oligonucleotide used for the gel shift analysis was Ig-κB 5′-AGTTGAGGGGACTTTCCCAGGC-3′. The Ig-κB oligonucleotide contains the consensus NF-κB binding site sequence. The sequence motifs within the oligonucleotides are underlined.

PMN adhesion assay.

PMN adhesion assay was performed as described (4). Briefly, HUVEC grown on 12-mm circular coverslips were stimulated with thrombin and washed extensively to remove residual thrombin before being labeled with 3 μM fluorescent (red) Cell Tracker dye for 30 min. Freshly isolated human neutrophils were stained with 5 μM fluorescent (green) Cell Tracker dye, coincubated with endothelial cells for 20 min, washed with PBS, and visualized using a fluorescent microscope. The adherent PMN were counted and expressed as PMN/0.8 mm2 of endothelial cell.

RESULTS

cAMP inhibits thrombin-induced ICAM-1 mRNA expression in endothelial cells.

We determined the effects of increasing intracellular cAMP on ICAM-1 mRNA expression in response to thrombin challenge of endothelial cells. We used forskolin, an adenylate cyclase activator, and DBcAMP, a synthetic cell-permeable cAMP analog, to raise the intracellular cAMP by two independent means. Northern blot analysis showed that pretreatment of HUVEC monolayers with forskolin or DBcAMP reduced thrombin-induced ICAM-1 transcript in a dose-dependent manner (Fig. 1, A and B). In another experiment, we determined whether cAMP can also inhibit ICAM-1 mRNA expression in response to TNF-α challenge of endothelial cells. These results showed that forskolin, in contrast to its effect on thrombin response, failed to prevent TNF-α-induced ICAM-1 mRNA expression (Fig. 1C).

Fig. 1.

Elevated cAMP concentration in endothelial cells inhibits thrombin-induced ICAM-1 mRNA expression. Confluent human umbilical vein endothelial cell (HUVEC) monolayers were pretreated with forskolin (FSK; A and C) or dibutyryl cAMP (DBcAMP; B) at the indicated concentrations. After 30 min, cells were challenged with thrombin (2.5 U/ml, A–C) for 3 h or TNF-α (100 U/ml, C) for 2 h. Total RNA was isolated and analyzed by Northern hybridization with a human ICAM-1 cDNA, which hybridizes to a 3.3-kb transcript. Blots were stripped and reprobed to determine glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression as a measure of RNA loading. Results are representative of 3 separate experiments.

cAMP inhibits thrombin-induced ICAM-1 cell surface expression and endothelial adhesivity toward PMN.

We next evaluated the effects of cAMP on thrombin-induced ICAM-1 cell surface expression and endothelial adhesivity toward naive PMN. FACS analysis showed that thrombin challenge of HUVEC resulted in increased ICAM-1 cell surface expression (Fig. 2Ab). Pretreatment of cells with forskolin or DBcAMP inhibited thrombin-induced ICAM-1 cell surface expression (Fig. 2A, c and e). In control experiments, forskolin or DBcAMP alone showed no effects on ICAM-1 cell surface expression (Fig. 2A, a and d). Because ICAM-1 expression induces PMN adhesion to endothelial cells (28, 29), we determined whether prevention of ICAM-1 cell surface expression interfered with endothelial adhesivity. We found that an increase in intracellular cAMP following pretreatment of cells with forskolin prevented the thrombin-induced endothelial adhesivity toward naive PMN (Fig. 2B).

Fig. 2.

A. Elevated cAMP concentration inhibits thrombin-induced ICAM-1 expression on endothelial cell surface. Confluent HUVEC monolayers were pretreated with FSK (20 μM, Aa and Ac) or DBcAMP (0.5 mM, Ad and Ae) for 30 min before being stimulated with thrombin (2.5 U/ml) for 12–15 h. ICAM-1 expression was quantitated by flow cytometry using MAb against ICAM-1 (BIRR0001) or MAb against IgG as described in materials and methods. Results are representative of 2 separate experiments. B: elevated cAMP concentration inhibits thrombin-induced endothelial adhesivity toward polymorphonuclear leukocytes (PMN). Confluent HUVEC monolayers were pretreated with FSK (20 μM) before being challenged with thrombin (5 U/ml). Expression of endothelial adhesiveness was determined by PMN adhesion assays as described in materials and methods. Data are means ± SE (n = 3 for each condition).

cAMP inhibits thrombin-induced NF-κB activity without affecting its DNA binding function.

Because NF-κB activation is essential for the thrombin-induced ICAM-1 gene transcription (28), we addressed the possibility that cAMP exerts its effect on ICAM-1 expression by inhibiting the NF-κB activity. HUVEC were cotransfected with pNF-κB-LUC containing five copies of consensus NF-κB sequence linked to a minimal adenovirus E1B promoter-luciferase reporter gene. As shown in Fig. 3, thrombin-induced NF-κB activity was markedly reduced in cells pretreated with forskolin.

Fig. 3.

Elevated cAMP concentration inhibits thrombin-induced NF-κB activity. HUVEC were transfected with NF-κB-LUC construct using the DEAE-dextran method as described (22). Cells were pretreated with FSK (20 μM) before stimulation for 8 h with thrombin (5 U/ml). Cell extracts were prepared and assayed for firefly and Renilla luciferase activities using Promega Biotech Dual Luciferase Assay System. The data are expressed as firefly/Renilla luciferase (LUC) activity. Data are means ± SE (n = 3 for each condition).

We also determined the effect of cAMP on thrombin-induced IκBα degradation, a requirement for NF-κB activation (6, 41). Western blot analysis showed that thrombin exposure of endothelial cells resulted in IκBα degradation (Fig. 4A). In contrast to forskolin's effect on NF-κB activity (Fig. 3), forskolin failed to prevent the thrombin-induced IκBα degradation (Fig. 4A). Because IκBα degradation results in nuclear translocation and DNA binding of NF-κB, we also evaluated the effects of cAMP on thrombin-induced nuclear uptake and DNA binding function of NF-κB. Pretreatment of cells with forskolin failed to prevent the nuclear translocation of RelA/p65 (Fig. 4B). EMSA showed that forskolin failed to inhibit the DNA binding activity of NF-κB (Fig. 4C), consistent with its lack of effect on nuclear uptake of RelA/p65 (Fig. 4B).

Fig. 4.

Elevated cAMP concentration fails to prevent thrombin-induced IκBα degradation, nuclear translocation, and NF-κB DNA binding activity. Confluent HUVEC monolayers were pretreated for 30 min with FSK (20 μM) before being challenged with thrombin (2.5 U/ml) for the indicated time periods. Cytoplasmic (A) and nuclear (B and C) extracts were prepared and assayed for IκBα degradation (A) and NF-κB nuclear translocation by Western blot analysis (B) and NF-κB DNA binding activity (C) by EMSA as described in materials and methods. Results are representative of 2–3 separate experiments.

cAMP inhibits thrombin-induced phosphorylation of RelA/p65.

Because activation of RelA/p65 requires phosphorylation of Ser536 in the transactivation domain (20, 44), we explored the possibility that cAMP inhibits thrombin-induced transcriptional activity of RelA/p65 by preventing its phosphorylation. We determined the effect of cAMP on the phosphorylation status of RelA/p65 using an antibody that detects RelA/p65 phosphorylated at Ser536. Coimmunoprecipitation studies demonstrated that thrombin challenge resulted in increased phosphorylation of RelA/p65 at Ser536, and this response was inhibited in cells treated with forskolin before the thrombin challenge (Fig. 5A).

Fig. 5.

A: elevated cAMP concentration and p38 MAPK inhibition prevent thrombin-induced Ser536 phosphorylation of RelA/p65. Confluent HUVEC monolayers were pretreated with FSK (20 μM) or SB-203580 (10 μM) for 30 min before being challenged with thrombin (5 U/ml) for 1 h. Cell lysates were immunoprecipitated (IP) with an antibody against p65, immunoblotted with an antibody against the phosphorylated (Ser536) form of RelA/p65. The blots were subsequently stripped and reprobed with an antibody against RelA/p65. Results are representative of 3 separate experiments. Elevated cAMP concentration inhibits thrombin-induced p38 MAPK activation. B: confluent HUVEC monolayers were pretreated with FSK (20 μM) for 30 min before being challenged with thrombin (5 U/ml) for 5 min. Total cell lysates (10 μg/lane) were separated by SDS-PAGE and immunoblotted with an antibody against the phosphorylated (Thr180/Tyr182) form of p38 MAPK. The blots were subsequently stripped and reprobed with an antibody against p38 MAPK. C: confluent HUVEC monolayers were pretreated with DBcAMP (0.5 mM) for 30 min before being challenged with thrombin (5 U/ml) for 5 min. Cell lysates were immunoprecipitated with an antibody against p38 MAPK, and in vitro kinase assays were carried out on immunoprecipitates using activating transcription factor (ATF)-2 as an exogenous substrate. Proteins were analyzed by SDS-PAGE and transferred to nitrocellulose membrane, and the phosphorylated form of ATF-2 was detected by autoradiography. The blots were subsequently stripped and reprobed with an antibody against p38 MAPK. Results are representative of 3 separate experiments.

cAMP inhibits thrombin-induced activation of p38 MAPK.

We have previously shown that p38 MAPK is a critical signaling intermediate required for thrombin-induced ICAM-1 transcription in endothelial cells (29). Our results showed that pretreatment of cells with SB-203580, a p38 MAPK inhibitor, prevented the thrombin-induced phosphorylation of Ser536 of RelA/p65 (Fig. 5A). These findings led us to investigate whether cAMP functions by interfering with the activation of p38 MAPK in response to thrombin challenge and thereby inhibits phosphorylation of Ser536 of RelA/p65 and its activation. Thus we determined the effects of cAMP on thrombin-induced p38 MAPK phosphorylation and found that thrombin activated p38 MAPK (Thr180/Tyr182) phosphorylation in a manner dependent on cAMP (Fig. 5B). Preincubation of cells with DBcAMP also prevented the thrombin-induced phosphorylation of p38 MAPK (data not shown). We next determined whether the effect of cAMP on p38 MAPK phosphorylation could be ascribed to its inhibition of p38 MAPK activation. In an in vitro kinase assay using ATF-2 as a substrate, we observed that the p38 MAPK immunoprecipitated from thrombin-challenged cells resulted in increased phosphorylation of ATF-2 (Fig. 5C), indicative of activation of p38 MAPK by thrombin. Pretreatment of cells with DBcAMP prevented the thrombin-induced p38 MAPK activation (Fig. 5C). Together, these data indicate that cAMP inhibits thrombin response by interfering with the activation of p38 MAPK.

DISCUSSION

The present study demonstrates that p38 MAPK is a critical target mediating the cAMP-dependent inhibition of thrombin-induced NF-κB activation and ICAM-1 expression in endothelial cells. We showed that inhibition of p38 MAPK by cAMP prevented the thrombin-activated phosphorylation of RelA/p65 and thus interfered with transcriptional competency of this DNA-bound NF-κB subunit.

The inhibitory effect of cAMP was only evident when thrombin was used as the agonist to elicit ICAM-1 expression. cAMP had no significant effect on the TNF-α-induced ICAM-1 expression. This latter finding agrees with studies showing that increased intracellular levels of cAMP failed to inhibit ICAM-1 expression in response to TNF-α challenge of endothelial cells (24, 26). However, we showed clearly that increased cAMP levels following incubation of cells with forskolin significantly reduced the thrombin-induced NF-κB-dependent reporter gene activity, indicating that cAMP prevented the activation of NF-κB and thus ICAM-1 expression. The reasons for the differential effects of cAMP in thrombin- vs. TNF-α-induced ICAM-1 expression are not clear but raise important questions of whether raising the cAMP concentration would be protective in inflammatory diseases associated with the generation of cytokines such as TNF-α. Whereas NF-κB was essential and sufficient for thrombin-induced ICAM-1 transcription (28), the TNF-α-induced ICAM-1 expression in endothelial cells required the cooperation of multiple transcription factors including NF-κB and CCAAT/enhancer binding protein (C/EBP) (19). A likely explanation for the failure of cAMP to prevent the TNF-α-induced ICAM-1 expression may be that increased intracellular cAMP concentration can independently result in activation of C/EBP (5, 13). Further support for this notion comes from studies (24, 26) in which elevated cAMP levels in endothelial cells inhibited the TNF-α-induced expression of genes encoding E-selectin and VCAM-1, which are both NF-κB dependent but do not require the cooperation of C/EBP (8), as is the case with ICAM-1. Studies have shown that transactivation of genes by NF-κB requires DNA binding secondary to phosphorylation and degradation of IκBα and translocation of RelA/p65 to the nucleus (3, 16, 41). We determined whether the increase in cAMP level could regulate NF-κB activation by influencing these events. Preincubation of cells with forskolin failed to prevent the NF-κB DNA binding activity induced by thrombin. Also, increased cAMP concentration did not inhibit the thrombin-induced IκBα degradation and translocation of RelA/p65 to the nucleus. Thus cAMP appears to exert its inhibitory effect on ICAM-1 expression without interfering with activation of NF-κB in the cytosol and its translocation to the nucleus.

Another important mechanism regulating NF-κB activity involves the phosphorylation of the DNA-bound RelA/p65 NF-κB subunit (2, 23). Studies showed that phosphorylation of RelA/p65 at Ser536 increases the transcriptional competency of NF-κB bound to the promoter (20, 44). We recently showed that inhibition of p38 MAPK by pharmacological and genetic approaches significantly reduced the thrombin-induced ICAM-1 expression secondary to inhibition of transactivation activity of the nuclear NF-κB but did not interfere with its cytosolic activation (29). Thus we addressed the possibility that elevation in cAMP levels could inhibit thrombin-induced transcriptional activity of NF-κB by preventing the phosphorylation of RelA/p65 at Ser536. Furthermore, we determined whether this effect of cAMP could be ascribed to inhibition of p38 MAPK activation. We observed, consistent with our hypothesis, that increase in intracellular cAMP concentration (elicited by pretreating cells with forskolin) prevented Ser536 phosphorylation of RelA/p65 in response to thrombin. We also found, interestingly, that cAMP prevented the activation of p38 MAPK and that this was in fact responsible for inhibiting the Ser536 phosphorylation of RelA/p65 induced by thrombin. In contrast, elevated cAMP levels failed to prevent the TNF-α-induced RelA/p65 phosphorylation in endothelial cells (24). Recently, Duran et al. (12) have reported that the TNF-α-induced RelA/p65 phosphorylation is mediated by PKC-ζ, suggesting that other kinases are also capable of transactivating the DNA-bound RelA/p65. Since PKC-ζ is required for TNF-α-induced NF-κB activation in endothelial cells (2, 27), it is possible that cAMP has no effect on PKC-ζ activation, which may thus explain why cAMP had no effect on TNF-α-induced RelA/p65 phosphorylation. Thus our findings demonstrate that the mechanism of action of cAMP in preventing thrombin-induced NF-κB activation involves inhibition of p38 MAPK activation, which in turn prevents phosphorylation of RelA/p65 bound to the ICAM-1 promoter.

In addition to the effect of p38 MAPK in inducing phosphorylation of RelA/p65A as discussed above, another possible mechanism by which p38 MAPK can mediate NF-κB activation is through phosphorylation of TATA binding protein (TBP), a subunit of transcription factor IID. Phosphorylation of TBP by p38 MAPK is necessary for TBP binding to the TATA box (7). Inhibition of phosphorylation of TBP was shown to reduce its binding to the TATA box and interaction with the NF-κB p65 subunit (7). Thus it is possible that inhibition of p38 MAPK by cAMP can abrogate ICAM-1 expression by preventing the phosphorylation of TBP. This idea has support from the findings of Delgado and Ganea (10) showing that vasoactive intestinal peptide and pituitary adenylate cyclase-activating polypeptide prevented the lipopolysaccharide-induced NF-κB activation through the cAMP-dependent inhibition of p38 MAPK phosphorylation of TBP.

Yet another explanation for the inhibitory effect of cAMP on NF-κB activation is the finding of Parry and Mackman (25) showing that cAMP promotes the phosphorylation of the CREB, and subsequent binding to the CREB-binding protein (CBP), which can, in turn, block NF-κB activation. CBP is a coactivator known to associate with RelA/p65 (15, 25). The presence of CBP in the nucleus may lead to a decrease in RelA-CBP complex formation and impaired NF-κB activation (25). Because the phosphorylation status of RelA/p65 is critical for its association with CBP (12, 46), it may be that cAMP-dependent inhibition of p38 MAPK activation and of phosphorylation of RelA/p65 can impair transcriptional competency of the bound NF-κB.

An increase in cellular cAMP levels induced by pharmacological agents pentoxifylline, rolipram, and amrinone suppresses expression of proinflammatory genes and is beneficial in inflammatory conditions such as autoimmune encephalomyelitis and acute cardiac allograft rejection (17, 18, 37). In this study, we have shown that elevation of endothelial cell cAMP levels inhibits NF-κB activation induced by thrombin, and this has important functional consequences in preventing ICAM-1 expression and endothelial adhesivity toward PMN. We have also shown that cAMP acts by targeting p38 MAPK activation and preventing the phosphorylation of the RelA/p65 subunit of NF-κB bound to the ICAM-1 promoter. Thus the present results provide a basis for the mechanism of anti-inflammatory action of cAMP, which may have implications in novel therapeutics targeting p38 MAPK.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants HL-67424, HL-46350, and HL-64573.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract