Persistent Activation of Nuclear Factor-κB by Interleukin-1β and Subsequent Inducible NO Synthase Expression Requires Extracellular Signal-Regulated Kinase
Abstract— The role of extracellular signal-regulated kinase (ERK) was studied in the signaling pathway by which interleukin-1β (IL-1β) increases the expression of inducible NO synthase (iNOS) in rat vascular smooth muscle cells. IL-1β induced a rapid and transient activation of nuclear factor-κB (NF-κB), followed by a prolonged activation of NF-κB that was required to induce iNOS expression. Either PD98059 or U0126, selective inhibitors of ERK activation, did not influence IL-1β-induced early activation but effectively reduced the prolonged activation of NF-κB and significantly reduced IL-1β induction of iNOS. Transfection with antisense, but not sense, phosphorothioate-modified oligodeoxynucleotides directed toward ERK also reduced IL-1β-induced prolonged NF-κB activation and iNOS expression. IκBβ, but not IκBα degradation, induced by IL-1β was markedly attenuated when ERK activation was inhibited and could be partially responsible for the persistent NF-κB activation. These data suggest that ERK activity is required for persistent NF-κB activation by IL-1β that is necessary for iNOS gene expression.
- extracellular signal-regulated kinase
- NO synthase
- nuclear factor-κB
- vascular smooth muscle cells
Interleukin-1β (IL-1β) is an important mediator of inflammatory responses and has been found in atherosclerotic plaques.1 IL-1β is known to be a potent inducer of NO production by increasing inducible NO synthase (iNOS) expression in vascular smooth muscle cells (VSMCs),2–4 but the signaling pathways leading to the induction of iNOS have not yet been fully characterized.
A variety of IL-1β-responsive genes, including iNOS, are regulated through the activation of nuclear factor-κB (NF-κB).3,5–7 It has been reported that interleukin-1 activates NF-κB through a signal cascade composed of interleukin-1R1, TRAF6, NF-κB-inducing kinase, IκB kinases, and IκB.8 In most unstimulated cells, NF-κB is sequestered in the cytoplasm by binding to IκB.7–10 Stimulation of cells with diverse agonists, including IL-1β, leads to phosphorylation and subsequently ubiquitination-dependent degradation of IκB proteins, accompanied by the translocation of NF-κB to the nucleus, where it binds to specific DNA sequences in the promoter region of target genes and activates transcription.
Several studies implicate the involvement of mitogen-activated protein kinases (MAPKs), including p38 MAPK and p44/42 MAPK (extracellular signal-regulated kinase [ERK]1/2), in the cytokine induction of iNOS expression.11,12 However, the mechanism by which the MAPK regulates iNOS gene expression remains unknown. We have recently reported that ERK, but not p38 MAPK, was involved in the IL-1β response leading to iNOS gene expression in cultured VSMCs.13 Inhibition of ERK activation markedly reduced IL-1β-induced iNOS expression and subsequent NO production. However, IL-1β-induced NF-κB nuclear translocation and DNA-binding activity were not altered by the inhibition of ERK activation when the nuclear extracts were prepared 1 hour after cytokine stimulation, suggesting that a unique signaling pathway leading to a relatively rapid activation of NF-κB may not be influenced by ERK. In the present study, we examined the temporal changes in NF-κB activation in response to IL-1β, and we found the requirement of ERK activity for producing a prolonged NF-κB activation that is necessary for iNOS induction.
DMEM/F-12, FCS, and Lipofectamine were purchased from Life Technologies. Recombinant human IL-1β (specific activity ≈1×108 U/mg) was from Genzyme. PD98059 and U0126 were from Calbiochem. Monoclonal antibody against iNOS was from Transduction Laboratories. Antibodies against phospho-p44/42 MAPK (Thr202/Tyr204), p42 MAPK, p44/42 MAPK, and IκBα were from New England Biolabs. Antibodies against NF-κB p65, NF-κB p50, IκBβ, OCT-1, and signal transducer(s) and activator(s) of transcription (STAT)-1 were from Santa Cruz. NF-κB consensus oligonucleotide was from Promega.
Rat VSMCs from the thoracic aorta were isolated and cultured as described previously.13 Cells between passages 4 and 8 were used for all experiments. At confluence, the cells were washed twice and then maintained in DMEM/F-12 with 0.1% FCS for 24 hours before use. Unless specified otherwise, 3 ng/mL IL-1β was added to the quiescent cells, and when PD98059 (or U0126) was used, it was added 1 hour before IL-1β and was maintained in the medium for the duration of the experiment. No difference in the cell viability in these experiments was observed, as monitored by use of trypan blue exclusion, measurement of total protein, or visual observation for changes in cell adherence.
Antisense phosphorothioate-modified oligodeoxynucleotides (ODNs, 5′-GCC GCC GCC GCC GCC AT-3′) directed against the initiation translation site of rat ERK1/214 and sense ODN (5′-ATG GCG GCG GCG GCG GC-3′) were synthesized by Gemini Biotech. VSMCs grown to ≈80% confluence in 6-well plates were washed twice and then incubated in DMEM/F-12 containing Lipofectamine (10 μg/mL) and 1 μmol/L antisense or sense ODNs for 6 hours. The cells were then washed once to remove Lipofectamine and incubated for 42 hours in DMEM/F-12 containing the same concentration of ODNs. The medium was then replaced with fresh DMEM/F-12 containing the same concentration of ODNs, and the cells were cultured with or without the addition of IL-1β for the designated times. Cell viability was maintained throughout this procedure.
Determination of Nitrite
The release of NO from cultures was assessed by determination of nitrite, a stable metabolite of NO, with Griess reagent as described previously.4
Western Blot Analysis
Whole-cell lysates or nuclear extracts were prepared, and Western blot analysis was performed as described previously.4 The images were obtained and analyzed by using the 420oe Scanning Densitometer (pdi, Inc).
Northern Blot Analysis
Total RNA was extracted, and Northern blot analysis was performed exactly as described previously.4 Randomly primed [α-32P]dCTP-labeled cDNA probes for iNOS4 and 18S ribosomal RNA (Ambion, Inc) were used in the hybridization.
Electrophoretic Mobility Shift Assay
Nuclear extracts were prepared, and DNA-binding activity was assessed by electrophoretic mobility shift assay (EMSA) with the use of NF-κB consensus oligonucleotide as described previously.4 To identify the components of the NF-κB-DNA complex, nuclear extracts were incubated for 30 minutes at room temperature with 4 μg of either NF-κB-specific or nonspecific antibodies in the binding buffer4 before a 20-minute incubation with [γ-32P]ATP-labeled NF-κB consensus oligonucleotide. For competitive binding experiment, a 50-fold molar excess of unlabeled NF-κB oligonucleotide was included in the reaction mixture.
Reverse Transcription-Polymerase Chain Reaction
The first-strand cDNA was synthesized from 1 μg total RNA by using random 9-mer primers and AMV reverse transcriptase (Takara Shuzo Co). Polymerase chain reaction (PCR) that used synthetic gene-specific primers for rat iNOS (forward 20-mer, 5′-GCT ACA CTT CCA ACG CAA CA-3′; reverse 20-mer, TGG GTG GGA GGG GTA GTG AT) amplified a 430-bp sequence between 2081 and 2510 of rat iNOS cDNA according to the following schedule: denaturation, annealing, and extension at 94°C, 60°C, and 72°C for 40 seconds, 30 seconds, and 1 minute, respectively, for 26 cycles. A parallel PCR of GAPDH was performed as a reference, with the same schedule and cycles as described above, by use of forward 20-mer, 5′-GCC ATC AAC GAC CCC TTC AT-3′, and reverse 20-mer, 5′-CGC CTG CTT CAC CAC CTT CT-3′, which amplified a 702-bp sequence between 88 and 789 of rat GAPDH cDNA. Equal volumes of iNOS and GAPDH PCR products were mixed, electrophoresed on a 2% agarose gel containing ethidium bromide, and visualized by UV-induced fluorescence.
Sustained ERK Activation Is Required for Induction of iNOS by IL-1β
IL-1β induced a biphasic activation of ERK1/2, as measured by immunodetection of phosphorylated ERK1/2 (Figure 1A). The first phase was transient, peaked at 30 minutes, and returned to the basal level at 1 hour. The second phase became apparent at 2 hours and was sustained at a high level for up to 24 hours after IL-1β addition. The transient and the sustained activation of ERK1/2 induced by IL-1β were completely inhibited by PD98059, an inhibitor of ERK kinase-1 (MEK-1), demonstrating that MEK-1 activation was required for the biphasic response (Figure 1B).
We have previously shown that inhibition of ERK phosphorylation by PD98059 reduced iNOS mRNA and protein levels in IL-1β-treated VSMCs.13 To further clarify the relationship between ERK activation and IL-1β-induced iNOS expression, PD98059 was added to the cell cultures at various time points before or after IL-1β addition. Western blot analysis (Figure 2A) showed clear induction of iNOS when cells were exposed to IL-1β for 24 hours. Inhibition of ERK activation by PD98059 reduced iNOS protein levels with similar potency when the inhibitor was added between 1 hour before and 3 hours after IL-1β addition. However, the inhibitory effect of PD98059 was progressively weaker when it was added between 6 and 9 hours after IL-1β stimulation. Northern blot analysis for iNOS (Figure 2B) showed that PD98059 reduced steady-state mRNA levels when the inhibitor was added to the cultures either 1 hour before or 3 hours after IL-1β treatment but that it was less effective when added after 6 hours, consistent with the alterations in iNOS protein levels. These results demonstrated that the early transient activation of ERK was insufficient for iNOS induction and that sustained ERK activity was required for IL-1β to activate iNOS gene transcription.
ERK Activation Is Required for Prolonged NF-κB DNA-Binding
NF-κB has been demonstrated to be essential for iNOS gene expression.3,5 In a previous study, inhibition of ERK activation by PD98059 showed little or no effect on IL-1β-induced NF-κB nuclear translocation and DNA-binding activity when the nuclear extracts were prepared 1 hour after cytokine stimulation.13 However, the results showing that PD98059 added even at 1 or 3 hours after IL-1β addition was still effective in suppressing iNOS gene expression (see Figure 2) suggested that the inhibition of ERK activation might influence the IL-1β-induced NF-κB DNA-binding in a time-dependent manner. To test this possibility, cells were exposed to IL-1β in the absence or in the presence of PD98059 for designated times before nuclear extracts were obtained. The NF-κB DNA binding (Figure 3, EMSA) increased progressively with time for up to 16 hours in the cells treated with IL-1β alone. Consistent with our previous report,13 PD98059 treatment did not significantly influence the IL-1β-induced NF-κB DNA-binding within 1 hour after IL-1β addition, but it had a marked effect on subsequent time points, with the inhibition being most pronounced at the later time points. Western blot analysis of nuclear proteins (Figure 3, bottom) showed that p65 localized in the nucleus was increased after the addition of IL-1β in a time-dependent manner during the first 3 hours, either in the absence or in the presence of PD98059. However, between 3 and 16 hours after the addition of IL-1β, the nuclear content of p65 was maintained at a high level in the absence of PD98059, but it decreased in the presence of the inhibitor. Nuclear p50 content was similarly increased during the first 3 hours in cells treated with or without PD98059. However in contrast to p65, nuclear p50 content further increased at the later time points in the cells treated with IL-1β alone, but this increase was attenuated in the presence of PD98059. The chart and the bar graphs in Figure 3 summarize the densitometric data of the time course for EMSA and the Western blot analysis and show the marked effect of PD98059 on the temporal pattern of NF-κB activation. A significant effect of PD98059 on NF-κB EMSA was seen at 3, 6, and 16 hours after IL-1β addition, and the effect on p65 and p50 translocation was also significant. Thus, on IL-1β stimulation, NF-κB translocation and DNA binding were significantly reduced at the later time points when PD98059 was present.
Identification of NF-κB Components Activated by IL-1β
Antibodies against NF-κB subunits were used in EMSA to identify the NF-κB components activated by IL-1β (data not shown). Nuclear extract from control cells showed no detectable DNA-binding activity; however, after IL-1β treatment, 2 protein-DNA complexes were observed. The upper one was abolished by the addition of either p65-specific or p50-specific antibodies, suggesting that the complex contained p65 and p50. The lower protein-DNA complex, most obvious in cells treated with IL-1β for 16 hours, may represent the DNA-binding activity of a p50 homodimer, inasmuch as the interaction was attenuated by addition of the p50-specific antibody but not the p65-specific antibody. Both bands were completely abolished by the addition of excess unlabeled NF-κB oligonucleotide. Anti-OCT-1 or anti-STAT-1 used as non-NF-κB-specific antibodies showed no effect on the protein-DNA complex formation.
Persistent Activation of NF-κB Is Required for iNOS Induction
A question raised is whether persistent activation of NF-κB is required for iNOS induction. Transient exposure of the cells to IL-1β for 1 hour, followed by a 23-hour incubation in the medium without the cytokine, did not induce iNOS expression, which was readily observed after 24 hours of continuous IL-1β treatment, as demonstrated by Western blot analysis (data not shown). To determine whether transient IL-1β treatment induced only a transient activation of NF-κB, cells were incubated with IL-1β for 1 hour and then incubated in cytokine-free medium for varying times (Figure 4A). NF-κB nuclear translocation and DNA binding that occurred after 1 hour of treatment with IL-1β disappeared rapidly once the cytokine had been removed from the medium. Removal of IL-1β from cultures after a 1-hour exposure allowed the initial activation but prevented the prolonged activation of NF-κB, suggesting a requirement for prolonged activation of NF-κB to induce iNOS gene expression. This was further demonstrated by measuring iNOS mRNA with reverse transcription (RT)-PCR (Figure 4B). iNOS mRNA was observed at 6 hours after IL-1β addition and increased with time thereafter if IL-1β was present continuously. However, no iNOS mRNA was detected during the 12-hour incubation period if IL-1β was removed from the cell cultures after a 1-hour exposure. These results further suggest that PD98059 inhibits the IL-1β induction of iNOS by a mechanism that is at least partially related to its inhibitory effect on the prolonged activation of NF-κB.
Inhibition of Prolonged NF-κB Activation and iNOS Expression by U0126 or ERK Antisense ODN
To further strengthen the evidence that ERK activation was required for IL-1β-induced prolonged NF-κB activation and subsequent iNOS expression, U0126 (a selective inhibitor of MEK-1 and MEK-2) and ERK antisense ODN (which has been shown to downregulate ERK1/2 synthesis) were used to block the ERK signaling pathway. Treatment of the cells with U0126 dramatically inhibited IL-1β-induced ERK phosphorylation, prolonged NF-κB activation, and iNOS expression (right panels in Figure 5A and 5B). As revealed by Western blot analysis, ERK antisense ODN treatment reduced ERK1 and ERK2 levels ≈50% compared with those of control cells. This reduction in total ERK levels reduced IL-1β-induced ERK phosphorylation, whereas sense ODN showed no obvious effect on either total ERK levels or IL-1β-induced ERK phosphorylation (Figure 5A, top and middle panels on left). IL-1β induction of iNOS was reduced by treatment of the cells with antisense, but not sense, ODN (Figure 5A, bottom left panel). EMSA (Figure 5B) showed that antisense ODN markedly reduced IL-1β-induced prolonged NF-κB DNA binding (at 16 hours) but had no effect on early activation (at 1 hour). Sense ODN did not affect IL-1β-induced NF-κB activation.
Inhibition of ERK Activation Attenuates IκBβ Degradation
NF-κB activation depends on the phosphorylation and degradation of IκB.7–10 Therefore, we examined the relationship between the activation of ERK and the degradation of IκB proteins after IL-1β stimulation. In addition, to determine whether the PD98059-induced decline of NF-κB DNA binding might result from changes in NF-κB availability, the content of NF-κB p65 in whole-cell lysates also was analyzed (Figure 6).
No obvious alterations were observed in cellular content of p65 after IL-1β treatment for the period examined, either in the absence or in the presence of PD98059. Cellular IκBα levels were significantly reduced at 30 minutes after exposure to IL-1β and then partially recovered thereafter during the 6-hour time period examined. PD98059 showed no obvious effect on the IL-1β-induced maximum reduction in cellular IκBα levels. IL-1β treatment resulted in a time-dependent decrease in cellular IκBβ levels with a maximum loss after 3 hours, coinciding with the increased phosphorylation of ERK in the later phase. Treatment of the cells with PD98059 partially prevented the loss of IκBβ caused by IL-1β, and the effect of PD98059 was more significant at 6 hours (P<0.01; see the bar graphs that summarize the densitometric analysis).
The present study reveals a novel mechanism by which ERK activation is required for IL-1β to induce iNOS gene expression in rat VSMCs. A biphasic activation of ERK occurred in response to IL-1β, and the second, more prolonged, phase was correlated with a concomitant prolonged activation of NF-κB. This prolonged activation of NF-κB included a sustained high level of nuclear p65/p50 heterodimer and a gradually increased nuclear p50/p50 homodimer, characterized by immunodetection procedures and EMSA. The requirement of ERK activity for IL-1β in the prolonged activation of p65/p50 and p50/p50 and of the subsequent induction of iNOS was demonstrated by inhibition of ERK phosphorylation with the use of 2 different inhibitors and by downregulation of ERK synthesis with ERK antisense ODN.
The reduction in the prolonged NF-κB activation caused by ERK inhibition was not due to insufficient availability of NF-κB p65, because there was no obvious alteration in cellular p65 content. It is known that IκBα is rapidly degraded after cytokine stimulation and rapidly resynthesized because of the presence of NF-κB-binding motif in the promoter of IκBα gene.15 This autoregulatory loop was suggested to limit NF-κB activation in a transient manner.15,16 The sustained reduction of IκBβ was suggested to contribute to the persistent NF-κB activation,16 and this possibility was reinforced in our experimental model involving iNOS gene expression. The selective effect of ERK that we observed on IκBβ, but not on IκBα, may indicate that the early activation of NF-κB is mediated through a mechanism requiring IκBα degradation that is not affected by ERK inhibition. In contrast, the more prolonged activation of NF-κB by IL-1β is initiated, at least partially, by an IκBβ degradation-mediated process that is attenuated when ERK is inhibited. These results may explain why inhibition of ERK activation did not affect IL-1β-induced early translocation and DNA-binding of NF-κB. However, either IκBα or IκBβ only specifically interacts with NF-κB containing p65 or c-Rel.16,17 Therefore, the effect of ERK on IκBβ degradation may be only a partial explanation for the activation of p65/p50 but not for the activation of p50/p50. In contrast to IκBα and IκBβ, BCL-3, which contains 7 ankyrin-like repeats with close homology to that of IκBα and IκBβ, has been reported to specifically interact with the p50 homodimer.17 Moreover, the generation of p50 may relate to a proteolysis of p105, the precursor of p50.18 Therefore, we cannot rule out the possibility that BCL-3 and p105 might also be selective targets of ERK and possibly influence NF-κB activation.
Persistent activation of NF-κB has been observed in human atherosclerotic lesions and in VSMCs after arterial injury,19–22 in which it may regulate the expression of numerous genes controlling cell differentiation, proliferation, apoptosis, and vascular remodeling. NF-κB activation is also involved in toxic shock, acute-phase responses, rheumatoid arthritis, asthma, cancer, and AIDS.22,23 Understanding the regulatory mechanisms controlling persistent NF-κB activation may be critical to the development of effective treatment for these diseases. Our findings demonstrating a role for ERK in prolonged NF-κB activation in VSMCs could provide a rationale for therapeutic intervention in vascular diseases thought to involve persistent activation of NF-κB during their pathogenesis.
This work was supported by National Institutes of Health grants HL-53471, HL-55001, and HL-55620 and by American Heart Association Research Award 0160251T.
Received July 25, 2001; revision accepted September 14, 2001.
Spink J, Cohen J, Evans TJ. The cytokine responsive vascular smooth muscle cell enhancer of inducible nitric oxide synthase: activation by nuclear factor-κB. J Biol Chem. 1995; 270: 29541–29547.
Jiang B, Haverty M, Brecher P. N-Acetyl-l-cysteine enhances interleukin-1β-induced nitric oxide synthase expression. Hypertension. 1999; 34: 574–579.
Xie QW, Kashiwabara Y, Nathan C. Role of transcription factor NF-κB/Rel in induction of nitric oxide synthase. J Biol Chem. 1994; 269: 4705–4708.
Roff M, Thompson J, Rodriguez MS, Jacque JM, Baleux F, Arenzana Seisdedos F, Hay RT. Role of IκBα ubiquitination in signal-induced activation of NFκB in vivo. J Biol Chem. 1996; 271: 7844–7850.
Silva JD, Pierrat B, Mary JL, Lesslauer W. Blockade of p38 mitogen-activated protein kinase pathway inhibits inducible nitric-oxide synthase expression in mouse astrocytes. J Biol Chem. 1997; 272: 28737–28380.
Larsen CM, Wadt KAW, Juhl LF, Andersen HU, Karlsen AE, Su MSS, Seedorf K, Shapiro L, Dinarello CA, Mandrup-Poulsen T. Interleukin-1β-induced rat pancreatic islet nitric oxide synthesis requires both the p38 and extracellular signal-regulated kinase 1/2 mitogen-activated protein kinases. J Biol Chem. 1998; 273: 15294–15300.
Jiang B, Brecher P. N-Acetyl-l-cysteine potentiates interleukin-1β induction of nitric oxide synthase: role of p44/42 mitogen-activated protein kinases. Hypertension. 2000; 35: 914–918.
Watanabe N, Iwamura T, Shinoda T, Fujita T. Regulation of NFKB1 proteins by the candidate oncoprotein BCL-3: generation of NF-κB homodimers from the cytoplasmic pool of p50-p105 and nuclear translocation. EMBO J. 1997; 16: 3609–3620.
Bourcier T, Sukhova G, Libby P. The nuclear factor κ-B signaling pathway participates in dysregulation of vascular smooth muscle cells in vitro and in human atherosclerosis. J Biol Chem. 1997; 272: 15817–15824.
Chen F, Castranova V, Shi X, Demers LM. New insights into the role of nuclear factor-κB, a ubiquitous transcription factor in the initiation of diseases. Clin Chem. 1999; 45: 7–17.