HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bode, J. G. Articles by Schaper, F. Search for Related Content PubMed PubMed Citation Articles by Bode, J. G. Articles by Schaper, F.

The Inhibitory Effect of IL-1 on IL-6-Induced ₂-Macroglobulin Expression Is Due to Activation of NF-B¹

Johannes G. Bode^2,*, Richard Fischer, Dieter Häussinger, Lutz Graeve^*, Peter C. Heinrich^3,* and Fred Schaper^*

^* Institut für Biochemie, Universitätsklinikum der Rheinisch-Westfälischen Technischen Hochschule Aachen, Aachen, Germany; and Klinik für Gastroenterologie, Hepatologie und Infektiologie, Medizinische Klinik der Heinrich-Heine Universität, Düsseldorf, Germany

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The cross-talk between the signal transduction of simultaneous acting cytokines largely determines the final impact of cytokines on their target genes. Both NF-B and STAT3 are transcription factors well known to be activated by many stimuli and to mediate transcriptional activation by binding to specific enhancer sequences. In this study, it is analyzed how IL-1 inhibits IL-6-induced transcriptional activation of the ₂-macroglobulin promoter. It is shown that IL-1 prevents STAT3 binding to the two STAT3-responsive sites within the ₂-macroglobulin promoter by association of IL-1-activated NF-B to this region. The observation that inhibition of IL-6-induced transcriptional activation of this promoter by IL-1 is reversed by cotransfection with I-B provides evidence that NF-B activation by IL-1 is responsible for inhibition of IL-6-mediated trans activation of the ₂-macroglobulin gene. Accordingly, cotransfection of the NF-B subunits p50 or p65 themselves inhibited activation of the ₂-macroglobulin promoter by IL-6. Introduction of point mutations in each of the two NF-B sites overlapping the two STAT3 binding sites within the ₂-macroglobulin promoter provides evidence that each of these two sites counteracts transcriptional activation via STAT3. Most interestingly, at least one functional NF-B consensus site is essential for the IL-6-induced transcriptional activation of the ₂-macroglobulin promoter. Additional data are provided indicating that the activation of NF-B by IL-1 is also responsible for the inhibition of other IL-6-inducible genes, such as the ₁-antichymotrypsin gene as well as the suppressor of cytokine signaling 3 gene, suggesting a more general relevance of this mechanism for transcriptional regulation.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The immediate set of inflammatory reactions of the organism to maintain homeostasis by counteracting impending challenges such as trauma, tissue injury, and infection characterizes the so-called acute-phase response. Its function is the isolation and neutralization of pathogens as well as the prevention of further pathogen entry.

The inflammatory cascade underlying the acute-phase response is largely controlled by the action of different mediators released under inflammatory conditions. Upon activation, blood monocytes and tissue macrophages release a set of primary inflammatory mediators such as IL-1 and TNF-, thereby inducing the synthesis and secretion of several secondary cytokines and chemokines such as IL-6 and IL-8 from macrophages, monocytes, and local stromal cells. Recruitment of other immune effector cells by chemotaxis then rapidly augments the local inflammatory response to counteract the inflammatory stimulus and to remove the cellular debris generated by any associated tissue damage (1).

One of the most extensively studied responses to an acute inflammatory stimulus is the change in the hepatic synthetic profile of acute-phase proteins (APP).⁴ IL-6 has been recognized to be the major mediator involved in the regulation and maintenance of the synthesis of most of these acute-phase proteins in the liver (2, 3).

The action of IL-6 on hepatocytes is mainly mediated via the Janus kinase (Jak)/STAT signal transduction cascade (4, 5), a key signaling system, involved in the signal transduction of numerous ILs, the IFNs, as well as a number of growth and differentiation factors (6, 7). Binding of these ligands to their appropriate receptors activates tyrosine kinases of the Jak family, followed by tyrosine phosphorylation, dimerization, and nuclear translocation of so-called STATs. In the nucleus, activated STAT dimers bind to specific enhancer sequences and modulate transcription of target genes. Many APP genes such as those coding for C-reactive protein (human), ₁-antichymotrypsin (₁ACT) (human), ₂-macroglobulin (₂M) (rat), and LPS-binding protein (human and rat) have been identified to be induced through STAT factors (8, 9, 10). The activation of Jak1 and STAT3 has been shown to be crucial for the transcriptional activation of these genes by IL-6 (4, 5, 11).

Furthermore, IL-6-stimulated gene induction can be modulated via other signal transduction pathways. Thus, STAT binding sites are often in close proximity to binding sites for other transcription factors such as NF-IL-6 (12), NF-B (13), AP-1 (14, 15), and GR (16), making a cooperative action of these factors with STATs in gene regulation most likely. Moreover, in the promoters of the rat ₂M gene and the human ₁ACT gene, STAT3 binding sites are arranged as a tandem (17, 18), suggesting that formation of multimers on clustered binding sites also represents a regulatory step in STAT-dependent gene activation. However, the exact function of these tandem motifs has yet not been uncovered.

On the other hand, recent results from several laboratories strongly indicate that IL-6-induced signaling and transcriptional activation are often modulated at the level of signal transduction upstream from the respective transcription factors. For example, LPS and the proinflammatory cytokine TNF- have been shown to inhibit IL-6-mediated STAT3 activation. This inhibition is most likely due to the induction of the de novo synthesis of the Jak inhibitor suppressor of cytokine signaling 3 (SOCS3) both by LPS and TNF- (19). A similar mechanism has been discovered for IFN- signaling, in which LPS inhibits IFN--dependent STAT1 activation also via the induction of SOCS3 (20).

IL-1 dose dependently inhibits the IL-6-induced synthesis and secretion of APP such as ₂M and fibrinogen in hepatocytes in primary culture (21). However, the underlying mechanism remained unclear.

Analyzing the mechanism of IL-1-mediated attenuation of IL-6 signal transduction, we show that NF-B activated by IL-1 counteracts IL-6-induced transcriptional activation of the ₂M promoter most likely by counteracting STAT3 DNA binding to its respective binding region within the ₂M promoter. Interestingly, although NF-B itself clearly inhibits STAT3 action, our data give further evidence that at least one intact NF-B binding site seems to be essential for IL-6-induced STAT3-dependent transcriptional activation of the ₂M promoter. In this context, activation of NF-B is assumed to be of more general relevance for the inhibitory effects of IL-1 on IL-6-induced gene induction, since data are provided, suggesting that the activation of NF-B by IL-1 also inhibits the induction of an IL-6-inducible ₁ACT reporter construct and the IL-6-mediated increase of SOCS3.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Materials

Restriction enzymes were purchased from Roche Molecular Biochemicals (Mannheim, Germany); Taq polymerase was from Hybaid (Heidelberg, Germany); oligonucleotides were obtained from MWG-Biotech (Ebersberg, Germany); SN50 and SN50 M (22) were from Calbiochem (Bad Soden, Germany); DMEM and DMEM/nut.mix F12 were from Life Technologies (Eggstein, Germany); FCS was from Seromed (Berlin, Germany); human rIL-6 and soluble IL-6R gp80 were prepared as described (23); the internal control plasmid DNA pCH110 was from Amersham Pharmacia Biotech (Uppsala, Sweden). Purified p50 was from Promega (Madison, WI); the specific Ab against p65 was from Upstate Biotechnology (Lake Placid, NY); and the Ab against STAT3 was kindly provided by W. Müller-Ester (Mainz, Germany).

Preparation, cultivation, and stimulation of cells

Isolated parenchymal cells were prepared from livers of 5- to 8-wk-old male Wistar rats by a collagenase perfusion technique. Cells were plated on collagen-coated culture dishes and maintained in Krebs-Henseleit medium supplemented with 6 mmol/L glucose in a humidified atmosphere of 5% CO₂ and 95% air at 37°C. After 2 h, medium was removed and the culture was continued for 24 h in DMEM containing 5% FCS, 0.1 mg/ml penicillin/streptomycin, 100 nmol/L insulin, 100 nmol/L dexamethasone, 30 nmol/L Na-selenite, and 1 µg/ml aprotinin.

The human hepatoma cells HepG2 were grown in DMEM/nut.mix F12 supplemented with 10% FCS, streptomycin (100 mg/L), and penicillin (60 mg/L). Medium was changed and adjusted to 5 ml 24 h before experiments were conducted.

Cells grown in a 100-mm dish were stimulated with IL-1 or IL-6 at the concentrations indicated. SN50 and SN50 M were dissolved in sterile water. Cells were preincubated with SN50 and SN50 M for 15 min at the concentrations indicated in the figure legends. Nuclear extracts were prepared as described by Andrews and Faller (24). Protein concentration was determined with a Bio-Rad (Munich, Germany) protein assay.

Electrophoretic mobility shift assay

EMSAs were performed as described previously (18). The protein/DNA complexes were separated on a 4.5% polyacrylamide gel containing 7.5% glycerol in 0.25-fold TBE (20 mM Tris base, 20 mM boric acid, 0.5 mM EDTA, pH 8) at 20 V/cm for 4 h. Gels were fixed in 10% methanol, 10% acetic acid, and 80% water for 1 h, dried, and autoradiographed. The double-stranded ³²P-labeled oligonucleotides used for EMSA are listed in Fig. 1. In addition to those depicted in Fig. 1, the following oligonucleotides were used: mutated m67SIE oligonucleotide from the c-fos promoter (m67SIE: 5'-GATCCGGGAGGGATTTACGGGAAATGCTG-3') (25) and an oligonucleotide comprising the proximal (p) STAT3 binding site from the ₂M promoter optimized for the binding of NF-B (pNF-B: 5'-GATCCTTCTGGGAATTCCTA-3') (26). For competition assays, unlabeled probes were used at 5, 10, 20, and 50 molar excess to the radioactive labeled probes. For supershift analyses, the nuclear extracts were preincubated with the respective Ab at 4°C for 20 min before EMSA procedures. For in vitro binding assay using purified NF-B p50, p50 protein was added in increasing amounts from 7, 12.5, 25, 50, to 100 ng to the nuclear extracts of IL-6-stimulated cells.

View larger version (71K): [in this window] [in a new window]   FIGURE 1. IL-1 inhibits STAT3 DNA binding to the 2M promoter fragment, but not its activation in general. HepG2 cells were preincubated with IL-1 for 10 min, as indicated, and subsequently stimulated with IL-6 for the time periods indicated. For the determination of STAT3 activation, nuclear extracts were prepared from these cells and equal amounts protein analyzed in EMSA with the STAT3-specific SIE probe (upper panel), as described in Materials and Methods. Protein DNA binding to the wild-type 2M promoter fragment (-193 to -147) is shown in the lower panel. STAT3/DNA complexes are indicated. II marks an additional DNA/protein complex appearing after stimulation with IL-1.

Standard cloning procedures were performed as outlined by Sambrook et al. (27). pGL32 M-215Luc contains the promoter region -209 to +8 of the ₂M gene fused to the luciferase-encoding sequence and was described previously (28). pGL3-hACT-359Luc contains the promoter region -379 to +25 of the ₁ACT gene fused to the luciferase-encoding sequence and was kindly provided by F. Horn (Leipzig, Germany) and described previously (29). Mutations in the ₂M promoter reporter construct were generated by PCR technique using appropriate oligonucleotides. The sequences of all constructs were controlled by sequencing using an ABI Prism automated sequencer (PerkinElmer, Norwalk, CT). The different point mutations introduced into the ₂M gene promoter fragment are summarized in Fig. 1.

Transfection procedure and reporter gene assay

For transfection of HepG2 cells, cells were grown on 60-mm dishes to 30% confluency and transfected in DMEM supplemented with 10% FCS. Calcium phosphate precipitation was performed with 3 µg reporter construct, 2 µg -galactosidase expression vector (pCR3lacZ; Amersham Pharmacia Biotech), and 4.5 µg I-B-encoding expression, as indicated in the figure legends. Transfections were adjusted with control vectors to equal amounts of DNA. Cells were incubated with the precipitate for 16 h, washed twice with PBS, and let for additional 10–24 h in fresh medium. For reporter gene assays, cells were stimulated for 16 h. Cell lysis and luciferase assays were conducted using the luciferase kit (Promega), as described by the manufacturer’s instructions. All expression experiments were done at least in triplicate. Luciferase activity values were normalized to transfection efficiency monitored by the cotransfected -galactosidase expression vector. Error bars are SD.

Chromatin immunoprecipitation

Chromatin immunoprecipitation was performed using the Chromatin Immunoprecipitation Assay kit from Upstate Biotechnology. A total of 8 x 10⁶ rat primary hepatocytes was seeded on 10-cm dishes and analyzed for NF-B DNA binding, according to the manufacturers’ instructions. NF-B/DNA complexes were precipitated with an Ab specific for the p65 subunit of NF-B (Santa Cruz Biotechnology, Santa Cruz, CA). Using PCR reaction, the purified chromatin precipitates were analyzed for the existence of the following rat ₂M promoter fragments: -12 to -419 and -103 to -283.

Total RNA isolation and Northern blot analysis

Total RNA was isolated using RNeasy mini kit (Qiagen, Hilden, Germany), as described by the manufacturer. A total of 10 µg total RNA was separated on 1% denaturing agarose gels and transferred to a NitroPlus transfer membrane (Micron Separations, Westboro, MA). The membranes were prehybridized for 2 h at 68°C in 10% dextran sulfate, 1 M NaCl, and 1% SDS, and hybridized overnight in the same solution with cDNA fragments labeled with Random Primed DNA Labeling Kit (Roche Molecular Biochemicals). Blots were exposed to Kodak X-OMAT AR-5 film at -70°C with intensifying screens.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
IL-1 inhibits IL-6-induced DNA binding of activated STAT3 by competition with NF-B at overlapping NF-B/STAT3 binding sites of the ₂M promoter

Whereas IL-1 is regarded as a proinflammatory cytokine, IL-6 exhibits many antiinflammatory activities. We have described previously that IL-1 counteracts the IL-6-induced synthesis of APP, such as ₂M and fibrinogen (21). However, the underlying molecular mechanism for this inhibitory effect of IL-1 is still unclear. Since STAT3 has been shown to be the major transcription factor involved in IL-6-induced APP synthesis (30), we analyzed whether the IL-6-induced STAT3 activation in human hepatoma (HepG2) cells is affected by a preincubation of these cells with IL-1 (Fig. 1). As shown in the upper panel of Fig. 1, preincubation with IL-1 for 10 min did hardly counteract STAT3 DNA binding to a STAT3-specific DNA probe deduced from the c-fos promoter (25). Thus, the IL-6-mediated activation of STATs is essentially unaffected by IL-1. However, IL-6-induced STAT3 DNA binding to a DNA fragment that corresponds to the sequence -193 to -147 of the ₂M promoter, containing a tandem STAT3-binding motif (Fig. 2A), was almost completely abolished by preincubation with IL-1 (Fig. 1, lower panel). Moreover, in parallel to the IL-1-dependent disappearance of the STAT3/DNA complex, a faster migrating complex appeared (indicated as (II) in Fig. 1, lower panel). These data strongly suggest that the inhibitory effect of IL-1 on IL-6-induced APP synthesis is rather due to a disturbed DNA binding of activated STAT3 to specific APP promoter elements than due to an inhibition of STAT3 activation. The observation that the disappearance of the STAT3/DNA complex is accompanied by the appearance of a faster migrating complex suggests that inhibition of IL-6-induced APP synthesis by IL-1 might be due to another protein competing STAT3 DNA binding. Presently, we have no explanation for the nature of the gel-shift band with intermediate mobility (open arrowhead). It is important to note that authentic oligonucleotides (Fig. 2) and not probes designed for optimal transcription factor binding were used in the present studies; these native oligonucleotides form protein/DNA complexes of lower affinity reflected in the much lower intensities in the EMSA.

View larger version (39K): [in this window] [in a new window]   FIGURE 2. Overlapping STAT3 and NF-B binding sites in the promoters of the 2M, the 1ACT, and the SOCS3 genes. A, The sequences summarized represent the location of mutations introduced into the STAT3 and NF-B binding sites of the 2M promoter. The effect of these mutations on promoter activation and STAT3 or NF-B DNA binding was analyzed in reporter gene assays as well as in EMSAs. Sequences representing STAT3 binding sites are framed; NF-B binding sites are represented as hatched bars. Nucleotides exchanged to affect the STAT3 or the NF-B DNA binding are underlined. In the graphic arts, the two STAT3 binding sites (distal and proximal STAT-responsive elements) are framed and depicted as boxes in light gray; the two putative overlapping NF-B binding sites are represented as hatched bars and depicted as black circles. The 2M promoter fragment -1 to -209 cloned 5' to the luciferase gene (black bar) of pBL3Luc is depicted as a dark gray bar. B, Representation of promoter fragments of the 1ACT and SOCS3 genes containing putative STAT3 and NF-B binding sites.

View larger version (30K): [in this window] [in a new window]   FIGURE 3. Binding of STAT3 and NF-B to the 2M promoter. A, Nuclear extracts prepared from HepG2 cells stimulated with IL-6 for 30 min (lanes 2–7) and prestimulated with IL-1 for 10 min (lanes 5–7) as indicated were analyzed for DNA-binding activities of STAT3 and NF-B in an EMSA with the wild-type 2M promoter fragment, as described in Fig. 1. To identify the proteins within the protein/DNA complexes, nuclear extracts were incubated with polyclonal Abs either specific for STAT3 (lanes 3 and 6) or for the p65 subunit of NF-B (lanes 4 and 7). B, Increasing amounts (7, 12.5, 25, 50, 100 ng) of purified NF-B p50 protein were added to nuclear extracts derived from HepG2 cells stimulated with IL-6 for 30 min. Protein samples were analyzed for STAT3 and NF-B DNA binding to the 2M promoter fragment in EMSA, as described in Fig. 1.

I-B impairs the inhibitory effect of IL-1 on IL-6-induced promoter activation, whereas overexpression of the NF-B subunits p65 and p50 inhibits IL-6-induced promoter activation

To test whether NF-B is responsible for the inhibitory effect of IL-1 on IL-6-induced expression of ₂M, we analyzed the effect of the NF-B inhibitor I-B on IL-6-induced ₂M promoter activation. We cotransfected HepG2 cells with I-B and an ₂M promoter-reporter construct and measured the IL-6-dependent induction of the reporter gene in the presence or absence of IL-1. As shown in Fig. 4A, cotransfection of I-B blocked the inhibitory effect of IL-1 on IL-6-induced activation of the ₂M promoter. Since overexpression of I-B is known to inhibit NF-B activation, these data indicate that activation of NF-B plays an important role for the inhibitory effect of IL-1 on the IL-6-mediated activation of the ₂M promoter.

View larger version (18K): [in this window] [in a new window]   FIGURE 4. Overexpression of I-B blocks the inhibitory effect of IL-1 on IL-6-induced activation of the 2M promoter, whereas the p65 and the p50 subunit of NF-B mediates an inhibitory effect on IL-6-induced activation of the 2M promoter. HepG2 cells were cotransfected with a reporter gene construct containing the 2M promoter (-209 to +8) fused to the firefly luciferase gene and a control plasmid or expression vector for A, I-B, and B, the p65 or the p50 subunit of NF-B, as indicated. An expression vector for -galactosidase was cotransfected for monitoring transfection efficiency. Two days after transfection, cells were preincubated with 100 U/ml IL-1, where indicated, and stimulated with or without IL-6 (100 U/ml) for 16 h, as shown. Luciferase activity in cellular extracts of these cells was determined and normalized to -galactosidase activity, as outlined in Materials and Methods. The scheme below the bar diagram represents the reporter gene construct used as described in Fig. 2.

Both NF-B binding sites within the ₂M promoter confer responsiveness of the promoter to the inhibitory activity of IL-1, but are also essential for IL-6-mediated promoter activation

The proximal as well as the distal STAT3 binding sites within the ₂M gene promoter overlap with potential NF-B/DNA binding sites (Fig. 2A; framed vs hatched boxes). Since we found that IL-1 exerts its inhibitory activity on the IL-6-induced ₂M gene expression through NF-B, we analyzed whether this activity of IL-1 could be overcome by mutating the NF-B sites within the ₂M promoter fragment. Therefore, point mutations within the proximal or distal NF-B consensus sequences were introduced in the ₂M promoter of the reporter constructs. As shown in Fig. 2A, two CA substitutions at positions -155 (mpNF) and -175 (mdNF) (underlined) were generated to affect NF-B binding to the ₂M promoter fragment. Minimal point mutations were used to keep encroachment of the rest of the promoter as low as possible and to leave the STAT3 binding sites intact. To analyze the changes in DNA affinity of NF-B or STAT3 achieved by the mutations introduced into the ₂M promoter, competition assays were chosen as a very reliable approach. Nuclear extracts from IL-6-stimulated HepG2 cells were used as a source for activated STAT3. Fig. 5A shows that a promoter fragment bearing both mutated NF-B sites (mdNF mpNF) is as efficient as a nonmutated DNA fragment to compete with the labeled DNA fragment for STAT3 binding. This demonstrates that neither the CA substitution at position -155 nor the one at -175 interferes with STAT3 binding to the promoter. The effect of these mutations on the binding of NF-B to the promoter fragment was analyzed in a similar assay, but with nuclear extracts of IL-1-stimulated cells as a source for NF-B (Fig. 5B). The nonmutated DNA fragment was very efficient to compete with the labeled DNA fragment for NF-B DNA binding. In contrast, a joint mutation of both NF-B binding sites in the promoter fragment (mdNF mpNF) led to a reduced competition indicative for a significantly reduced affinity to NF-B. The affinity of NF-B to the promoter fragment was less affected by a single CA point mutation at position -155 (mpNF) or -175 (mdNF). This moderate effect of a single point mutation is probably due to the other (remaining) intact NF-B binding site. These data confirm that the point mutations introduced into the two NF-B binding sites overlapping with the STAT3 sites do not influence STAT3 binding, but interfere with NF-B binding to the ₂M promoter element.

View larger version (33K): [in this window] [in a new window]   FIGURE 5. Function of the individual NF-B sites for mediating the inhibitory effect of IL-1 on IL-6-induced activation of the 2M promoter. Binding of NF-B and STAT3 to promoter fragments containing mutated NF-B sites was analyzed in EMSA by competition experiments. A, Ten micrograms of nuclear extracts from HepG2 cells stimulated with IL-6 (100 U/ml) for 30 min. Cells were analyzed for binding of STAT3 to the labeled wild-type 2M promoter fragment (-193 to -147), as described in Fig. 1. For competition experiments, the extracts were preincubated with a 5-, 10-, 20-, or 50-fold molar excess of the unlabeled 2M promoter fragment or the (mdNF mpNF) oligonucleotide containing point mutations within both NF-B binding sites (see Fig. 2). B, Ten micrograms of nuclear extracts from IL-1 (100 U/ml; 30 min)-stimulated HepG2 cells were used to demonstrate binding of NF-B to an optimized proximal NF-B binding site of the 2M promoter (pNF-B). The effect of mutations of the NF-B binding sites in the 2M promoter fragment (-193 to -147) on binding of NF-B was analyzed in competition experiments using unlabeled 2M promoter fragments where none (wt), the proximal (mpNF), the distal (mdNF), or both (mdNF mpNF) NF-B sites are mutated. Therefore, the nuclear extracts were incubated with a 5-, 10-, 20-, or 50-fold molar excess of the indicated unlabeled oligonucleotides before the EMSA. The scheme on top of the radiograms represents the oligonucleotides used for competition. STAT3- and NF-B-DNA complexes are indicated by arrowheads. C, The mutations used for the experiments presented in A and B were also introduced into the 2M promoter-reporter construct and analyzed for promoter activation. Transfection procedures and stimulation of HepG2 cells as well as reporter gene assays were performed, as described in Fig. 4. The left part of the figure represents the promoter-reporter constructs used. D, The 2M promoter fragment containing mutations within both the proximal and the distal NF-B site was analyzed for binding new proteins due to mutations introduced. EMSA was performed with the (mdNF mpNF) oligonucleotide and for comparison with the wt promoter fragment, as described in A. E, NF-B was analyzed for binding the endogenous 2M promoter in cells that have not been stimulated with IL-1. Untreated (lanes 2 and 4) or IL-6-stimulated (lanes 3 and 5) primary rat hepatocytes were analyzed in respect to chromatin immunoprecipitation using a specific NF-B Ab (lanes 4 and 5); as control the Ab was omitted in lanes 2 and 3. The origin of the coprecipitated endogenous DNA fragment was proven by PCR with two sets of DNA primers corresponding to the rat 2M gene promoter (upper and lower panels).

Since the presence of at least one intact NF-B/DNA binding site seems to be indispensable for proper promoter function, we next analyzed whether NF-B binds to the STAT-responsive region of the endogenous gene in cells that have not been stimulated with IL-1. Therefore, we performed chromatin immunoprecipitation from primary rat hepatocytes using an NF-B-specific Ab. The origin of the coprecipitated endogenous DNA fragment was proven by PCR with two sets of DNA primers corresponding to the rat ₂M gene promoter regions from -12 to -419 (Fig. 5E, upper panel) and -103 to -283 (Fig. 5E, lower panel). As visible in lanes 4 and 5 of Fig. 5E, NF-B binds prior to and after stimulation with IL-6 to the endogenous ₂M promoter, supporting the idea that NF-B binding to the promoter is essential for proper promoter function.

The inhibitory effect of IL-1 on the promoters containing a unique mutated distal or proximal NF-B site was completely blocked by cotransfected I-B (Fig. 6), which is in line with the observation for the ₂M wild-type promoter (Fig. 4A). This further demonstrates that IL-1 is able to exert its NF-B-mediated inhibitory function on the ₂M promoter through a single NF-B binding site within the promoter.

View larger version (28K): [in this window] [in a new window]   FIGURE 6. I-B overcomes the IL-1-dependent inhibitory activities on the 2M promoter induction mediated through both NF-B sites. 2M promoter-reporter constructs containing a mutated proximal (mpNF) or distal (mdNF) NF-B binding site were analyzed in reporter gene assays in the presence or absence of coexpressed I-B, as described in Fig. 4. The schemes below the bar diagram represent the reporter constructs used.

Both NF-B binding sites in the ₂M promoter act negatively on gene induction mediated through both STAT3 binding sites

Both the proximal as well as the distal NF-B binding site confer responsiveness of the ₂M promoter to the inhibitory activity of IL-1 through NF-B (Figs. 5C and 6). We next analyzed whether the NF-B binding sites affect gene induction mediated through the individual STAT3 binding sites (Fig. 7, C and D). Therefore, promoter constructs were generated containing a single functional STAT3 binding site in combination with a single functional NF-B binding site (Fig. 2A). The mutations within the STAT3 response elements were verified not to inhibit NF-B binding (Fig. 7A), but to inhibit STAT3 DNA binding (Fig. 7B). The promoter fragment with both STAT3 binding sites mutated (mdST mpST) competed for NF-B binding with the labeled NF-B probe as efficient as the wild-type ₂M promoter fragment (wt; Fig. 7A), demonstrating that the mutation does not interfere with NF-B DNA binding. In contrast, the mdST/mpST-mutated oligonucleotide did not compete for STAT3 binding to the labeled promoter fragment (Fig. 7B; right panel), confirming that this oligonucleotide does not contain any functional STAT3 binding site. Both DNA fragments containing a single mutated STAT3- and NF-B-binding motif ((mdNF mpST) and (mdST mpNF)) competed for STAT3 DNA binding with the labeled promoter fragment, demonstrating that both fragments are still capable of binding STAT3. The (mdNF mpST) oligonucleotide was found to be less efficient for competition than the (mdST mpNF) oligonucleotide. This is in line with the described observation that STAT3 binds with higher affinity to the proximal than to the distal STAT3-responsive element of the ₂M promoter (18).

View larger version (32K): [in this window] [in a new window]   FIGURE 7. Both NF-B sites display their IL-1-induced inhibitory function on both STAT3 binding sites of the 2M promoter. Binding of NF-B and STAT3 to promoter fragments containing mutated STAT3-responsive elements was analyzed in EMSA by competition experiments. A, Similar as described in Fig. 5B. Ten micrograms of nuclear extracts prepared from IL-1 (100 U/ml; 30 min)-stimulated HepG2 cells were used to demonstrate binding of NF-B to the proximal NF-B site (pNF-B) of the 2M promoter. The effect of mutations in the STAT3/DNA binding sites in the 2M promoter on NF-B binding was analyzed in competition experiments using unlabeled 2M promoter fragments where none (wt) or both STAT3/DNA binding sites (mdST mpST) were mutated. Therefore, the nuclear extracts were incubated with a 5-, 10-, 20-, and 50-fold molar excess of the indicated unlabeled oligonucleotides before the EMSA. B, Similar as described in Fig. 5A. Ten micrograms of nuclear extracts from IL-6 (100 U/ml; 30 min)-stimulated HepG2 cells were analyzed for binding of STAT3 to the labeled 2M promoter fragment (-193 to -147). For competition, the nuclear extracts were preincubated with a 5-, 10-, 20-, or 50-fold molar excess of the indicated unlabeled oligonucleotides. The schemes on top of the radiograms represent the oligonucleotides used for competition. STAT3- and NF-B-DNA complexes are indicated by arrowheads. C, Mutants of the 2M promoter containing only the wild-type distal NF-B binding site and one of both functional STAT3-responsive elements, as depicted in the left scheme of the figure, were analyzed in reporter gene assays for their responses to IL-1, as described in Fig. 4. Therefore, in addition to the mutation within the proximal NF-B site (CA -155), an additional TC substitution was introduced in the proximal (TC -165) or the distal (TC -187) STAT3-responsive element of the 2M promoter-reporter construct. D, Mutants of the 2M promoter containing only the wild-type proximal NF-B binding site and one of both functional STAT3/DNA binding sites, as depicted in the left scheme of the figure were analyzed in reporter gene assays, as is described in Fig. 4.

A similar observation was made for the NF-B binding site overlapping with the proximal STAT3 binding site. Neither the combined mutation of the distal NF-B (CA -175) site with the distal STAT3 binding site (TC -187) nor disruption of the distal NF-B site plus the proximal STAT3 (TC -165) site led to an IL-1-insensitive promoter (Fig. 7D). The observation that the ₂M promoter is much more sensitive to a mutation in the proximal than in the distal STAT3-responsive element further demonstrates the importance of the proximal STAT binding site and confirms previous findings (18).

Down-regulation of other IL-6-inducible genes by IL-1

Tandem STAT3 binding sites are not unique for the ₂M promoter, but are also found in the regulatory elements of other IL-6-inducible genes such as the ₁ACT and the SOCS3 gene promoter. Within these promoter regions, similar to the ₂M gene promoter, putative NF-B binding sites overlap with STAT3 binding sites (Fig. 2B). Thus, it is well possible that the investigated regulatory mechanism concerning the ₂M promoter is of more general importance, i.e., the ₁ACT and the SOCS3 genes are similarly modulated as the ₂M gene. The following experiments were conducted to give support to this idea.

As shown in Fig. 8A, a reporter construct containing a promoter fragment (-379 to +25) of the ₁ACT gene could be activated by stimulation of transfected HepG2 cells with IL-6, but was also negatively affected by costimulation with IL-1. As found for the ₂M gene promoter, IL-1 is suggested to exert its inhibitory activity on the analyzed part of the ₁ACT promoter also through NF-B since expression of I-B at least partially blocked its inhibitory effect.

View larger version (29K): [in this window] [in a new window]   FIGURE 8. NF-B is involved in the inhibitory effect of IL-1 on IL-6-induced activation of the 1ACT promoter and SOCS3-mRNA. A, HepG2 cells were cotransfected with a reporter gene construct containing the 1ACT promoter (-379 to +25) fused to the firefly luciferase gene and a control plasmid or expression vector for I-B, as indicated. An expression vector for -galactosidase was cotransfected for monitoring transfection efficiency. Reporter gene activation was determined as described in Fig. 4. B, HepG2 cells were pretreated with or without IL-1 (100 U/ml) for 10 min and subsequently stimulated with IL-6 (100 U/ml) for the times indicated. C, HepG2 cells were pretreated with IL-1 (lanes 3–6) for 10 min and stimulated with IL-6 (lanes 2–5) for 1 h. To inhibit nuclear translocation of NF-B, cells were incubated 10 min before IL-1 pretreatment with SN50 (50 µg/ml, lane 4) or SN50 M (50 µg/ml, lane 5) as control. Cells were harvested, and total RNA was isolated and subjected to Northern blot analysis of SOCS3-mRNA and GAPDH-mRNAs, as described in Materials and Methods.

In conclusion, these data lead us to the speculation that the dual function of NF-B as a positive and negative regulator of gene expression, described in this study in more detail for the ₂M promoter, might also be true for other promoters such as those of the ₁ACT and the SOCS3 genes.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Activation of NF-B is one of the major events following the onset of an inflammatory response mainly initiated by proinflammatory cytokines such as IL-1 and TNF-. On the other hand, activation of STAT factors is an important part of the action of cytokines and IFNs released during the inflammatory response (1). In the past, it became increasingly evident that there exists an extensive and complex cross-talk between various signal transduction cascades initiated by simultaneously acting cytokines, largely modulating cellular responses toward a single stimulus.

For example, it is well documented that IL-1, TNF-, as well as LPS counteract IL-6 signal transduction and IL-6-mediated gene induction (19, 20, 21). Several mechanisms for the negative regulation of IL-6 signal transduction have been identified: 1) The protein tyrosine phosphatase SH2-containing protein tyrosine phosphatase 2 is activated by IL-6 and counteracts initiation of IL-6 signaling (28, 32, 33), although the direct molecular targets of SH2-containing protein tyrosine phosphatase 2 still have to be determined. 2) Protein inhibitors of activated STATs have been cloned and found to bind and inhibit tyrosine-phosphorylated STAT transcription factors (34, 35). 3) SOCS proteins have been recognized as IL-6-induced feedback inhibitors of the Jak kinases (36, 37, 38). 4) Nuclear tyrosine phosphatases have been shown to inactivate STATs (39). 5) The proteasome has been found to degrade STAT factors (40).

Very recently, data have been presented suggesting that TNF- and LPS act as inhibitors of IL-6 signaling through the induction of SOCS3 (19, 20). Analyzing the mechanism underlying the IL-1-mediated attenuation of IL-6 signal transduction, it was observed that IL-1, in contrast to TNF- and LPS (19), did not induce SOCS3 gene expression (visible in Fig. 8, B and C) and hardly affected IL-6-mediated STAT3 activation (Fig. 1, upper panel). In this study, we identified NF-B as a mediator of IL-1-dependent negative regulation of IL-6-inducible genes (Figs. 4, 5C, 6, 7C, 7D, and 8). NF-B exerts its negative regulatory function on the ₂M promoter by counteracting DNA binding of STAT3 at overlapping STAT3/NF-B binding sites (Figs. 3; 5, A, B, D, and E; and 7, A and B). We propose that this activity is responsible for the inhibitory effect of IL-1 on IL-6-induced ₂M promoter activation. Surprisingly, although NF-B acts as a competitive inhibitor for STAT3 DNA binding, at least one intact NF-B consensus site was found to be crucial for the STAT3-dependent activation of the ₂M gene promoter (Fig. 5C). Furthermore, additional data are provided, suggesting that the activation of NF-B by IL-1 also inhibits the induction of an ₁ACT reporter gene construct (Fig. 8A) as well as of the SOCS3 gene (Fig. 8, B and C). Thus, competition between NF-B and STAT3 seems to be of more general relevance for inhibition of IL-6-induced gene induction by IL-1.

Cross-talk between STAT factors and NF-B has also been described for other promoters, for example, the one of the IFN regulatory factor-1 gene. This gene promoter was shown tobe synergistically activated by TNF- and IFN- (41, 42, 43). The synergism was suggested to be due to the independent interaction of the involved transcription factors with components of the basal transcription machinery. Prolactin-activated STAT5B inhibits NF-B-dependent IFN regulatory factor-1 promoter activation by squelching limited coactivators (44). These modes of cross-talk between STAT factors and NF-B are in contrast to our observations for the ₂M promoter, in which NF-B and STAT3 compete for overlapping DNA binding sites. More similar to the conditions described for the ₂M promoter, in this work is the regulation of the E-selectin gene promoter by STAT6 and NF-B. At this promoter, IL-4-induced STAT6 binding counteracts TNF--induced expression of the E-selectin gene via competition with NF-B for DNA binding at overlapping STAT6/NF-B binding sites (45). Thus, in this case, the roles for STAT and NF-B are reversed when compared with the ₂M promoter.

As shown in this study, at least one nonmutated NF-B consensus site is crucial for promoter activation. ₂M promoter constructs with both NF-B binding sites mutated displayed almost no basal or IL-6-inducible activity (Fig. 5C). Thus, it might be that activated NF-B at low concentrations, not high enough to counteract STAT3 DNA binding, further represents an important constituent of the transcription machinery involved in STAT3-mediated transcriptional activation. In this respect, it is important to note that the NF-B subunit p65 has been shown to specifically engage CBP/p300 for maximal transcriptional stimulation of the IL-6 gene promoter by its histone acetyltransferase activity (46). Indeed, we were able to show NF-B binding to the endogenous ₂M promoter in both unstimulated and IL-6-treated primary rat hepatocytes (Fig. 5E). We did not detect direct NF-B binding to DNA in nuclear extracts from cells stimulated with IL-6 alone in EMSA (Fig. 3A). This suggests that under conditions that allow STAT3-dependent gene induction, NF-B has no high affinity to the NF-B binding site of the promoter element. However, DNA binding of NF-B became visible after elimination of STAT3 DNA binding by interfering STAT3 Abs, indicating that NF-B is also present in nuclear extracts from cells stimulated with IL-6 alone (Fig. 3A). Since removal of STAT3 obviously enables NF-B to bind the DNA in the gel retardation assay, these data emphasize the idea that the ratio of STAT3 and NF-B is critical for the affinity of these transcription factors to bind the ₂M promoter element. This idea is further supported by the observation that increased amounts of activated NF-B after IL-1 stimulation led to a loss of STAT3 DNA binding (Figs. 1 and 3). Moreover, prevention of the IL-1-induced NF-B activation blocks the inhibitory effect of IL-1 on STAT3-dependent IL-6-mediated activation of the ₂M promoter (Fig. 4A), an effect that is conferred by each of both NF-B motifs (Fig. 6). In line with these data is the observation that expression of the NF-B subunits p65 or p50 mimics IL-1-dependent signal attenuation (Fig. 4B).

Another possibility to explain the IL-1-mediated repression of the ₂M promoter would be the specific induction of promoter binding by negative regulatory p50 homodimers (47) in IL-1-stimulated cells. Such a putative IL-1-mediated switch from p50/p65 to p50/p50 promoter binding is considered to be unlikely since p65 has been demonstrated to be present within the NF-B/DNA complex induced by IL-1 (Fig. 3A) and, again, its overexpression of p65 alone mimics the inhibitory effect of IL-1 (Fig. 4B).

Further preliminary evidence for a competition of STAT3 and NF-B for binding the ₂M promoter was given by Zhang and Fuller (26). However, these authors concentrated on the isolated proximal STAT3 binding site and did not analyze a larger part of the ₂M promoter containing the complete tandem motif. Furthermore, this observation was not followed up with respect to its functional implications for inhibition of IL-6-induced gene expression by IL-1, LPS, or TNF-.

In our study, we analyzed an extended part of the ₂M promoter containing both STAT binding sites. The arrangement of tandem STAT sites has been shown in many promoters to mediate binding of tandem STAT dimers, which is strengthened by association through the N-terminal domain of the STATs (8). This appears to be important at promoters in which one site does not fit the consensus sequence. The ₂M promoter meets these conditions. As shown in Fig. 7, C and D, the proximal STAT3 site, which fits the STAT consensus better than the distal STAT site, is much more potent to mediate IL-6-dependent promoter activation than the distal site. Furthermore, both sites act somehow synergistically on the promoter. Thus, the relevance of individual binding sites should be analyzed in context of the other binding sites within the ₂M promoter.

In general, the loss of promoter activity of mutants lacking both NF-B consensus sequences (Fig. 5C) might be due to the introduction of new binding sites for unidentified inhibitory proteins within the mutated promoter element. However, we exclude this possibility since mutation of a single NF-B site did not reduce, but rather increased promoter activity (Figs. 5C, and 7, C and D). Furthermore, no additional protein/DNA complexes were observed in EMSA with the corresponding DNA element (Fig. 5D).

Finally, one could speculate that there is another, unidentified protein binding to the NF-B site that is responsible for working synergistically with STAT3 to activate transcription. The unspecific band visible in all EMSA performed with the ₂M promoter fragment could represent a candidate protein. Thus, mutation of both NF-B consensus sequences should eliminate DNA binding of this protein. This explanation seems also to be unlikely since the mutations introduced into the promoter fragment did not affect DNA binding of this protein (Fig. 5D).

In summary, activation of NF-B is shown to be a crucial negative regulatory step by which the proinflammatory cytokine IL-1 controls STAT3-dependent gene induction by IL-6. Considering the fact that IL-6 displays distinct antiinflammatory properties, it is attractive to speculate that proinflammatory mediators such as LPS, TNF-, or IL-1 down-regulate IL-6 signaling to enforce the inflammatory response. Being aware that antiinflammatory activities have been described for several APP induced by IL-6 (48, 49, 50), this hypothesis becomes even more conclusive.

	Acknowledgments

 
The I-B expression vector, p65 expression vector, and the SOCS3 cDNA were kindly provided by K. Brand (Munich, Germany), M. Nourbakhsh and H. Hauser (Braunschweig, Germany), and D. Hilton (Parkville, Australia), respectively. We gratefully acknowledge the generous supply of human rIL-1 from D. Boraschi (L’Aquila, Italia). We thank W. Frisch and M. Ruhl for technical assistance.

	Footnotes

 
¹ This work was supported by the Deutsche Forschungsgemeinschaft (Bonn), the Sonderforschungsbereich 542 "Molekulare Mechanismen Zytokin-gesteuerterEntzündungsprozesse: Signaltransduktion und Pathophysiologische Konsequenzen" and the Sonderforschungsbereich 575 "Experimentelle Hepatologie," and the Fonds der Chemischen Industrie (Frankfurt).

² Current address: Klinik für Gastroenterologie, Hepatologie und Infektiologie, Medizinische Klinik der Heinrich-Heine Universität, 40255 Düsseldorf, Germany.

³ Address correspondence and reprint requests to Dr. Peter C. Heinrich, Institut für Biochemie, Klinikum der RWTH Aachen, Pauwelsstrae 30, D-52074 Aachen, Germany. E-mail address: heinrich{at}rwth-aachen.de

⁴ Abbreviations used in this paper: APP, acute-phase protein; ₁ACT, ₁-antichymotrypsin; ₂M, ₂-macroglobulin; Jak, Janus kinase; SOCS, suppressor of cytokine signaling.