Transcription factor nuclear factor kappaB regulates the inducible expression of the human B1 receptor gene in inflammation.

Expression of the bradykinin B1 receptor gene is up-regulated in vascular smooth muscle cells (VSMCs) in response to a variety of inflammatory stimuli. We isolated the 5'-flanking region of the human bradykinin B1 receptor gene and examined its promoter activity by transient transfection analysis. This region (-2582 to +34) showed promoter activity inducible by lipopolysaccharide (LPS), tumor necrosis factor alpha (TNF-alpha), and interleukin-1beta (IL-1beta) in VSMCs. Further deletion analysis revealed that constructs containing 111 base pairs of 5'-flanking sequence were sufficient for transcriptional induction. Mutagenesis of a nuclear factor kappaB (NF-kappaB)-like site at -64 to -55 abolished most of the LPS, TNF-alpha, and IL-1beta inducibility, whereas a mutation of a cyclic AMP response element at -50 to -43 markedly reduced the basal promoter activity, and a mutation of the activator protein 1 (AP-1) site at -78 to -72 had minimal effects. Nuclear extracts from LPS, TNF-alpha, and IL-1beta-treated VSMCs, IL-1beta-treated human hepatoma HepG2, and human lung fibroblast IMR-90 cells showed strong inducible binding activity to the NF-kappaB-like site by gel shift assays. These results demonstrated that NF-kappaB-like nuclear factor was involved in the inducible expression of the human bradykinin B1 receptor gene during inflammatory processes.

Kinins are biologically active peptides that are formed locally after tissue damage and inflammatory stimuli from precursor kininogens by limited proteolysis (1). Kinins exert a broad spectrum of physiological and pathological effects, including smooth muscle contraction, vasodilation, increased vascular permeability, and pain induction, by binding to their cell surface receptors (2). Two subtypes of kinin receptors, B 1 and B 2 , have been characterized on a pharmacological basis (3). Whereas the B 2 receptor is responsive to the intact kinins, bradykinin (BK) 1 and Lys-BK (kallidin), the B 1 receptor has a higher affinity for the carboxypeptidase metabolites of kinins, des-Arg 9 -BK and des-Arg 10 -kallidin (4). Expression cloning of B 1 and B 2 receptor cDNAs confirms the existence of these two subtypes and reveals that they both belong to the G proteincoupled superfamily of receptors with seven transmembrane domains (5)(6)(7)(8)(9)(10)(11). Activation of the kinin receptors leads to generation of inositol phosphate and transient increases in intracellular Ca 2ϩ levels via coupling to phospholipase C, involving, at least, the G i , G q/11 , and G 13 proteins (12)(13)(14)(15).
The bradykinin B 2 receptor is constitutively expressed in a variety of tissues and cultured cell lines and mediates most of the in vivo effects usually assigned to kinins (16). The bradykinin B 1 receptor, however, is not present to any significant extent in normal tissues. It is expressed in a limited number of cultured cell types, such as embryonic lung fibroblasts, vascular smooth muscle cells, and endothelial cells (17)(18)(19). B 1 receptor-mediated responses are up-regulated in a time and protein synthesis-dependent process (4). Recent studies in animal models of hyperalgesia and in B 1 receptor knockout mice with induced peritoneal inflammation suggest that the B 1 receptor may play an important role in the pathogenesis of chronic inflammatory diseases (20 -22).
The genes encoding B 1 receptors have been cloned from the rat 2 and human (23,24). The human B 1 receptor gene was located close to the B 2 receptor gene within the same chromosomal region at 14q32 (23). The B 1 receptor is induced following tissue injury, following prolonged in vitro incubation of tissues, or upon exposure in vivo or in vitro to proinflammatory mediators, such as lipopolysaccharide (LPS) (4). Whereas cytokines, including IL-1␤, IL-2, and IL-8 but not TNF-␣ or IL-6, induced the B 1 receptor, continuous incubation with the glucocorticoid dexamethasone or the protein synthesis inhibitor cycloheximide suppressed the induction in vitro (25,26). To understand the molecular mechanism underlying the regulated expression of the B 1 receptor gene, we isolated and characterized the promoter of the human B 1 receptor gene. In this study, we present evidence that nuclear factor B (NF-B) is involved in the dynamic regulation of human B 1 receptor gene expression.

EXPERIMENTAL PROCEDURES
Cell Culture-Rat vascular smooth muscle cells (VSMCs) were explanted from a media layer of the thoracic aorta of male Sprague-Dawley rats (200 -250 g) according to the method of Biro et al. (27) and maintained in Dulbecco's modified Eagle's medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (HyClone, Logan, UT), 100 units/ml penicillin, 100 g/ml streptomycin sulfate, and 25 mM Hepes, pH 7.4. These cells exhibit a "hill and valley" growth pattern and are characterized by positive immunostaining with monoclonal antibodies against smooth muscle ␣-actin (28). Experiments using VSMCs were performed between passages 3 and 10. Human hepatoma HepG 2 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and antibiotics. Human diploid lung fibroblasts, IMR-90 cells obtained from ATCC (Rockville, MD) were cultured as * This work was supported by National Institutes of Health Grants HL 29397 and HL 52196. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The  1 Receptor Gene-A human lambda genomic library (Stratagene, La Jolla, CA) was screened with an oligonucleotide probe, 5Ј-AGA AAA CTC CTC CAA AAG CAG CTC TCA-3Ј, specific for exon I of the human bradykinin B 1 receptor gene (GenBank TM accession no. U30271) as described previously (11,23). Positive clones were picked for further characterization. Lambda phage DNA was digested and subjected to Southern blotting to identify fragments containing exon I and possibly the 5Ј-flanking sequence of the human B 1 receptor gene, which were in turn subcloned into pBluescript KS II vector (Stratagene) and sequenced on both strands using a dsDNA cycle sequencing system (Life Technologies, Inc.).
Construction of Reporter Plasmids-A 2.6-kb fragment spanning nucleotides Ϫ2582 to ϩ34 relative to the transcription start site (29) was amplified by polymerase chain reaction (PCR) and cloned upstream of a firefly luciferase gene in the promoterless plasmid pGL2-Basic (Promega, Madison, WI). The resulting construct was named p-2582Luc. The same fragment as above was also inserted into pGL2-Basic in the reverse orientation, giving rise to p2582Luc. A series of deletion constructs containing truncated 5Ј-flanking fragments of the human B 1 receptor gene was made by PCR. Constructs containing these 5Ј-progressively removed fragments were named p-1473Luc, p-825Luc, p-589Luc, p-251Luc, p-111Luc, and p-43Luc. Mutant constructs were as follows (lowercase indicates mutation): p-2582mALuc, with the AP-1 site (Ϫ78 to Ϫ72), mutated from AGACTCA to AGAgctc; p-2582mNLuc, with 2 bp in the NF-B-like site (Ϫ64 to Ϫ55), mutated from GGCAATCCCC to ctCAATCCCC; and p-2582mCLuc, with the cAMP response element (CRE) (Ϫ50 to Ϫ43), mutated from TGACATCA to TGctAgCA. Mutants were created using a PCR site-directed mutagenesis using p-2582Luc as the template (30). A double mutant construct, p-2582mCNLuc, with both the CRE and NF-B-like sites mutated as above, was prepared with the same method using p-2582mNLuc as the PCR template. Another construct, p1900Luc, with a 1.9-kb PCR-generated fragment upstream of the translation start codon ATG as described previously (23), was also produced. All constructs were sequenced to confirm the product fidelity.
Transfection and Luciferase Assays-Twenty-four hours before transfection, BAECs, HepG 2 cells, or VSMCs were seeded at a concentration that would achieve 60 -70% confluence in 12-well tissue culture plates. Liposome-mediated cotransfection using a mixture of 400 l of Opti-MEM, 4.8 l of Lipofectin (for BAECs) or LipofectAMINE (for HepG 2 and VSMCs), 0.6 g of the reporter construct, and 0.2 g of pCMV-␤gal (Clontech Laboratories, Inc.) (as a control for measuring transfection efficiency) per well was carried out as recommended by the manufacturer (Life Technologies, Inc.). The transfection mixture was replaced with normal growth medium 6 h later. Forty-eight hours after transfection, the cells were harvested. The ␤-galactosidase activity in cell extracts was determined by an enzymatic assay with chlorophenol red-␤-D-galactopyranoside as a substrate, and the luciferase activity was measured using a liquid scintillation counter (Packard Instrument Co.) and the Luciferase Assay system (Promega) according to the manufacturers' instructions. Luciferase activity in each sample was normalized to that of ␤-galactosidase.
For the induction assay, transfection was scaled up using 25-cm 2 culture flasks with 4.0 g of the indicated reporter construct, 1.3 g of pCMV-␤gal control, and 21 l of LipofectAMINE. Six hours later, the transfection mixture was replaced with 5 ml of normal growth medium. After overnight growth, cells were trypsinized and reseeded equally into six wells in a 12-well plate. Fifty-six hours after transfection, the cells were treated by a change of serum-free medium with or without mediators as indicated in the figure legends for the times indicated before harvesting the cells for luciferase and ␤-galactosidase assays.
Reverse Transcription-PCR Southern Analysis-Confluent cells were preincubated with cycloheximide (CHX) and dexamethasone (DEX) for 1 h, and then stimulated for the specified time by IL-1␤, TNF-␣, or LPS and harvested. Total RNA was extracted with Trizol reagent (Life Technologies, Inc.) or the method described by Gough (31) and then treated with RNase-free DNase I for 1 h at 37°C. Reverse transcription was performed with 2 g of total RNA and 100 pmol of random hexamers in a total volume of 20 l to produce first-strand cDNA. The quality of RNA samples was evaluated by reverse transcription-PCR using ␤-actin-specific primers (sense primer, 5Ј-GAA CCC TAA GGC CAA CCG TG-3Ј; antisense primer, 5Ј-TGG CAT AGA GGT CTT TAC GG-3Ј). The levels of BK B 1 receptor cDNAs were evaluated by PCR/ Southern blot semiquantitative method. PCR experiments were performed with 1 l of the first-strand cDNA in a 50-l reaction mixture. Rat B 1 receptor cDNA was amplified with the sense primer (5Ј-AAG ACA GCA GTC ACC ATC-3Ј) that bound in the first exon and the antisense primer (5Ј-GAC AAA CAC CAG ATC GGA-3Ј) that bound in the second exon (GenBank TM accession no. U66107). 2 Human B 1 receptor cDNA was amplified with the sense primer (5Ј-AGA AAA CTC CTC CAA AAG-3Ј) that bound in the first exon and the antisense primer (5Ј-GTA GAT TTC TGC CAC GT-3Ј) that bound in the third exon (7). Amplification protocol was as follows: denaturation at 94°C for 60 s, annealing at 55°C for 50 s, and extension at 72°C for 45 s. All PCRs were linear up to 25-30 cycles. PCR products were resolved on 0.8% agarose gels and transferred to Hybond nylon filters (Amersham Corp.). Hybridization conditions were the same as described (23). Oligonucleotides 5Ј-AAG ACT GGG ACC TGC TGT AT-3Ј, 5Ј-TAC GGC CTG TGA CAA TG-3Ј, and 5Ј-CGC ACG ATT TCC CTC TCA GC-3Ј were endlabeled with [␥-32 P]ATP and T4 polynucleotide kinase (Life Technologies, Inc.) and used as nested specific probes for rat and human B 1 receptor genes and ␤-actin, respectively.
Nuclear Extract Preparation-Confluent cells were washed once with serum-free medium and then treated with IL-1␤ (2 ng/ml), TNF-␣ (10 ng/ml), or LPS (10 g/ml) or left untreated as controls. Nuclear proteins were then extracted by a modified method of Dignam et al. (32). Briefly, after being washed with ice-cold phosphate-buffered saline (PBS), cells were scraped into 1.2 ml of cold PBS, pelleted, and resuspended in 500 l of Buffer A (10 mM Hepes-KOH, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM DTT, 0.2 mM phenylmethylsulfonyl fluoride) by tapping the tube. The cells were allowed to swell on ice for 10 min and vortexed vigorously for 10 s after adding 25 l of 10% Nonidet P-40 (Sigma). The homogenate was centrifuged for 10 s, and the nuclear pellet was resuspended in 100 l of cold Buffer C (20 mM Hepes-KOH, pH 7.9, 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM DTT, 0.2 mM phenylmethylsulfonyl fluoride) and incubated for 20 min on ice with shaking. The nuclear extract was centrifuged at 14,000 ϫ g for 5 min at 4°C, and the supernatant was aliquoted and stored at Ϫ80°C. Protein concentrations were determined by the Bradford method (Bio-Rad) (33).
Electrophoretic Mobility Shift Assay-A double-stranded oligonucleotide containing a wild type NF-B-like site (Ϫ73 to Ϫ50, 5Ј-CAC TTT TGC GGC AAT CCC CAC AAT-3Ј), was labeled with [␥-32 P]ATP as a probe. Ten g of nuclear extract was incubated for 20 min at room temperature with 20,000 cpm of the 32 P-labeled oligonucleotide probe in a total volume of 25 l containing 12 mM Hepes, pH 7.9, 60 mM KCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM DTT, 2 g poly(dI-dC), and 15% glycerol. For competition assays, a 100-fold molar excess of unlabeled specific or mutant competitors was added to the binding reactions and incubated for 10 min at room temperature before the probe was added. Protein-DNA complexes were resolved on a 4% native polyacryamide gel in Tris/glycine/EDTA buffer at pH 8.5. The sequence of mutant oligonucleotide competitor had two nucleotide variations (variations shown in lowercase) in the NF-B-like site: 5Ј-CAC TTT TGC ctC AAT CCC CAC AAT-3Ј. The sequence of the competitor containing the NF-B motif from the promoter of IRF-1 transcription factor was 5Ј-TTC GGG CCG GGG AAT CCC GCT AAG-3Ј (34).

Induction of Bradykinin B 1 Receptor Gene Expression-It
has been shown that stimulation of human lung fibroblast IMR-90 cells with IL-1␤ increased both the levels of the B 1 receptor mRNA and the number of B 1 receptors (7,17,29). We examined whether expression of the B 1 receptor gene was induced in rat VSMCs and human hepatoma HepG 2 cells following IL-1␤ stimulation. Because preliminary studies revealed that serum in the medium increased the B 1 receptor message, we washed cells once with serum-free Dulbecco's modified Eagle's medium before incubation with mediators in serum-free medium to eliminate the stimulatory effects of serum. The B 1 receptor mRNA level was low in unstimulated VSMCs, but it was markedly increased by treatment with IL-1␤ for 3 h (Fig. 1). In contrast, the same dose of IL-1␤ failed to induce B 1 receptor mRNA expression in HepG 2 cells after prolonged incubation (data not shown). We next examined the levels of the B 1 receptor mRNA in VSMCs incubated with other stimuli and/or CHX and DEX. Proinflammatory agents LPS and TNF-␣ were also able to up-regulate the B 1 receptor mRNA. The protein synthesis inhibitor CHX augmented the effects of IL-1␤ and LPS, whereas CHX alone induced an increase in B 1 receptor message. The synthetic glucocorticoid DEX reduced B 1 receptor mRNA levels in unstimulated cells and suppressed up-regulation by LPS. In contrast to the B 1 receptor, the level of ␤-actin transcript remained the same irrespective of the presence of various mediators (Fig. 1).
Potential Regulatory Elements in the 5Ј-Flanking Region of the Human B 1 Receptor Gene-We previously cloned a BamHI fragment of the human B 1 receptor gene containing the sole coding exon (23). The fragment of about 1.9 kb, upstream of the translation start site ATG in the BamHI fragment, was found to have a significant basal promoter activity in HepG 2 cells (23). In a preliminary study, however, we found that the fragment was not able to confer IL-1␤ or LPS inducibility on the reporter gene in VSMCs and HepG 2 cells. As demonstrated above, B 1 receptor gene expression is up-regulated in response to stimuli by LPS and IL-1␤. Thus, the 1.9-kb fragment could not represent the promoter for the B 1 receptor gene, at least in VSMCs under inflammatory stimuli. Later, we and other investigators found that the human B 1 receptor mRNA in IMR-90 cells includes two noncoding exons, indicating that the 1.9-kb fragment is not the 5Ј-flanking region of the gene (24). In this study, we isolated the 5Ј-flanking region of the B 1 receptor gene for elucidating the molecular mechanism of up-regulation of B 1 receptor expression. Sequence determination on both strands was extended 1500 bp compared with the longest reported sequence (24). In addition to the consensus TATA box, several other potential transcription factor binding sites, including Sp1, AP-1, NF-B, and CRE, were found in the 5Јflanking region by computer search (Fig. 2).
Inducible Promoter Activity of the 5Ј-Flanking Region of the Human B 1 Receptor Gene-To assess whether the 5Ј-flanking region of the B 1 receptor gene has inducible promoter activity, transient transfection was performed in VSMCs and HepG 2 cells. In VSMCs, LPS (10 g/ml), IL-1␤ (2 ng/ml), and TNF-␣ (10 ng/ml) increased 3-4-fold the luciferase activity of p-2582Luc, which contained the longest 5Ј-flanking region (Fig.  3, A and B). p1900Luc, with a basal promoter activity of about FIG. 1. Reverse transcription-PCR Southern blot analysis of rat bradykinin B 1 receptor mRNA. VSMCs were washed once with serum-free medium and then incubated for 3 h with serum-free medium (control), 10 g/ml LPS, 2 ng/ml IL-1␤, 10 ng/ml TNF-␣, 10 g/ml CHX, 10 g/ml CHX plus 10 g/ml LPS, 10 g/ml CHX plus 2 ng/ml IL-1␤, 1 M DEX, 1 M DEX plus 10 g/ml LPS, or 10% fetal bovine serum (FBS). CHX and DEX were applied to the cells 1 h before LPS and IL-1␤ were applied. Total RNA was prepared, 2 g of total RNA was used in the reverse transcription reaction, and PCR was conducted for 30 cycles (B 1 receptor) or 25 cycles (␤-actin), followed by Southern blot analysis as described under "Experimental Procedures." 1 ⁄7 of that of p-2582Luc, showed no inducibility of the luciferase activity in the presence of LPS, IL-1␤, or TNF-␣. The luciferase activity of p-2582Luc in HepG 2 cells was about 1 ⁄5 that of VSMCs (with respect to promoterless pGL2-Basic) and was also inducible by IL-1␤ with about 1.8-fold induction (Fig. 3C).
On the contrary, p1900Luc had basal promoter activity more than 8-fold higher than p-2582Luc in HepG 2 cells, although it was not responsive to IL-1␤. The plasmid p2582Luc, containing the 5Ј-flanking sequence inserted in the reverse orientation, showed minimal luciferase activity and was not responsive to FIG. 3. Inducible promoter activity of the 5-flanking region of the human bradykinin B 1 receptor gene. A, a series of deletion constructs, containing nucleotides Ϫ2582 to Ϫ43 bp of the 5Ј-flanking sequence of the human bradykinin B 1 receptor gene and extending to ϩ34 relative to the transcription start site, and a 1.9-kb fragment upstream of the translation codon ATG in exon III were generated and used in this study. B, VSMCs were transiently transfected with p-2582Luc or p1900Luc, divided into groups, and then treated with control medium (4 h), 10 g/ml LPS (4 h), 2 ng/ml IL-1␤ (3 h), or 10 ng/ml TNF-␣ (3 h). C, HepG 2 cells were transfected as in B, divided into groups, and then treated for 3 h with control medium or 2 ng/ml IL-1␤. The cells were harvested for both luciferase and ␤-galactosidase assays. Transfections were normalized to the ␤-galactosidase control activity, and the promoter activity is expressed as luciferase activity relative to that of the p-2582Luc control in the same cells. The data are presented as means Ϯ S.E. from three independent experiments, each performed in duplicate. IL-1␤, TNF-␣, and/or LPS in VSMCs or HepG 2 cells (data not shown).
To determine whether the induction of promoter activity is sensitive to antioxidants and steroidal and nonsteroidal antiinflammatory drugs, we incubated VSMCs transfected by p-2582Luc with pyrrolidine dithiocarbamate (PDTC), DEX, or sodium salicylate before stimulation with LPS. As shown in Fig. 4, DEX inhibited the basal activity of p-2582Luc and prevented LPS induction, whereas PDTC or sodium salicylate had no apparent effects on the basal activity but suppressed the LPS inducibility.
Cellular Specificity of the Human B 1 Receptor Gene Promoter-A series of deletion constructs (Fig. 3A), including p1900Luc, were transfected into BAECs, HepG 2 cells, and VSMCs to investigate the basal cellular specificity of the B 1 receptor gene promoter. Due to variations in transfection efficiency, it is difficult to make an accurate comparison of the basal activities of the same construct in different cells. We therefore compared the basal activities of the series of deletion constructs within each cell line. In VSMCs and BAECs, which are known to express B 1 receptors (19), the series of deletion constructs exhibited a similar profile of basal activities (Fig. 5): the 5Ј-flanking region had higher basal promoter activity than the 1.9-kb fragment, and the p-111Luc showed a level comparable (about 80%) to that observed with the p-2582Luc, implying that the upstream sequence of 2400 bp in p-2582Luc has little contribution to its basal activity. However, in HepG 2 cells, the deletion constructs in the 5Ј-flanking region had a profile of relative basal activity different from those in VSMCs and BAECs (Fig. 5). Constructs containing truncated 5Ј-flanking sequences had significantly lower basal promoter activities than that of p1900Luc, and the p-111Luc showed a higher basal level than p-2582Luc, indicating negative regulatory elements residing between Ϫ2582 and Ϫ111.
Identification of LPS-Responsive Elements-To delineate LPS-responsive elements within the 5Ј-flanking region of the human B 1 receptor gene, transient transfection of VSMCs was performed with the series of deletion constructs in the presence or absence of LPS. Fig. 6 shows that the induction of promoter activity upon stimulation with LPS was evident in all constructs except for p-47Luc. The level of induction by LPS in p-111Luc was similar to all other constructs containing additional upstream sequences up to 2582 bp. The promoter activity of p-47Luc was about 4.5% that of p-2582Luc and showed no induction with LPS treatment. The data do not indicate that upstream elements are LPS-responsive but rather that an LPS-responsive region located at Ϫ111 to Ϫ47 in the 5Ј-flanking portion mediates the LPS induction of the promoter activity. Fig. 2 shows the presence of a CRE site, an NF-B-like site and an AP-1 site in the LPS-responsive region. In addition, CHX superinduced B 1 receptor mRNA, whereas PDTC, an NF-B activation inhibitor, reduced the LPS inducibility, suggesting the involvement of NF-B in the induction of B 1 receptor gene expression. To test this hypothesis, we performed electro- FIG. 4. Effects of dexamethasone, pyrrolidine dithiocarbamate, and sodium salicylate on the promoter activity of p-2582Luc in VSMCs. VSMCs were transiently transfected with p-2582Luc and divided into groups. The cells were then treated for 4 h with control medium, 1 M DEX, 20 mM sodium salicylate, 240 M PDTC, 10 g/ml LPS, or a combination of two of the above as indicated. DEX, PDTC, and sodium salicylate were applied to the cells 1 h before LPS was applied. The cells were harvested for luciferase assay. The promoter activity is expressed as luciferase activity relative to that of the control. The data are presented as means Ϯ S.E. from two independent experiments, each performed in duplicate.

NF-B-like Site Mediating the Induction by LPS and IL-1␤-
FIG. 5. Cellular specificity of the human bradykinin B 1 receptor gene promoter. Deletion constructs were cotransfected into BAECs, rat VSMCs, and human hepatoma HepG 2 cells as described under "Experimental Procedures." The cells were harvested for both luciferase and ␤-galactosidase assays. Transfections were normalized to the ␤-galactosidase control activity, and the promoter activity is expressed as luciferase activity relative to that of p-2582Luc in the VSMCs. The data are presented as means Ϯ S.E. from three independent experiments, each performed in duplicate. phoretic mobility shift assays. A probe spanning the promoter from Ϫ73 to Ϫ50 containing the wild type NF-B-like site was used. A faint binding activity was detectable in unstimulated VSMCs, and it was increased upon stimulation by LPS, IL-1␤, and TNF-␣ (Fig. 7A). The binding complexes were entirely competed by cold probes and by an oligonucleotide containing a canonical NF-B site, but not by the mutant oligonucleotide (Fig. 7B) or oligonucleotides containing the C/EBP, OCT-1, NF/CTF, or AP-1 binding site (Stratagene) (data not shown), indicating the specificity of the NF-B-like binding. With nuclear extracts from IMR-90 and HepG 2 cells, the same specific binding complexes induced by IL-1␤ were also observed, indicating that the NF-B-like site is also functional in IMR-90 and HepG 2 cells (Fig. 7A). To further determine the functional importance of the NF-B-like site, we mutagenized this site in construct p-2582mNLuc. The sequence was changed from 5Ј-GGC AAT CCC C-3Ј to 5Ј-CTC AAT CCC C-3Ј, identical to the mutant oligonucleotide competitor used in the gel shift assay. When p-2582mNLuc was transfected into VSMCs, mutation of the NF-B-like site almost abolished the induction of promoter activity by LPS, IL-1␤, and TNF-␣ (Fig. 8). A similar result was obtained with HepG 2 cells: the observed IL-1␤ inducibility of p-2582Luc was completely abrogated by the mutation (data not shown), indicating that the NF-B-like site appears to be necessary for the induction by IL-1␤. To support the key role of the NF-B-like site in the induction of B 1 receptor gene expression, we also mutagenized the CRE and AP-1 sites. The p-2582m ALuc exhibited almost the same promoter activity as the wild type p-2582Luc. On the other hand, the CRE mutant p-2582mCLuc was still responsive, but with a diminished induction and at a lower basal level. As expected, the double mutant p-2582mCNLuc was not responsive to LPS and IL-1␤ and demonstrated a much lower basal activity (Fig. 8). DISCUSSION Kinin receptors were classified into two subtypes, B 1 and B 2 (3). Whereas the B 2 receptor is constitutively expressed, the B 1 receptor is highly induced following tissue injury or inflammatory stimuli (4). However, the molecular mechanism of this induction is not understood. In the present study, we characterized the promoter of the human bradykinin B 1 receptor gene. We first showed that expression of the bradykinin B 1 receptor gene is inducible in primary cultured rat vascular smooth muscle cells but not in human hepatoma HepG 2 cells. We then demonstrated that the 5Ј-flanking region of the human B 1 receptor gene, extending from Ϫ2582 to ϩ34, conferred  Fig. 3A. The cells were then treated for 3 h with control medium or 5 g/ml of LPS, and the luciferase activity was measured. Relative luciferase activity was determined as in Fig.  5. The data are presented as means Ϯ S.E. from two independent experiments, each performed in duplicate. full responsiveness to LPS and IL-1␤ by transient transfection of the fragment-driven luciferase reporter gene. Functional analysis showed that a 111-bp sequence from the transcription initiation site, which contained the NF-B-like binding site, was sufficient for the induction of transcription in VSMCs in response to LPS, IL-1␤, and TNF-␣. Exposure to LPS, IL-1␤, and TNF-␣ increased the formation of NF-B-like complexes with this element, as demonstrated by gel shift assays. In contrast, mutation of this NF-B-like site abolished most of the LPS, IL-1␤, and TNF-␣ inducibility of the B 1 receptor promoter construct in the same cells. Taken together, these data for the first time suggest that the NF-B-like nuclear factor is involved in the bradykinin B 1 receptor gene induction process.
NF-B is a pleiotropic transcription factor involved in the regulation of many genes implicated in the immune response and inflammatory processes (35). In resting cells, NF-B is sequestered in the cytosol by association with inhibitory proteins of the I-B family. Stimulation by agents such as TNF-␣, LPS, and IL-1␤ initiates a phosphorylation-dependent proteolytic degradation of I-B, allowing active NF-B to translocate into the nucleus and inducing transcription by binding to defined promoter elements (35). The B 1 receptor was suggested to play an important role in the actions of kinins in chronic inflammation (4). The involvement of NF-B-like nuclear factor in the transcriptional induction of the B 1 receptor gene is consistent with the role for NF-B in the activation of genes mediating inflammation.
As noted, NF-B is critical for the inducible expression of genes involved in inflammation, and we therefore tested the effects of pyrrolidine dithiocarbamate and sodium salicylate, inhibitors of NF-B activation (34,36). We did not observe apparent inhibitory effects of both agents on the basal transcriptional activity of the human B 1 receptor gene promoter, but we did observe some inhibitory effects on induction by LPS. This lack of effect might be attributed to an NF-B-like activity constitutively expressed or induced by serum in VSMCs (37,38), because PDTC and sodium salicylate seemed to block activation of NF-B only before stimulation, and they did not alter the NF-B binding activity (34,36).
Glucocorticoids have been used as anti-inflammatory drugs (39). It was reported that the synthetic glucocorticoid dexamethasone inhibited the stimulant effect of LPS on the responsiveness to des-Arg 9 -BK in isolated rabbit aortic strips (25,26). We examined the effect of dexamethasone on the induction of the B 1 receptor in VSMCs. Dexamethasone decreased the expression of the B 1 receptor mRNA under basal conditions, which is in agreement with the observation that continuous exposure to dexamethasone inhibited the development of the contractile response to B 1 agonists in rabbit aortic strips and human umbilical vein rings (25,40) and suppressed the LPS inducibility (see Fig. 1). Our transient transfection experiments demonstrated that dexamethasone down-regulated the basal transcriptional activity of the human B 1 receptor gene promoter and inhibited the induction by LPS in VSMCs. Thus, dexamethasone inhibited, at least in part, B 1 receptor expression at the transcriptional level.
Both in vitro and in vivo studies have suggested that cytokines play an important role in the B 1 receptor induction process (4,25,26). Our study has shown that the protein synthesis inhibitor cycloheximide enhanced B 1 induction by LPS and IL-1␤, indicating that the induction of the B 1 receptor by LPS and IL-1␤ does not require new protein synthesis and that IL-1 is not essential for the LPS induction, although VSMCs produce IL-1 in response to endotoxin (41). The enhanced effect by cycloheximide is generally attributed to superinduced NF-B activity due to blocking of the synthesis of the inhibitory protein I-B (42). In addition, our results showing that recombinant human TNF-␣ up-regulates B 1 receptor gene expression and increases NF-B-like binding activity in VSMCs are in contradiction with the finding that TNF-␣ had no effect on the FIG. 8. Site-directed mutagenesis of the AP-1, CRE, and NF-B-like sites in the human bradykinin B 1 receptor gene promoter. The AP-1 site, the CRE, or the NF-B-like site was mutated as indicated in the context of deletion constructs. Transfection into VSMCs proceeded as described under "Experimental Procedures." The cells were then treated with control medium, 10 g/ml LPS for 4 h, or 2 ng/ml IL-1␤ for 3 h, and the luciferase activity was measured. Relative luciferase activity was determined as in Fig. 6. spontaneous development of the response to kinins in vitro (26). In addition to potential species and experimental system differences, one hypothesis that might explain the discrepancy is that TNF-␣ may have inhibitory effects in the posttranscription process.
Early studies showed that the 1.9-kb fragment upstream of the translation start codon ATG in exon III (which covers intron II, exon II, and part of intron I) had a high basal promoter activity in HepG 2 cells based on CAT assays (23). On the basis of this finding, we inferred that this fragment might be the promoter for the human B 1 receptor gene. Based on their finding in transformed IMR-90 cells, Yang and Polgar (29) thought that intron II of the human B 1 receptor gene might function as an alternative promoter. We included this 1.9-kb fragment in the current study, inserted upstream of the luciferase reporter instead of the CAT reporter that was used previously (23). This fragment exhibited a strong basal promoter activity in HepG 2 cells, which was 6 -8-fold higher than that of the 2582-bp 5Ј-flanking promoter. It is of interest to note that the basal promoter activity of this fragment in VSMCs is 6-fold lower than that of the 2582-bp 5Ј-flanking promoter. Because the 5Ј-flanking promoter was able to confer IL-1␤ inducibility in both VSMCs and HepG 2 cells and the 1.9-kb fragment was not, it is not likely that the 1.9-kb fragment serves as the promoter in VSMCs and HepG 2 cells under inflammatory conditions. Whether and under what conditions this 1.9-kb fragment or part of it may function as an alternative promoter still needs further investigation.
The NF-B-like site is functional in VSMCs as well as in IMR-90 and HepG 2 cells. IMR-90 cells are known to express endogenous B 1 receptors under basal conditions (17). Because of the low transfection efficiency, we did not use IMR-90 cells for the transient transfection study. Whether the identified NF-B-like site mediates IL-1␤ inducibility of B 1 receptor expression in IMR-90 cells is not clear. Because we could not detect B 1 receptor expression in HepG 2 cells even when NF-B-like nuclear factor was activated, activation of NF-B-like activity seemed to be insufficient for B 1 receptor expression, at least in HepG 2 cells. Some other (tissue-specific) factors must be involved, especially under basal conditions. The participation of serum in the medium in the regulation of B 1 receptor expression makes the search more complicated. We identified an AP-1 site and a CRE in close proximity to the NF-B-like site (Fig. 2). The AP-1 site and the CRE seem to play a minor role in the inducibility of the B 1 receptor gene. Although the AP-1 site is less functional, at least in VSMCs, the CRE seems to have an important role in the basal transcriptional activity of the B 1 receptor gene promoter according to site-directed mutagenesis. However, we did not observe any stimulatory effects of the cAMP analog, dibutyryl-cAMP, on B 1 receptor mRNA induction and on human B 1 receptor gene promoter activity (data not shown). We also failed to observe any complex formation using the CRE site as a probe. Even so, we can not exclude the possibility that NF-B may interact with the AP-1 or CRE binding proteins as reported by other investigators (43,44).
In conclusion, we have described the role of NF-B-like nuclear factor in the regulation of the inducible expression of the B 1 receptor gene. Understanding the regulatory mechanisms of the B 1 receptor gene by LPS and IL-1␤ should provide impor-tant insight for developing therapies to antagonize B 1 receptormediated detrimental effects in inflammatory diseases.