INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE KININS [peptides related to bradykinin (BK)] are known to activate two types of G protein-coupled receptors: B₁ and B₂ (30). The pharmacological profile of kinins supports their role as inflammatory mediators: local vasodilation, exudation, and pain are produced after the injection of BK into tissues. Other suspected roles of the kallikrein-kinin system are related to cardiovascular physiology, where in some respects kinin effects are antagonistic to those of the renin-angiotensin system. The natriuretic and diuretic effects of endogenous kinins are documented by several approaches, including B₂ receptor (B₂R) gene inactivation in mice: these animals are prone to salt-sensitive hypertension (1). Endogenous kinins may also limit ischemic damage in the peripheral circulation (17) and mediate flow-dependent arterial dilation in the human heart (19). Vascular endothelial cells and endothelium-derived mediators [nitric oxide (NO) and eicosanoids] are important in kinin inflammatory and circulatory effects. A widely used class of antihypertensive drugs, the ANG I-converting enzyme (ACE) inhibitors, blocks one of the major inactivation pathways for kinins (24). The contribution of kinins to the therapeutic and side effects of ACE inhibitors has been widely studied (24).

The two types of kinin receptors are defined by a series of pharmacological criteria, and the corresponding genes were isolated and characterized (30). Lys-des-Arg⁹-BK (des-Arg¹⁰-kallidin) is the optimal agonist sequence acting on the human and rabbit B₁ receptors (B₁Rs), whereas BK and Lys-BK in low concentrations stimulate B₂Rs. Selective antagonists are very useful for defining kinin receptor subtypes. The metabolically resistant peptide icatibant (Hoe-140; D-Arg[Hyp³,Thi⁵,D-Tic⁷,Oic⁸]BK) has been widely used in recent years (21) and is an effective tool in several mammalian species. Icatibant is a competitive antagonist in human tissues and cells, but there is evidence that this drug is a noncompetitive, nonequilibrium antagonist for the rabbit B₂Rs in contractility assays (32) and in COS-1 cells expressing the cloned rabbit B₂R (5). This species-specific effect of icatibant is characterized by a time-dependent loss of receptors in the presence of the antagonist and by a low reversibility of the antagonist-receptor interaction (22). The discovery of the B₁R antagonist prototype, [Leu⁸]des-Arg⁹-BK, was followed by the development of more potent and stable peptide antagonists such as B-9858 (Lys-Lys-[Hyp³,Igl⁵,D-Igl⁷,Oic⁸]des-Arg⁹-BK) (16). A large body of evidence shows that the B₁Rs are generally not expressed in normal physiological conditions but are rapidly induced after some types of injury, such as the injection of bacterial lipopolysaccharide (LPS); the cytokine network and mitogen-activated protein (MAP) kinases have been implicated in B₁R induction (23, 30, 33).

It is not known whether endogenous levels of kinins or of endothelium-derived mediators released by kinins can regulate the expression of kinin receptors. In vivo ACE inhibition, postulated to potentiate endogenous kinins, could also influence kinin receptor expression. The two genes corresponding to the B₁R and B₂R are located in the same region of human chromosome 14q32 (4, 8, 26), suggesting a possible coordinated regulation of gene expression. The rabbit genes for the two receptors' subtypes have been isolated and characterized (5, 27), enabling a more detailed investigation of their regulation. Thus several treatments pertaining to the above-mentioned issues have been applied to rabbits, which were the source of tissues for B₁R and B₂R mRNA level measurements, to the testing of functional responses to agonists of both receptor types, and to other experiments (B₂R immunohistochemistry and electrophoretic mobility shift studies of a promoter region of the B₁R gene).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Drugs. The following compounds were purchased from Sigma Chemical (St. Louis, MO): N^G-nitro-L-arginine methyl ester (L-NAME), diclofenac sodium, enalapril maleate, histamine dihydrochloride, captopril, and BK. Phenylephrine (PE) was manufactured by Sabex (Boucherville, PQ, Canada), and iloprost was a gift from Berlex (Lachine, PQ, Canada). Sar-[D-Phe⁸]des-Arg⁹-BK is a metabolically stable B₁R agonist (3). The kinin receptor antagonists for B₁R and B₂R, B-9858 and icatibant, respectively, were gifts from Laboratoires Fournier (Daix, France). LPS, extracted from Escherichia coli serotype O111:B4, was purchased from DIFCO (Detroit, MI).

Treatment of animals. Groups of four New Zealand White rabbits of either gender, weighing 1.5-2.2 kg, were used as a source of tissues for all experiments. Each rabbit was weighed before the treatments, which consisted in the injection of the following drugs: L-NAME (30 mg/kg, an NO synthesis inhibitor), diclofenac (5 mg/kg, a cyclooxygenase inhibitor), icatibant (50 µg/kg, a B₂R antagonist), B-9858 (50 µg/kg, a B₁R antagonist), enalapril maleate (2.8 mg/kg, an ACE inhibitor), or iloprost (30 µg/kg, a prostacyclin mimetic). Drugs or their vehicle (saline) was injected into the marginal ear vein every 8 h for 48 h in a fixed volume of 0.5 ml/kg. Drug doses were based on previously described experimental systems in which they were found to be effective in rabbits [as shown for enalapril (6), L-NAME (3), diclofenac (35), iloprost (12), icatibant (9), and B-9858 (18)]. LPS-treated rabbits were injected with 25 µg/kg LPS 8 h before they were killed; this treatment is known to induce vascular B₁Rs (30, 37). Icatibant, LPS, B-9858, and diclofenac do not acutely change the blood pressure at the selected or higher dose levels in rabbits (9, 18, 35, 37). L-NAME at 30 mg/kg induces a persistent hypertension in conscious rabbits (25), and iloprost and enalapril treatments may induce a minor hypotensive effect (12, 40). Of these treatments, LPS and/or ACE blockade are documented as increasing blood immunoreactive kinins in anesthetized rabbits (36). B₂R blockade alone with icatibant did not change immunoreactive kinin concentration in rat blood or tissues (7, 11) but reportedly increased immunoreactive BK in the renal interstitial fluid of uninephrectomized dogs (38). All rabbits were consecutively killed by CO₂/O₂ asphyxiation, and several organs (heart, abdominal aorta, kidney, duodenum, and psoas muscle) were quickly removed, frozen in liquid nitrogen, and kept at 80°C until RNA isolation. The thoracic aorta and the jugular veins were immediately used in the contractility studies (2 separate tissues of each type were used per animal). Serum was also sampled in enalapril-treated and control rabbits at the time the animals were killed to monitor the drug effect by using enzyme activity (ACE activity test kit based on the substrate [³H]hippuryl-Gly-Gly, Hycor Biomedical, Irvine, CA).

Total RNA isolation. Total RNA was isolated using the TRIzol reagent (GIBCO-BRL), as described by the manufacturer. Briefly, ~1 g of each tissue was ground in a mortar, and the resulting powder was resuspended in 10 ml of TRIzol. The suspension was then homogenized using a glass tissue grinder and centrifuged for 10 min at 10,000 g. The supernatant was extracted with 2 ml of chloroform, and after centrifugation the aqueous phase was mixed with 10 ml of isopropyl alcohol. The resulting pellet was washed with 5 ml of 75% ethanol and finally resuspended in 100 ml of sterile water. All total RNA samples were kept at 80°C until use.

Semiquantitative duplex RT-PCR. The RT-PCR experiments were based on a method published by Dukas et al. (13). Briefly, 100 µg of total RNA were treated for 30 min at 37°C in 20 mM Tris · Cl (pH 8.4), 2 mM MgCl₂, 50 mM KCl, and 60 U of RNA guard (Pharmacia Biotech, Baie d'Urfé, PQ, Canada) with 2 U of DNase I (amplification grade; GIBCO-BRL) to remove traces of genomic DNA. The mixes were then extracted with phenol-chloroform, and the RNA was ethanol precipitated and resuspended in 100 µl of sterile water. DNase I-treated RNA (2.5 µg) was denatured at 94°C for 3 min in 50 mM Tris · Cl (pH 8.3), 75 mM KCl, 3 mM MgCl₂, and 250 ng of oligo(dT)₁₅. Then, 1 mM dithiothreitol (DTT), 50 U of RNA guard, 1 mM each dNTP, and 200 U of Maloney's murine leukemia virus RT (GIBCO-BRL) were added, and the mix was incubated for 10 min at room temperature, then for 75 min at 37°C. One-tenth of the RT reactions was further used for PCR amplification. Each 50-µl PCR reaction contained 2 µl of RT mix, 10 mM Tris · Cl (pH 8.0), 50 mM KCl, 2 mM MgCl₂, 2% DMSO, 60 µM each dNTPs, 250 ng of each of the B₁R or B₂R primers, 25 ng of the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) primers (needed for the amplification of an internal standard), and 1.5 U of Taq polymerase (GIBCO-BRL). The following oligonucleotides were utilized as PCR primers: 5'-TGTCCCGGCCGAGTCACTGTG-3' and 5'-GCACCAGCACGCTGTAGCGGT-3' were used as sense and antisense primers, respectively, for the amplification of a specific rabbit B₁R fragment; the primer sequences were selected from MacNeil et al. (27) and on the basis of our unpublished results concerning the genomic organization of the rabbit B₁R, inasmuch as the sense primer covers an intron-exon junction located 11 nucleotides upstream from the ATG codon (exactly as in the human B₁R gene), thus eliminating any possibility of genomic DNA amplification during the RT-PCR reaction. 5'-GCGTCTTCTGCCTGCACAAGAGC-3' and 5'-GGCCCTCCTCTCCGTCTGGATCT-3' were used as sense and antisense primers, respectively, for the amplification of a specific rabbit B₂R fragment (primer sequences selected from Ref. 5); 5'-CACCATCTTCCAGGAGCGAGATCC-3' and 5'-GTCTTCTGGGTGGCAGTGATGGC-3' were used as sense and antisense primers, respectively, for the amplification of a specific rabbit GAPDH fragment (primer sequences selected from Ref. 2). The samples were denatured initially for 3 min at 94°C and then submitted to 24 cycles of PCR (45 s at 94°C, 45 s at 64°C, and 75 s at 72°C) and a 10-min final elongation step at 72°C. The number of cycles was chosen to keep the PCR-amplified DNA in the exponential phase of amplification. One-tenth of each of the PCR reactions was run on a 1% agarose gel in 1× Tris-acetate-EDTA buffer and then transferred to BrightStar-Plus nylon membrane (Ambion). The membranes were prehybridized for 1 h at 65°C in a buffer containing 6× saline-sodium citrate, 5× Denhardt's solution, 0.5% SDS, and 100 µg/ml salmon sperm DNA; then rabbit B₁R or B₂R random-primed ³²P-labeled probe (10⁶ cpm/ml) was added, and hybridization was carried out for 8-16 h at 65°C. The membranes were repeatedly washed (final wash 0.1× saline-sodium citrate-0.1% SDS at 65°C) and exposed to Biomax MS (Kodak) autoradiographic films. The membranes were then stripped in boiling 0.1% SDS and rehybridized with a rat GAPDH probe used as an internal standard. The resulting autoradiograms were scanned with a ScanJet 6 scanner (Hewlett-Packard) and analyzed with the 1D-main densitometry software (AAB Software).

Contractility studies. The isolated thoracic aortas were cut into rings (~4 mm diameter, 3-4 mm long) and suspended between a metal hook and a thread loop under a tension of 2 g in 5-ml tissue baths containing oxygenated (95% O₂-5% CO₂) and warmed (37°C) Krebs solution (23). The composition of Krebs solution was (mM) 117.5 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.18 MgSO₄, 2.5 CaCl₂, 25.0 NaHCO₃, and 5.5 D-glucose. The jugular veins were cut into 2-cm strips and suspended under a tension of 1 g in 5-ml tissue baths containing oxygenated and warmed Krebs solution to which 1 µM captopril was added (32). Changes of the vascular tone were measured using isometric force transducers (model 52-9545, Harvard) coupled with chart recorders.

Preparation of crude nuclear extracts. Crude nuclear extracts were prepared from kidneys of LPS-treated and control (saline-treated) rabbits as described previously (20), with some modifications. Briefly, ~1 g of kidney tissue (frozen in liquid nitrogen) was rapidly triturated with a mortar and pestle. Ten milliliters of 0.3 M sucrose in homogenizing buffer (15 mM HEPES, pH 7.9, 60 mM KCl, 15 mM NaCl, 0.15 mM spermine, 0.5 mM spermidine, 14 mM -mercaptoethanol, 0.5 µM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, and 1 mg/ml pepstatin) were added to the powdered tissue, and the suspension was homogenized with 10 strokes in a Dounce manual-type tissue grinder. Two hundred microliters of 10% Nonidet P-40 were added to the mixture (final concentration 0.5%), and the homogenization was repeated. The homogenate was layered on top of a 10-ml cushion of 0.9 M sucrose in homogenizing buffer and centrifuged for 10 min at 3,500 rpm and 4°C in a Sorvall HB-4 rotor. The nuclear pellet was resuspended in 10 ml of 0.3 M sucrose in homogenizing buffer and 0.2% Nonidet P-40 and consecutively homogenized and centrifuged as described above with use of the same sucrose cushion and centrifugation conditions. The nuclear pellet was then resuspended in 1 ml of buffer containing 20 mM HEPES-KOH, pH 7.9, 100 mM NaCl, 1.5 mM MgCl₂, 0.5 mM EDTA, 0.5 mM DTT, 50% glycerol, 0.5 mM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, and 1 µg/ml pepstatin and dialyzed against the same buffer at 4°C for several hours. After the protein concentration was measured, small aliquots of the nuclear extracts were quickly frozen and stored at 70°C.

Electrophoretic mobility shift assays. Two double-stranded oligonucleotides, one (5'-CACTTTTGCGGCAATCCCCAC- AAT-3') containing a putative wild-type nuclear factor-B (NF-B) site at positions 73 to 50 from the promoter region of the human B₁R gene and the other (5'-CACTTTTCAATCCCCACAAT-3') containing a mutated NF-B sequence (underlined), were designed essentially as described previously (33) and used as probes for mobility-shift analyses. The double-stranded oligonucleotides were labeled with [-³²P]ATP and eluted on 6% polyacrylamide gels containing 1× Tris-borate-EDTA (1× = 89 mM Tris, 89 mM borate, 1 mM EDTA). The eluted fragments were concentrated by lyophilization and purified by passage through Sephadex G-50 spin columns (Pharmacia Biotech). A similar strategy has been used for the isolation of nonlabeled double-stranded oligonucleotides used as cold competitors in the gel-shift assays. Approximately 20,000 counts/minute of ³²P-labeled oligonucleotides were incubated for 20 min at room temperature in 10-ml reaction volumes containing 5-10 µg of nuclear extract, 1 µg of poly(dI-dC) (Pharmacia Biotech), 300 µg/ml BSA fraction V (Sigma-Aldrich, Oakville, ON, Canada), 10 mM HEPES, pH 7.9, 50 mM KCl, and 5 mM DTT. Competitor DNA was added 10 min before the addition of the labeled probe. For supershift assays, 2 µg of different antibodies were incubated with the nuclear reaction mixture for 30 min at room temperature before the probe was added. The antibodies used were GADD 153 monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), which recognizes a C/EBP-related nuclear protein (CHOP-10), and NF-B p50 subunit polyclonal antibody (goat origin, concentrated formulation; Santa Cruz Biotechnology), used separately or together with NF-B p65 subunit monoclonal antibody (Transduction Laboratories, Lexington, KY). The DNA-protein complexes were analyzed by electrophoresis on 8-10% polyacrylamide gels containing 0.25× Tris-borate-EDTA at room temperature and a constant voltage of 250 V. Gels were fixed, dried, and autoradiographed with intensifying screens.

B₂R immunohistochemistry. Parts of tissue samples used for RNA isolation or bioassays were fixed using 10% buffered Formalin phosphate for histological examination. Paraffin-embedded tissue sections (5 µm thick) were prepared, and preliminary tests showed that the chosen antibody could be used only if an antigen-retrieving procedure was applied to sections (5 min of high-temperature heat denaturation in 0.01 M citrate buffer, pH 6.0) (34). Tissue sections were immunostained overnight at room temperature with the anti-B₂R monoclonal antibody (clone 20, Transduction Laboratories, dilution 1:100). This antibody has been raised against the COOH-terminal 15-mer peptide of the human B₂R (see DISCUSSION); a synthetic peptide corresponding to the immunogen has been purchased from Transduction Laboratories to test the specificity of the antibody tissue binding (coincubation of the peptide, 10 µg/ml, with the antibody on some sections). The monoclonal antibody staining was revealed by horseradish peroxidase-coupled goat anti-mouse IgG (Sigma Chemical) that was allowed to react for 15 min at 25°C with the Immunopure Metal Enhanced diaminobenzidine substrate (Pierce). Endogenous peroxidase was initially inhibited in tissue sections by the Immunoperoxidase Suppressor reagent (Pierce) used as directed.

Statistical analysis. Statistical analysis was performed using the Kruskal-Wallis test followed by the Mann-Whitney test with the InStat 2.0 computer program (GraphPad Software).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Multiplex RT-PCR analysis of kinin receptor mRNA in organs from treated rabbits. The B₁R and B₂R mRNA expression levels in several rabbit organs were determined using a semiquantitative RT-PCR assay; the results are presented in Table 1. A baseline mRNA expression was detected in each organ for both receptors (normalized to 1 in Table 1). The 8-h LPS treatment was the most consistent in inducing B₁R expression over the baseline (significant in the kidney, duodenum, and striated muscle). A weaker but still significant increase in the B₁R mRNA expression was also monitored in the duodenum after 48-h treatments with iloprost or the B₁R antagonist B-9858. The B₂R mRNA expression was not significantly upregulated by any of the treatments in all organs studied. Notably, prolonged B₂R blockade by icatibant did not result in a rebound production of the B₂R mRNA. On the contrary, the B₂R mRNA expression was even suppressed by icatibant treatment in the duodenum and heart. Such organ-specific responses were found with other drugs: enalapril and diclofenac tended to reduce B₂R mRNA concentration in the heart and duodenum; iloprost exerted a similar effect in the duodenum and L-NAME in the heart. LPS had no consistent effect on the B₂R mRNA expression in the sampled organs (inhibition only in the heart). Significant suppression of the B₁R mRNA levels was also found in some organs after various treatments (with iloprost in the duodenum, heart, and aorta and with B-9858 in the aorta).

Vascular contractility mediated by B₁R and B₂R as a function of treatments applied in vivo to rabbits. Sigmoidal concentration-effect curves for each tested agonist (Figs. 1 and 2) were characterized by the EC₅₀ and the maximal absolute contraction amplitude (percentage of internal standard; Table 2). The recorded aortic ring responses to the B₁R agonist Sar-[D-Phe⁸]des-Arg⁹-BK were recorded within the 1st h of tissue isolation. LPS treatment induced a definite state of responsiveness to this peptide compared with saline-treated controls, but all the other treatments produced very low levels of maximal responses similar to the saline-treated animals (Fig. 1; Table 2). The rabbit jugular vein stimulated with the B₂R agonist BK revealed a similar state of sensitivity and maximal response in all groups, except for the group treated with icatibant. In this case, the maximal effect was significantly decreased, and the EC₅₀ value increased relative to values from the saline control animals (Table 2, Fig. 2). Tissues from icatibant-treated rabbits were not exposed to the drug in vitro, inasmuch as the last drug injection was applied 8 h before the animal was killed.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Initial responsiveness of aortic rings derived from control (saline vehicle iv) or lipopolysaccharide (LPS)-treated rabbits (25 mg/kg iv 8 h before they were killed) to an agonist of B1 receptor (B1R). Cumulative concentration-effect curve of B1R agonist Sar-[D-Phe8]des-Arg9-bradykinin (BK) was established after a short in vitro incubation (45 min) to minimize influence of isolation and tissue incubation on responses. Values (means ± SE of 8 determinations from 4 animals in each group) are expressed as a percentage of an internal contractile standard, maximal effect of phenylephrine, established in each tissue. See Table 2 for statistical analysis.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Initial responsiveness of external jugular vein strips derived from control (saline vehicle iv), LPS-treated (25 mg/kg iv 8 h before animals were killed), or icatibant-treated rabbits to an agonist of B2 receptor (B2R; see Table 1 for drug dose and administration schedules). Cumulative concentration-effect curve of B2R agonist BK was established after a short in vitro incubation (60 min). Values (means ± SE of 8 determinations from 4 animals in each group) are expressed as a percentage of an internal contractile standard, maximal effect of histamine, established in each tissue. See Table 2 for statistical analysis.

Treatment effects unrelated to kinin receptors. In the four rabbits treated with enalapril, the serum ACE activity was reduced to 11.4 ± 6.3 U/ml compared with 197.6 ± 31.4 U/ml in controls (n = 4, P = 0.03, Mann-Whitney test). Unexpectedly, some of the treatments altered the effects of the reference agonists in the contractility tests (Table 2): L-NAME, both kinin receptor antagonists, enalapril, and iloprost treatments significantly decreased the EC₅₀ values for the ₁-adrenoceptor agonist PE, whereas that of histamine was reduced by the diclofenac treatment in the jugular vein.

Electrophoretic mobility shift analysis of a putative NF-B binding motif in the kinin B₁R gene promoter. Recently, it has been suggested that transcription factor NF-B regulates the inducible expression of the human B₁R gene in inflammation, inasmuch as an NF-B binding motif from the promoter region of the human B₁R gene (positions 73 to 53, relative to the transcription initiation site) was shown to be implicated in the interleukin-1-induced upregulation of this receptor gene (33). We have performed electrophoretic mobility shift assays to investigate whether NF-B is involved in the induction of B₁R expression in the LPS-treated rabbits. Two double-stranded oligonucleotides, one (5'-CACTTTTGCGGCAATCCCCACAAT-3') containing the putative wild-type NF-B site at positions 73 to 50 from the promoter region of the human B₁R gene and the other (5'-CACTTTTCAATCCCCACAAT-3') containing a mutated NF-B sequence (underlined), were used as ³²P-labeled probes for mobility-shift analyses in combination with crude nuclear extracts isolated from kidney samples of control and LPS-treated rabbits. Electrophoretic mobility shift assays revealed a specific protein-DNA complex when nuclear extracts from LPS-treated rabbits and the wild-type probe were used (Fig. 3A); this complex was considerably weaker with the mutated probe and absent when nuclear extracts from control animals were used (Fig. 3B). The specificity of these protein-DNA interactions was confirmed by progressive inhibition of binding in the presence of a 20- and 100-fold molar excess of the same unlabeled probe used as a competitor. To further determine whether this complex is generated by an NF-B factor binding to the probe, antibody supershifts were performed using a mix of two antibodies: NF-B p50 polyclonal antibody and NF-B p65 monoclonal antibody. As shown in Fig. 3, these antibodies failed to supershift the protein-DNA complex, suggesting that another factor (or other factors) differing from classical NF-B binds to this promoter region in our experimental system. The same antibodies, when used separately, also failed to supershift the nuclear factor from LPS-treated rabbits (data not shown). Because the sequencing motif of this promoter domain also resembles a CHOP binding site and CHOP represents a stress-responsive transcriptional regulator (39), we performed similar supershift experiments using the CHOP-specific monoclonal antibody GADD 153 and again failed to detect any supershift, indicating that the CHOP protein is not involved in the formation of the protein-DNA complex in this promoter region (Fig. 3).

View larger version (54K): [in this window] [in a new window]   Fig. 3.   Electrophoretic mobility shift and supershift assays of a putative nuclear factor-B (NF-B) binding motif in kinin B1R gene promoter. A: gel-shift analyses with use of crude nuclear extracts (NE) from kidneys of LPS-treated rabbits. B: gel-shift analyses with use of crude nuclear extracts from kidneys of control rabbits. Arrow, specific protein-DNA complex. Ab, antibody; CHOP, C/EBP-related nuclear protein.

B₂R immunohistochemistry. A series of experiments were performed to validate the staining obtained with the anti-B₂R monoclonal antibody. COS-1 cells were transiently transfected with an expression vector for the cloned rabbit B₂R, as previously described (5), transferred onto glass slides, and submitted to the antigen retrieval treatment. Many of the treated cells stained positively (Fig. 4A), whereas the nontransfected COS-1 cells were not stained under the same conditions (Fig. 4B). In rabbit kidney tissue sections, the bulk of the immunostaining was localized in tubular epithelium, but other structures such as blood vessels were also stained (Fig. 4C, arrowhead). In the rabbit duodenum, the staining was shared between the epithelium and the underlying structures that include glands and smooth muscle; in the rabbit heart, endothelial cells of small blood vessels were the most intensively stained cells (Fig. 4I, arrowheads). In the psoas muscle, the staining was also prominent in microvessels (data not shown). Two types of controls were used to interpret the staining: the omission of the monoclonal antibody revealed a low background for each organ (Fig. 4, D, G, and J); there was also a massive loss of staining in the presence of the immunogen peptide of human sequence (Fig. 4, E, H, and K).

View larger version (166K): [in this window] [in a new window]   Fig. 4.   Immunohistochemistry of rabbit B2R. A: staining of COS-1 cells transfected with an expression vector for rabbit B2R. B: lack of specific staining of untransfected COS-1 cells. C, F, and I: immunostaining of B2Rs in tissue sections of normal rabbit kidney, duodenum, and heart (left ventricle), respectively. D, G, and J: background staining of secondary peroxidase-labeled antibody in absence of primary antibody in same organs. E, H, and K: competition of B2R immunostaining by coincubation with immunogen peptide. Magnification ×400 (A and B), ×100 (C-H), and ×200 (I-K).

View larger version (132K): [in this window] [in a new window]   Fig. 5.   Immunohistochemistry of rabbit B2R in tissue sections derived from saline or icatibant-treated rabbits. Magnification ×600.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our experimental data indicate that LPS was the only applied treatment to consistently induce B₁R expression (functional response in aorta and mRNA induction in kidney, duodenum, and psoas muscle). The B₁R mRNA concentration was slightly, but not significantly, elevated in the aorta of LPS-treated animals. It was shown previously that there is an induction of B₁R mRNA expression in the cardiac and aortic tissues in the rabbit 3 h after LPS injection [Northern blots of poly(A)⁺ RNA] (30, 31). Thus the 8-h period between LPS injection and organ harvesting may be too long to record a persistent rise in receptor mRNA in some organs. The weaker induction of B₁R mRNA expression in the duodenum on treatment with the B₁R antagonist B-9858 or with the prostacyclin mimetic iloprost remains unexplained.

The B₁R may be one of the most profoundly regulated G protein-coupled receptors, to the point of being completely inducible (30). Constructions based on the human B₁R gene promoter transfected into vascular smooth muscle cells are strongly upregulated by interleukin-1 or bacterial LPS, apparently via an NF-B-like site located proximal to the TATA box and transcription initiation site (33). Our data from the electrophoretic mobility shift assays with use of a human B₁R promoter fragment containing this NF-B-like motif and nuclear extracts from kidneys of LPS-treated rabbits and from control (saline-treated) animals have indicated the presence of a specific factor binding to this domain only in the LPS-treated animals, inasmuch as the mutated motif showed a much weaker binding activity (Fig. 3). The lack of supershift on application of NF-B-specific antibodies indicates that NF-B is not involved in the formation of the specific DNA-protein complex in the region studied. CHOP (GADD 153), a member of the C/EBP transcription factor family, forms heterodimers with homolog factors and regulates the transcription of some genes in response to stress, notably after LPS treatment (39). By performing similar supershift analyses, we have investigated whether the CHOP protein is involved in binding the promoter region, because CHOP is phosphorylated and activated by the p38 MAP kinase, which was previously shown to participate in postisolation B₁R induction in the rabbit aorta (23). In addition, the DNA consensus sequence for the binding of CHOP heterodimer, RRRTGCAATMCCC (where R is A or G and M is A or C) (39), is strikingly similar to the putative proximal NF-B promoter motif that confers inducibility in rat vascular smooth muscle cells transfected with a reporter gene construction (see MATERIALS AND METHODS) (33). The results of these studies were also negative, indicating that the p38 MAP kinase regulation of the B₁R gene may not involve CHOP binding to this proximal promoter region.

The B₂Rs are constitutively expressed in a wide variety of tissues, but there is some evidence that B₂R expression is transcriptionally regulated in cultured cells (30). It is remarkable that systemic and persistent blockade of the B₂Rs with icatibant does not induce a compensatory mechanism involving B₁R upregulation, as assessed by B₁R mRNA concentration or function in the aorta. A shorter-term treatment with icatibant (3 h) also failed to upregulate B₁R mRNA in rabbit hearts (31). The observed downregulation of B₂R mRNA expression in some organs in animals treated with enalapril, icatibant, iloprost, or diclofenac (Table 1) may suggest kinin- or prostanoid-induced regulation of B₂R expression, at least in some tissues; however, pairs of treatments that were designed to exert opposing physiological effects on tissues (enalapril-icatibant and diclofenac-iloprost) did not influence kinin receptor expression in opposite directions in any organ, thus failing to support general regulatory feedback that would link kinin receptor expression to the actions of kinins or prostanoids. The similar suppression of B₂R mRNA in the heart and duodenum by enalapril, postulated to increase kinin concentration, and icatibant, capable of blocking kinin actions on preformed B₂Rs, is a paradoxical and unexplained observation. A possible common effect of both treatments is to increase tissue kinins, at least in some organs (36, 38), but it is not clear whether increased kinin concentrations would exert any tissue effect when B₂Rs are blocked by icatibant. The present data do not document a reciprocal switch between B₁R and B₂R expression, inasmuch as the B₂R mRNA is not consistently suppressed in organs where the B₁R is induced.

Enalapril treatment applied to rabbits is associated with a 94% inhibition of the circulating ACE activity at the end of the treatment, representing a higher level of inhibition than that observed in human patients treated with ACE inhibitors and assessed using the same test (29). This clear drug effect was not associated with B₁R induction in any of the tested organs but, rather, with a suppression of B₂R mRNA expression in the intestine and heart. A shorter treatment with the active metabolite enalaprilat (3 h) also failed to induce B₁R mRNA synthesis in the rabbit heart (31). Similarly, hemodynamic analyses applied to anesthetized rabbits failed to detect functional evidence for cardiovascular B₁R in captopril- or enalapril-pretreated animals (10). These observations further disprove B₁R induction by endogenous kinins, inasmuch as there is some evidence that ACE inhibition in rabbits is associated with an increased concentration of endogenous kinins (36). However, the effect of ACE inhibition and of a superimposed cardiovascular pathology (e.g., hypertension with end organ damage) on kinin receptor regulation remains to be evaluated.

One of the limitations of the present in vivo approach concerns the cell type(s) expressing kinin receptors. These cell types may vary from one organ to the other, as shown by the B₂R immunohistochemistry studies. Thus epithelial elements are prominent in the kidney and intestine, as reported previously (15, 28), whereas the vasculature may be the major contributor to the mRNA signal in other organs such as the heart and striated muscle. Tissue and cell type-specific forms of kinin receptor regulation are likely to exist, and the present results are compatible with this idea, as indicated by the B₂R mRNA expression in heart tissue, which appears to be sensitive to many influences (Table 1). The treatments may even affect unrelated receptor populations, as suggested by some effects on responses to PE and histamine (Table 2). Another limitation concerns the RT-PCR technique applied, since physiologically irrelevant levels of receptor mRNA could be amplified in some cases by this nonlinear technique. Indeed, the aortic contractility data suggest that the B₁R is not expressed on any of the systemic treatments except LPS, although the corresponding mRNA is detected in all groups. Artifacts due to kinin or mediator generation during the procedure used to kill the animals and tissue preparation may also influence these results.

Several types of controls support the staining of a relevant rabbit B₂R population by a monoclonal antibody raised against a 15-mer from the human B₂R (ERQIHKLQDWAGSRQ) that is not conserved in the B₁R sequence but is relatively well conserved in the corresponding rabbit B₂R region (ERQIHKLPEWTRSSQ) (5). Notably, staining was shown in heterologous cells transfected with an expression vector coding for the rabbit B₂R (Fig. 4A), and a soluble form of the immunogen peptide competed for antibody staining (Fig. 4). The immunohistochemistry studies identified the major cell types expressing B₂Rs, but the tissue fixation technique, paraffin sections, and antigen retrieval scheme are not likely to reveal the subcellular localization of receptors (15, 34). The various treatments applied to rabbits did not overtly change the immunostaining of the examined organs, except for the icatibant treatment (Fig. 5), which is also the only treatment to depress the ex vivo contractility of the jugular vein significantly. Icatibant is a nonequilibrium, noncompetitive antagonist of the rabbit B₂Rs (32), meaning that it binds the receptor in a poorly reversible or nonreversible manner (22). The molecular determinants of this pharmacological behavior are not precisely known, but the receptor and the drug structures are critical, inasmuch as icatibant may be competitive for B₂Rs in other species and other B₂R antagonists are competitive in rabbits (32). The present data suggest that the receptor population in vivo is cleared more rapidly than it is replaced in icatibant-treated rabbits, as evidenced by the disappearance of the B₂R COOH-terminal epitope recognized by the antibody 8 h after the last injection of the drug (Fig. 5). The residual functional contractile response of the jugular vein after the icatibant treatment (Fig. 2) suggests that there is a large receptor reserve in this organ, inasmuch as the same tissues appear to be massively depleted of immunoreactive receptors (Fig. 5).

In summary, several treatments (e.g., B₂R blockade and ACE inhibition) failed to upregulate the B₁Rs, further supporting the association of the B₁R expression to tissue injury situations and immunopathology. B₁R induction by LPS is correlated to the presence in renal tissue of unidentified nuclear factor(s) binding to a previously identified motif in the B₁R gene promoter. B₂R mRNA expression may be downregulated in some organs by several of the applied treatments, but the data did not support generally applicable regulatory mechanisms involving endogenous kinins, prostanoids, or NO. The expression of the two kinin receptor genes appears to be distinctly regulated. Icatibant, a nonequilibrium, noncompetitive antagonist for the rabbit B₂R, can deplete the B₂R population in vivo in the rabbit.

Perspectives