Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The classification of human receptor (R) subtypes for bradykinin (BK)-related peptides (kinins) as the B₁ receptor (B₁R) and the B₂ receptor (B₂R) is now supported by extensive pharmacologic and genetic evidence (Hess et al., 1992; Menke et al., 1994; Marceau et al., 1998). The potency order of agonists is typical: the B₁R is stimulated by kinin fragments without the C-terminal arginine residue, optimally Lys-des-Arg⁹-BK (des-Arg¹⁰-kallidin), whereas the B₂R is stimulated by the native kinin sequences, BK and Lys-BK (kallidin). Specific antagonists further differentiate B₁R and B₂R types. Several peptide and nonpeptide B₂R antagonists are presently known (Hock et al., 1991; Aramori et al., 1997). Lys-[Leu⁸]-des-Arg⁹-BK is the prototype high-affinity B₁R antagonist (Marceau et al., 1998). The B₁R is a peculiar example of an inducible G protein-coupled receptor, with a low or null level of expression in physiological conditions, but strong induction following various forms of tissue injury, including isolation and in vitro incubation of smooth muscle preparations (Marceau et al., 1998).

Some genetic polymorphisms of human kinin receptor genes have been recently described (see Fig. 1 for details). A three exon structure of the human B₂R gene has been proposed, with the coding sequence in exons 2 and 3 (Kammerer et al., 1995; Lung et al., 1997). An exon 1 (noncoding) polymorphism in which alleles differ by a 9-bp deletion, designated () versus the (+) complete sequence, has been proposed to be of clinical significance (Lung et al., 1997). The () allele determines a splice variant presumably more stable relative to the action of RNases, and appears to confer a higher level of expression, as it is always present in the most symptomatic cases of C1 inhibitor deficiency (hereditary angioedema with angioedema crises determined by the activation of plasma kallikrein and other blood proteases). Thus, the B₂R () allele is proposed to modulate in a dominant manner the penetrance of the basic genetic defect in this disorder, C1 inhibitor deficiency, that is by itself presumably linked to explosive kinin formation and consecutive increased vascular permeability. However, this polymorphism failed to predict the most common side effect of angiotensin-converting enzyme inhibitors, the nonproductive cough (Zee et al., 1998). A B₂R gene polymorphism affecting the coding region (exon 2, C¹⁸¹  T) determines an Arg¹⁴  Cys substitution in the extracellular N-terminal domain (Braun et al., 1995). Other polymorphisms of the B₂R gene have been located in the 3' untranslated region of exon 3 (Braun et al., 1995) or in the postulated promoter region (Braun et al., 1996; Erdmann et al., 1998); one of the promoter polymorphisms appears to be frequent (C/T at 58; Braun et al., 1996).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the structure of the human B1R and B2R genes and the location of the polymorphisms studied. Exons are indicated with boxed areas (designated E1 to E3); hatched areas represent coding regions. The approximate length of some of the longer introns are indicated. SSCP, single strand conformation polymorphism.

The three-exon structure of the human B₁R gene was recently established (Fig. 1; Bachvarov et al., 1996), and two allelic polymorphisms were discovered in this gene, but none so far in the coding region (Bachvarov et al., 1998b). An A¹⁰⁹⁸  G polymorphism, located 35 nucleotides downstream of the stop codon and 14 nucleotides upstream of the polyadenylation signal, introduces a TaqI restriction site but appears to be clinically neutral. A second, and more frequent polymorphism consists of a single base substitution (G⁶⁹⁹  C) in a positive control region of the promoter (position relative to the major transcription initiation site; this generates an AciI restriction site). The C allele is significantly less prevalent in patients with renal failure or inflammatory bowel disease (Bachvarov et al., 1998a,b).

Although the mechanical properties and sensitivity of the human umbilical vein collected postpartum are not as consistent as those of many preparations of animal origin, it was felt by some investigators that this tissue of fetal origin could be useful to characterize the human forms of kinin receptors (Marceau et al., 1994; Gobeil et al., 1996; Sardi et al., 1997, 1998, 1999). Kinins contract the isolated human umbilical vein via preformed B₂Rs responsive to BK, and via inducible B₁Rs sensitive to both des-Arg⁹-BK and Lys-des-Arg⁹-BK (Marceau et al., 1994; Gobeil et al., 1996; Sardi et al., 1997, 1998, 1999). Randomly sampled umbilical cords should segregate into various known genotypes of the characterized allelic pairs corresponding to kinin receptor genes. We have exploited the contractile response of the umbilical vein mediated by both B₁Rs and B₂Rs to evaluate the contribution of the frequent polymorphic alleles to the variability of responses. We hoped to identify functional consequences of recently found genetic polymorphisms, in particular of those claimed to promote functional alterations even under the heterozygote genotype (Lung et al., 1997; Bachvarov et al., 1998a,b).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Metabolic Stability of B₁ Agonists. Sar-[D-Phe⁸]des-Arg⁹-BK is a high-affinity B₁R agonist that is completely resistant to blood aminopeptidase, kininases I and II (angiotensin I converting enzyme) and kidney-neutral endopeptidase (Drapeau et al., 1991, 1993). To reduce response variability originating from drug distribution or metabolism in the preparation, we evaluated B₁R-dependent contractility using the most metabolically stable agonist available. The metabolic resistance of two B₁R agonists, Lys-des-Arg⁹-BK and Sar-[D-Phe⁸]des-Arg⁹-BK, was evaluated using biological material of human origin relevant for the pharmacologic assay, namely, fresh umbilical vein tissue prepared as described below. The tissues (142-176 mg of fresh tissue per tube) were incubated for 2 h in the presence of Sar-[D-Phe⁸]des-Arg⁹-BK or Lys-des-Arg⁹-BK (100 µg/ml) dissolved in 1 ml of filtered (0.22 µm) Krebs' solution, with periodical sampling. HPLC analysis of pure peptides or samples from metabolic studies was performed using a Waters HPLC equipped with a 746 data module and a 486 UV detector set at 214 nm. Separations were achieved with a Vydac 10 m (3.9 × 300 mm) reversed phase C₁₈ column using a linear gradient of 5 to 65% acetonitrile/TFA 0.05%/water at 2 ml/min over a period of 20 min. For degradation studies, 10 to 25 µl of each aliquot was injected, and rates of peptide metabolism were calculated from the decrease of peptide substrate concentration. Integration of peak areas and quantification of peptide substrate were calculated automatically by the data module.

Kinin Receptor-Mediated Contractility in Rings of the Human Umbilical Vein. Rings of human umbilical vein from 100 individuals were prepared and mounted in organ baths for contractility studies as described previously (Marceau et al., 1994), except that 2 g of resting tension was applied to the rings. Changes in tension were recorded using isometric force transducers (model 52-9545; Harvard Bioscience, South Natick, MA). The tissues were equilibrated for 2 h before starting the experiments.

DNA Isolation and Genetic Testing. Genomic DNA was isolated from approximately 0.5-g pieces of umbilical cord, as described (Strauss, 1995). Polymerase chain reaction-restriction fragment length polymorphism (RFLP) analysis of the 3' flanking region bearing a TaqI polymorphism and of the 5' flanking (promoter) region bearing an AciI polymorphism of the human B₁R gene was performed as described (Bachvarov et al., 1998b). The analysis of the B₂R gene polymorphisms in exons 1 and 2 (coding and noncoding regions) and in the promoter region was performed using exactly the same polymerase chain reaction primers and conditions indicated by the authors of the original articles (Braun et al., 1995, 1996; Lung et al., 1997). Alleles and genotypes are designated either by the (+) and () symbols for the B₂R gene insertion/deletion polymorphism, or by the initial of the polymorphic base for the four other polymorphisms (single base substitutions). Homozygotic and heterozygotic individuals are readily recognized in each one of these tests by the final electrophoresis migration patterns.

Drugs. Lys-des-Arg⁹-BK was purchased from Peninsula Laboratories (Belmont, CA), and BK was obtained from Bachem (Torrance, CA). Sar-[D-Phe⁸]des-Arg⁹-BK and Ac-Lys-[Leu⁸]des-Arg⁹-BK were either produced locally (Drapeau et al., 1991, 1993) or supplied by Dr. D. Regoli (Sherbrooke, Canada). LF16.0687 ((1-[[2,4-dichloro-3-[(2,4-dimethylquinolin-8-yl)oxy]methyl]phenyl]sulfonyl]-N-[3-[[4-(aminoiminomethyl)phenyl] carbonylamino]propyl]-2(S)-pyrrolidinecarboxamide, mesylate salt) was a gift from Laboratoires Fournier, Daix, France. 5-HT (creatinine sulfate complex) was purchased from Sigma.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Metabolism of B₁R Agonists in the Human Umbilical Vein. HPLC determinations proved that Sar-[D-Phe⁸]des-Arg⁹-BK was metabolized more slowly than the corresponding natural sequence Lys-des-Arg⁹-BK following exposure to umbilical vein fragments (Fig. 2A). The 1st h of metabolism is compatible with first order kinetics, with calculated half-lives of 57 and 15 min for Sar-[D-Phe⁸]des-Arg⁹-BK and Lys-des-Arg⁹-BK, respectively. Control experiments showed that these peptides dissolved in Krebs' solution were stable when incubated at 37°C for 2 h in the absence of an enzyme source (not shown). Metabolites arising from cleavage of substrate by umbilical veins were observed as peaks with retention times longer than the added peptide substrates (less polar peptides, not shown), but were not further characterized. The analog Sar-[D-Phe⁸]des-Arg⁹-BK was used to functionally monitor the population of B₁Rs in this preparation to reduce the response variability caused by drug distribution or metabolism in the preparation.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   B1R agonists and the human umbilical vein. A, metabolic breakdown of 100 µg/ml solutions of Lys-des-Arg9-BK or Sar-[D-Phe8]des-Arg9-BK by human umbilical vein fragments. Peptide concentrations were determined by HPLC based on calibration curves established with authentic standards (see Materials and Methods). B, effect of kinin receptor antagonists on the umbilical vein contractile response to Sar-[D-Phe8]des-Arg9-BK. The B1R antagonist Ac-Lys-[Leu8]des-Arg9-BK (Ac-Lys-Leu) or the B2R antagonist LF16.0687 were introduced 30 min before constructing the concentration-response curves. C, increasing maximal contractile effect (Emax) of the B1R agonist Sar-[D-Phe8]des-Arg9-BK on human umbilical vein rings over time. The Emax values were assessed 2 and 6 h after the start of the tissue incubation in each tissue represented in the pharmacogenetic analysis (see text). D, concentration-response curve for the contractile effect of the B1R agonist Sar-[D-Phe8]des-Arg9-BK on human umbilical vein rings. The concentration-response curve was constructed 6 h after the beginning of in vitro incubation. In B, values are the means, and vertical bars the S.E. values of 7 to 13 determinations from different donors. In C and D, values are the means, and vertical bars are the S.E. values of 100 determinations from the 100 different donors represented in the pharmacogenetic study.

Preliminary Pharmacologic Studies on the Human Umbilical Vein. Ancillary experiments established that Sar-[D-Phe⁸]des-Arg⁹-BK has no activity on the B₂Rs present in the umbilical vein preparation, as the B₂R antagonist LF16.0687 (1 µM) failed to antagonize the effect of this agonist (Fig. 2B). In the same set of experiments, the B₁R antagonist Ac-Lys-[Leu⁸]des-Arg⁹-BK (5 µM) shifted the concentration-response curve of Sar-[D-Phe⁸]des-Arg⁹-BK to the right in a surmountable manner. Conversely, BK is a selective B₂R agonist in the human umbilical vein because it is antagonized in a surmountable manner by LF16.0687, but not by Ac-Lys-[Leu⁸]des-Arg⁹-BK (Fig. 3A).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of the B2R agonist BK on the human umbilical vein. A, effect of kinin receptor antagonists on the umbilical vein contractile response to BK. The B1R antagonist Ac-Lys-[Leu8]des-Arg9-BK (Ac-Lys-Leu) or the B2R antagonist LF16.0687 were introduced 30 min before constructing the concentration-response curves. Values are the means, and vertical bars the S.E. values of 7 to 13 determinations from different donors. B, concentration-response curve for the contractile effect of the B2R agonist BK on human umbilical vein rings constructed 4 h after the beginning of in vitro incubation. In B, values are the means, and vertical bars the S.E. values of 100 determinations from the 100 different donors represented in the pharmacogenetic study.

4

View this table: [in this window] [in a new window]   TABLE 1 Contractility parameters as a function of genotype distributions for kinin receptor gene polymorphisms Alleles and genotypes were designated either by the (+) and () symbols for the B2R gene insertion/deletion polymorphism, or by the initial of the polymorphic base for the four other polymorphisms.

Intraindividual Variability. When expressed in units of force (g), the maximal contractile effects of 5-HT and BK are highly correlated in the sampled preparations (r = .93, P < 10⁴, n = 112). The maximum response to the B₁R agonist Sar-[D-Phe⁸]des-Arg⁹-BK is more variable, but still correlated to that of 5-HT (r = .79, P < 10⁴). Thus, contraction was normalized using the maximal contraction caused by 5-HT as an internal standard to reduce the variability related to different mechanical properties of individual tissue preparations. As for the variability related to the tissue donors, the duplicate 6-h E_max values for Sar-[D-Phe⁸]des-Arg⁹-BK and 4-h E_max for BK, both expressed as a percentage of the maximal response to 5-HT, were significantly correlated in 12 umbilical veins for which duplicate rings were obtained, (r = .64, P = .02; r = .72, P = .008, respectively, for the two peptides).

Effect of Allelic Polymorphism of the Kinin Receptor Genes on the Contractile Action of Kinins. The results of the analyses of the five kinin receptor polymorphisms studied in the random neonatal population are summarized in Table 1. The Hardy-Weinberg equilibrium applied to the genotype distribution of all polymorphisms in the sample of tissue donors, and there was no significant linkage disequilibrium between any polymorphism pair in the sample. Statistical analysis of the E_max and EC₅₀ values revealed genotype-specific significant differences only in two cases. The (/) genotype of a frequent insertion/deletion (+/) polymorphism of the B₂R exon 1 was associated with increased contractile efficiency (lower median EC₅₀, higher median E_max at 6 h) of the B₁R agonist, Sar-[D-Phe⁸]des-Arg⁹-BK, relative to that of the most frequent genotype (+/), but with no significant effect on BK-induced contractility (Table 1, Fig. 4). A B₂R exon 2 polymorphism (C¹⁸¹  T) influenced selectively the potency of BK (median EC₅₀ of 38.8 nM when the T allele was present in heterozygotes, compared with 12.1 nM for the CC genotype; Table 1, Fig. 5). The homozygote TT genotype was not present in the sample.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effect of the exon 1 insertion/deletion B2R polymorphism on the response of the human umbilical vein to the B1R agonist Sar-[D-Phe8]des-Arg9-BK. Statistical analysis of these data is presented in Table 1.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Effect of the exon 2 B2R polymorphism on the response of the human umbilical vein to the B2R agonist BK. Statistical analysis of these data is presented in Table 1.

699

58

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The human umbilical vein preparation possesses a B₂R population mediating a stable contractile response to BK as a function of time (Marceau et al., 1994). Additional evidence supporting a role for B₂R mediating BK contraction is the competitive antagonism exerted by B₂R-selective drugs as icatibant, WIN 64338 (Marceau et al., 1994), or LF16.0687 (present results). This preparation also contains an inducible B₁R population (Gobeil et al., 1996; Sardi et al., 1997, 1998, 1999). Sar-[D-Phe⁸]des-Arg⁹-BK activates the B₁R population, as shown by the effect of a B₁R-selective antagonist and the increased agonist E_max as a function of incubation time. The E_max increase is a B₁R-specific phenomenon, as the concentration-response curves for BK repeatedly constructed at times 2, 4, and 6 h are identical (Marceau et al., 1994). Protein synthesis inhibition with cycloheximide prevented the up-regulation of contractile response to the B₁R agonist des-Arg⁹-BK in this preparation (Sardi et al., 1997), as did several other drugs (brefeldin A, tunicamycin, dexamethasone); others enhanced the contractility up-regulation (lipopolysaccharide, interleukin-1, tumor necrosis factor-; Sardi et al., 1998, 1999). Sar-[D-Phe⁸]des-Arg⁹-BK possesses the distinct advantage over natural B₁R agonists of being metabolically stable. In a binding assay using rabbit B₁Rs, this analog is 7-fold more potent than des-Arg⁹-BK, but 11-fold less potent than Lys-des-Arg⁹-BK (Levesque et al., 1995). The human umbilical vein responds to Sar-[D-Phe⁸]des-Arg⁹-BK with an intermediate potency between the optimal B₁R agonist, Lys-des-Arg⁹-BK, and the shorter sequence des-Arg⁹-BK (Gobeil et al., 1997). The stability of Sar-[D-Phe⁸]des-Arg⁹-BK is not absolute, as it is slowly hydrolyzed into undetermined products in contact with human cells, and it is metabolized in vivo in rabbits, although much more slowly than Lys-des-Arg⁹-BK (Audet et al., 1997). As the response to B₁R agonists was found to be particularly variable in the umbilical vein preparation, a stable agonist and a normalization of contractility relative to an internal standard for each tissue (the maximal response to 5-HT) were used to reduce the variability associated with drug distribution and metabolism, or with the cell contractile apparatus. This approach is further justified by a relatively low intravein variability when the E_max values recorded at 4 or 6 h for kinin receptor agonists are compared in duplicate tissue preparations.

These experiments were based on the assumption that the variable responses to agonists of both kinin receptors could be partially determined genetically, and that such a genetic influence would not be obscured by nongenetic factors that could contribute to response variability in umbilical veins (e.g., gestational maturity, fetal sex, storage duration and conditions, intracord variability, and postisolation gene regulation concerning the B₁Rs). The fact that the rarer alleles are not represented in the homozygous form in the sampled individuals for three of five polymorphisms further limits our study. Thus, the genetic influence on phenotype should be particularly important, if detected under such unfavorable conditions.

The B₁R gene promoter polymorphism (G⁶⁹⁹  C) has been discovered recently and has been found to exhibit an altered frequency in patients with a history of end stage renal failure (Bachvarov et al., 1998b). A deficit of the C allele was observed in the patient group and in some etiologic subgroups, in comparison with healthy volunteers. This has been independently confirmed using a familial approach in a Polish population (Zychma et al., 1999). A similar and even more profound alteration of allele frequencies was found in patients with inflammatory bowel disease (Bachvarov et al., 1998a). Thus, the human B₁R gene may influence the clinical course or the complications of inflammatory processes affecting different organs, in line with the inflammatory effects of kinins. A 821-bp promoter fragment of the B₁R gene containing the C⁶⁹⁹ allele mediated a 40% increase in the expression of a reporter gene in transfected IMR-90 human fibroblasts, compared with the same fragment with the G⁶⁹⁹ allele (Bachvarov et al., 1998b). A recent study applying similar approaches (B₁R promoter fragments fused to a reporter gene and transiently transfected in IMR-90 cells) has identified a positive control element (nucleotides 733 to 685) that includes the polymorphic site (Yang et al., 1998). Induction of the B₁R by interleukin-1 and lipopolysaccharide in rat vascular smooth muscle may rather be conferred by the promoter region proximal to the TATA box and transcription initiation site (Ni et al., 1998). Our results do not indicate that the C allelic variant determines an altered function of the B₁Rs induced in human vascular smooth muscle by postisolation incubation. As mentioned above, there is significant evidence that the B₁R gene promoter functions in a cell-specific manner, and the influence of the promoter C/G polymorphism on B₁R regulation in various cell types capable of expressing these receptors (smooth muscle cells, endothelial cells, enterocytes, lymphocytes, etc.; Marceau et al., 1998) is currently unknown. In addition, protein-DNA interactions were not detected in the immediate vicinity of the clinically significant B₁R promoter polymorphism (G⁶⁹⁹  C) based on the sensitive technique ligation-mediated polymerase chain reaction applied to IMR-90 or human vascular smooth muscle cells (Angers et al., 2000). Thus, the protective effect of the C allele relative to the incidence of some pathologies may not be associated with an altered function of the B₁Rs in human vascular smooth muscle, or the functional difference may be too small to be detected by the method applied. Similarly, the B₂R promoter polymorphism, recently found to be predictive for essential hypertension in populations of Japanese and African Americans (the T allele is protective; Mukae et al., 1999; J. V. Gainer, N. Brown, M. Bachvarova, L. Bastien, I. Maltais, F. Marceau, and D. R. Bachvarov, submitted), appears to be neutral when the contractile response of the umbilical vein to BK is considered in our study based on Caucasian tissue donors (despite a trend for a decreased EC₅₀ for the TT genotype).

The exon 1 polymorphism of the B₂R gene has been claimed to be of clinical significance (see introduction; Lung et al., 1997). Surprisingly, we have found that the "pathogenic" allele () under the homozygote form was associated with an increased contractile effect of the B₁R agonist, but exerted no effect on contraction caused by the B₂R agonist. The only significant difference, (/) versus (+/), may be a consequence of larger samples (25 versus 50 individuals), as the comparison between the (+/+) and (/) genotypes would suffer from more statistical uncertainty (25 individuals in each sample). It is particularly clear in this case that the polymorphism is not causal, but it rather may be in linkage disequilibrium with a functionally important polymorphism of the B₁R gene. This explanation is plausible because both kinin receptor genes lie close to each other in the same chromosomal region (14q32; Ma et al., 1994; Chai et al., 1996; Bachvarov et al., 1998b). Previous analyses have failed to detect linkage disequilibrium between the B₂R exon 1 (+/) polymorphism and the only suspected functionally important B₁R polymorphism (the promoter AciI restriction fragment length polymorphism; Bachvarov et al., 1998a). Therefore, the (+/) polymorphism may rather be in linkage disequilibrium with an unknown polymorphism of the B₁R gene that determines an altered function in smooth muscle. It is intriguing whether the interaction of the B₂R exon 1 polymorphism with hereditary angioedema can be dependent on a functional alteration of the inducible B₁R, as the presentation of this disease consists of isolated crises.

The only coding sequence polymorphism represented in this study affects exon 2 of the B₂R. This polymorphism determines the Arg¹⁴  Cys substitution (Braun et al., 1995) and has a modest effect on the potency of BK under the heterozygote form. The affected N-terminal extracellular domain coded in exon 2 of the gene is not believed to be important for B₂R function, as this domain can be truncated without loss of affinity for [³H]BK or alteration of signal transduction (Abd Alla et al., 1996). However, these data support a functional alteration selective for the B₂R agonist BK, perhaps a loss of affinity, when Cys¹⁴ is present. Interestingly, the T allele has been recently proposed to protect in a dominant manner from the development of end stage renal failure (Zychma et al., 1999).

In conclusion, based on the methods used, most of the frequent polymorphic alleles of the kinin receptor genes appear to be functionally neutral, or determine function alterations that are too subtle to detect using a human vascular smooth muscle preparation. One of the positive findings of this study, the (+/) insertion/deletion of B₂R exon 1 predicting the response to the B₁R agonist, supports the idea that both kinin receptors are clustered in the same genetic locus. The finding of two polymorphisms that apparently influence the responses to kinins suggests that part of the pharmacodynamic variability of the human umbilical vein can be attributed to the genetic background. These results should be confirmed using larger populations and alternate approaches (e.g., radioligand binding). If novel polymorphisms are discovered in the future in the kinin receptor gene locus, the matched DNA and pharmacologic data banks established by this study might be used again to assess their functional importance.