Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The nonapeptide bradykinin (H-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-OH; BK) and related kinins, such as kallidin (Lys-BK), are produced by the catalytic action of kallikrein enzymes on plasma and tissue precursors termed kininogens (Bhoola et al., 1992). Kinins are potent autacoid peptides involved in various processes of physiological and pathological relevance, such as smooth muscle contraction, vasodilation, increased vascular permeability, recruitment of inflammatory cells, induction of pain and cell division (Hall, 1992).

Kinins produce their effects in mammals by stimulating two distinct receptor types, termed B₁ and B₂ receptors (Regoli and Barabé, 1980), both of which have been cloned (Eggerickx et al., 1992, Hess et al., 1992, Menke et al., 1994). The B₁ kinin receptor has a low level of expression in normal tissues but undergoes a marked up-regulation (de novo synthesis) during tissue trauma/inflammation (Marceau, 1995). On the other hand, the B₂ kinin receptor is constitutively expressed by a variety of cell types, (Regoli and Barabé, 1980) and is ready to transduce the signals delivered by newly formed kinins in plasma and tissues.

Owing to the putative role of kinins in mediating pain and inflammation, there is a remarkable interest in developing potent and selective kinin receptor antagonists. These efforts have been especially successful in the case of the B₂ kinin receptor. A first generation of B₂ receptor antagonists was obtained through the insertion of D-Phe at position 7 of the BK sequence (Vavrek and Stewart, 1985). In the early 1990s, the introduction of nonnatural amino acids led to the discovery of a very potent linear peptide antagonist Hoe 140 or Icatibant ([D-Arg⁰,Hyp³,Thi⁵,DTic⁷,Oic⁸]-BK) (Hock et al., 1991; Wirth et al., 1991). Because of its potency, metabolic stability and suitability for in vivo investigations, Icatibant has been an instrumental tool in assessing the pathophysiological role of B₂ kinin receptors in both animals and humans. In particular, clinical studies with Icatibant have proven a role of the endogenous kinins, acting via B₂ receptors in regulating coronary circulation (Groves et al., 1995) and in the pathophysiology of allergic reactions (Austin et al., 1994), rhinitis (Proud et al., 1995), and asthma (Akbary et al., 1996). Potent analogs of Icatibant, such as the linear peptide antagonist B9430 ([D-Arg⁰,Hyp³,Igl⁵,DIgl⁷,Oic⁸]-BK), have also been developed by other groups (Stewart et al. 1996). More recently, nonpeptide B₂ kinin receptor antagonists have also been discovered, such as WIN64338 ({[4-[[2[[bis(cyclohexylamino) methylene]amino]-3-(2-naphthale nyl)-1-oxopropyl]amino]phenyl]methyl]tributyl phosphonium chloride HCl} (Salvino et al., 1993, Sawutz et al., 1994) and, more recently, FR173657 [(E)-3-(6-acetamido-3-pyridyl)-N-[N-[2,4-dichloro-3-[(2-methyl-8-quinolinyl)oxymethyl]phenyl]-N-methylaminocarbonylmethyl]acrylamide] (Aramori et al., 1997; Asano et al., 1997) and LF 16.0335 (1-[[3-[2,4-dimethylquinolin-8-yl)oxymethyl]-2,4-dichloro-phenyl]sulfonyl]-2(S)-[[4-[4-aminoiminomethyl)phenyl-carbonyl]piperazin-1-yl]carbonyl]pyrrolidine) (Pruneau et al., 1998).

Although the discovery of nonpeptide ligands offers some advantages in the development of kinin receptor antagonists suitable for treatment of human diseases, the unprecedented level of potency of Icatibant for the human B₂ kinin receptor has not yet been achieved with nonpeptide ligands. Therefore, a proper structural knowledge of the bioactive conformation of Icatibant could be of great help in the rational design of novel nonpeptide kinin receptor antagonists. Interestingly, it has been proposed that the C-terminal tetrapeptide of Icatibant undergoes a -turn arrangement for the interaction with B₂ kinin receptors (Guba et al., 1994). With the aim of probing this hypothesis, we designed a series of Icatibant analogues in which the C-terminal region of the molecule has been cyclized to constrain it to the -turn conformation. In the present study, we describe the pharmacological profile of MEN 11270 or H-D-Arg-Arg-Pro-HypGly-Thi-c(Dab-DTic-Oic-Arg)c(7-10), a cyclized analog of Icatibant, by performing radioligand binding studies and comparing its affinity and selectivity at the human B₂ kinin receptor with that of a number of other ligands (agonists and antagonists). The antagonist activity of MEN 11270 was assessed in the human umbilical vein, an in vitro preparation which is suitable for functional characterization of B₁ and B₂ kinin receptor antagonists (Gobeil et al., 1996).

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. ³H-BK (specific activity, 114 Ci·mmol¹) and ³H-Lys-[des-Arg⁹]-BK (specific activity, 92 Ci·mmol¹) were provided by DuPont NEN (Hertfordshire, UK). BK, Lys-[des-Arg⁹]-BK, and Lys-[des-Arg⁹,Leu⁸]-BK were obtained from Peninsula (St. Helens, UK), whereas [Hyp³,Tyr(Me)⁸]-BK and [des-Arg⁹]-BK were obtained from Novabiochem (Läufelfingen, Switzerland). All B₂ receptor antagonists used were synthesized at Menarini Ricerche (Florence, Italy). Leupeptin was obtained from Boehringer Mannheim (Mannheim, Germany) and DL-2-mercaptomethyl-3-guanidoethylthiopropanoic acid was obtained from Calbiochem (La Jolla, CA). L-glutamine, trypsin plus EDTA, and HEPES were obtained from Gibco Ltd. (Paisley, Scotland). GF/B glass fiber filtermats were provided by Brandel (Semat, Gaithersburg, MD). All other material and culture reagents were obtained from Sigma (St. Louis, MO).

Cell Culture and Membrane Preparation. WI38 fibroblasts (American Type Culture Collection, Rockville, MD) were used up to passage number 23 and were grown to confluence in minimum essential Eagle's medium with 0.1 mM nonessential amino acids, 1.0 mM sodium pyruvate, 2 mM L-glutamine, and 10% fetal bovine serum. Cells were cultured in 175-cm² flasks and maintained in a humidified atmosphere at 37°C with 5% CO₂. The cells were subcultured every 4 to 6 days at a ratio of 1:2 by using 0.25% trypsin and 1 mM EDTA to detach them.

1

1

Binding Experiments. The buffer used for binding experiments was N-tris[hydroxymethyl]methyl-2-aminoethanesulfonic acid (10 mM, pH 7.4) containing 1,10-phenanthroline (1 mM), bacitracin (140 µg·ml¹), and BSA (1 g·l¹). The binding assay was performed in polypropylene tubes using a final volume of 0.5 ml. Nonspecific binding was defined as the amount of labeled ligand bound in the presence of 1 µM Lys-[des-Arg⁹]-BK or BK. Membrane concentrations of 60 µg·ml¹ and 30 µg·ml¹ were chosen for binding experiments with ³H-Lys-[des-Arg⁹]-BK and [³H]-BK, respectively. At these protein concentrations, the specific binding was approximately 75 to 90% of the total binding for ³H-Lys-[des-Arg⁹]-BK and 80 to 90% of the total binding for ³H-BK. Two to 3% of the total added radioactivity was bound to the membranes for each radioligand. An incubation time of 60 min at 4°C (unless stated otherwise) was used for both radioligands throughout the study (saturation and competition studies) (Phagoo et al., 1996). With every batch of membranes, a saturation curve to the tritiated agonist was performed to allow for a correct elaboration of competition studies. The chosen radioligand concentration was 0.3 nM both for B₁ and B₂ radioligands.

Analysis of Binding Data. Saturation and competition data were processed according to the method of Munson and Rodbard (1980), initially by means of the EBDA software and then by LIGAND, to determine the maximum binding site density (B_max), affinity constant (K_D), equilibrium inhibition constants (K_i) and to test the significance of the binding site models. All values are given as mean ± S.E.M. for the stated number of experiments (n). Each experimental determination was obtained in duplicate on a different batch of membrane preparation.

Human Isolated Umbilical Vein. Human umbilical cords (n = 13) from healthy women, 22 to 40 years old, were collected after spontaneous delivery at term and placed in cold (4°C) Krebs' solution. The lapse of time between the delivery and the experiment was on average 4 h (1-10 h). In the laboratory, the middle segment of the cord (7-8-cm long) was placed in Krebs' solution at room temperature and, within 30 min, the vein was dissected free of surrounding tissue. The endothelium was rubbed off using a cotton swab. The tissues were cut into spiral strips (2-cm long, 3-mm wide) and suspended in 10-ml organ baths containing warm (37°C), oxygenated (95% O₂, 5% CO₂) Krebs' solution (118 mM NaCl , 4.7 mM KCl , 2.5 mM CaCl₂ , 0.5 mM MgCl₂, 1.2 mM KH₂PO₄, 1.2 mM MgSO₄, 25 mM NaHCO₃, and 10 mM glucose). The human umbilical vein strips were stretched to a resting tension of 2g. Changes in tension were measured isometrically with Grass FT03 force transducers (Grass Instruments, Quincy, MA) and recorded on a Linseis (model L2005) multichannel chart recorder (Linseis Inc., Princeton Jct., NJ). In all experiments, the kininase II inhibitor captopril was added to the Krebs' solution at a 1 µM concentration. Before testing any agent, the tissues were allowed to stabilize for 120 to 150 min, resting tension was readjusted every 15 min. The experiments began with the application of 100 mM KCl to measure the responsiveness of the preparations. The antagonist activity of MEN 11270 was evaluated by measuring the cumulative concentration-response curves to BK or Lys-[des-Arg⁹]-BK in the absence and presence of the antagonist. The antagonist was applied 15 min before the agonist.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Binding Studies at the Human Kinin B₂ Receptor. ³H-BK (0.02-5 nM) saturation isotherm was fitted according to a one-site model with a K_D of 123 ± 17 pM (Hill slope, 0.92 ± 0.03, n = 7) and a B_max of 485 ± 78 fmol·mg¹ of protein. In competition experiments, the binding of a panel of kinin receptor-selective agonists and antagonists indicated that the binding sites labeled by ³H-BK correspond to a classic B₂ kinin receptor (Fig. 2a; Table 1). In fact, a high-affinity binding was detected with BK and the selective B₂ kinin receptor agonist [Hyp³,Tyr(Me)⁸]-BK, whereas only a negligible binding affinity was measured for [des-Arg⁹]BK and Lys-[des-Arg⁹]-BK.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Chemical structure of MEN 11270 [H-D-Arg-Arg-Pro-Hyp-Gly-Thi-c(Dab-DTic-Oic-Arg)c(7-10)].

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Inhibition of 3H-BK binding to WI38 membranes by a range of B1- and B2-selective ligands. Specific 3H-BK binding was plotted against increasing concentration of agonists (a): BK (), [Hyp3, Tyr(Me)8]-BK (), Lys-[des-Arg9]-BK (), and [des-Arg9]-BK (); antagonists (b): Icatibant (), MEN 11270 (), B9430 (), B9858 (×), FR173657 (), and WIN 64338 (*). [3H]-BK (0.3 nM) and WI38 membranes (30 µg·ml1) were incubated as described in Experimental Procedures in the presence of increasing concentrations of displacer. The data shown represent the mean ± S.E.M. of at least three separate determinations in duplicate.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Icatibant (a), MEN 11270 (b), and FR173657 (c) displacement curves of the 3H-BK binding (0.3 nM) to WI38 membranes performed at 4°C () or 20°C (]). The data shown represent the mean ± S.E.M. of at least three separate determinations in duplicate.

Binding Studies at the Human Kinin B₁ Receptor. ³H-Lys[des-Arg⁹]-BK (0.03-3 nM) bound to WI38 cell membranes in a specific and saturable manner. The Scatchard transformation of the specific binding was consistent with a one-site binding model. The experiments indicated a K_D value of 175 ± 40 pM (n = 5) with a B_max of 104 ± 8 fmol·mg¹ of protein and a Hill slope of 1.02 ± 0.05.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Inhibition of 3H-Lys-[des-Arg9]-BK binding to WI38 membranes. Specific 3H-Lys-[des-Arg9]-BK binding was plotted against increasing concentrations of Lys-[des-Arg9]-BK (), Lys-[Leu8,des-Arg9]-BK (), Icatibant (), MEN 11270 (), and BK (). 3H-Lys-[des-Arg9]-BK (0.3 nM) and WI38 membranes (60 µg·ml1) were incubated as described in Experimental Procedures in the presence of increasing concentrations of displacer. The data shown represent the mean ± S.E.M. of three separate determinations in duplicate.

Receptor Selectivity and Ion Channel-Binding Affinity. The affinity of MEN 11270 was determined for many receptors and ion channels by using conventional radioligand-binding techniques. MEN 11270 showed no relevant binding affinity (pIC₅₀ < 5.5) at the angiotensin 1, angiotensin 2, calcitonin gene-related peptide, interleukin-1, tumor necrosis factor , chemokine receptor type 1, chemokine receptor type 2, neurokinin 1, neurokinin 2, neurokinin 3, neuropeptide Y, thromboxane A₂, prostaglandin I₂, opioid-like receptor 1, serotonin, muscarinic, neurotensin, somatostatin, bombesin, and vasoactive intestinal peptide receptors. MEN 11270 had no measurable binding affinity (pIC₅₀ < 5.5) for Ca²⁺ channels (L-type: dihydropiridine, verapamil, diltiazem sites; N-type), K⁺ channels (ATP-, voltage-, and Ca²⁺-dependent), Na⁺ channel (site 2), or Cl (picrotoxinin) channel.

Organ Bath Studies. BK (0.001-100 nM) produced concentration-dependent contraction of the human umbilical vein with a pEC₅₀ value of 8.3 (cf. 95% CLs, 8.0-8.5; n = 7). The administration of MEN 11270 (10-100-1000 nM, contact time 15 min) did not produce any motor effect per se but, in the presence of MEN 11270, a concentration-dependent rightward shift of the curve to the agonist was observed, without depression of E_max (Fig. 5a). The Schild plot was compatible with competitive antagonism with an extrapolated pA₂ value of 8.14 ± 0.22 (n = 7) with the slope being not significantly different from unity (0.95 ± 0.11; Fig. 5b).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   a, concentration-response curves to BK in the human umbilical vein in the absence and presence of various concentrations of MEN 11270: BK (), 10 nM MEN 11270 (), 100 nM MEN 11270 (), and 1000 nM MEN 11270 (). Contact time of the antagonist was 15 min. b, Schild plot of MEN 11270 against BK. Abscissa, log of the molar concentration of the antagonist. Ordinate, log of the concentration ratio 1 (CR  1) of the agonist. Values represent the mean ± S.E.M. of seven experiments.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Concentration-response curves to Lys-[des-Arg9]-BK in the human umbilical vein in the absence () and presence () of 10 µM MEN 11270. Contact time of the antagonist was 15 min. Values represent the mean ± S.E.M. of five experiments.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The linear decapeptide Icatibant, a potent and selective antagonist at the B₂ kinin receptor, has been previously studied by combining NMR spectroscopy and molecular dynamic simulations to define its conformational behavior into a hydrophilic/hydrophobic environment (Guba et al., 1994). The results of this analysis had suggested the presence of a II'-turn motif in the C-terminal part of the sequence involving the Ser-DTic-Oic-Arg segment (Guba et al., 1994). This conformational feature was defined as being critical for the high-affinity interaction of Icatibant with the receptor (Guba et al., 1994; Kyle, 1995). In our previous studies in the field of tachykinins, we successfully used a 14-membered cyclic lactam to constrain a -turn conformation in a hexapeptide sequence (Pavone et al., 1995). To apply a similar strategy to Icatibant, we designed a constrained analog in which the Ser-DTic-Oic-Arg segment was fixed in its turn conformation throughout the formation of an intramolecular lactam bridge. To obtain the 14-membered ring, we substituted the serine residue with a diaminobutiric (Dab) residue, since the serine side chain of Icatibant is not critical for high affinity at the human B₂ kinin receptor (L. Quartara, R. Ricci, S. Meini, R. Patacchini, A. Giolitti, S. Amadesi, C. Rizzi, A. Rizzi, K. Varani, P. A. Borea, C. A. Maggi, and D. Regoli, in preparation). Cyclization was performed between the carboxyl group of the C-terminal arginine residue and the side chain of the Dab residue. The resulting compound, MEN 11270 [H-D-Arg-Arg-Pro-Hyp-Gly-Thi-c(Dab-DTic-Oic-Arg)c(7-10), Fig. 1], is a potent and selective antagonist at the human B₂ kinin receptor.

WI38 human fetal lung fibroblasts constitutively express both B₁ and B₂ kinin receptors (Phillips et al., 1992; Webb et al., 1994; Phagoo et al., 1996) and are representative of several of the effects produced by BK, which acts almost exclusively at the B₂ kinin receptor type. It has been reported that activation of the B₂ receptor in these cells leads to a tyrosine kinase activity, which is involved in prostaglandin production (Jong et al., 1993) and induces interleukin-1 expression (Pan et al., 1996).

We used membranes of WI38 cells to probe the affinity and selectivity of MEN 11270 at both the kinin receptor types, compared to that of other peptide and nonpeptide ligands.

With few exceptions, the affinity values obtained in our binding experiments with several peptide and nonpeptide antagonists are comparable to the values reported in other studies.

Phagoo et al. (1996) reported a K_i of 0.021 nM for Icatibant, corresponding to a pK_i of 10.6 in competing the ³H-BK binding. In a different fibroblast cell line (CCD-16), Icatibant inhibited the binding of the radioligand B₂ receptor antagonist ³H-NPC17731 with a comparable affinity (K_i of 0.05 nM, pK_i 10.3) (Zhang and Codd, 1998).

Cortech researchers have described two antagonists, B9430 and B9858, the first one possessing mixed activity at both the B₁ and B₂ receptors and the second one selective for the B₁ receptor (Burkard et al., 1996; Stewart et al., 1996). These authors obtained a pIC₅₀ of 9.6 and 7.9 for B9430 and B9858, respectively, in inhibiting the ³H-BK binding to cells transfected with the human B₂ receptor, and apparent pA₂ values of 8.6 for B9430 and <5 for B9858 in the human ileum assay. Our data confirm the affinity data presented previously by these authors. It was previously reported that Icatibant had a pA₂ amounting to 8.36 in the same bioassay (Zuzack et al., 1996), comparable to the pA₂ value obtained in the human umbilical vein by Gobeil et al. (1996).

FR 173657 has been reported to inhibit with high affinity (IC₅₀, 2.9 and 8.9 nM, respectively) the binding of ³H-BK to membranes of IMR90 fibroblasts and Chinese hamster ovary cells transfected with cDNA coding for the human B₂ kinin receptor, respectively (Aramori et al., 1997; Asano et al., 1997). In the initial set of experiments, performed at 4°C, the estimated affinity of FR 173657 for the B₂ kinin receptor in WI38 cells was surprisingly low (pK_i 7.6) as compared to the values reported by Aramori et al. (1997) or by Asano et al. (1997). However, these authors did not specify at which temperature they performed the assay. We therefore repeated the experiments at 20°C and, quite surprisingly, observed that the binding affinity of FR 173657 for the B₂ kinin receptor was increased by about 1 log unit when the incubation temperature was raised to 20°C. Our results are comparable to those obtained by Gessi et al. (1997) who, working at room temperature and after an incubation time of 60 min, reported a pK_i of 8.7 for FR173657 in inhibiting the ³H-BK binding in the human umbilical vein membranes.

It is intriguing to note that the estimate of the affinities of the two peptide antagonists, MEN 11270 and Icatibant, was unaffected by the increase in the binding temperature. This observation suggests a different type of interaction with the human B₂ kinin receptor by peptide and nonpeptide ligands. However, the increase in affinity for FR173657, following an increase in the temperature of binding conditions, is unique of this molecule; in fact, the affinity of the other nonpeptide antagonist WIN64338, which also has quite a low affinity at the human B₂ kinin receptor, did not change if measured at 4 and 20°C (data not shown). The fact that FR 173657 better inhibits BK from its binding site at a higher temperature could indicate the importance of thermodynamic factors which enable this nonpeptide ligand to achieve the proper bioactive conformation for its interaction with the B₂ kinin receptor. On the other hand, it is also known (Testa et al., 1987) that a ligand binding in which nonpolar or lipophilic groups lose their contact with water molecules to form new hydrophobic interactions are favored by an increase in temperature. In this case, an entropy-driven increase in the binding of FR 173657, which, contrary to WIN 64338, is a lipophilic molecule with no charged groups, may be facilitated at higher temperatures. We cannot rule out the possibility that, at low temperature, FR 173657 selects a G protein-uncoupled conformer of the B₂ receptor which is different from that selected by FR 173657 at higher temperature. Experiments with a radiolabeled form of this nonpeptide antagonist are needed to answer these questions.

It is interesting to note that, in the human isolated umbilical vein, both peptide and nonpeptide antagonists possess a comparable competitive antagonist activity: Icatibant was previously reported to have an apparent pA₂ of 8.42 (Gobeil et al., 1996), FR173657 a pA₂ of 8.22 (Rizzi et al., 1997), and MEN 11270 has a pA₂ of 8.14 (this study). Notably, the apparent pA₂ value of FR173657 in human umbilical vein assay is in good keeping with the pK_i value estimated for this ligand in binding at WI38 cell membranes, whereas the pK_i values of Icatibant and MEN 11270 are at least one order of magnitude higher than their apparent pA₂ values. The reasons for this discrepancy, which selectively affects the peptide antagonists, are not clear at present, although the existence of a binding paradox (higher binding affinities of kinin receptor antagonists versus their apparent affinity in functional assay) has been repeatedly reported in the literature (Hall, 1992; Ransom et al., 1992; Burkard et al., 1996; Gobeil et al., 1996; Gessi et al., 1997; Rizzi et al., 1997). Notably the binding paradox (as defined above) reported to occur in some systems for bradykinin has been shown to depend at least in part from the ionic strength of the medium used for binding experiments (Ransom et al., 1992). Whether a similar factor could be involved in the quantitative discrepancies between K_i and pA₂ values reported in the present study remains to be established.

A low but sizable affinity of Icatibant at the human B₁ kinin receptor has already been observed (Menke et al., 1994; Burkard et al., 1996; Phagoo et al., 1996; Bastian et al., 1997). The fact that MEN 11270 maintains a comparable affinity to the B₁ kinin receptor indicates that the postulated C-terminal -turn arrangement of Icatibant is also compatible with an interaction of these ligands at the B₁ receptor.

Our results confirm a very low affinity of the BK derivative lacking of the C-terminal arginine as compared to the Lys-[des-Arg⁹]-BK, as already reported at both the native or cloned human B₁ kinin receptor (Menke et al., 1994; Austin et al., 1997; Bastian et al., 1998).

In conclusion, in the present study we have shown that the pharmacological profile of MEN 11270 is comparable to that of Icatibant, supporting the idea that a constriction of the C-terminal sequence in a -turn arrangement is able to preserve a high-affinity for interaction with the human B₂ kinin receptor.

The discovery of a constrained peptide molecule, such as MEN 11270, acting as a high-affinity antagonist at the B₂ kinin receptor is a helpful starting-step for designing new low-molecular weight molecules.