Structure activity relationships for bradykinin antagonists on the inhibition of cytokine release and the release of histamine☆1

Peptides 21 (2000) 527–533

Structure activity relationships for bradykinin antagonists on the inhibition of cytokine release and the release of histamine夞 Siegmund Reissmanna,*, Felipe Pinedaa, Gabriele Vietinghoffb, Heinz Wernerb, Lajos Gerac, John M. Stewartc, Inge Paegelowb a

Institute of Biochemistry and Biophysics, Biological Faculty, Friedrich-Schiller University, 07743 Jena, Germany Institute of Experimental and Clinical Pharmacology and Toxicology, University of Rostock, 18057 Rostock, Germany c Department of Biochemistry, University of Colorado School of Medicine, Denver, CO 80262, USA

b

Received 11 August 1999; accepted 13 January 2000

Abstract Highly potent bradykinin antagonists were found to inhibit bradykinin-induced release of cytokines but to stimulate histamine release. Both actions show structural requirements completely different from those for bradykinin B1 and B2 receptors, indicating that the release of some cytokines from spleen mononuclear cells and of histamine from rat mast cells is not mediated by these receptors. Most potent bradykinin antagonists release histamine at lower concentrations than does bradykinin itself. Dimers of bradykinin antagonists are the most potent compounds for histamine release. In contrast to enhanced histamine release, potent inhibition of cytokine release enhances the applicability of these compounds as anti-inflammatory drugs. Many of the peptides designed for high B2-receptor antagonism were found to be compared by their concentrations far more potent for inhibition of cytokine release than for smooth muscle contraction. Thus, for some antagonists inhibition of cytokine release was detected at concentrations as low as 10⫺15 M. The rational design of peptide and nonpeptide bradykinin antagonists for therapeutic use requires not only knowledge about the potency but also knowledge about the structure–activity relationships of such important side effects as cytokine and histamine release. © 2000 Elsevier Science Inc. All rights reserved. Keywords: Bradykinin receptor antagonists; Inhibition of cytokine release; Histamine release; Mast cells; Structure activity relationships; Spleen mononuclear cells; Tanned erythrocyte electrophoretic mobility (TEEM) test; QSAR

1. Introduction Bradykinin receptor antagonists are of great interest as potential drugs for treatment of inflammation processes Abbreviations: Aca, adamantane carboxyl-; bApG, N,N-bis(3-aminopropyl)-glycine; BK, bradykinin; Chg, ␣-cyclohexylglycine; Cpg, ␣-cyclopentylglycine; DDD, dodecanedioyl; EGS, ethylene glycol bis-succinyl-; Dhq, 2-dehydroquinuclidine-3-carboxyl; f5f, pentafluorophenylalanine; Gun guanidyl-; Hyp, trans hydroxyproline; Igl, ␣-(2-indanyl)glycine; Mosi, methoxy-suberimido; MPIV, 2.4-methanoproline; Nchg, N-cyclohexylglycine; Nc7g, N-cycloheptylglycine; Nig, N-(2-indanyl)glycine; NMF, Nmethylphenylalanine; Oic, octahydroindole-2-carboxylic acid; Sub, suberyl:-CO-(CH2)6-CO-; Suim, suberimidyl: -C(¢NH)-CH2)6-C(¢NH)-; Thi, ␤-(2-thienyl)-alanine; Tic, 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid. 夞 This work was supported by the Deutsche Forschungsgemeinschaft (Re-853/2-1) and US NIH grant (HL-26284). * Corresponding author. Tel.: ⫹49-3641-949-350; fax: ⫹49-3641-949352. E-mail address: [email protected] (S. Reissmann).

such as allergic rhinitis, septic shock, trauma, and bronchitis and asthma [28], and in the therapy of some kinds of cancer [8]. Since the first antagonist was published in 1985, new generations were developed with antagonistic potencies enhanced by many orders of magnitude [16,30] and with increased stability against enzymatic degradation [32]. In addition to peptide antagonists, nonpeptide bradykinin receptor antagonists have been developed [1–5,9,17] in recent years. Because of the lack of a rationale for the conversion of bradykinin agonists into antagonists, we started a systematic search for new antagonist structures by amino acid replacements at different sequence positions. We found two new lead structures for antagonists, both types without any amino acid replacement at position 7, one type with replacement of proline at position 2 by N-methyl-phenyl-alanine [25,26], and a second type with replacement of phenylalanine at position 5 by dehydrophenylalanine [14]. In some of

0196-9781/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved. PII: S 0 1 9 6 - 9 7 8 1 ( 0 0 ) 0 0 1 7 1 - 6

528

S. Reissmann et al. / Peptides 21 (2000) 527–533

Table 1 Effect of bradykinin and analogs on histamine release and on inhibition of cytokine release Number Code

Structure ⴚ1

Histamine-release TEEM-test (M) GPI EC50 (␮M) (n) Inhibition pA2

1 2 3 4

Bradykinin J30 J201 HOE-140

1 Arg Arg Arg DArg Arg

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

B9340 B10004 B10008 B10010 B10036 B10038 B9430 B10206 B9330 B9966 B9992 B9976 B9984 B9986 B10042 B10150 B9978 B10214 B10044 B8994

DArg DArg DArg DArg Arg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg DArg

Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg

25 26 27 28 29 30 31 32

B9560 B9562 B9870-1 B9870-2 B10052 B10054 B10084 B10092

DArg DArg DArg DArg DArg DArg DArg DArg

Arg Arg Arg Arg Arg Arg Arg Arg

33 34 35 36 37

Bradykinin (1–8) B9066 B10046 B10050 Lys B9858 Lys

DArg DArg Lys Lys

Arg Arg Arg Arg Arg

Aca Dhg Mosi SUIM ␧-Sub(Lys) ␣-DDD(Lys) bApG (Gun)2-bApG

0

2 3 4 5 6 7 8 Pro Pro Gly Phe Ser Pro Phe Pro Pro Gly Phe Ser DPhe Phe NMF Pro Gly Phe Ser Pro Phe Pro Hyp Gly Thi Ser DTic Oic Peptides with additional modifications Pro Hyp Gly Thi Ser DIgl Oic Pro Hyp Gly Igl Ser DTic Nc7G Pro Hyp Gly Thi Ser DIgl Nc7G Pro Hyp Gly Igl Ser DIgl Nc7G Pro Hyp Gly Igl Ser DTic Nc6G Pro Hyp Gly Igl Ser DTic Nc6G Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DF5F Nc7G Pro Hyp Gly Thi Ser DTic Nig NMF Hyp Gly Thi Ser DIgl Oic Oic Hyp Gly Thi Ser DTic Oic Igl Hyp Gly Igl Ser DIgl Oic DLOic Hyp Gly Igl Ser DIgl Oic Oic Pro Gly Igl Ser DIgl Oic Oic Hyp Gly Thi Ser DTic Nc6G NMF Hyp Gly Igl Ser DF5F Oic Pro Igl Gly Igl Ser DIgl Oic Pro Igl Gly Igl Ser DF5F Oic Pro Hyp Gly Igl Ser DF5F Oic Pro MeP Gly CpG Ser DCpG CpG Acylated analogs and dimers Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic Pro Hyp Gly Igl Ser DIgl Oic [des Arg9]-Analogs Pro Pro Gly Phe Ser Pro Phe Pro Hyp Gly CpG Ser DTic CpG Pro Hyp Gly Igl Ser DTic ChG Pro Hyp Gly Igl Ser DTic ChG Pro Hyp Gly Igl Ser DIgl Oic

9 Arg Arg Arg Arg

68.5 ⫾ 11.4 (4) 19.4 ⫾ 4.1 (5) 16.6 ⫾ 6.3 (7) 50.1 ⫾ 22.3 (6)

10⫺10 10⫺12 10⫺12

5.9 Agonist 8.4

Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg Arg

22.4 ⫾ 8.7 (3) 7.8 ⫾ 2.0 (3) 21.0 ⫾ 8.0 (3) 3.9 ⫾ 2.7 (3) 10.7 ⫾ 3.2 (3) 8.2 ⫾ 1.7 (3) 17.0 ⫾ 5.2 (3) 0.53 ⫾ 0.39 (3) 38.7 ⫾ 5.7 (3) 5.5 ⫾ 1.4 (3) 6.3 ⫾ 1.8 (3) 4.4 ⫾ 2.8 (3) 2.7 ⫾ 0.3 (3) 6.0 ⫾ 1.1 (3) 3.97 ⫾ 0.93 (3) 0.99 ⫾ 0.70 (3) 0.65 ⫾ 0.4 (3) 0.48 ⫾ 0.19 (4) 17.0 ⫾ 5.2 (3) 39.8 ⫾ 18.8 (4)

10⫺14 10⫺15 10⫺13 10⫺15 10⫺10 10⫺15 10⫺15 10⫺14 10⫺13 10⫺14 10⫺14 10⫺13 10⫺10 10⫺14 10⫺11 10⫺15 10⫺15 10⫺15 10⫺15 10⫺14

7.9 7.4 7.1 7.7 7.8 7.7 7.9 7.8 7.8 6.9 7.5 6.8 6.8 8.1 7.8 7.9 5.0 8.1 8.4 6.5

Arg 8.2 ⫾ 3.8 (3) Arg 6.0 ⫾ 1.0 (4) Arg 5.3 ⫾ 1.5 (3) Arg)2 0.09 ⫾ 0.05 (3) Arg)2 0.04 ⫾ 0.01 (3) Arg)2 0.022 ⫾ 0.007 (3} Arg 1.42 ⫾ 0.53 (3) Arg 0.54 ⫾ 0.31 (3)

10⫺13 10⫺11 10⫺14 10⫺13 10⫺10 10⫺14 10⫺14 10⫺14

7.8 7.6 8.0 7.6 7.3 7.2 8.1 8.7

— — — — —

10⫺11 10⫺12 10⫺15 10⫺12 10⫺13

— I(P) 5.3 5.1 5.6

No release No release 14.7 ⫾ 4.5 (3) 0.57 ⫾ 0.16 (3) 3.9 ⫾ 2.8 (3)

Residues in bold type are different from the native bradykinin sequence. pA2 is the negative logarithm to base 10 of the molar concentration of antagonist that makes it necessary to double the concentration of BK needed to elicit the original submaximal response [18].

the analogs listed in Table 1, the replacement of proline at position 7 by hydrophobic D-amino acids was combined with replacement of proline at position 2 by N-substituted amino acids. Additionally, amino acids at other positions were replaced to enhance as well as biologic potency for the B2- or B1-receptor and enzymatic stability. Histamine and some cytokines are compounds that strongly activate inflammation processes. Because bradykinin releases cytokines from mononuclear cells [23] and histamine from rat mast cells [31], it is important to know the effects of bradykinin antagonists on both of these releasing actions. Thus, the histamine release could lead to undesirable side effects. Knowledge about structure–activ-

ity relationships for both actions, the inhibition of the cytokine release, and the histamine release should help development of bradykinin antagonists as therapeutics. It seems to be of importance to develop compounds without histamine releasing activity. In this article we describe the effects of a group of potent bradykinin antagonist analogs on histamine and cytokine release. Structure–activity relationships for these actions were compared with those for guinea pig ileum, the classic smooth muscle organ used for bradykinin assay. Quantitative structure–activity relationships (QSAR) for inhibition of cytokine release of a group of bradykinin analogs were calculated by the Free–Wilson method.

S. Reissmann et al. / Peptides 21 (2000) 527–533

529

2. Materials and methods

2.4. Spleen cell cultures

2.1. Peptides

A total of 1.5 ⫻ 106 spleen cells derived from CBA mice were incubated with the optimal agonist dose of BK (10⫺7 M) in serumfree Eagle’s minimal essential medium for 4 h at 37°C. To test the effect of the potential antagonists, cells were preincubated for 30 min at 23°C before the addition of the optimal BK dosage. Cell-free supernatants were prepared by centrifugation, partially purified by gel filtration (AcA 54 column 0.9 ⫻ 60 cm), and added to the targets cells. After incubation for 1 h at 23°C, mobility was measured as described above. The inhibitory effect of the potential antagonists is given as the concentration [M], which is able to reduce the BK-induced lymphokine release more than 90%. Each value represents the mean taken from two experiments in duplicate cultures.

Peptides were synthesized by the solid phase method using special coupling methods for sterically hindered nonproteinogenic amino acids. Detailed procedures for synthesis, purification, and chemical characterization, together with pharmacological activities on the smooth muscle organs, rat uterus, guinea pig ileum, and in certain cases lung strip have been published [10,12,13]. 2.2. Histamine release Female Wistar rats (200 –250 g) were killed by cervical dislocation under light ether anesthesia. 15 ml of physiological salt solution (containing in mM: NaCl, 137; KCl, 2.7; CaCl2, 1.8; MgCl2, 1.0; Glucose, 5.5; HEPES, 20; adjusted to pH 7.2 with NaOH) were injected into the peritoneal cavity. After gentle abdominal massage the intraabdominal fluid from three animals was pooled, centrifuged (350 ⫻ g for 5 min), and resuspended; the final cell suspension contained about 3– 6% mast cells. Aliquots of these peritoneal cells (50 000 mast cells/assay) were incubated for 20 min at 37°C with various concentrations of the peptides. The reaction was terminated by placing the samples in ice. After centrifugation at 350 ⫻ g for 10 min, the histamine content of the supernatant was measured spectrofluorometrically as described by Shore et al. [27]. The amount of histamine released to the medium was expressed as a percentage of the total amount of histamine in the cells at the start of experiment and was corrected for spontaneous release (usually 4 –5%). Total histamine content was determined by 6 min boiling of aliquots of mast cells from the same animals in each experiment. The EC50 is the concentration of the peptide that releases 50% of the total histamine content of the mast cells (representation of mean ⫾ SEM of the EC50 values calculated for each experiment, n ⫽ 3–5). 2.3. Tanned erythrocyte electrophoretic mobility (TEEM) test The test system is based on the determination of socalled ‘charge-changing’ lymphokines released from activated mononuclear cells by means of sulfosalicylic acidstabilized tanned sheep red blood cells as target indicator cells [23]. Cell electrophoresis measurement of the indicator cells was performed using a commercial cytopherometer (Zeiss, Oberkochen, FRG). The electrophoretic mobility time of the target cells was recorded by microscopic observation using an electronic timer and plotter. The slowing effect (percent) was calculated using targets or targets plus peptides (to exclude side effects) as controls.

2.5. Quantitative structure–activity relationship (QSAR) in the TEEM test Data for 36 nona- and decapeptide analogs of BK were used for the Free–Wilson analysis of QSAR, as described by us [24]. The negative logarithm of the lowest concentration that inhibits the BK-induced effect in the TEEM test, pCmin, was used, and had values from 6 to 15. The training set included 16 compounds containing substituents that appear only once (single point variables) in the series. Although these compounds should be, for statistical reasons, excluded from the analysis, they were kept due to the relatively small total number of compounds available. In our previous studies [24], we could show that introduction of compounds containing single point variables does not affect substantially the results of the analysis. However, the contribution of the respective substituents to activity could be only roughly estimated. In this study BK was used as the parent compound and its sequence was divided into 10 segments, which were numbered as follows: 0 1 2 3 4 5 6 7 8 9 H– Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-OH Substituents were labeled by the abbreviated name of the corresponding group and their position in the polypeptide chain. All but positions 1, 4, and 6 vary in the peptide set used for this study. There are 32 variables (substituents), among which 17 are single point ones. A few of the remaining variables are more or less uniformly distributed among the peptides, as is frequently the case for sets not specifically designed and therefore only partially suitable for QSAR.

3. Results The release of histamine from rat mast cells and inhibition of cytokine release from mouse spleen cells by the bradykinin analogs are listed in Table 1. The analogs tested

530

S. Reissmann et al. / Peptides 21 (2000) 527–533

Table 2 Bradykinin analogs used in the Free–Wilson analysis and their observed and predicted antagonistic activity in the TEEM-test Code number

Structure

B9426 B9984 B10004 B10008 B10010 B10038 B9560 B9562 B9870 B9966 B9976 B9978 B9986 B9992 B10036 B10042 B10044 B9932 B10046 B10050 B9452 B9066

DARG-[Hyp3,Thi5,DIgl7]-BK DArg-[D,LOic2,Hyp3,Igl5,DIgl7,Oic8]-BK DArg-[Hyp3,Igl5,DTic7,Nc7g8]-BK DArg-[Hyp3,Thi5,DIgl7,Nc7g8]-BK DArg-[Hyp3,Igl5,DIgl7,Nc7g8]-BK DArg-[Hyp3,IgI5,DTic7,Nc7g8]-BK Aca-DArg-[Hyp3,Igl5,DIgl7,Oic8]-BK Dhq-DArg-[Hyp3,Igl5,DIgl7,Oic8]-BK Mosi-DArg-[Hyp3,Igl5,DIgl7,Oic8]-BK DArg-[NMF2,Hyp3,Thi5,DIgl7,Oic8]-BK DArg-[Igl2,Hyp3,Igl5,DIgl7,Oic8]-BK DArg-[Igl3,5,DIgl7,Oic8]-BK DArg-[D,LOic2,Igl5,DIgl7,Oic8]-BK DArg-[Oic2,8,Hyp3,Thi5,DTic7]-BK Arg-[Hyp3,Igl5,DTic7,Nc7g8]-BK DArg-[D,LOic2,Hyp3,Thi5,DTic7,Nchg8]-BK DArg-[Hyp3,Igl5,Df5f7,Oic8]-BK Lys-Lys-[Hyp3,Igl5,DIgl7,Oic8,des-Arg9]-BK DArg-[Hyp3,Igl5,DTic7,Chg8,des-Arg9]-BK Lys-Lys-[Hyp3,Igl5,DTic7,Chg8,des-Arg9]-BK DArg-[Hyp3,Thi5,DTic7,Nig8,des-Arg9]-BK DArg-[Hyp3,Cpg5,8,DTic7,des-Arg9]-BK des-Arg9-BK [NMF2]-BK [NMF7]-BK [DNMF7]-BK [Hyp3,Thi5,DNMF7]-BK [DNMF2]-BK [NMF3]-BK [DNMF3]-BK [DPhe7]-BK DArg-[NMF2]-BK DArg-[NMF7]-BK DArg-[Hyp3,Thi5,DNMF7]-BK DArg-[Hyp3,Thi5,DTic7,Oic8]-BK DArg-[Hyp3,DNMF7,NMF8]-BK

J201 J203 J200 J204 J202 J210 J211 J30 J219 J212 J205 HOE-140 J229

pCmin Obs.

Pred.

14 10 15 13 15 15 13 11 14 14 13 15 14 14 10 11 15 13 15 12 13 12 11 13 12 11 12 12 9 9 10 13 11 6 12 6

13 11 14 13 14 14 14 11 14 13 14 16 13 13 10 11 14 14 14 14 13 11 12 12 12 9 10 12 9 9 12 12 12 10 13 7

were not designed for studies on these biologic activities, but for high antagonist activity on the bradykinin B2 receptor, combined with a high degree of resistance to enzymatic degradation. QSAR results for inhibition of cytokine release reported in Table 2 include analogs containing N-methyl phenylalanine for which the inhibitory activity was published previously [26]. Biologic activities in pharmacological tests on smooth muscle preparations with B1 and B2 receptors are published elsewhere, together with the synthetic procedure [6,10 –13,30]. 3.1. Histamine release Histamine release for all tested compounds was found to be enhanced relative to bradykinin. Especially the acylated dimers SUIM(DArg[Hyp3,Igl5,DIgl7,Oic8]-BK)2 (28) and ␣-DDD(Lys-DArg[Hyp3,Igl5,DIgl7,Oic8]-BK)2 (30) show high activity; these compounds are more than 1000 times as

potent as bradykinin. The B1 receptor antagonists DArg [Hyp3,Cpg5,DTic7,Cpg8,desArg9]-BK (34), DArg[Hyp3,Igl5, DTic7,Chg8,desArg9]-BK (35), Lys-Lys[Hyp3,Igl5,DTic7, Chg8, desArg9]-BK (36) and Lys-Lys[Hyp3,Igl5,DIgl7,Oic8, desArg9]-BK (37), which lack the C-terminal arginine, also have enhanced releasing activity compared to [desArg9]-BK (33), which does not stimulate histamine release. 3.2. Inhibition of cytokine release The new highly potent B1 and B2 receptor antagonists, mostly having multiple replacements of proteinogenic by nonproteinogenic amino acids, were extremely potent inhibitors of cytokine release, with some antagonists showing activity at concentrations as low as 10⫺15 M. Some of the analogs also show agonist action in concentrations 106– 1010 higher than for antagonism of cytokine release, but this effect is in most cases nonspecific. Release of these ‘chargechanging’ lymphokines was quantitatively measured by the tanned erythrocyte electrophoretic mobility test (TEEM). The lymphokines released by bradykinin and [des-Arg9]bradykinin were identified as interleukins 1, 6, 2, 3, and soluble interleukin 2 receptor [23]. Cytokine release evoked by bradykinin and its agonist analogs showed a bell-shaped dose dependence in the range 10⫺8–10⫺6 M, and was inhibited by bradykinin antagonists. In contrast to the enhanced histamine release, the very potent inhibition of cytokine release favours application of these new antagonists as antiinflammatory drugs. Many of the compounds designed for high B2-receptor antagonism provide inhibitory constants for cytokine release up to three orders of magnitude higher than for the earliest antagonists [D-Phe7]-BK and [NMF2]-BK [23,26]. Replacements in all sequence positions contribute to the enhanced inhibitory activity, but in a different manner. In the C-terminal sequence at positions 5, 7, and 8, such hydrophobic amino acids such as Tic, Oic, Nig, Igl, Nch, Nc7g Cpg, Chg, and F5F contribute to inhibition of cytokine release. Arginine at position 9 seems to be without influence, as demonstrated by the B1 receptor antagonists, the des-Arg9 analogs DArg[Hyp3,Cpg5,DTic7,Cpg8,desArg9]-BK (34), DArg[Hyp3, Igl5,DTic7,Chg8,desArg9]-BK (35), Lys-Lys[Hyp3,Igl5,DTic7, Chg8,desArg9]-BK (36) and Lys-Lys [Hyp3,Igl5,DIgl7,Oic8, desArg9]-BK (37). Even without the C-terminal Arg, inhibition was detected at concentrations starting from 10⫺15 M (comp. 35). The precise influence of amino acid replacements in the N-terminal sequence is difficult to estimate from the set of analogs used, because the replacements are in all cases combined with other replacements in the C-terminal part. The only difference between compounds Arg[Hyp3,Igl5, DTic7,Nc6g8]-BK (9) and DArg[Hyp3,Igl5,DTic7,Nc6g8]BK (10) is the configuration of arginine at sequence position 0; L-Arginine gives a peptide having an inhibitory potency five orders of magnitude lower than that of the D-arginine analog.

S. Reissmann et al. / Peptides 21 (2000) 527–533

Comparing compounds DArg-Arg[Igl2,Hyp3,Igl5,DIgl7, Oic8]-BK (16), DArg-Arg[DLOic2,Hyp3,Igl5,DIgl7,Oic8]BK (17) and DArg-Arg[Oic2,Igl5,DIgl7,Oic8]-BK (18), the high inhibitory activity of 17 indicates that the amino acid Oic at position 2 can be more effective combined at position 3 with Pro than with hydroxyproline. In the acylated analogs tested, the acyl residue shows no significant influence on the inhibition of cytokine release. To find correlations between the interaction of the analogs with bradykinin receptors of the B1- and B2-type and the histamine release or inhibition of the cytokine release we compared the pA2 value for antagonistic potency on the B2-receptor mediated contraction of the guinea pig ileum to the concentrations needed for histamine release or inhibition of bradykinin induced cytokine release. Table 1 shows that there exists no correlation between the antagonistic potency on smooth muscle contraction and the inhibition of the cytokine release. Thus for the selected following compounds from Table 1, the pA2-values and the corresponding concentrations for inhibition of the bradykinin induced cytokine release are: comp. 4: 8.4 ⫻ 10⫺12 M, comp. 8: 7.7 ⫻ 10⫺15 M, comp. 9: 7.8 ⫻ 10⫺10 M, comp. 10: 7.7 ⫻ 10⫺15 M, and comp. 21: 5.0 ⫻ 10⫺15 M. It is seen that the potent B2-receptor antagonist comp. 4 needs a three orders of magnitude higher concentration to inhibit the cytokine release than the weak B2-receptor antagonist comp. 21. The same situation was found for the histamine release. Compound 4 with a pA2-value of 8.4 requires a concentration of 50.1 ␮M to release 50 percent of the total histamine amount, whereas compound 21 with a relatively low affinity to the B2-receptor (pA2 ⫽ 5.0) needs for the same effect only a concentration of 0.65 ␮M. The comparison to the B1-receptor affinity suffers from the reduced number of available dates. Only for some of the analogs listed in Table 1, pIC50-values from binding experiments on the B1-receptor-containing human lung fibroblasts cell line IMR 90 or from a functional assay on the isolated rabbit aorta are available from the literature [6,13,21]. Nevertheless, the structural requirements for the affinity to the B1-receptor and the influence on histamine and cytokine release seems to be different. Thus, compound 34 with an pIC50 value of 9.3 inhibits the cytokine release in a concentration of 10⫺12 M while compound 11 with a pIC50 of 7.9 acts even in a concentration of 10⫺15 M. Despite their binding affinity to the B1-receptor, compounds 33 (pIC50 ⫽ 5.6) and 34 (pIC50 ⫽ 9.3) are not able to evoke the histamine release. From all these dates together, we conclude that there is no correlation between B2- and B1-receptor affinity on the one hand and histamine releasing activity or inhibition of the cytokine on the other hand. 3.3. Quantitative structure-activity relationships in the TEEM test Results of QSAR calculations on 36 bradykinin analogs for inhibition of cytokine release are given in Table 2 and

531

Fig. 1. Observed antagonist activity of the bradykinin analogs on the TEEM test (–logCmin) versus calculated activity from the QSAR model derived by the Free–Wilson Method using a set of 36 compounds (stepwise backwards multiple linear regression, n ⫽ 36, s ⫽ 1.46, r ⫽ 0.839, F9,26 ⫽ 6.88).

plotted in Fig. 1. The calculated contributions of the different substituents to antagonist activity in the TEEM test are listed in Table 3. During the stepwise, backward regression, 23 variables were removed as not significant. The correlation level achieved was relatively low (n ⫽ 36, s ⫽ 1.46, r ⫽ 0.84, F9,26 ⫽ 6.88); see Fig. 1. By these calculations, most substitutions should either reduce the (assumed) potency of the parent compound, BK, or have no influence on it (not significant contributions). Just replacement of Phe at position 5 by Thi or Igl is expected to enhance antagonistic activity in the TEEM test. However, due to the poor statistical quality of the QSAR model mentioned above, these results should be taken with caution. Although bradykinin does not show an antagonistic efTable 3 Calculated Free–Wilson substituent contributions to the antagonistic activity of bradykinin analogs in the TEEM-test relative to those of Bk Substituent1

Contribution2

Substituent1

Contribution2

DArg0 Lys-Lys0 Aca-DArg0* Dhq-DArg0* Mosi-DArg0* Arg0* D,LOic2 NMF2 DNMF2* Igl2* Oic2* Hyp3 Igl3* NMF3* DNMF3* Thi5

NS NS NS ⫺2.91 (3.13) NS ⫺3.91 (3.13) ⫺2.54 (1.92) NS NS NS NS ⫺1.62 (1.72) NS ⫺3.12 (3.18) ⫺3.12 (3.18) 2.66 (1.99)

Igl5 Cpg5* DIgl7 DTic7 DNMF7 NMF7 Df5f7* DPhe7* Oic8 Nc7g8 Chg8 Nchg8* Cpg8* Nig8* NMF8* OH9 Backbone

3.41 (1.83) NS NS NS ⫺3.49 (1.72) NS NS NS NS NS NS NS NS NS NS NS 12.12 (1.04)

1

Single point variables are labeled by an asterisk; 2 NS, variable excluded as not significant; 95% confidence intervals in brackets.

532

S. Reissmann et al. / Peptides 21 (2000) 527–533

fect in the TEEM test, it was chosen as the parent compound. The only consequence of this choice is that the calculated substituent contributions will refer to those of the corresponding residues in the native peptide, for which the QSAR model will predict an (hypothetical) inhibitory activity, but their ranking will remain unaffected. In our opinion, this is not a strong drawback, as the Fujita-Ban approach does not require explicitly the presence of the parent molecule among the compounds of the training set and, on the other hand, the switch from one type of biological activity (agonism) to another (antagonism) cannot be explained within the Free–Wilson model.

[Hyp3,Igl5,DIgl7,Oic8]-BK)2 (28), and ␣-DDD(Lys-DArg [Hyp3,Igl5,DIgl7,Oic8]-BK)2 (30), which are under investigation as potential anticancer drugs. Most antagonists that strongly inhibit bradykinin-induced smooth muscle contraction are also active inhibitors of cytokine release. Some compounds antagonized bradykinin-evoked cytokine release at concentrations as low as 10⫺15 M, and amino acid replacements throughout the peptide chain contribute to this enhanced inhibitory activity. The calculated contributions of the different substituents indicate that replacement of Phe at position 5 by Thi or Igl has the greatest effect on antagonist activity in the TEEM test for cytokine release.

4. Discussion Acknowledgments The rational design of peptide and nonpeptide bradykinin antagonists for therapeutic use requires not only knowledge about the potency and receptor-bound conformation but also knowledge about enzymatic degradation, absorption, distribution, elimination, and side effects of the potential drug. Important side effects of bradykinin antagonists are their influence on the release of histamine and cytokines, substances that are involved in inflammatory processes. Stimulation of histamine release works against the antiinflammatory action of bradykinin receptor antagonists and is undesirable. Most peptide bradykinin antagonists tested here release histamine from rat mast cells in concentrations lower and in some cases many orders of magnitude lower than does bradykinin itself. Rat data, however, may not be applicable to human histamine release. Studies of histamine release from human basophils, lung mast cells, and skin mast cells by early bradykinin antagonists showed activity of those peptides only in skin mast cells, and their potency was low [20]. Additional studies are needed on potent bradykinin antagonists in human mast cells. There is no correlation between histamine release and B2 receptor interaction as indicated by activity on guinea pig ileum, but receptor–independent activation of G-proteins has been demonstrated in rat peritoneal mast cells for bradykinin, mastoparan, and substance P. A proposed mechanism of this activation is based on the basic amphipathic nature of the peptides having a resemblance to the third cytosolic loop of the receptor for formyl peptides [19]. In our previous studies, we found that besides the presence of positively charged amino acids, elongation of the nonapeptide at the N-terminus gives compounds with high histamine-releasing activities [31]. Acylation, which reduces the N-terminal basic charge, was expected to decrease histamine-releasing potency, but the acylated decapeptides Aca-DArg-Arg[Hyp3,Igl5,DIgl7,Oic8]-BK (25), Dhq-DArg-Arg[Hyp3,Igl5,DIgl7,Oic8]-BK (26), Mosi-DArgArg[Hyp3,Igl5,DIgl7,Oic8]-BK (27), bApG-DArg-Arg[Hyp3, Igl5,DIgl7,Oic8]-BK (31) and (Gun)2-bApG-DArg-Arg [Hyp3,Igl5,DIgl7,Oic8]-BK (32) are potent. The most potent compounds in this test are the dimers, SUIM(DArg

We thank I. Agricola, B. Schilling, and H. Rohde for skillful technical assistance. References [1] Abe Y, Kayakiri H, Satoh S, et al. A novel class of orally active non-peptide bradykinin B2-receptor antagonists. Construction of the basic framework. J Med Chem 1998;41:564 –78. [2] Abe Y, Kayakiri H, Satoh S, Inoue T, Sawada Y, Inamura N, Asano M, Hatori C, Sawai H, Oku T, Tanaka H. A novel class of orally active non-peptide bradykinin B2-receptor antagonists. Overcoming the species difference between guinea pig and man. J Med Chem 1998;41:4053– 61. [3] Abe Y, Kayakiri H, Satoh S, et al. A novel class of orally active non-peptide bradykinin B2-receptor antagonists. Discovering bioisosteres of the imidazol [1,2-a]pyridine moiety. J Med Chem 1998;41: 4061–79. [4] Aramori I, Zenkoh J, Morikawa N, et al. Nonpeptide mimic of bradykinin with long-acting properties at the bradykinin B2 receptor. Mol Pharmacol 1997;52:16 –20. [5] Asano M, Hatori C, Inamura N, Sawai H, Hirosumi J, Fujiwara T, Nakahara K. Effects of a nonpeptide bradykinin B2 receptor antagonist, FR 167344, on different in vivo animal models of inflammation. Br J Pharmacol 1997;122:1436 – 40. [6] Burkhard M, Zuzack JS, Jones St Francis M, Whalley ET, Stewart JM, Gera L. Comparative profile of novel potent bradykinin antagonists at human B1 and B2 receptors. Immunopharmacology 1996;33: 186 –90. [7] Chakravarty S, Mavunkel BJ, Goehring RR, Kyle DJ. Novel bradykinin receptor antagonists from a structurally directed non-peptide combinatorial library. Immunopharmacology 1996;33:61–7. [8] Chan D, Gera L, Helfrich B, Helm K, Stewart J, Whalley E, Bunn P. Novel bradykinin antagonist dimers for the treatment of human lung cancers. Immunopharmacology 1996;33:201– 4. [9] Dankwardt SM, Ferla S, Krstenansky JL, Bhakta S, Ostrelich H, Jarnagin K. Nonpeptide bradykinin antagonist analogs based on a model of a Sterling-Winthrop nonpeptide bradykinin antagonist overlapped with cyclic hexapeptide bradykinin antagonist peptides. Bioorg Med Chem Lett 1997;7:1921– 6. [10] Gera L, Stewart JM. A new class of bradykinin antagonists containing indanylglycine. Immunopharmacology 1996;33:174 –7. [11] Gera L, Stewart JM. Novel highly potent bradykinin antagonists containing penta-fluoro-phenylalanine. Proceedings of the Fifteenth American Peptide Symposium. Tam JP, Kaumaya PTP, editors. London: Kluwer Academic Publ, 1999, pp. 730 –2.

S. Reissmann et al. / Peptides 21 (2000) 527–533 [12] Gera L, Stewart JM, Whalley E, Burkard M, Zuzack JS. A new class of potent bradykinin antagonist dimers. Immunopharmacology 1996; 33:178 – 82. [13] Gera L, Stewart JM, Whalley ET, Burkard M, Zuzack JS. New bradykinin antagonists having very high potency at B1 receptors. Immunopharmacology 1996;33:183–5. [14] Greiner G, Dornberger U, Paegelow I, Schoelkens BA, Liebmann C, Reissmann S. Bradykinin antagonists with dehydrophenylalanine analogues at position 5. J Pept Sci 1998;4:92–100. [15] Hsieh KH, Stewart JM. Cyclic and linear bradykinin analogues: implications for B2 antagonist design. J Peptide Res 1999;54:23–31. [16] Hock FJ, Wirth K, Albus U, Linz W, Gerhards HJ, Wiemer G, Henke ST, Breipohl G, Ko¨nig W, Scho¨lkens BA. HOE-140, a new, potent and long acting bradykinin pantagonist: in vitro studies. Br J Pharmacol 1991;102:769 –73. [17] Inamura N, Asano M, Kayakiri H, Hatori C, Oku T, Nakahara K. Characterization of FR173657, a novel, non-peptide BK2 antagonist, in vitro and in vivo. Can J Physiol Pharmacol 1997;75:622– 8. [18] Jenkinson DH, Bernard EA, Hoyer D, Humphrey PPA, Leff P, Shankley NP. International Union of Pharmacology Committee on Receptor Nomenclature and Drug Classification. IX. Recommendations on terms and symbols in quantitative pharmacology. Pharm Rev 1995; 47:255– 66. [19] Klinker J, Seifert R. Rezeptor-unabhaengige Aktivierung von GProteinen. Pharmazie in Unserer Zeit 1995;24:250 – 63. [20] Lawrence ID, Warner JA, Cohan VL, Lichtenstein LM, Kagey– Sobotka A, Vavrek RJ, Stewart JM, Proud D. Induction of histamine release from human skin mast cells by bradykinin analogs. Biochem Pharmacol 1989;38:227–33. [21] Marceau F, Hess JF, Bachvarov DR. The B1 receptors for kinins. Pharmacol Rev 1998;50:357– 86.

533

[22] Mousli M, Hugli TE, Landry Y, Bronner C. Peptidergic pathway in human skin and rat peritoneal mast cell activation. Immunopharmacology 1994;27:1–11. [23] Paegelow I, Werner H, Reissmann S. Effects of bradykinin and bradykinin analogues on spleen cells of mice. Eur J Pharmacol 1995; 279:211– 6. [24] Pineda De Castro LF, Reissmann S. QSAR in bradykinin antagonists. Inhibition of the bradykinin induced contraction of the isolated rat uterus and guinea pig ileum. Quant Struct Act Relat 1995;14:249 –57. [25] Reissmann S, Greiner G, Seyfarth L, Paegelow I, Werner H, Vietinghoff G, Boeckmann S, Schulz E, Wartner U, Gera L. A new type of bradykinin B2 receptor antagonists: Bradykinin analogs with N-alkyl amino acids at position 2. Immunopharmacology 1996;33:73– 80. [26] Reissmann S, Schwuchow C, Seyfarth L, De Castro LFP, Liebmann C, Paegelow I, Werner H, Stewart JM. Highly selective bradykinin agonists and antagonists with replacement of proline residues by N-methyl-D- and L-phenylalanine. J Med Chem 1996;39:929 –36. [27] Shore PA, Burkhalter A, Cohn VH. A method for the fluorometric assay of histamine in tissues. J Pharmacol Exp Ther 1959;127:182– 6. [28] Stewart JM. Bradykinin antagonists: Development and applications. Biopolymers (Pept Sci) 1995;37:143–55. [29] Stewart JM, Gera L. A new class of bradykinin antagonists containing indanylglycine. Immunopharmacology 1996;33:174 –7. [30] Stewart JM, Gera L, Hanson W, Zuzack JS, Burkard M, McCullough R, Whalley ET. A new generation of bradykinin antagonists. Immunopharmacology 1996;33:51– 60. [31] Vietinghoff G, Paegelow I, Reissmann S. Induction of histamine release from rat mast cells by bradykinin analogues. Peptides 1996; 17:1467–70. [32] Whalley ET, Hanson WL, Stewart JM, Gera L. Oral activity of peptide bradykinin antagonists. Can J Physiol Pharmacol 1997;75:629 –32.