- Email: [email protected]
Endothelin analogs which distinguish vasoconstrictor and vasodilator ETB receptors
Life Sciences, Vol. 56, NO. 15 pp. 1251-1256, 1995 Copyright 0 1995 Else&r Science Ltd Printed in the USA. All rights reserved cm-3205/95 $950 + .@I
Pergamon
0024-3205(95)00070-4
ENDOTHELIN VASOCONSTRICTOR
ANALOGS WHICH DISTINGUISH AND VASODILATOR ETB RECEPTORS
Sesha Natarajans, John T. Hunts,*, Stephen Festins,, Randy Serafinot, Suzanne Moreland*
Rongan ZhangS, and
Departments of Chemistry5 and Pharmacology*, Bristol-Myers Squibb Pharmaceutical Institute, P.O. Box 4000, Princeton, NJ 08543-4000, USA (Received
in final form January
Research
18, 1995)
summarv [PenlTll, Nle7, Glu9, Ala’*]-Sarafotoxin S6b (BMS-184696) and [Penl,“, Nle7, Glu”, Leu18]-sarafotoxin S6b (BMS-184697) were synthesized with the aim of preparing ETB receptor antagonists. BMS- 184696 was a potent ETA antagonist, an extremely potent vasoconstictor ETB agonist, and a non-competitive vasodilator ETB antagonist with no agonist activity. BMS-184697 was a potent ETA antagonist, a potent vasoconstrictor ETB agonist, and a vasodilator ETB agonist with moderate potency. The ability of BMS-184696 to activate the vasoconstrictor ETB receptor but not the vasodilator ETB receptor, despite having high affinity binding to the vasodilator ETB receptor as evidenced by its antagonist activity, strongly suggests the existence of ETB receptor subtypes. Key Words: endothclin,
sarafotoxin
analogs,
vasodilator,
vasoconstrictor
ET,
receptor
In the seminal paper describing the isolation of endothelin (ET) in 1988, both vasodilator and vasoconstrictor actions were reported for this peptide (1). In the intervening time, extensive research has helped to clarify the receptor pharmacology of the ET family and the related sarafotoxins. The ETA receptor subtype (2), which preferentially binds ET- 1 and ET-2 over ET-3, is generally believed to mediate vasoconstriction. The ETB receptor subtype (3), which has equal affinity for ET-l, ET-2 and ET-3, appears to mediate either vasodilation (4) or vasoconstriction (5), depending upon the tissue type. The different pharmacological responses which can be elicited by stimulation of the ETB receptor have led to disagreement as to whether ETB receptor subtypes exist or whether there is only a single receptor subtype which can cause different cellular responses depending upon the cell type. Agents which could distinguish between these possibilities would be valuable tools. Among the members of the ET and sarafotoxin families, sarafotoxin S6c (S6c) is unique in its extremely high selectivity for ETB receptors (6). Compared to the non-selective agonist sarafotoxin S6b (Shb), S6c contains 4 amino acid replacements (Ser* j Thr, Lys4 ti Asn, Lys9 + Glu, Tyr*3 a Asn). Pharmacological characterization of S6b/S6c chimeric peptides demonstrated that it is the Lysy = Glu substitution which leads to loss of ETA affinity, making [Glu9]-S6b a selective ETB agonist (7). We have demonstrated that bis-penicillamine ET analogs which contain substitutions for Asp’* provide ETA receptor antagonists (8). With the expectation that a similar strategy would allow the preparation of ETB selective antagonists, we prepared bis-penicillamine [Gl@]-S6b analogs which contain substitutions for Asp ‘8. In this report, we describe the synthesis of these analogs and their
*To whom correspondence and reprint requests should be addressed at Department of Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute, P.O. Box 4000, Princeton, NJ 085434000, USA. (609) 252-4989, FAX (609) 252-6804. Email, [email protected]
1252
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
pharmacological properties. The behavior of these peptides in vasodilator and vasoconstrictor preparations suggests that their actions are mediated by different subtypes of ETR receptors.
ETR
Methods [Pent~lr, Nle7, Glu”, Ala18]-S6b (BMS-184696) and [Penr*tl, Nle’, Glu9, Leu18]-S6b (BMS1X4697) were prepared using previously described procedures (8). Peptides were purified to > 95% homogeneity as analyzed by HPLC and were characterized by FAB mass spectrometry and amino acid analysis. ET- 1, ET-2, and ET-3 were purchased from Peptides International and S6c from Peninsula Laboratories. Stock solutions were prepared immediately before use by dissolving the peptides in a small amount of 1 M acetic acid which was subsequently adjusted to the appropriate volume with water containing 0.05% bovine serum albumin. Binding experiments were performed as previously described (9). New Zealand White rabbit carotid arteries and saphenous veins were handled as described elsewhere (5,lO). Wistar rats were killed by inhalation of CO2. The aortae were carefully removed, cut into rings, and mounted at 5 grams preload for isometric force recording. The presence of a functional vascular endothelium was confirmed by contracting the tissues with 30 nM norepinephrine, then quantifying the relaxation observed upon addition of 1 PM acetylcholine. Only rings that relaxed more than 80% in response to acetylcholine were used. The EC50 values were calculated by linear regression analysis. KR values were determined by analysis of the dose-ratios (11). Schild analysis was used to calculate pA2 values.
Results and Discussion ]Penl.tl, Nle’, Glu”]-S6b analogs were prepared containing either Ala (BMSlX4696) or Leu (BMSlX4697) in place of Asp t8. This bis-penicillamine structure forces the almost exclusive formation of the natural l-15, 3-11 disulfide isomer (12). The combination of a I,1 I-bispenicillamine and an l&position Ala or Leu, which in ET-l produced ETA antagonists (X), was incorporated into [Glu”]-S6b, an ETR selective framework, with the aim of producing ETR antagonists. Both peptides were tested for: binding affinity in preparations containing either ETA or ETR receptor subtypes, vasoconstrictor activity in ETA- or vasoconstrictor ETR-containing tissues, and ETR-mediated vasodilator activity. Using an ETA receptor preparation (rat A10 vascular smooth muscle cell membranes), both BMS- 1X4696 (Ki = 6.6 f 3.4 nM) and BMS- 184697 (Ki = 20 + 2.8 nM) were high affinity ligands. These ETA affinities are somewhat higher than that reported for [Glu9]-S6b (IC50 = 115 nM (7)) but considerably higher than those reported for S6c (1.2 yM (7); > 5 PM (6)). We have previously shown that vasoconstriction of rabbit carotid artery rings is mediated by ETA receptors and that S6c, at concentrations as high as 300 nM, does not contract this tissue (5). As expected, neither [Glu9]Shb analog showed any vasoconstrictor activity at the highest concentration tested (3 PM BMS1X4696 and 10 PM BMS-lX46Y7). As shown in Figure 1, each peptide was able to competitively antagonize the contractile effects of ET- 1 in the rabbit carotid artery (BMS- 1X4696, pA2 = 7.3, slope = -0.Xx; BMS-lX46Y7, PA;! = 6.3, slope = -1.05). BMSlX46Y6 was a particularly high affinity ligand for the ETR receptors present in rat cerebellum (Ki = 0.20 + 0.01 nM), with binding affinity essentially identical to S6c (Ki = 0.20 f 0.01 nM). BMS-184697 showed weaker but nonetheless substantial affinity for the cerebellar ETR subtype (Ki = 22 do2.2 nM). We have shown that rabbit saphenous vein contains vasoconstrictor receptors which have ETe-like characteristics, including high sensitivity to the agonist S6c (5). As shown in Figure 2, both BMS-184696 and BMS-184697 were potent constrictors of this tissue, with EC50 values roughly in line with their respective ETR Ki values in rat cerebellar tissue (Et& = O.Y5 + 0.15 nM BMS-184696 and 230 + 20 nM BMS-184697). The EC50 value for S6c was 0.13 f 0.03 nM in this assay.
-II
1253
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
-9 -10 [ET- I], log M
-8
-9 -IO [ET-l], log M
-II
-7
-8
-7
Fig. 1 Cumulative concentration-response curves to ET- 1 obtained in rabbit carotid artery rings in the absence (0) and presence of 0.3 (O), 1 (W), and 3 (A) PM BMS184696 (A) or 1 (O), 3 (m), and 10 (A) PM BMS- 184697 (B). Data are plotted as mean + SEM; n = 3 - 4 rings from different rabbits.
-II
-10
-9
-8
-7
-6
-5
[Pcptidc], log M
Fig. 2 Cumulative concentration response curves to ET-l (0) BMS-184696 (a), and BMS-184697 (A) obtained in rabbit saphenous vein rings. Data are plotted as mean f SEM; n = 4 - 8 rings from different rabbits. Vasodilator ETn receptors are located on vascular endothelium and their activation can lead to dilation of underlying vascular smooth muscle cells by the release of nitric oxide (NO), prostacyclin, or other vasodilatory substances (13). We developed an assay to test for agonists and antagonists of the vasodilator ETB receptor. The agonist assay monitors the ability of a test compound to cause relaxation of contractions elicited by 40 nM norepinephrine (NE, EC& in rat aortic rings with intact endothelium. In this assay, S6c caused a maximum of approximately 50% relaxation even in rings in which acetylcholine caused 100% relaxation. The antagonist assay compares the concentration
Constrictor and Dilator ET, Receptors
1254
Vol. 56, No. 15, 19%
relaxation curves for S6c in the absence and presence of test compound; test compounds are added 20 minutes before the rings are contracted with 40 nM NE. When force plateaus, the concentrationresponse curve for S6c is initiated. The antagonist assay is relatively specific for ETB receptor antagonists because even at high concentrations none of the following agents blunted the S6c-elicited relaxation: the selective ETA receptor antagonist BMS-182874 (14) (30 PM), the calcium channel blocker nifedipine (0.1 PM), the thromboxane A2 receptor antagonist SQ 30,741 (1 PM), or the angiotensin AT1 receptor antagonist SR 47436 (15) (10 PM). Removal of the endothelium or treatment of the aorta with NO synthesis inhibitors L-NNA or L-NMMA or with the guanyl cyclase inhibitor methylene blue abolished the S6c-induced relaxations suggesting that NO release from the vascular endothelium was responsible for the ETg-mediated relaxation (data not shown).
-11
-10
-9
-8
-7
-6
-5
-11
-10
-9
-x
-7
[SAC],log M
[Peptides]. log M Fig. 3
A. Cumulative concentration response curves to S6c (0), BMS-184696 (Cl), and BMS- 184697 (A) obtained in rat aortic rings. B. Cumulative concentration response curves to S6c obtained in rat aortic rings in the absence (0) and presence of 0.1 (0) 0.3 (M), and 1 (A) nM BMS-184696. Data are plotted as mean + SEM; n = 4 - 8 rings from different rats. When tested as an agonist, BMS-184696 showed essentially no vasodilatory activity at concentrations up 10 nM (Figure 3A). In a separate experiment, a single 1 PM concentration of BMS-184696 also did not evoke relaxation in the NE-contracted rat aorta (data not shown). In contrast, BMS-184697 was a full agonist with moderate potency (EC50 = 82 + 33 nM) compared with S6c (EC50 = 0.22 I!Z0.05 nM). Because BMS-184696 bound to the ETB receptor in rat cerebellar membranes with a high affinity, but did not behave as an agonist for inducing relaxation of the rat aorta, we examined its ability to blunt the relaxation elicited by S6c. When tested as an antagonist, BMS- 184696 blocked the S6c response in a basically all-or-none fashion, producing no block at 0.1 nM but a full block at 0.3 nM (Figure 3B). Several explanations for the non-competitive nature of the vasodilatory antagonist behavior of BMS-184696 must be considered. While it is possible that the BMS-184696 is a nonspecific vasodepressant or alpha-receptor blocker, this seems unlikely because the compound did not blunt NE contractions when it was tested as an agonist. Another possibility is that BMS- 184696 stimulated constrictor receptors as well as relaxant receptors in the rat aorta, i.e., that it functionally antagonized its own actions. However BMS-184696 is clearly not an ETA agonist and we have no evidence for constrictor ETB receptors in the rat aorta (10). Finally, BMS-184696 may have interacted directly with S6c, however such an interaction is not apparent in the rabbit saphenous vein assay. BMS- 1846Y6 is a potent ETA antagonist, an extremely potent vasoconstrictor ETB agonist, and a non-competitive vasodilator ETB antagonist with little or no agonist activity. BMS-184697 is a potent ETA antagonist, a vasoconstrictor ETn agonist, and a vasodilator ETB agonist of moderate
1255
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
potency. The differential ability of two potent vasoconstrictor ETB agonist peptides to either activate (BMS-184697) or not activate and in fact antagonize (BMS-184696) vasodilator ETB receptor preparations strongly suggests that ETB receptor subtypes exist. The putative ETB receptor subtypes may in fact be products of the same gene, since evidence to date supports the existence of only two ET receptor genes accounting for the ETA and ETB receptor subtypes. If this is the case, the existence of subtypes may be due to different post-translational modification, different G protein populations in endothelial versus smooth muscle cells or alternate transcripts. TABLE 1. Biological Activity of BMS- 184696 and BMS- 184697 BMS- 184697 BMS- 184696 ETA Receptor
Ki BCso KB
ETn Receptor Ki EC50 - contraction ECso - relaxation KB - relaxation Data, in nM, shown as mean f SE.
6.6 + 3.4 > 3000 77f
10
0.20 f 0.01 0.95 + 0.15 > 1000 non-competitive
20 + 2.8 > 10000 460 f 43 22 f 2.2 230 f 20 82+33 -
Relatively little published evidence supports the existence of ETB receptor subtypes. IRL 103X was reported to discriminate between the two ETB receptors (16), but unfortunately that paper was retracted because results differed among various batches of IRL 1038 (17). RES-701-l may distinguish between ETn receptor subtypes because it blocks the relaxant effects of ET in rat aorta more effectively than the contractile effects of S6c in rabbit saphenous vein (18.19). PD 142893 has been shown to block the constrictor effects of S6c in rabbit pulmonary artery (pA2 = 5.85 (20)). These data were confirmed and extended in a report claiming that 10 pM PD 142893 had essentially no effect when ET-I was used as the constrictor stimulus in pulmonary artery, however 1 pM PD 142893 strongly antagonized the vasodilation elicited by S6c in rat isolated perfused mesentery (2 1j. The large difference in potency for PD 142893 blockade of constrictor and dilator effects has been suggested as evidence for ETB receptor subtypes. Using the assays described herein, we have been unable to duplicate this difference in potency. In our hands, PD 142893 blocks the constrictor response to S6c in rabbit saphenous vein (KB = 9.9 _+3.8 PM) at concentrations similar to those required to blunt the S6c dilator response in rat aorta precontracted with NE (KB = 21 + 14 PM). Thus, either PD 142893 does not discriminate between ETB receptor subtypes or the receptor mediating dilation in the rat aorta is different from that in the rat mesentery. Structure-activity studies of the ETA receptor using full length endothelin peptides has demonstrated the critical nature of Asp’s to receptor activation (X,22,23). This study further confirms this requirement but also indicates that an acidic residue at position 18 is not a required feature of either a vasoconstrictor ETB agonist (both BMS-184696 and BMSlX4697) or a vasodilator ETn agonist (BMS- 184697). In summary, in this study we attempted to prepare ETB selective antagonists by incorporating into ETn selective peptides substitutions which led to ETA antagonists. While only partially successful, the behavior of these peptides in BTn-containing tissues provides evidence that the ETn receptors which mediate vasocontraction in the rabbit saphenous vein and vasodilation in the rat aorta are of different subtypes. Acknowledgements The authors thank Mr. Eddie C.-K. Liu for expert technical assistance critical reading of the manuscript.
and Dr. Maria Webb for a
1256
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
References
1. 2. 3. 4. _5 . 6. 7. x. Y. IO. II. 12. 13. 14. 15. 16. 17. 1X. 19. 20. 21. 22. 23.
M. YANAGISAWA, H. KURIHARA, H. KIMURA, Y. TOMOBE, M. KOBAYASHI, Y. MITSUI, Y. YAZAKI, K. GOT0 and T. MASAKI, Nature m 411-415 (1988). H. ARAI, S. HORI, I. ARAMORI, H. OHKUBO and S. NAKANISHI, Nature 34x 730-732 (1990). T. SAKURAI, M. YANAGISAWA, Y. TAKUWA, H. MIYAZAKI, S. KIMURA, K. GOT0 and T. MASAKI, Nature 348 732-735 (1990). R. TAKAYANAGI, K. KITAZUMI, C. TAKASAKI, K. OHNAKA, S. AIMOTO, K. TASAKA, M. OHASHI and H. NAWATA, FEBS Letters 282 103-106 (1991). S. MORELAND, D.M. MCMULLEN, C.L. DELANEY, V.G. LEE and J.T. HUNT, Biochem. Biophys. Res. Comm. 184 1OO-106 (1992). D.L.J. WILLIAMS, K.L. JONES, D.J. PETTIBONE, E.V. LIS and B.V. CLINESCHMIDT, Biochem. Biophys. Res. Comm. 175 556-561 (1991). C. TAKASAKI et al., Eur. J. Pharmacol. 19x 165-169 (1991). J. T. HUNT et al., Bioorganic Med. Chem. 159-65 (1993). M.L. WEBB, E.C.-K. LIU H. MONSHIZADEGAN, C.C. CHAO, J. LYNCH, S.M. FISHER and P.M. ROSE, Mol. Pharmacol. 44 959-965 (1993). S. MORELAND, D.M. MCMULLEN, 8. ABBOA-OFFEI and A. SEYMOUR, Br. J. Pharmacol. 112 704-708 (1994). E. ARIENS and J. VAN ROSSUM, Arch. Int. Pharmacodyn. Ther. 110 275-299 (1957). J.T. HUNT, V.G. LEE, E.C.K. LIU, S. MORELAND, D. MCMULLEN, M.L. WEBB and M. BOLGAR, Int. J. Peptide Protein Res. 42 249-258 (1993). J.R. VANE and R.M. BOTTING, J. Physiol. Pharmacol. 44 (Suppl. 215-36 (1993). P.D. STEIN et al., J. Med. Chem. 37 329-331 (1994). C. CAZAUBON et ul., J. Pharmacol. Exp. Ther. 265 826-834 (1993). S.A. SUDJARWO, M. HORI, M. TAKAI, Y. URADE, T. OKADA and H. KARAKI, Life Sciences 53 43 l-437 ( 1993). Y. URADE, Y. FUJITANI, K. ODA, T. WATAKABE, I. UMEMURA, M. TAKAI, T. OKADA , K. SAKATA and H. KARAKI, FEBS Letters 342 103 (1994). H. KARAKI, S. A. SUDJARWO, M. HORI, T. TANAKA and Y. MATSUDA, Eur. J. Pharmacol. 262 255-259 (1994). S. A. SUDJARWO, M. HORI, T. TANAKA, Y. MATSUDA, T. OKADA and H. KARAKI, Biochem. Biophys. Res. Comm. 200 627-633 (1994). W. L. CODY et al., J. Med. Chem. 3 3301-3303 (1992). T.D. WARNER, G.H. ALLCOCK, R. CORDER and J.R. VANE, Br. J. Pharmacol. 110 7777X2 (1993). K. NAKAJIMA et al., Biochem. Biophys. Res. Comm. 163 424-429 (I 9X9). J.T. HUNT, V.G. LEE P.D. STEIN, A. HEDBERG, E.C.-K. LIU, D. MCMULLEN and S. MORELAND, Bioorganic Med. Chem. Lett. 133-38 (1991).
Pergamon
0024-3205(95)00070-4
ENDOTHELIN VASOCONSTRICTOR
ANALOGS WHICH DISTINGUISH AND VASODILATOR ETB RECEPTORS
Sesha Natarajans, John T. Hunts,*, Stephen Festins,, Randy Serafinot, Suzanne Moreland*
Rongan ZhangS, and
Departments of Chemistry5 and Pharmacology*, Bristol-Myers Squibb Pharmaceutical Institute, P.O. Box 4000, Princeton, NJ 08543-4000, USA (Received
in final form January
Research
18, 1995)
summarv [PenlTll, Nle7, Glu9, Ala’*]-Sarafotoxin S6b (BMS-184696) and [Penl,“, Nle7, Glu”, Leu18]-sarafotoxin S6b (BMS-184697) were synthesized with the aim of preparing ETB receptor antagonists. BMS- 184696 was a potent ETA antagonist, an extremely potent vasoconstictor ETB agonist, and a non-competitive vasodilator ETB antagonist with no agonist activity. BMS-184697 was a potent ETA antagonist, a potent vasoconstrictor ETB agonist, and a vasodilator ETB agonist with moderate potency. The ability of BMS-184696 to activate the vasoconstrictor ETB receptor but not the vasodilator ETB receptor, despite having high affinity binding to the vasodilator ETB receptor as evidenced by its antagonist activity, strongly suggests the existence of ETB receptor subtypes. Key Words: endothclin,
sarafotoxin
analogs,
vasodilator,
vasoconstrictor
ET,
receptor
In the seminal paper describing the isolation of endothelin (ET) in 1988, both vasodilator and vasoconstrictor actions were reported for this peptide (1). In the intervening time, extensive research has helped to clarify the receptor pharmacology of the ET family and the related sarafotoxins. The ETA receptor subtype (2), which preferentially binds ET- 1 and ET-2 over ET-3, is generally believed to mediate vasoconstriction. The ETB receptor subtype (3), which has equal affinity for ET-l, ET-2 and ET-3, appears to mediate either vasodilation (4) or vasoconstriction (5), depending upon the tissue type. The different pharmacological responses which can be elicited by stimulation of the ETB receptor have led to disagreement as to whether ETB receptor subtypes exist or whether there is only a single receptor subtype which can cause different cellular responses depending upon the cell type. Agents which could distinguish between these possibilities would be valuable tools. Among the members of the ET and sarafotoxin families, sarafotoxin S6c (S6c) is unique in its extremely high selectivity for ETB receptors (6). Compared to the non-selective agonist sarafotoxin S6b (Shb), S6c contains 4 amino acid replacements (Ser* j Thr, Lys4 ti Asn, Lys9 + Glu, Tyr*3 a Asn). Pharmacological characterization of S6b/S6c chimeric peptides demonstrated that it is the Lysy = Glu substitution which leads to loss of ETA affinity, making [Glu9]-S6b a selective ETB agonist (7). We have demonstrated that bis-penicillamine ET analogs which contain substitutions for Asp’* provide ETA receptor antagonists (8). With the expectation that a similar strategy would allow the preparation of ETB selective antagonists, we prepared bis-penicillamine [Gl@]-S6b analogs which contain substitutions for Asp ‘8. In this report, we describe the synthesis of these analogs and their
*To whom correspondence and reprint requests should be addressed at Department of Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute, P.O. Box 4000, Princeton, NJ 085434000, USA. (609) 252-4989, FAX (609) 252-6804. Email, [email protected]
1252
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
pharmacological properties. The behavior of these peptides in vasodilator and vasoconstrictor preparations suggests that their actions are mediated by different subtypes of ETR receptors.
ETR
Methods [Pent~lr, Nle7, Glu”, Ala18]-S6b (BMS-184696) and [Penr*tl, Nle’, Glu9, Leu18]-S6b (BMS1X4697) were prepared using previously described procedures (8). Peptides were purified to > 95% homogeneity as analyzed by HPLC and were characterized by FAB mass spectrometry and amino acid analysis. ET- 1, ET-2, and ET-3 were purchased from Peptides International and S6c from Peninsula Laboratories. Stock solutions were prepared immediately before use by dissolving the peptides in a small amount of 1 M acetic acid which was subsequently adjusted to the appropriate volume with water containing 0.05% bovine serum albumin. Binding experiments were performed as previously described (9). New Zealand White rabbit carotid arteries and saphenous veins were handled as described elsewhere (5,lO). Wistar rats were killed by inhalation of CO2. The aortae were carefully removed, cut into rings, and mounted at 5 grams preload for isometric force recording. The presence of a functional vascular endothelium was confirmed by contracting the tissues with 30 nM norepinephrine, then quantifying the relaxation observed upon addition of 1 PM acetylcholine. Only rings that relaxed more than 80% in response to acetylcholine were used. The EC50 values were calculated by linear regression analysis. KR values were determined by analysis of the dose-ratios (11). Schild analysis was used to calculate pA2 values.
Results and Discussion ]Penl.tl, Nle’, Glu”]-S6b analogs were prepared containing either Ala (BMSlX4696) or Leu (BMSlX4697) in place of Asp t8. This bis-penicillamine structure forces the almost exclusive formation of the natural l-15, 3-11 disulfide isomer (12). The combination of a I,1 I-bispenicillamine and an l&position Ala or Leu, which in ET-l produced ETA antagonists (X), was incorporated into [Glu”]-S6b, an ETR selective framework, with the aim of producing ETR antagonists. Both peptides were tested for: binding affinity in preparations containing either ETA or ETR receptor subtypes, vasoconstrictor activity in ETA- or vasoconstrictor ETR-containing tissues, and ETR-mediated vasodilator activity. Using an ETA receptor preparation (rat A10 vascular smooth muscle cell membranes), both BMS- 1X4696 (Ki = 6.6 f 3.4 nM) and BMS- 184697 (Ki = 20 + 2.8 nM) were high affinity ligands. These ETA affinities are somewhat higher than that reported for [Glu9]-S6b (IC50 = 115 nM (7)) but considerably higher than those reported for S6c (1.2 yM (7); > 5 PM (6)). We have previously shown that vasoconstriction of rabbit carotid artery rings is mediated by ETA receptors and that S6c, at concentrations as high as 300 nM, does not contract this tissue (5). As expected, neither [Glu9]Shb analog showed any vasoconstrictor activity at the highest concentration tested (3 PM BMS1X4696 and 10 PM BMS-lX46Y7). As shown in Figure 1, each peptide was able to competitively antagonize the contractile effects of ET- 1 in the rabbit carotid artery (BMS- 1X4696, pA2 = 7.3, slope = -0.Xx; BMS-lX46Y7, PA;! = 6.3, slope = -1.05). BMSlX46Y6 was a particularly high affinity ligand for the ETR receptors present in rat cerebellum (Ki = 0.20 + 0.01 nM), with binding affinity essentially identical to S6c (Ki = 0.20 f 0.01 nM). BMS-184697 showed weaker but nonetheless substantial affinity for the cerebellar ETR subtype (Ki = 22 do2.2 nM). We have shown that rabbit saphenous vein contains vasoconstrictor receptors which have ETe-like characteristics, including high sensitivity to the agonist S6c (5). As shown in Figure 2, both BMS-184696 and BMS-184697 were potent constrictors of this tissue, with EC50 values roughly in line with their respective ETR Ki values in rat cerebellar tissue (Et& = O.Y5 + 0.15 nM BMS-184696 and 230 + 20 nM BMS-184697). The EC50 value for S6c was 0.13 f 0.03 nM in this assay.
-II
1253
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
-9 -10 [ET- I], log M
-8
-9 -IO [ET-l], log M
-II
-7
-8
-7
Fig. 1 Cumulative concentration-response curves to ET- 1 obtained in rabbit carotid artery rings in the absence (0) and presence of 0.3 (O), 1 (W), and 3 (A) PM BMS184696 (A) or 1 (O), 3 (m), and 10 (A) PM BMS- 184697 (B). Data are plotted as mean + SEM; n = 3 - 4 rings from different rabbits.
-II
-10
-9
-8
-7
-6
-5
[Pcptidc], log M
Fig. 2 Cumulative concentration response curves to ET-l (0) BMS-184696 (a), and BMS-184697 (A) obtained in rabbit saphenous vein rings. Data are plotted as mean f SEM; n = 4 - 8 rings from different rabbits. Vasodilator ETn receptors are located on vascular endothelium and their activation can lead to dilation of underlying vascular smooth muscle cells by the release of nitric oxide (NO), prostacyclin, or other vasodilatory substances (13). We developed an assay to test for agonists and antagonists of the vasodilator ETB receptor. The agonist assay monitors the ability of a test compound to cause relaxation of contractions elicited by 40 nM norepinephrine (NE, EC& in rat aortic rings with intact endothelium. In this assay, S6c caused a maximum of approximately 50% relaxation even in rings in which acetylcholine caused 100% relaxation. The antagonist assay compares the concentration
Constrictor and Dilator ET, Receptors
1254
Vol. 56, No. 15, 19%
relaxation curves for S6c in the absence and presence of test compound; test compounds are added 20 minutes before the rings are contracted with 40 nM NE. When force plateaus, the concentrationresponse curve for S6c is initiated. The antagonist assay is relatively specific for ETB receptor antagonists because even at high concentrations none of the following agents blunted the S6c-elicited relaxation: the selective ETA receptor antagonist BMS-182874 (14) (30 PM), the calcium channel blocker nifedipine (0.1 PM), the thromboxane A2 receptor antagonist SQ 30,741 (1 PM), or the angiotensin AT1 receptor antagonist SR 47436 (15) (10 PM). Removal of the endothelium or treatment of the aorta with NO synthesis inhibitors L-NNA or L-NMMA or with the guanyl cyclase inhibitor methylene blue abolished the S6c-induced relaxations suggesting that NO release from the vascular endothelium was responsible for the ETg-mediated relaxation (data not shown).
-11
-10
-9
-8
-7
-6
-5
-11
-10
-9
-x
-7
[SAC],log M
[Peptides]. log M Fig. 3
A. Cumulative concentration response curves to S6c (0), BMS-184696 (Cl), and BMS- 184697 (A) obtained in rat aortic rings. B. Cumulative concentration response curves to S6c obtained in rat aortic rings in the absence (0) and presence of 0.1 (0) 0.3 (M), and 1 (A) nM BMS-184696. Data are plotted as mean + SEM; n = 4 - 8 rings from different rats. When tested as an agonist, BMS-184696 showed essentially no vasodilatory activity at concentrations up 10 nM (Figure 3A). In a separate experiment, a single 1 PM concentration of BMS-184696 also did not evoke relaxation in the NE-contracted rat aorta (data not shown). In contrast, BMS-184697 was a full agonist with moderate potency (EC50 = 82 + 33 nM) compared with S6c (EC50 = 0.22 I!Z0.05 nM). Because BMS-184696 bound to the ETB receptor in rat cerebellar membranes with a high affinity, but did not behave as an agonist for inducing relaxation of the rat aorta, we examined its ability to blunt the relaxation elicited by S6c. When tested as an antagonist, BMS- 184696 blocked the S6c response in a basically all-or-none fashion, producing no block at 0.1 nM but a full block at 0.3 nM (Figure 3B). Several explanations for the non-competitive nature of the vasodilatory antagonist behavior of BMS-184696 must be considered. While it is possible that the BMS-184696 is a nonspecific vasodepressant or alpha-receptor blocker, this seems unlikely because the compound did not blunt NE contractions when it was tested as an agonist. Another possibility is that BMS- 184696 stimulated constrictor receptors as well as relaxant receptors in the rat aorta, i.e., that it functionally antagonized its own actions. However BMS-184696 is clearly not an ETA agonist and we have no evidence for constrictor ETB receptors in the rat aorta (10). Finally, BMS-184696 may have interacted directly with S6c, however such an interaction is not apparent in the rabbit saphenous vein assay. BMS- 1846Y6 is a potent ETA antagonist, an extremely potent vasoconstrictor ETB agonist, and a non-competitive vasodilator ETB antagonist with little or no agonist activity. BMS-184697 is a potent ETA antagonist, a vasoconstrictor ETn agonist, and a vasodilator ETB agonist of moderate
1255
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
potency. The differential ability of two potent vasoconstrictor ETB agonist peptides to either activate (BMS-184697) or not activate and in fact antagonize (BMS-184696) vasodilator ETB receptor preparations strongly suggests that ETB receptor subtypes exist. The putative ETB receptor subtypes may in fact be products of the same gene, since evidence to date supports the existence of only two ET receptor genes accounting for the ETA and ETB receptor subtypes. If this is the case, the existence of subtypes may be due to different post-translational modification, different G protein populations in endothelial versus smooth muscle cells or alternate transcripts. TABLE 1. Biological Activity of BMS- 184696 and BMS- 184697 BMS- 184697 BMS- 184696 ETA Receptor
Ki BCso KB
ETn Receptor Ki EC50 - contraction ECso - relaxation KB - relaxation Data, in nM, shown as mean f SE.
6.6 + 3.4 > 3000 77f
10
0.20 f 0.01 0.95 + 0.15 > 1000 non-competitive
20 + 2.8 > 10000 460 f 43 22 f 2.2 230 f 20 82+33 -
Relatively little published evidence supports the existence of ETB receptor subtypes. IRL 103X was reported to discriminate between the two ETB receptors (16), but unfortunately that paper was retracted because results differed among various batches of IRL 1038 (17). RES-701-l may distinguish between ETn receptor subtypes because it blocks the relaxant effects of ET in rat aorta more effectively than the contractile effects of S6c in rabbit saphenous vein (18.19). PD 142893 has been shown to block the constrictor effects of S6c in rabbit pulmonary artery (pA2 = 5.85 (20)). These data were confirmed and extended in a report claiming that 10 pM PD 142893 had essentially no effect when ET-I was used as the constrictor stimulus in pulmonary artery, however 1 pM PD 142893 strongly antagonized the vasodilation elicited by S6c in rat isolated perfused mesentery (2 1j. The large difference in potency for PD 142893 blockade of constrictor and dilator effects has been suggested as evidence for ETB receptor subtypes. Using the assays described herein, we have been unable to duplicate this difference in potency. In our hands, PD 142893 blocks the constrictor response to S6c in rabbit saphenous vein (KB = 9.9 _+3.8 PM) at concentrations similar to those required to blunt the S6c dilator response in rat aorta precontracted with NE (KB = 21 + 14 PM). Thus, either PD 142893 does not discriminate between ETB receptor subtypes or the receptor mediating dilation in the rat aorta is different from that in the rat mesentery. Structure-activity studies of the ETA receptor using full length endothelin peptides has demonstrated the critical nature of Asp’s to receptor activation (X,22,23). This study further confirms this requirement but also indicates that an acidic residue at position 18 is not a required feature of either a vasoconstrictor ETB agonist (both BMS-184696 and BMSlX4697) or a vasodilator ETn agonist (BMS- 184697). In summary, in this study we attempted to prepare ETB selective antagonists by incorporating into ETn selective peptides substitutions which led to ETA antagonists. While only partially successful, the behavior of these peptides in BTn-containing tissues provides evidence that the ETn receptors which mediate vasocontraction in the rabbit saphenous vein and vasodilation in the rat aorta are of different subtypes. Acknowledgements The authors thank Mr. Eddie C.-K. Liu for expert technical assistance critical reading of the manuscript.
and Dr. Maria Webb for a
1256
Constrictor and Dilator ET, Receptors
Vol. 56, No. 15, 1995
References
1. 2. 3. 4. _5 . 6. 7. x. Y. IO. II. 12. 13. 14. 15. 16. 17. 1X. 19. 20. 21. 22. 23.
M. YANAGISAWA, H. KURIHARA, H. KIMURA, Y. TOMOBE, M. KOBAYASHI, Y. MITSUI, Y. YAZAKI, K. GOT0 and T. MASAKI, Nature m 411-415 (1988). H. ARAI, S. HORI, I. ARAMORI, H. OHKUBO and S. NAKANISHI, Nature 34x 730-732 (1990). T. SAKURAI, M. YANAGISAWA, Y. TAKUWA, H. MIYAZAKI, S. KIMURA, K. GOT0 and T. MASAKI, Nature 348 732-735 (1990). R. TAKAYANAGI, K. KITAZUMI, C. TAKASAKI, K. OHNAKA, S. AIMOTO, K. TASAKA, M. OHASHI and H. NAWATA, FEBS Letters 282 103-106 (1991). S. MORELAND, D.M. MCMULLEN, C.L. DELANEY, V.G. LEE and J.T. HUNT, Biochem. Biophys. Res. Comm. 184 1OO-106 (1992). D.L.J. WILLIAMS, K.L. JONES, D.J. PETTIBONE, E.V. LIS and B.V. CLINESCHMIDT, Biochem. Biophys. Res. Comm. 175 556-561 (1991). C. TAKASAKI et al., Eur. J. Pharmacol. 19x 165-169 (1991). J. T. HUNT et al., Bioorganic Med. Chem. 159-65 (1993). M.L. WEBB, E.C.-K. LIU H. MONSHIZADEGAN, C.C. CHAO, J. LYNCH, S.M. FISHER and P.M. ROSE, Mol. Pharmacol. 44 959-965 (1993). S. MORELAND, D.M. MCMULLEN, 8. ABBOA-OFFEI and A. SEYMOUR, Br. J. Pharmacol. 112 704-708 (1994). E. ARIENS and J. VAN ROSSUM, Arch. Int. Pharmacodyn. Ther. 110 275-299 (1957). J.T. HUNT, V.G. LEE, E.C.K. LIU, S. MORELAND, D. MCMULLEN, M.L. WEBB and M. BOLGAR, Int. J. Peptide Protein Res. 42 249-258 (1993). J.R. VANE and R.M. BOTTING, J. Physiol. Pharmacol. 44 (Suppl. 215-36 (1993). P.D. STEIN et al., J. Med. Chem. 37 329-331 (1994). C. CAZAUBON et ul., J. Pharmacol. Exp. Ther. 265 826-834 (1993). S.A. SUDJARWO, M. HORI, M. TAKAI, Y. URADE, T. OKADA and H. KARAKI, Life Sciences 53 43 l-437 ( 1993). Y. URADE, Y. FUJITANI, K. ODA, T. WATAKABE, I. UMEMURA, M. TAKAI, T. OKADA , K. SAKATA and H. KARAKI, FEBS Letters 342 103 (1994). H. KARAKI, S. A. SUDJARWO, M. HORI, T. TANAKA and Y. MATSUDA, Eur. J. Pharmacol. 262 255-259 (1994). S. A. SUDJARWO, M. HORI, T. TANAKA, Y. MATSUDA, T. OKADA and H. KARAKI, Biochem. Biophys. Res. Comm. 200 627-633 (1994). W. L. CODY et al., J. Med. Chem. 3 3301-3303 (1992). T.D. WARNER, G.H. ALLCOCK, R. CORDER and J.R. VANE, Br. J. Pharmacol. 110 7777X2 (1993). K. NAKAJIMA et al., Biochem. Biophys. Res. Comm. 163 424-429 (I 9X9). J.T. HUNT, V.G. LEE P.D. STEIN, A. HEDBERG, E.C.-K. LIU, D. MCMULLEN and S. MORELAND, Bioorganic Med. Chem. Lett. 133-38 (1991).