1142
J. Med. Chem. 1996, 39, 1142-1147
Antagonistic Properties of Centrally Truncated Analogs of [D-Trp32]NPY A. Balasubramaniam,*,† M. Ujhelyi,‡ M. Borchers,† Y. Huang,† W. Zhai,† Y. Zhou,§ M. Johnson,§ S. Sheriff,† and J. E. Fischer† Division of Gastrointestinal Hormones, Department of Surgery, Colleges of Medicine and Pharmacy, University of Cincinnati, Cincinnati, Ohio 45267-0558, and Center for Pharmaceutical Biotechnology and Department of Medicinal Chemistry, College of Pharmacy, University of Illinois at Chicago, Chicago, Illinois 60612-7230 Received July 21, 1995X
We have previously shown that [D-Trp32]NPY can competitively antagonize NPY-induced feeding in rats (Balasubramaniam et al. J. Med. Chem. 1994, 37, 811-815). This peptide, however, did not bind to SK-N-MC cells with Y-1 receptors. Since centrally truncated NPY analogs have been shown to bind Y-1 receptors, we synthesized similar analogs of [D-Trp32]NPY and investigated their Y-1 (SK-N-MC) and Y-2 (SK-N-BE2) receptor affinities and their properties in human erythroleukemia (HEL) cells. None of the analogs with D-Trp32 mobilized intracellular calcium, [Ca2+]i, in HEL cells. Although Des-AA6-24[Aoc6]NPY and the corresponding D-Trp32 analog exhibited no affinity to Y-1 receptors, Des- AA7-24[Aoc6,D-Trp32]NPY (6) exhibited weak binding. Replacing Pro5 in 6 with D-Ala to stabilize the central chain reversal, and hence the antiparallel alignment of the N- and C-terminal regions known to be important for Y-1 binding, resulted in an analog, Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY (7), which exhibited moderate antagonist potency in attenuating NPY effects on cAMP and [Ca2+]i in SK-N-MC and HEL cells, respectively. This analog also shifted the dose-response curve of NPY on blood pressure in anesthetized rats. Deletion of only the 7-17 and/or the incorporation of N-Me-Ala5, a superior β-turn stabilizer, in 7 did not improve the Y-1 receptor affinity. Des-AA7-24[D-Ala5,Gly6,DTrp32]NPY exhibited an affinity similar to that of 7, suggesting that a long spacer arm is not necessary for efficient Y-1 receptor interaction. Locking the antiparallel alignment via a 2/26 or 2/27 lactam bridge did not improve the binding. Finally, replacement of D-Ala5 in 7 with D-Trp dramatically increased both the binding and the antagonistic potencies. Modeling based on the avian pancreatic polypeptide X-ray structure suggested that analogs which have the Nand C-terminal regions in close proximity might exhibit good binding, and that the D-Trp32 substitution may induce a β-turn that could be important for exhibiting antagonism. A systematic investigation has resulted in the development of relatively potent Y-1 receptor antagonists. Further structure-activity studies with these compounds and those previously reported by us and other investigators should result in the development of long-acting and receptor selective antagonists. Introduction Neuropeptide Y (NPY) is a 36-residue peptide amide isolated originally from porcine brain.1 It is widely distributed in both the central and peripheral nervous system, and is now considered to be one of the most abundant peptides present in the mammalian brain and the heart. The sequence homology and the overlap of the biological functions of NPY with peptide YY (PYY) isolated from the intestine and the pancreatic polypeptide (PP) in the pancreas have led to the classification of these peptides into a family of homologous hormones. Moreover, NPY and PYY have been suggested to have a structure similar to that of aPP, consisting of a 1-8 polyproline type II helix segment kept antiparallel to the 16-32 R-helical region by the central 9-14 loop. The last four residues of aPP adopt a flexible turn conformation in the crystal structure.2 However, solution NMR studies by us3 and others4 revealed that the helical structure of NPY extends to the 36th residue, and also * Author to whom correspondence should be addressed at: Division of Gastrointestinal Hormones, Department of Surgery, College of Medicine, University of Cincinnati, Cincinnati, OH 45267-0558. TEL: (513) 558-3819. FAX: (513) 558-3474. E-mail: Ambi.bala.uc.edu. † College of Medicine, University of Cincinnati. ‡ College of Pharmacy, University of Cincinnati. § College of Pharmacy, University of Illinois at Chicago. X Abstract published in Advance ACS Abstracts, February 1, 1996.
0022-2623/96/1839-1142$12.00/0
no evidence for the antiparallel alignment of the polyproline region and the helical segment was found. Since the initial isolation of neuropeptide Y (NPY), the pharmacological properties of this hormone have been vigorously investigated in many laboratories. These studies have characterized NPY as one of the most potent orexigenic and vasopressor peptides isolated to date (see ref 2 for a review). NPY receptors mediating these actions have been broadly classified into three main groups, Y-1, Y-2, and Y-3.5,6 However, there is mounting evidence for the existence of receptors with unique properties that would not permit classification of them into any of the known receptor categories.7,8 To date, the NPY receptors have been classified on the basis of the pharmacological actions and/or the sequence requirements to elicit a particular effect. For example, although both NPY and PYY interact with both Y-1 (postsynaptic) and Y-2 (presynaptic) receptors with nearly equal affinity, only Y-2 receptors can interact with C-terminal partial sequences with high affinity. On the other hand Y-3 receptors exhibit higher affinity for NPY than PYY. While [Leu31,Pro34]NPY and NPY13-36 have been found to be useful selective agonists of Y-1 and Y-2 receptors, respectively, the identification of potent and selective antagonist agents is lacking. Development of receptor selective antagonists is one of the important © 1996 American Chemical Society
NPY Receptor Antagonists
Journal of Medicinal Chemistry, 1996, Vol. 39, No. 5 1143
Table 1. Properties of Truncated Analogs of [D-Trp32]NPY IC50 (µM)a
no.
peptide
1 2 3 4 5 6 7 8 9 10 11 12 13
NPY [Leu31,Pro34]NPY [D-Trp32]NPY106 Des-AA6-24[Aoc5]NPY Des-AA6-24[Aoc5,D-Trp32]NPY Des-AA7-24[Aoc6,D-Trp32]NPY Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY Des-AA7-24[N-Me-Ala5,Aoc6,D-Trp32]NPY Des-AA7-17[N-Me-Ala5,Aoc6,D-Trp32]NPY Des-AA7-24[D-Ala5,Gly6,D-Trp32]NPY Des-AA7-24[D-Trp5,32,Aoc6]NPY cyclo(2/26)des-AA8-24[Glu2,D-Ala6,Aoc7,Lys26,D-Trp32]NPY cyclo(2/27)des-AA7-24[Asp2,D-Ala6,D-Lys27,D-Trp32]NPY
MH+
(calcd)
4254.8 (4254.7) 4222.5 (4223.7) 4341.3 (4339.8) 2248.5 (2349.7) 2335.2 (2334.8) 2430.5 (2431.9) 2405.3 (2405.9) 2419.8 (2419.8) 3244.5 (3244.8) 2320.2 (2321.7) 2519.7 (2520.9) 2507.5 (2508.0) 2325.8 (2326.7)
agonist antagonist Y-1 (SK-N-MC) Y-2 (SK-N-BE2) activityb activityc 0.0006 ( 0.0004 0.0010 ( 0.0008 NI g10 >10 4.60 ( 2.10 2.00 ( 0.90 1.63 ( 0.20 g10 1.10 ( 0.80 0.10 ( 0.06 2.20 ( 1.0 2.90
0.0001 ( 0.000 2.00 ( 0.50 0.35 ( 0.20 1.50 0.49 ( 0.23 0.013 ( 0.006 2.20 ( 2.00 0.50 ( 0.37 0.30 ( 0.01 1.20 ( 0.75 0.10 ( 0.07 0.09 ( 0.05 0.65
100 100 NT NT NT 0 0 0 NT 0 0 0 NT
NT NT NT 95% homogeneous by analytical reversed phase chromatography. [D-Trp32]NPY, which has been shown by us to antagonize NPY-induced feeding in rats,11 failed to bind to Y-1 receptor, but exhibited moderate Y-2 receptor affinity (Table 1). Although previous studies indicate that Y-1 receptors mediate NPY-induced feeding,22 our observation with [D-Trp32]NPY is suggestive of the involvement of a receptor with Y-2 binding characteristics in NPYinduced feeding. This inference is also in agreement with the finding by Gehlert et al.23 that rat hypothalamus exhibits Y-2 binding characteristics. However, a
number of Y-2 ligands, NPY(13-36), NPY(17-36), and NPY(18-36) as well as centrally truncated analog DesAA6-24[Aoc5]NPY (4) in Table 1, neither stimulated feeding nor antagonized NPY-induced feeding in rats (Balasubramaniam & Chance, unpublished results). It appears therefore that feeding effects of NPY are mediated by yet another unidentified receptor system. Krstenansky et al.24 and Beck et al.25 demonstrated that the centrally truncated analogs of NPY exhibit good agonist potencies in a number of systems. Subsequently, Fishman and co-workers21 showed that one of these analogs bound and inhibited cAMP synthesis in SK-N-MC cells with good potencies. We therefore synthesized similar truncated analogs of [D-Trp32]NPY, investigated their Y-1 and Y-2 receptor affinities, and determined their effects on [Ca2+]i in HEL cells. Although Des-AA6-24[Aoc5]NPY (4) and the corresponding D-Trp32 analog (5) did not bind to Y-1 receptors, Des-AA7-24[Aoc5,D-Trp32]NPY (6) exhibited some degree of binding to the Y-1 receptors. The failure of analogs 4 and 5 to display any receptor affinity may be due to the improper orientation of the N- and C-terminal regions because, on the basis of their investigations with cyclic analogs, Krstenansky et al.24 and subsequently Rivier and co-workers26 have suggested that N- and C-terminal regions should be in close proximity for efficient binding. Similar conclusions have also been drawn on the basis of modeling studies using the avian pancreatic polypeptide (aPP) X-ray structure.2 On the basis of these observations, we reasoned that the stabilization of the antiparallel alignment in 6, Tyr1Pro-Ser-Lys-Pro5---Aoc---Arg25-His-Tyr-Ile-Asn-Leu30Ile-D-Trp32-Arg-Gln-Arg35-Tyr-NH2, may increase its receptor affinity. Although Pro5 in 6 may be contributing toward the antiparallel alignment, it is also known that Pro residues may destabilize the conformation due to conformational isomerization.27 Since D-residues at the i + 1 positions are known to stabilize the β-turn, we replaced the Pro5 with D-Ala, and as expected this analog, Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY (7), exhibited increased receptor affinity (Table 1). Moreover, 7 did not exhibit any agonist activity, and at a concentration of 1.0 and 10 µM shifted the dose-response curves of NPY on cAMP synthesis and [Ca2+]i in SK-N-MC and HEL cells, respectively (Figure 1), suggesting that 7 is a competitive antagonist of NPY in Y-1 receptors.
1144
Journal of Medicinal Chemistry, 1996, Vol. 39, No. 5
Figure 1. Antagonism of NPY effects on isoproterenol (10 µM)-stimulated cAMP synthesis by SK-N-MC cells (A) and intracellular calcium mobilization by fura-2-loaded HEL cells (B) by Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY, 7. Representative concentration-response curves of NPY (O), antagonist (0), and NPY in the presence of the antagonist (b) are shown for both SK-N-MC (A) and HEL (B) cells. Concentration of antagonist: 1.0 µM for SK-N-MC cells and 10 µM HEL cells. cAMP studies in SK-N-MC cells were carried out according to the published procedures of Fishman and co-workers (ref 21).
Since 7 exhibited a moderate antagonistic potency in HEL cells, and because the activity in HEL cells correlated well with blood pressure responses in rats,28 we also investigated the antagonist properties of 7 in anesthetized rats. In preliminary studies, 7 (100 nmol/ kg) exhibited neither agonist nor antagonist activities (not shown). However, at a dose of 200 nmol/kg, 7 significantly inhibited the blood pressure responses induced by 0.5, 1.0, and 3.0 nmol/kg doses of NPY (Figure 2A). The fact that NPY (10 nmol/kg) surmounted the antagonistic effects of 7 confirmed the competitive nature of this antagonism. This attenuation of the NPY-induced blood pressure responses by 7 is not due to the desensitization of NPY receptors because, in control experiments, a comparable NPY dose-resonse curve was obtained in rats which had previously received random doses of NPY (not shown). Experiments also revealed that the antagonistic effects of 7 lasted more than 40 min (Figure 2B). To understand the structural features required for binding and exhibiting antagonistic properties at Y-1 receptors, we did modeling studies based on the X-ray structure of aPP as described in the Experimental Section.11 These preliminary investigations suggested
Balasubramaniam et al.
Figure 2. (A) Antagonism of NPY-induced systolic blood pressure response in anesthetized rats by 0.20 µmol/kg of DesAA7-24[D-Ala5,Aoc6,D-Trp32]NPY, 7. Dose-response curves of NPY in the absence (O) and presence (b) of antagonists are shown. The change in blood pressure at each NPY dose in the presence and the absence of antagonist were statistically compared using a paired t test. (B) Determination of the duration of the antagonistic effect of Des-AA7-24[D-Ala5,Aoc6,DTrp32]NPY on NPY-induced blood pressure response in anesthetized rats. Following one 0.20 µmol/kg dose of the antagonist at time 0, repeated 0.001 µmol/kg doses of NPY were administered for every 10 min for 90 min. The change in systolic blood pressure response at each time point is shown separately for three rats. See Experimental Sections for details.
that the N- and C-terminal regions are improperly oriented in 4 and 5, while they are far apart in 6 despite the correct alignment. On the other hand, the N- and C-terminal regions were found to be in close proximity in Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY (7) which exhibited moderate antagonistic potency (Figure 3). These findings are consistent with previous observations2,21,24,26 and suggest that a stable turn structure in the central region may be very important for efficient interaction with Y-1 receptors. It is also tempting to speculate that as in the hypothalamus NPY receptors,11 a turn structure in the C-terminal region induced by D-Trp32 may be important for exhibiting antagonistic properties at Y-1 receptors. Further studies are required to validate this hypothesis. We performed further structure-activity studies with 7 to improve the receptor affinity and determine the structural requirements for binding to Y-1 receptors. Since N-Me-Ala has been reported to be a superior β-turn stabilizer,29 we replaced D-Ala5 in 7 with N-MeAla. However, the receptor affinity and the antagonist potency of this analog, Des-AA7-24[N-Me-Ala5,Aoc6,DTrp32]NPY (8), did not improve compared to 7. Rivier and co-workers have shown that the central residues may play an important role in binding to Y-1 receptors.26 Accordingly, we synthesized Des-AA7-17[N-MeAla5,Aoc6,D-Trp32]NPY (9) containing seven more of the central residues (AA18-24). Although our modeling studies suggested that the N- and C-terminal regions could be properly oriented, this analog exhibited hardly any binding to Y-1 receptors. It is possible that the destabilization of the antiparallel conformation due to
NPY Receptor Antagonists
Journal of Medicinal Chemistry, 1996, Vol. 39, No. 5 1145
Figure 3. Stereoview of the Des-AA7-24[D-Ala5,Aoc6,D-Trp32]NPY model based on the aPP crystal structure (ref 2). Model was constructed as described in the Experimental Section.
the cis-trans isomerization induced by N-Me-Ala30 may contribute to the lower affinity of 9. The importance of the long spacer to link the N- and C-terminal regions was investigated by replacing Aoc6 with Gly. This analog, Des-AA7-24[D-Ala5,Gly6,D-Trp32]NPY (10) in fact exhibited greater receptor affinity and an antagonist potency comparable to that of 7. It appears therefore that in contrast to previous reports24-26 a long spacer group in the central region is not critical for binding to Y-1 receptors. Modeling, in fact, suggests that a shorter linker could provide a more stable structure. A number of investigations have found D-Trp to stabilize β-turns and impart greater agonist/antagonist potencies (e.g. refs 11, 31, 32). Therefore, we believed that the central chain reversal involved in the antiparallel alignment of N- and C-terminal regions could be strengthened by substituting D-Trp for D-Ala5. This analog, Des-AA7-24[D-Trp5, 32,Aoc6]NPY (11), antagonized the effects of NPY on Ca2+ in HEL cells and displayed a dramatic improvement in Y-1 receptor affinity compared to 7 (Table 1). Locking the antiparallel alignment through cyclization via lactam bridges26 or disulfide bonds24 have also been reported to enhance NPY receptor affinities. Therefore, we synthesized similar analogs with D-Trp32, cyclo(2/26)des-AA8-24[Glu2,D-Ala6,Aoc7,Lys26,D-Trp32]NPY (12) and cyclo(2/ 27)des-AA7-24[Asp2,D-Ala6,Aoc7,D-Lys27,D-Trp32]NPY (13). Although the cyclization points appear to be appropriate from modeling studies, these analogs, 12 and 13, did not exhibit much improvement compared to 7 and were much less potent than 11. Some of the analogs, 6, 11, and 12, also exhibited good Y-2 receptor affinity (Table 1). Moreover, 6 appears to be selective to Y-2 receptors. It is, therefore, desirable to investigate the functional properties of these analogs in Y-2 model systems. However, we were unable to classify the analogs into agonist or antagonist peptides with confidence because NPY inhibited the cAMP synthesis in SK-N-BE2 by
