J.Med. Chem. 1996,38,3454-3461
3484
Expedited Articles Cyclic Lactam a-Melanotropin balogues of Ac-Nle4-cyclo{Asp5,~Phe7,Lys10] a-Melanocyte-StimulatingHormone-(4- 10)-NH2 with Bulky Aromatic Amino Acids at Position 7 Show High Antagonist Potency and Selectivity at Specific Melanocortin Receptors Victor J. Hruby,*lt Dongsi Lu,* Shubh D. Sharma; Ana de L. Castrucci,' Robert A. Kesterson,g Fahad A. Al-Obeidi,+Mac E. Hadley,' and Roger D. Cone**$ Departments of Chemistry and Cell Biology and Anatomy, University of Arizona, Tucson, Arizona 85721, and Vollum Institute for Advanced Biomedical Research, Oregon Health Sciences University L-474, Portland, Oregon 97201 Received May 18,1995@
The cloning of the melanocyte-stimulating hormone (MSH) and adrenocorticotropic hormone (ACTH) receptors (MC1-R and MC2-R, respectively) recently has led to the identification of three additional melanocortin receptors, MC3-R, MCCR, and MC5-R. The MC2 receptor primarily recognizes only ACTH peptides, but the other four receptors all recognize a-melanocyte-stimulating hormone (a-MSH) and potent a-MSH agonists such as [Nle4,~-Phe7]a10)-NH2as well as ACTH. The absence MSH-NH2 and A~-Nle~-c[Asp~,~-Phe~,Lys~~]a-MSH-(4of any known physiological role for these new receptors, expressed both in the brain (MC3-R and MC4-R) and throughout a number of peripheral tissues (MC5-R),has necessitated a search for potent and receptor selective agonists and antagonists. We report here that analogues of in the superpotent cyclic agonist analogue Ac-Nle4-c[Asp5,~-Phe7,Lys101a-MSH-(4-10)-NH~, which a bulky aromatic amino acid is substituted in the 7-position, can produce potent and selective antagonists for melanocortin receptors. Thus, the ~-p-iodophenylalanine~-containing analogue A~-Nle~-c[Asp~,~-Phe(pI)~,Lys~~]a-MSH-(4101-NH2 is a potent antagonist (pA2 = 10.3) in the classical frog skin (Rana pipiens) assay (MC1-R), as is the D-2'-na~hthylalanine~ (DNal(2)7)-containinganalogue A~-Nle~-c[Asp~,~-Nal@)~, Ly~~~la-MSH-(4-10)-NH2 (pAz > 10.3). Interestingly, the D-p-chloro- and ~-p-fluorophenylalanine~-containing analogues lacked antagonist activities at all melanotropin receptors, and both exhibited full agonist potency in the frog skin assay. The activity of these analogues also was examined at four mammalian melanocortin receptors. Interestingly, A~-Nle~-c[ksp~,(D-Nal(2)~, Lysl01a-MSH-(4- 10)-NH2was found to be a potent antagonist of the MC4-R (pA2 = 9.3) with minimal agonist activity, a less potent antagonist of the MC3-R (pAz = 8.3) with minimal agonist activity, and a full agonist was found of the MC1 and MC5 receptors. Surprisingly, Nle4-c[Asp5,~-Phe(pI)7,Lys101a-MSH to be a potent agonist at the cloned human MC1-R (EC50 = 0.055 nM) and mouse MC1-R (EC50 = 0.19 nM) but had potent antagonist activities at the human MC4-R (pA2 = 9.7) and human MC3-R (pAz = 8.3) with significant partial agonist activities = 0.57 and 0.68 nM, respectively) as well. Thus, highly potent and receptor selective antagonist analogues can arise from substitution of the D-Phe7 residue with a bulky aromatic amino acid. These analogues can be used to help determine the functional roles of these receptors.
Introduction While pharmacological methods have been traditionally used t o define receptor subtypes, receptor-cloning experiments have often led t o the discovery of novel receptor subtypes within many receptor families. Following the cloning of the melanocyte-stimulating hormone (MSHF and adrenocorticotropic hormone (ACTH) receptor gene^,^,^ for example, three unique yet related genes were identified that also encoded functional, highaffinity receptors for the melanocortin (MSWACTH) * Authors to whom reprints should be addressed at the Department of Chemistry, University of Arizona (V.J.H., chemistry and frog skin assays), and Vollum Institute (R.D.C., cloned melanocortin receptor assays). Department of Chemistry, University of Arizona. Department of Cell Biology and Anatomy, University of Arizona. 8 Oregon Health Sciences University L-474. Abstract published in Advance ACS Abstracts, August 15, 1995.
* +
@
peptides4-13. Labeled numerically in the order of their discovery, the melanocorth-3, melanocortin-4, and melanocortind receptor genes have been demonstrated to be expressed primarily in the hypothalamus, midbrain, and brain stem (MC3-R and MC4-R) or in a wide distribution of peripheral tissues (MC5-R). a-Melanotropin (a-MSH, Ac-Ser-Tyr-Ser-Met-GluHis-Phe-Arg-Trp-Gly-Lys-Pro-Val-NHz) was among the first peptide hormone to be isolated and to have its structure determined. This hormone plays an important biological role in pigmentation,14J5and numerous central nervous system (CNSI-related activities also have been proposed for this hormone.14-17 Extensive structure-activity relationships have established the central role of a-melanotropin in pigmentation in vertebrates, and super potent, enzymatically stable, super prolonged acting agonist analogues such as [Nle4,~-
OQ22-2623/95/1838-3454$09.00/00 1995 American Chemical Society
Cyclic Lqctam a-Melanotropin Analogues
Journal of Medicinal Chemistry, 1995, Vol. 38, No. 18 3455
Table 1. Agonist and Antagonist Activities of Cyclic Melanotropin Analogues at Various Melanocortin 1 Receptors EC60 values (nM) compound frog skin assay mMC1-R assay 0.10 f 0.03 1.3 f 1.4 a-MSH 1,A C - N ~ ~ ~ - C [ A ~ ~ ~ , D - P ~ ~ ( ~ F ) ~ , L (SHU9128) ~ ~ ~ O ] ~ - M0.10 SHf( 3.5 ~ - ~ O ) - N H ~0.026 ~ f 0.010 0.0095 f 0.0053 2, Ac-Nle4-c[Asp5,~-Phe(pCl)~,Lys1~]a-MSH-(4-lO)-NH~ (SHU9203) 2.0 f 0.8 antagonist p A 2 = 10.3 0.19f 1.3 3,A~-Nle~-c[Asp~,~-Phe(pI)~,Lys~~]a-MSH-(4-10)-NH~ (SHU8914) antagonist p A 2 z 10.5 0.039 f 0.029 4, A~-Nle~-c[Asp~,~-Na1(2)~,Lys~~Ia-MSH-(4-10)-NH~ (SKu9119)
hMC1-R assay 0.091 f 0.070 0.016f 0.003 0.005 f 0.004 0.055 f 0.031 0.036f 0.012
a The complete structure of 1 is Ac-Nle4-~p-His-~-Phe(pF)-Arg-Trp-L;rslo-NH2. Compounds 2-4 are the same except for the amino acid in the 7-position.
the synthesis was accomplished on a p-methylbenzhydrylamine resin using an Nu-Boc strategy and via cyclization of the macrocyclic lactam rings on the solid support using (benzotriazol-l-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP)reagent in the presence of diisopropylethylamine as described previously.20 The cyclic, partially protected peptide was deprotected and cleaved from the resin by treatment with HF-anisole for 45 min a t 0 0C.20The resulting crude peptides were purified by reversed-phase HPLC and characterized by fast atom bombardment mass spectrometry and amino acid analysis. Purity was assessed by analytical RP-HPLC and thin layer chromotography (TLC) in three different solvent systems (see the Experimental Section for details). Biological Assay Methods. The analogues were assayed for agonist and antagonist activity using the classical frog skin (Rana pi pi en^)^^ and, in a few cases, the lizard skin (Anolis c a r ~ l i n e n s i s )bioassays. ~~ The analogues also were assayed for agonist and antagonist activities at cloned mammalian melanocortin receptors using a newly developed CAMP-dependent colorimetric 8-galactosidase assay2 (see the Experimental Section for details). This assay uses clonal 293 cell lines expressing human MSH receptor (hMC1-R), human MC3 receptor, human MC4 receptor, and mouse MC5 Results receptor which were transfected with a P-CRE(P-galactosidase) construct using a Cap04 method33 (see the Design and Synthesis. The antagonists reported Experimental Section for details). Antagonist pA2 here are based on the super potent and super prolonged values in all assays were determined using the method acting cyclic ladam-containing agonist Ac-Nle4-c[Asp5,Dof S ~ h i l d . ~ ~ Phe7,Lys10]a-MSH-(4-10)-NH~. Structure-activity and The results for the R . pipiens, human, and mouse conformation-a~tivity~~~~~~~~ studies have led us to MSH (MC1) receptors for the four ~-Phe~-substituted propose a bioactive conformation for a-MSH a t the analogues of A~-Nle~-c[Asp~,~-Phe~,Lys~~]a-MSH-(4 classical pigment cell r e ~ e p t o r . ~These ~t~~ studies indilO)-NHz, namely, the ~ - P h e ( p F )D-Phe(pC1I7, ~, D-Phecate that the side chain residues of His6,Phe7,Ar8,and 1-4, respectively, are (pII7, and ~ - N a 1 ( 2 analogues )~ Trpg are critical for agonist activity. Using strategies given in Table 1. The comparative results for the four we have previously suggested for peptide hormone and mammalian melanocortin receptors are given in Table neurotransmitter antagonist d e v e l ~ p m e n t we , ~ ~have 2. sought ways to disrupt the proposed bioadive conformation necessary for transduction while maintaining strong The analogue A~-Nle~-c[Asp~,~-Phe(pI)~,Lys~~la-MSH binding to the inactive form of the receptor. One of (4-10) (3;Table 1)has only minimal agonist activity these approaches has involved modification of the in the frog skin ( R . pipiens) assay but, as shown in critically important Phe7 residue by a variety of bulky Figure 1,is a potent inhibitor of the biological response aromatic amino acid residues at this position. We report of a-MSH. Evaluation of the dose-response displacehere that substitution of D-2'-naphthylalanine (D-Nalment curves (Figure 1)showed that 3 was an exception(2)) and D-p-iodophenylalaninein position 7 of the potent ally potent antagonist analogue (pAz = 10.3, Table 1) cyclic agonist analogue A~-Nle~-c[Asp~,~-Phe~,Lys~~]ain this in vitro melanocortin 1receptor assay. InterestMSH44- l O ) - N H z produced potent antagonist analogues ingly, the analogue was a potent agonist (ECm = 0.60 at certain melanocortin receptors. nM, data not shown) in the lizard (A. carolinensis) skin assay, a highly potent agonist at the human MC1-R Synthesis of the peptide analogues 1-4 (Table 1)was (EC5o = 55 pM, Table 11,and a modestly potent agonist accomplished by solid-phase methods of peptide chema t the mouse MC1-R (EC50 = 0.19 nM, Table 1). The istry similar to those previously reported for the syncyclic analogue Ac-Nle4-c[Asp5,~-Nal(2)7,Lys101a-MSHthesis of Ac-Nle4-c[Asp5,~-Phe7,Lys101a-MSH-(4-10)(4- 10)-NHz (4) which has the bulky aromatic amino NH2 using an automatic peptide synthesizer. Briefly,
Phe7]a-MSH18and the cyclic lactam analogue Ac-Nle4c[A~p~,~-Phe~,Lys~~la-MSH-(410)-NH219~20 have been developed and widely used in biological studies related to the role of a-melanotropin in pigmentation. The effects of a-MSH on pigmentation are mediated by the MC1-R expressed specifically on the surface of melanocytes. Similarly the MC2-R is involved in the regulation of adrenal steroidogenesis by ACTH. However, given the complexity of expression of the MC3, MC4, and MC5 receptors, it has not been possible to identify any simple correlations between these receptors and the reported biological activities of the melanocortin peptides. Consequently, potent and receptor specific agonists and especially antagonists would be extremely valuable tools for the determination of the physiological roles of the MC3, MC4, and MC5 receptors. Though the extensive structure-activity relationships mentioned above have provided much information on agonist activity related to pigmentary effects (see, for example, refs 14-17,21), until recently there have been only a few report^^^-^^ on the development of a-MSH antagonists. We report here on the discovery of two highly potent and selective antagonists for certain amphibian MC1 receptors and for the mammalian neural MC3 and MC4 receptors.
5458 Journal of Medicinal Chemistry, 1995, Vol. 38, No. 18
Hruby et al.
Table 2. ECeo Values (pM) for D-Phe7-SubstitutedAc-Nle4-c[Aapa,~-Phe7,Lys10~-MSH~(4-10~~~ Andogues at the Different Melanocrotin Receptors ECm values (pM) hMSH1-R hMC3-R hMC4-R dC5-R a-MSH 91 f 69 669 f 365 210 f 57 807 f 125 17 f 18 [Nle4,~-Phe71a-MSH-NH2 23 f 7 132 f 31 NJY 19 f 14 1, SHU9128 (pF) 16 f 3 191 f 9 1360 f 649 117 f 70 2, SHU9203 (pC1) 5f4 63 f 26 18 f 14 55 f 31 1134 f 197 3, SHU8914 (PI) 573 f 357 684 f 227 partial agonist partial agonist partial agonist p A 2 = 8.3 p A 2 = 9.7 4, sHu9119 (~-Na1(2)) 36 f 12 2813 f 575 no activity 434 f 260 full agonist antagonist partial agonist p A 2 = 8.3 p A 2 = 9.3 a ND = not determined.
Modification of D-~henylalanine~ at the para position or substitution of the phenyl group with a naphthyl group (compound 4, Table 1)had little effect on agonist activity of the compounds at the human MSH receptor (hMC1-R) or human MC5 receptor (Figure 2). In contrast, replacement of the D-Phe7 with D-Phe(pI)17 dramatically reduced agonist activity at the MC3 and MC4 receptors (Figure 2). Both compounds are partial I agonists with greatly increased ECSOS (Table 2). Inter0 ; , , ,; , , 1 10-1110-1110.1010.0 10.8 10.7 104 10.6 estingly, the ~-Phe(pCl)~-substituted analogue 2 was a Log [a-luSH] (M) full agonist and actually was more potent than [Ne4,DFigure 1. Demonstration that A~-Me~-c[Asp~,~-Phe(pI)~,Phe71a-MSHat every receptor except the hMC3-R. The Lysl01a-MSH-(4-10)-NH2(3)is a potent antagonist of a-MSH ~-Phe(pF)~-substituted analogue 1 likewise was a very in the frog skin MC1 receptor assay system. Antagonism of potent full agonist in all assays (Table 2, Figure 2). the dose-response curve of a-MSH by (A), (A)and The two compounds acting as weak partial agonists lo-' (0) M antagonist 3 is demonstrated by the rightward shift at the MC3-R and MC4-R (analogues 3 and 4) were then of the dose- response curve of a-MSH. examined for antagonist activity at these receptors (Figure 3). As can be seen, the ~-Nal(2)~-substituted acid D-Nd(2) in position 7 also exhibited potent antagocyclic lactam analogue 4 is a potent antagonist of the nist activity in the frog skin MC1-R assay (pA2 L 10.5, MC3-R and MC4-R, with p A 2 values of 8.3 and 9.3, Table 1)and potent agonist activity at the mouse and respectively (Table 2). Very little agonist activity is seen human MC1-R receptors (Table 1). Interestingly, the D-p-fluorophenylalanine-and D-p-chlorophenylalanine- with these compounds at the hMC4-R. In contrast, the p-iodo-substituted compound 3 is also a potent antagocontaining analogues 1and 2 (Table 1)were both potent nist but retains significant partial agonist activity, agonists at all melanocortin 1receptors. stimulating cAMP-dependent fi-galadoeidase activity to In the frog skin bioassay, it was noted that the 50% of maximal levels at the hMC4-R and hMC3-R as antagonist activity of both 3 and 4 was prolonged in a well (Figure 3). dose-dependent manner. The antagonist effects of both Competition binding experiments (see the Experianalogues can be blocked by pretreating the frog skin mental Section for details) were perfomed to determine with the potent prolonged acting agonist [Nle4,D-Phe7]aif iodination of the phenylalanine aromatic ring or MSH. Concentrations of the antagonists at 10-6-10-7 replacement of the phenyl ring with a naphthyl ring had M generally produced an irreversible antagonism in any effect on the affinity of the cyclic lactam compounds that, following removal of analogues from the assay for the MC3 MC4 receptors (Figure 4, Table 3). No medium, subsequent challenges with a-MSH failed to significant alteration in IC50 values was observed relaactivate frog MC1 receptors for several hours aRer tive to those calculated for [Nle4,~-Phe71-a-MSH. removal of the antagonists. The antagonism of frog skin MC1 receptor was specific for a-MSH since the skins Diecumion would still maximally darken in response to theophylline (a phosphodiesterase inhibitor). These observations These exciting and intriguing results provide new may prove of great importance if any of these antagoinsights into antagonist stfilcture-activity relationships nists were to prove to be of clinicdphysiological usefor melanocortin receptors and point to different reMlness. quirements for antagonist activity at different pigmenA newly described cAMP-dependent colorimetric j3-gatary receptors in different species and at different lactosidase assay2 was used to determine the agonist melanocortin receptors in the same species. Ip previous and antagonist activities of the new cyclic lactam studies from our laboratory, we demonstrated that Acderivatives of a-MSH at cloned mammalian melanocor[D-Trp7,~-Phe101a-MSH-(7-10)-NHz was a very weak antagonist in both the frog skin and lizard skin meltin receptors. This assay utilizes a /3-galactosidase anocortin receptor assays (pA2 = 4.8 and 5.7, respecreporter gene fused to a CAMP-regulated promoter to detect changes in intracellular CAMP downstream of tively), and other closely related analogues were either receptor activation and was useful in these studies since weak agonists or weak antagonists but only at one of it has been shown that all the melanocortin receptors the two receptors.23 In another study, the linear 4-10 a-MSH analogue Ac-Nle-Asp-Trp-D-Phe-Nle-Trp-Lyscouple to the effector adenylyl ~ y c l a s e . ~ - l ~ 501
Journal of Medicinal Chemistry, 1995, Vol. 38, No. 18 3457
Cyclic Lactam a-Melanotropin Analogues A NDP-a-MSH
I
f
0
o SHU9203(pU)
-
A NPP-a-MSH
B1
A SHU8914(p-I)
9-2
i'
hMC1-R
i
L .. A
0.0
-R .ll
.10
a
.9
I
a
.I
-8
LoalcomDoundlIM)
DB
a-MSH NDP-a-MSH SHU9203(pU)
la i'
SHU91lS(D-Nd)
0.50
0.25
-R
..
0 . 0 0 1 I
.13
42
.ti
.io
.o
Log(Compound](h4)
.e
.7
.e
I
ma-w
0.8
0.4
0.3 0.2 0.1
