Analogs of α-melanocyte stimulating hormone with high agonist potency and selectivity at human melanocortin receptor 1b: The role of Trp9 in molecular recognition†
Dedicated to the memory of Dr. Elkan R. Blout in whose lab, at Harvard Medical School, I (M.A. Bednarek) had the privilege to work in the early eighties. At that time Prof. E. R. Blout (frequently called ERB by his coworkers) had numerous responsibilities as a Professor at Harvard Medical School, and also as Dean of Academic Affairs at the Harvard School of Public Health and Treasurer at the National Academy of Sciences. Every week for a half an hour, however, he gave undivided attention to each member of his laboratory, his postdoctoral students and visiting scientists. We all looked forward to those brief but so rewarding interactions. He expected us to be independent scientists, concise and clear about what we wanted to accomplish. He discussed our efforts and judgments, provided criticism and reassurance. And all the time we were absolutely in awe of this brilliant scientist, teacher, mentor, manager, statesman and businessman. He was our rock, always calm, friendly, and kind. We felt very fortunate to be part of ERB's world, participate in his research endeavors, and observe his extraordinary interactions with scientists, managers, business people and politicians. It was certainly a unique experience reaching far beyond the typical postdoctoral training. I treasure memories of Dr. Elkan R. Blout from the lab at Harvard Medical School. Equally warmly, I treasure memories of ERB and his family from the lab gatherings at the Harvard and Commonwealth Clubs or at his house.
Abstract
α-Melanocyte stimulating hormone (αMSH), Ac-Ser1-Tyr2-Ser3-Met4-Glu5-His6-Phe7-Arg8-Trp9-Gly10-Lys11-Pro12-Val13-NH2, is an endogenous agonist for the melanocortin receptor 1 (MC1R), the receptor found in the skin, several types of immune cells, and other peripheral sites. Three-dimensional models of complexes of this receptor with αMSH and its synthetic analog NDP-αMSH, Ac-Ser1-Tyr2-Ser3-Nle4-Glu5-His6-D-Phe7-Arg8-Trp9-Gly10-Lys11-Pro12-Val13-NH2, have been previously proposed. In those models, the 6–9 segment of the ligand was considered essential for the ligand-receptor interactions. In this study, we probed the role of Trp9 of NDP-αMSH in interactions with hMC1bR. Analogs of NDP-αMSH with various amino acids in place of Trp9 were synthesized and tested in vitro in receptor affinity binding and cAMP functional assays at human melanocortin receptors 1b, 3, 4, and 5 (hMC1b,3-5R). Several new compounds displayed high agonist potency at hMC1bR (EC50 = 0.5–5 nM) and receptor subtype selectivity greater than 2000-fold versus hMC3-5R. The Trp9 residue of NDP-αMSH was determined to be not essential for molecular recognition at hMC1bR. © 2007 Wiley Periodicals, Inc. Biopolymers 89: 401–408, 2008.
This article was originally published online as an accepted preprint. The “Published Online” date corresponds to the preprint version. You can request a copy of the preprint by emailing the Biopolymers editorial office at [email protected]
INTRODUCTION
α-Melanocyte stimulating hormone (αMSH), Ac-Ser1-Tyr2-Ser3-Met4-Glu5-His6-Phe7-Arg8-Trp9-Gly10- Lys11-Pro12-Val13-NH2, has been originally recognized as an endogenous peptide which affects pigment formation and granule dispersion in the skin of lower vertebrates.1-3 In mammals, this peptide has subsequently been implicated in physiological processes such as learning and memory, blood pressure, pigmentation, immune modulation, weight homeostasis, and others.1-3 In the 1990s, four receptors were cloned for which αMSH is a high affinity endogenous agonist (melanocortin receptors 1, 3, 4, and 5; MC1,3-5R).4-7 Each of these G-protein coupled-receptors displays a specific pattern of tissue distribution in the CNS and peripheral sites.4 For example, the melanocortin receptor 1 (MC1R) has been detected on the surface of several types of skin cells (melanocytes, keratinocytes, sebocytes, and others) and on various melanoma cells, and found to control the relative amounts of eumelanin and pheomelanin in mammals.4, 8 Variants of MC1R which do not exhibit any function are associated with fair skin, poor tanning, propensity to freckle and increased skin cancer.7, 9-12 MC1R has also been detected in other peripheral sites: the pituitary, testis, corpus luteum, placenta, endothelial cells, glioma cells, astrocytes, and in low levels in the brain.4, 5 Its presence on the surface of various immune cells such as macrophages, fibroblasts, monocytes, mast cells, and neutrophils dendritic cells, suggests this receptor's involvement in endogenous control of some inflammatory processes.5, 13
Human MC1R is activated equally well by αMSH, and ACTH, and to a lesser degree by βMSH and γMSH.2, 14 At the mouse receptor, αMSH is more potent than ACTH. The His6-Phe7-Arg8-Trp9 segment of αMSH, present also in the structures of βMSH, γMSH and ACTH, has been identified as necessary for effective interactions with MC receptors.9 The same segment was found to be critical to binding and potency of most synthetic, peptidic ligands at the melanocortin receptors.15-20
Hypothetical three-dimensional models of complexes of the melanocortin receptor 1 with αMSH, and its two synthetic analogs NDP-αMSH and MTII, were proposed.21-26
Ac-Ser1-Tyr2-Ser3-Nle4-Glu5-His6-D-Phe7-Arg8-Trp9-Gly10- Lys11-Pro12-Val13-NH2 (NDPαMSH).
Ac-Nle4-cyclo-(Asp5-His6-D-Phe7-Arg8-Trp9-Lys10)-NH2 (MTII).
In the construction of these models, bacteriorhodopsin served as a template for MC1R. Docking of the ligands into the hMC1R structure was based on the understanding that the His6-Phe7-Arg8-Trp9 segment of αMSH, or His6-D-Phe7-Arg8-Trp9 segment of NDP-αMSH and MTII, is indispensable to the formation of stable complexes with hMC1R. The N- and C-terminal ends of the ligands were proposed to face the outside environment. Of these models, that of the complex between NDP-αMSH and hMC1R was studied in greater detail.26 The site-directed mutagenesis of hMC1R combined with computer-assisted molecular-dynamic simulation and energy minimization studies on the complex, suggested several amino acid residues of this receptor which are the most likely involved in the interactions with the His6-D-Phe7-Arg8-Trp9 segment of NDP-αMSH. Acidic residues of MC1R: Glu,94 Asp,117 and Asp,121 together with Phe280 and Asn,281 were postulated to form a binding pocket which may accommodate the basic side chain of Arg8 of NDP-αMSH, and the Phe residues 175, 179, 195, 196, 257, 258, and 288, and Tyr residues 182, 183 of this receptor were postulated to form another binding pocket with which the side chains of Phe7 and Trp9 of the “essential core” of the ligand may interact, probably through stacking of the aromatic rings. To verify these hypotheses, binding affinity and agonist potency of NDP-αMSH at several mutants of hMC1R was measured.26 In some of those mutants, one of the aromatic residues of hMC1R listed above was replaced with Ala, and such single alteration to the structure of hMC1R had a negligible effect on binding affinity and agonism of NDP-αMSH. It was speculated that the lack of one, relatively weak, hydrophobic ligand-receptor interaction may be compensated by a network of other aromatic ligand-receptor interactions. Indeed, a much larger effect was observed when NDP-αMSH was tested at mutants of hMC1R in which two or three aromatic residues were replaced with Ala.26
In the present study, we probed the role of Trp9 of NDP-αMSH in the interactions with hMC1bR through the ligand structure-function studies. Analogs of NDP-αMSH with various aliphatic, hydrophilic or hydrophobic, conformationally flexible or constrained amino acids in place of Trp9 were tested in vitro at hMC1bR in the affinity binding and cAMP activation assays and counter-screened in the same assays at the human MC receptors 3, 4, and 5. Some of the compounds displayed high agonist potency at hMC1bR (EC50s from 0.5 to 5 nM) and were unable to activate efficiently the human MC 3, 4, and 5 receptors even at micromolar concentrations.
MATERIALS AND METHODS
Peptide Synthesis, Purification, and Characterization
Elongation of peptidyl chains on p-methylbenzhydrylamine resin (Boc-synthesis, 431A ABI peptide synthesizer), deprotection and cleavage of peptides from the resin with HF, and purification of the crude products by high-pressure liquid chromatography were performed as described (in detail) previously.16 A standard gradient system of 10–100% buffer B in 30 min (G1) was used for analysis; buffer A was 0.1% trifluoroacetic acid in water and buffer B was 0.1% trifluoroacetic acid in acetonitrile. The second gradient system used for analysis was 0–100% buffer B in 30 min (G2); buffer A was 0.1% trifluoroacetic acid in water and buffer B was 0.1% trifluoroacetic acid in methanol. The chromatographically homogenous compounds were analyzed by electrospray mass spectrometry (Hewlett Packard series 1100 MSD spectrometer).
Competitive Binding Assays
Binding activity of compounds was measured using membranes from Chinese hamster ovary (CHO) cells expressing the cloned melanocortin receptors. Binding reactions contained membranes, 200 pM [125I]NDP-αMSH (New England Nuclear), and increasing concentrations of unlabelled test compounds from 0.05 nM to 20 μM. Reactions were incubated for 1.5 h and then filtered as described previously.16 Binding data were analyzed using GraphPad curve-fitting software. Active peptides were evaluated in three or more independent experiments.
cAMP Assays
Agonist activities of all compounds were measured using CHO cells expressing the cloned melanocortin receptors (see Ref.16 for details). Cells were detached from tissue culture flasks, collected by 5-min centrifugation, and resuspended in Earle's Balanced Salt solution [Life Technologies, Gaithersburg, molecular dynamics (MD)] with addition of 10 mM HEPES (pH 7.5), 5 mM MgCl2, 1 mM glutamine, and 1 mg/ml bovine serum albumin. Compounds from 0.003 to 5000 nM concentration together with 0.6 mM isobutylmethylxanthine were incubated at room temperature with dissociated cells for 40 min and lysed with 0.1M HCl to terminate the assay (SMP 001J) or with PerkinElmer detection buffer (SMP-004). cAMP was quantitated by Perkin Elmer Life Sciences (NEN) (Boston, MA) SMP-001J or SMP-004 Flashplate cAMP assay. Activation by compounds was compared to the maximum response to αMSH. Active peptides were evaluated in three or more independent experiments.
RESULTS
Analogs of αMSH (26) and NDP-αMSH (1) listed in Tables I and 2 were prepared by solid phase syntheses as previously described (see Ref.16 and the Experimental Section.) (Throughout this report, the numbering of the amino acid residues in αMSH, Ac-Ser1-Tyr2-Ser3-Met4-Glu5-His6-Phe7- Arg8-Trp9-Gly10-Lys11-Pro12-Val13-NH2, has been retained for all linear and cyclic peptides). They were evaluated for their binding affinities to the human melanocortin receptors 1b, 3, 4, and 5 in the competitive binding assays using the radiolabeled ligand [125I]-NDP-αMSH and for their agonist potency in cAMP assays employing the CHO cells expressing these receptors (see Ref.16). The human melanocortin 1b receptor (hMC1bR) possesses pharmacological properties virtually identical to that of its isoform, human melanocortin receptor 1a (hMC1aR).27 Functional antagonism of peptides discussed in this study was not determined.
| No. | Compound | Binding Assay IC50a (nM) | Selectivity | cAMP Assay EC50b (nM) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MC-1bR | hMC-3R | hMC-4R | hMC-5R | 3/1b | 4/1b | 5/1b | MC-1Br | hMC-3R | hMC-4R | hMC-5R | ||
| NDP-αMSH: Ac-Ser1-Tyr-Ser-Nle4-Glu-His-D-Phe7-Arg-Trp9-Gly-Lys-Pro-Val13-NH2 (1) | ||||||||||||
| 1 | NDP-αMSH | 0.39 ± 0.01 | 0.78 ± 0.02 | 1.6 ± 0.03 | 0.27 ± 0.01 | 2 | 4 | 0.7 | 0.87 ± 0.04 | 0.29 ± 0.006 | 0.58 ± 0.01 | 0.37 ± 0.006 |
| 2 | Ala9 | 4.1 ± 0.3 | >5000 | >5000 | >5000 | 2.4 ± 0.14 | >1000 (34%) | 22% at 2.5 μM | >1500 (26%) | |||
| 3 | Aib9 | 1.3 ± 0.1 | 4800 ± 100 | 1900 ± 190 | >10,000 | 3690 | 1460 | 1.9 ± 0.0.7 | 390 ± 28 (57%) | >2500 (53%) | >2500 (51%) | |
| 4 | Val9 | 2 ± 0.08 | >5000 | 3500 ± 400 | >5000 | 1750 | 3 ± 0.5 (74%) | 1200 ± 290 (52%) | 23% at 2.5 μM | >1500 (30%) | ||
| 5 | Cha9 | 8.1 ± 0.7 | 490 ± 90 | >6000 | 1400 ± 360 | 60 | 170 | 4.2 ± 0.37 | 130 ± 9 (74%) | 36% at 5 μM | 1500 ± 40 (41%) | |
| 6 | Lys9 | 5.1 ± 1 | 39% at 5 μM | 41% at 5 μM | >2000 | 0.58 ± 0.08 (74%) | >1500 | 13% at 2 μM | >1500 | |||
| 7 | Glu9 | 3700 ± 400 | >20,000 | >20,000 | >20,000 | 1300 ± 210 | 0% at 10 μM | 0% at 10 μM | 1% at 10 μM | |||
| 8 | Phe9 | 0.64 ± 0.08 | 2.9 ± 1.2 | 3.4 ± 1.1 | 3.4 ± 0.3 | 4 | 5 | 5 | 4.2 ± 2.9 | 3.2 ± 0.01 | 6.7 ± 1.7 | 6.9 ± 1 |
| 9 | Pip9 | 1.1 ± 0.2 | >10,000 | 800 ± 400 | 4600 ± 1700 | 730 | 4180 | 2.5 ± 0.11 | 2000 ± 190(50%) | 830 ± 40 (38%) | >2500 (39%) | |
| 10 | Pro9 | 5.3 ± 0.4 | 25% at 5 μM | 17% at 5 μM | 33% at 5 μM | 2.9 ± 0.5 | >1500 (31%) | 6% at 5 μM | 15% at 2.5 μM | |||
| 11 | Oic9 | 1.2 ± 0.04 | 2400 ± 200 | >1000 | 820 ± 33 | 2000 | 680 | 1.1 ± 0.18 | 180 ± 14 (74%) | 34% at 2.5 μM | 440 ± 35 (84%) | |
| 12 | Tic9 | 22 ± 1 | >10,000 | >10,000 | >10,000 | 22 ± 6.4 | 9% at 5μM | 2% at 5 μM | 8% at 5 μM | |||
| 13 | Sar9 | 3.9 ± 0.2 | >10,000 | >10,000 | 7600 ± 100 | 1950 | 2.1 ± 0.35 | 2100 ± 380 (28%) | 6% at 5 μM | 2400 ± 350 (33%) | ||
| 14 | Gly9 | 2.1 ± 0.3 | 2400 ± 40 | >2000 | >1000 | 1140 | 2.8 ± 0.54 | >3000 (32%) | >3000 (21%) | >3000 (36%) | ||
| 15 | Ava9-10 | 5.1 ± 0.5 | >5000 | >5000 | 3500 ± 190 | 680 | 2.9 ± 0.21 | 11% at 2 μM | 0% at 2 μM | >1500 | ||
| 16 | D-Ala9 | 4.2 ± 1.3 | >5000 | >5000 | >5000 | 1.2 ± 0.42 | 770 ± 160 (60%) | 13% at 2.5 μM | >1500 (30%) | |||
| 17 | D-Pip9 | 1.2 ± 0.2 | 1700 ± 600 | 1700 ± 300 | 1600 ± 300 | 1400 | 1400 | 1330 | 1 ± 0.1 | 110 ± 20 | 2300 ± 580 | 1000 ± 140 |
| 18 | desTrp9 | 16 ± 1 | >2500 | >2500 | >2500 | 9.6 ± 0.71 | 2% at 2 μM | 1% at 2 μM | 9% at 1 | |||
| 29 | D-Nal7,Pro9 | 2.7 ± 0.4 | 640 ± 70 | 1500 ± 360 | 47 ± 3 | 240 | 560 | 17 | 3.9 ± 0.71 (66%) | >1500 | >1500 | 380 ± 100 |
| 20 | D-Nal7,Pip9 | 2.8 ± 0.2 | 420 ± 60 | 350 ± 10 | 30 ± 5 | 150 | 130 | 11 | 2.5 ± 0.36 (60%) | >1500 | 1% at 2.5 μM | 100 ± 14 |
| 21 | D-Phe(pCI)7,Pip9 | 0.73 ± 0.06 | 230 ± 100 | 480 ± 110 | 44 ± 12 | 320 | 660 | 60 | 0.51 ± 0.08 | 160 ± 28 (70%) | 38% at 2.5 μM | 160 ± 7 |
| 22 | D-Phe(pF)7,Pip9 | 2.1 ± 0.3 | 3700 ± 1800 | 3900 ± 900 | 1300 ± 180 | 1760 | 1860 | 620 | 2.6 ± 0.82 | 1300 ± 110 (63%) | >5000 | 2100 ± 180 (70%) |
| 23 | D-Tyr(Me)7,Pro9 | 0.71 ± 0.13 | 4800 ± 100 | 8000 ± 1100 | 680 ± 95 | 6760 | 11,300 | 960 | 1.1 ± 0.17 (86%) | 550 ± 78 (65%) | 26% at 10 μM | 1900 ± 420 |
| 24 | D-Tyr(Me)7,Pip9 | 0.38 ± 0.03 | >8000 | 240 ± 140 | 220 ± 40 | 630 | 580 | 2.4 ± 0.53(85%) | 590 ± 180 (75%) | 2200 ± 1000 (36%) | 780 ± 51 | |
| 25 | D-4,4′Bip7,Pip9 | 0.06 ± 0.03 | 72 ± 40 | 540 ± 80 | 5.4 ± 3.2 | 1200 | 9000 | 90 | 0.28 ± 0.05 (72%) | 35 ± 5.8 (65%) | 23% at 5 μM | 3.2 ± 0.78 |
| αMSH: Ac-Ser1-Tyr-Ser-Met4-Glu-His-Phe7-Arg-Trp9-Gly-Lys-Pro-Val13-NH2 (26) | ||||||||||||
| 26 | αMSH | 3.9 ± 0.9 | 19 ± 2 | 19 ± 2 | 120 ± 19 | 5 | 5 | 30 | 3.4 ± 0.1 | 1.1 ± 0.05 | 1.9 ± 0.08 | 16 ± 0.5 |
| 27 | Pro9 | 3400 ± 100 | 2% at 5 μM | 8% at 5 μM | 1% at 5 μM | 1100 ± 190 (54%) | 0% at 2.5 μM | 4% at 5 μM | 0% at 2.5 μM | |||
| 28 | Pip9 | 350 ± 70 | >10,000 | >10,000 | >10,000 | 91 ± 2.1 | >2500 | 0% at 5 μM | 4% at 5 μM | |||
| 29 | Nle4,Pip9 | 44 ± 1 | >5000 | >5000 | >5000 | 50 ± 0.7 | >3000 | 0% at 5 μM | 3% at 5 μM | |||
| 30 | D-Phe7,Pip9 | 9.1 ± 0.3 | >10,000 | >2300 | >10,000 | 5.6 ± 1.2 | >2500 | 4% at 5 μM | 7% at 5 μM | |||
- Data are the average of three or more independent determination together with the standard error of the mean.
- a Concentration of peptide at 50% specific binding or, the percentage of inhibition (relative to [125I]-NDP-αMSH) observed at a given peptide concentration (μM).
- b Concentration of peptide at 50% maximum cAMP accumulation and/or the percentage of cAMP accumulation (relative to[125I]-NDP-αMSH) observed at the designated μM concentration.
| No. | Compound | Binding Assaya IC50 (nM) | cAMP Assayb EC50 (nM) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| MC1bR | hMC3R | hMC4R | hMC5R | MC1bR | hMC3R | hMC4R | hMC5R | ||
| 31 | Ac-Nle-Glu-His-D-Phe-Arg-Pip-Gly-Lys-Pro-Val-NH2 | 20 ± 2 | 7% at 10 μM | >5000 | 27% at 10 μM | 14 ± 2 | >5000 | 23% at 10 μM | >5000 |
| 32 | Ac-Ser-Tyr-Ser Nle-Glu-His-D-Phe-Arg-Pip-Gly-NH2 | 15 ± 3 | 37% at 10 μM | >10,000 | >10,000 | 9 ± 1 | >4000 | 29% at 10 μM | >5000 |
| 33 | Ac-Nle-Glu-His-D-Phe-Arg-Pip-Gly-NH2 | 220 ± 30 | 4% at 10 μM | 30% at 10 μM | 9% at 10 μM | 48 ± 18 | 4% at 5 μM | 4% at 5 μM | 5% at 5 μM |
| 34 | Ac-His-D-Phe-Arg-Pip-Gly-NH2 | 17% at 10 μM | 9% at 10 μM | 21% at 10 μM | 22% at 10 μM | 18% at 10 μM | 0% at 5 μM | 0% at 10 μM | 0% at 10 μM |
| 35 | Ac-Nle-Glu-His-D-Phe-Arg-Pip-NH2 | 160 ± 15 | 10% at 10 μM | 27% at 10 μM | 23% at 10 μM | 110 ± 20 | 0% at 5 μM | 3% at 10 μM | 11% at 10 μM |
| 36 | Ac-Nle-His-D-Phe-Arg-Pip-NH2 | 240 ± 21 (76%) | 0% at 10 μM | 17% at 10 μM | >2000 | >5000 | 0% at 5 μM | 0% at 10 μM | 0% at 10 μM |
| 37 | Ac-His-D-Phe-Arg-Pip-NH2 | 25% at 10 μM | 3% at 10 μM | 28% at 10 μM | 28% at 10 μM | >5000 | 1% at 5 μM | 0% at 10 μM | 0% at 10 μM |
- a Concentration of peptide at 50% specific binding or, the percentage of inhibition (relative to [125I]-NDP-αMSH) observed at a given peptide concentration (μM).
- b Concentration of peptide at 50% maximum cAMP accumulation and/or the percentage of cAMP accumulation (relative to[125I]-NDP-αMSH) observed at the designated μM concentration.
Binding and activation data for NDP-αMSH (1), αMSH (26), and their full-length analogs with different amino acid residues in position 9 (Trp) have been compiled in Table I. Most of these compounds were moderate potency ligands for hMC3-5R or were not able to activate these receptors even at micromolar concentrations. With the exception of 7 and 27, they were all, however, nanomolar agonists at hMC1bR. In the structure of the first analog of NDP-αMSH, Trp9 was replaced with Ala (2). This peptide was a full 2.4 nM agonist at hMC1bR and weak partial agonist at hMC3,5R, and practically inactive at hMC4R. Similarly, compounds with aliphatic, hydrophobic, or hydrophilic side chains in position 9, Aib, Val, cyclohexylalanine (Cha), and Lys (3–6), were potent agonist at hMC1bR and showed weak agonistic properties at hMC3,5R, but were not able to activate hMC4R. In contrast, peptide 7 with the acidic side chain: Glu9-NDP-αMSH was merely a micromolar agonist at hMC1bR and practically inactive at hMC3-5R.
Incorporation of an aromatic residue such as Phe in position 9 of NDP-αMSH restored ability of this peptide to form stable complexes with hMC3-5R. Peptide 8, Phe9-NDP-αMSH was a 2- to 30-fold less potent agonist at the MCRs than NDP-αMSH and equally nonselective.
In an attempt to diminish flexibility of the NDP-αMSH peptide chain, the sterically constrained Pip (6-membered ring, pyridine-2-carboxylic acid, pipecolic acid) was incorporated in position 9, Pip9-NDP-αMSH (9). This peptide was only about threefold less potent than NDP-αMSH (1) at hMC1bR. Unlike the parent compound, peptide 9 was a weak agonist at hMC3-5R, and hence of higher receptor subtype selectivity for hMC1bR. Its analog with the 5-membered ring of Pro in the same position, Pro9-NDP-αMSH (10) showed comparable pharmacological profile at the hMC1b,3-5R.
Compounds 11 and 12 were analogs of NDP-αMSH with cyclic amino acids in position 9 as well. The Oic9 peptide (11), (Oic: octahydroindole-2-COOH, an analog of Pro) retained the agonist potency of NDP-αMSH at hMC1bR (EC50 = 1.1 nM) and was a partial agonist at hMC3,5R, and inactive at hMC4R at micromolar peptide concentrations. Similarly, Tic9-NDP-αMSH (Tic: tetrahydroquinoline-3-COOH, an aromatic analog of Pip) (12) remained a full agonist at hMC1bR, albeit about 20-fold weaker than NDP-αMSH. This peptide was not able to form stable complexes with hMC3-5R at micromolar concentrations.
In the aforementioned analogs of NDP-αMSH (9–12) with cyclic residues: Pro, Pip, Oic, or Tic, the free αN-H in position 9 was not available for the formation of hydrogen bonds. To differentiate the impact of a ring from that of the secondary α-amino group in position 9 on binding and activation at hMC1b,3-5R, the Sar9 and Gly9 peptides were tested. In the structure of Sar9-NDP-αMSH (13) a side chain and the free αN-H were absent in position 9. Yet, the Sar9 analog showed agonist potency at hMC1bR similar to that of Pip9-NDP-αMSH (9), and comparable receptor subtype selectivity. In the second analog: Gly9-NDP-αMSH (14), the free αN-H was present in position 9 and this peptide (14) as well displayed a pharmacological profile at hMC1b,3-5R comparable to the Pip9 analog.9 Compound 14 was a 3 nM agonist at hMC1bR and about 1000-fold selective versus other studied receptor subtypes. In peptide 15, the Trp9-Gly10 segment of NDP-αMSH was replaced with the conformationally flexible γ-aminovaleric acid (Ava). The Ava9-10 peptide lacked the amide bond between residues 9 and 10 but similarly displayed high agonist potency at the hMC1bR and was practically inactive at hMC3-5R.
Taken together, the in vitro affinity binding and receptor activation data for analogs of NDP-αMSH with Pip9, Sar9, Gly9 or Ava9-10 indicated that absence of a side chain and free αNH or CO in position 9 has minimal effect on interactions with hMC1bR but is detrimental to the formation of stable complexes with hMC3-5R.
In analogs 16 and 17, the chirality of residue 9 was reversed and this change was not detrimental to binding and activation at hMC1bR. Also, at the other MC receptors, for example, D-Ala9-NDP-αMSH16 showed the pharmacological profile similar to that of the L-Ala9 compound (2). The new analog 16 was a 1.2 nM agonist at hMC1bR, a partial weak agonist at hMC3/5R and practically inactive at hMC4R. Additionally, agonist potency of the D-Pip9 peptide (17) was measured to be similar to that of its L-Pip9 counterpart 9 (compound 17 displayed 15-fold lower selectivity with respect to hMC3R).
Structure-function studies described thus far has clearly shown that diverse structural changes in position 9 of NDP-αMSH do not affect (or only minimally) interactions with hMC1bR but are detrimental to molecular recognition at hMC3-5R. Consequently, our next analog tested was compound 18 in which the Trp9 residue was omitted from the peptide chain of NDP-αMSH. Rather unexpectedly, this shorter analog, desTrp9-NDP-αMSH, was a 10 nM full agonist at hMC1bR. Similarly to many analogs with modified residue 9, listed in Table I, peptide 18 was not able to form stable complexes with hMC3-5R even at micromolar peptide concentration.
Compounds 19–25 were designed to evaluate the role of D-Phe7 in interactions of the NDP-αMSH analogs modified in position 9 (Pip or Pro) with hMC1bR. In these peptides, D-Phe7 was replaced with an aromatic amino acid such as D-2′-naphthylalanine [D-Nal(2′)], D-Phe(pCl), D-Phe(pI), D-Tyr(Me) or D-4,4′-biphenylalanine (D-4,4′Bip). It was anticipated that the Pro9 and Pip9 analogs of Nle4,D-Nal(2′)7-αMSH (designated NDN-αMSH, Ref.28) might have similar in vitro properties to those of the analogous NDP-αMSH-derived peptides. Yet when tested, D-Pro9-NDN-αMSH (19) and D-Pip9-NDN-αMSH (20) were partial agonists at hMC1bR (ca. 60% activation at 5 nM) and moderate potency agonists at hMC5R (EC50 about 100–400 nM).
Compounds with D-Phe(pCl), D-Phe(pI), D-Tyr(Me) or D-4,4′Bip in position 7 (21–25) were high affinity agonists at hMC1bR (EC50 from 0.5 to 3.5 nM) but of lower selectivity with respect to hMC3-5R. Most of them were partial agonists at hMC3R. Moreover, compound 25, D-4,4′Bip,7Pip9-NDP-αMSH was a subnanomolar partial agonist at hMC1bR and a 3 nM full agonist at hMC5R.
Several analogs of the endogenous ligand, αMSH (26), with modified position 9 were also evaluated at hMC1b,3-5R (27–30). Compounds 27 and 28 were less potent at hMC1bR than their NDP-αMSH counterparts (8, 9). For example, Pro9-αMSH (27) was a weak agonist at hMC1bR (EC50 about 1 μM) which was unable to activate the hMC3-5R even at micromolar peptide concentrations. Another analog of αMSH (28), that with the slightly larger, 6-membered ring of Pip in position 9, was 10-fold more potent at hMC1bR than the Pro9 peptide (27) and was also practically inactive at the counter screened receptors, hMC3-5R.
In the subsequent two Pip9 analogs of αMSH, Nle was in place of Met4 and D-Phe in place of Phe7, 29, and 30, respectively. Collectively, such alternations in the structure of αMSH are known to account for high agonist potency of NDP-αMSH (1) at hMC1,3-5R. The new peptides: Nle4, Pip9-αMSH (29) and D-Phe7,Pip9-αMSH (30) displayed higher agonist potency at hMC1bR (EC50 about 50 nM and EC50 about 5.6 nM, respectively) than compound 28, Pip9-αMSH, EC50 = 91 nM, and similarly, were unable to interact efficiently with hMC3-5R.
To evaluate a role of the residues which are external to the essential core of the NDP-derived peptides in interactions with hMC1b,3-5R, truncated analogs of 9 were tested (see Table II). These shorter peptides were not able to activate the MC3-5R even at micromolar concentrations. Compounds 31–33, 35, and 36, with the His6-D-Phe7-Arg8-Pip9 segment extended at the N-terminus by one or several residues were moderately potent agonists at hMC1bR (EC50s from 5 to 240 nM). Two analogs with Ac-His6 at their N-termini were practically inactive at hMC1bR (34, 37).
DISCUSSION AND CONCLUSIONS
In the last several years, extensive structure-function studies on αMSH have mainly focused on the development of potent and selective hMC4R agonists, compounds that might be useful in the treatment of obesity.14, 29-32 Numerous analogs of αMSH were prepared and tested in vitro in the affinity binding and functional assays at the hMC4R and at the other melanocortin receptors present in the brain—hMC3R and hMC5R. Frequently, the new compounds were not evaluated at MC1R, the receptor found in many peripheral sites, the skin and some immune cells.
Briefly, it was established that the His6-Phe7-Arg8-Trp9 segment of αMSH, known to be essential for the hormonal activity in the frog and lizard skin bioassays,20 is also necessary for the efficient binding to the human MC3-5R and activation of these receptors by the synthetic αMSH-derived ligands. The same residues were later described as critical to molecular recognition at MC1R as well. This conclusion was drawn from the in vitro analysis of analogs of αMSH in which each amino acid in turn was replaced with Ala (tested for tyrosinase activity on cultured B16 murine melanoma cells expressing mouse MC1 receptor)10 and of the Ala-substituted analogs of the core tetrapeptide Ac-His6-D-Phe7-Arg8-Trp9-NH2 (evaluated for binding and function at the mouse MC1,3-5R).32
Consequently, in this study we expected a peptide lacking the Trp side chain in position 9, Ala9-NDP-αMSH (2) to be a poor ligand for hMC1b,3-5R. However when 2 was evaluated in the affinity binding and functional assays at the cloned human melanocortin receptors 1b, 3, 4, and 5, it displayed high agonist potency at hMC1bR (EC50 = 2.4 nM) but was a very weak, partial agonist at the human MC receptors 3 and 5 and practically inactive at the human MC receptor 4. Inability of this peptide to interact effectively with hMC3-5R was consistent with the previous observation that the aromatic side chain in position 9 (Trp) is necessary for molecular recognition of the αMSH-derived peptides at those receptors. The high agonist potency of Ala9-NDP-αMSH (2) at hMC1bR was, however, inconsistent and thus rather unanticipated. Peptide 2 was a slightly less potent agonist at hMC1bR than NDP-αMSH (about threefold) and this difference in potency could stem from the absence of one relatively weak hydrophobic interaction in the complex of Ala9-NDP-αMSH with hMC1bR. It also could reflect the previously proposed26 compensatory effect of a network of other aromatic ligand-receptor interactions. Alternatively, the high agonist potency of Ala9-NDP-αMSH at hMC1bR could suggest that in the ligand-receptor complexes, the aromatic side chain of Trp9 is at a distance to the Phe and Tyr residues of hMC1R that doesn't allow for effective interactions among the aromatic side chains of the ligand and receptor. It could also be concluded that in the complexes of NDP-αMSH (1) with hMC1bR, the side chain of Trp9 faces the outside environment. This notion, however, appears to be not supported by the micromolar agonist potency at hMC1bR observed for a compound with Glu in place of Trp9, Glu9-NDP-αMSH (7). Weak agonist properties of 7 at hMC1bR seems to suggest that in the ligand-receptor complexes the side chain of Glu9 is in a close proximity to the acidic groups present on the receptor and is repelled by them. Yet Lys9-NDP-αMSH (6), with the basic side chain in the same position available for ionic interactions with those putative acidic groups of the receptor, displayed binding affinity at hMC1bR similar to that of Ala9-NDP-αMSH2 and other analogs with aliphatic residues in position 9, Aib, Val, Cha (4–6).
Alternatively, the ionic interactions between the basic residues of NDP-αMSH (for example His6) and the side chain of Glu9 may affect biologically relevant conformers of peptide 7. Evaluation of these analogs of NDP-αMSH at the appropriately designed mutants of hMC1R perhaps could verify some of these speculations.
Replacement of Trp9 with another aromatic residue such as Phe had a minor effects on the binding affinity and agonist potency of NDP-αMSH at hMC1bR compound 8. Also at hMC3-5R, the new peptide 8 was only 3- to 10-fold less potent agonist than the parent peptide. This emphasized, yet again, that an aromatic side chain in position 9 of NDP-αMSH is necessary for molecular recognition at hMC3-5R.
The flexible peptide chain of NDP-αMSH (1) may adopt various conformations at the receptor site. In this study, an attempt to stabilize some conformers of this chain by the replacement of Trp9 with a sterically constrained amino acid such as Pro, Pip, Oic, or Tic yielded peptides 9-12 of potency at hMC1bR similar to that of the parent compound. Also, other structural changers in position 9 of NDP-αMSH: reversal of chirality or replacement of the Trp9-Gly10 segment with γ-aminovaleric acid (Ava, mimics in length this dipeptide unit) resulted in peptides of high agonist potency at hMC1bR (15–17). Even an analog of NDP-αMSH lacking residue 9 in its chain, desTrp9-NDP-αMSH (18), was a 10 nM full agonist at hMC1bR. Hence, these observations indicated that Trp9 is not required for efficient interactions of NDP-αMSH with hMC1bR, but is necessary for molecular recognition at hMC3-5R. These observations also suggested that the His6-D-Phe7-Arg8 segment of the essential core of NDP-αMSH is largely responsible for the formation of stable complexes with hMC1bR. The residues located at the N-terminus of this unit may perhaps also contribute to the receptor binding and/or facilitate an appropriate alignment of His6-D-Phe7-Arg8 at the receptor site. Previously it was shown17, 18, 28 that, for example, the side chains in positions 4 and 5, which are N-terminal to the essential core of NDP-αMSH and MTII, are essential to high agonist potency at hMC5R.
Among the analogs of NDP-αMSH tested in this study, Ava9-10-NDP-αMSH (15) and desTrp9-NDP-αMSH (18) emerged as the full, high potency agonists at hMC1bR which were not able to activate the hMC3-5R even at micromolar concentrations.
Ac-Ser-Tyr-Ser-Nle-Glu-His6-D-Phe7-Arg8-Ava9-10-Lys11-Pro- Val-NH2 (15, Ava9-10-NDP-αMSH).
Ac-Ser-Tyr-Ser-Nle-Glu-His6-D-Phe7-Arg8–Gly10-Lys11-Pro- Val-NH2 (18, desTrp9-NDPα-MSH).
Koikov et al.33, 34 tested a number of N-acylated short peptides encompassing the essential core of NDP-αMSH at hMCRs and observed that one lacking Trp9 in its structure: Ph-(CH2)3-CO-His6-D-Phe7-Arg8-NH2, designated LK-394, was a 5 nM agonist at hMC1R and a weak partial agonist at hMC3R and hMC4R (11 and 24% activation, respectively.) Binding and functional data for LK-394 at hMC5R were not reported. It was proposed that the N-terminal phenyl group of this peptide may interact with an alternative, putative, hydrophobic binding site on hMC1R and, hence, contribute to the stabilization of the ligand-receptor complexes.33, 34
Truncated analogs of Pip9-NDP-αMSH (9) evaluated in this study were noticeably less potent agonists at hMC1bR than the parent compound (31–37). In the structures of these compounds, an additional aromatic moiety was not present in the proximity of the amino group of His6 and this may explain their lower potency at hMC1bR. Similarly, lack of an additional aromatic “anchor” close to the amino group of His6 may explain a weak micromolar agonism at the mouse MC1R of the previously reported Ac-His6-D-Phe7-Arg8-Ala9-NH2.32
Another short analog of NDP-αMSH lacking Trp9 in its structure was recently measured by us to be a potent and selective hMC5R agonist: Ac-Nle4-Glu5-Oic6-D-4,4Bip7-Pip8-NH2 (Oic: octahydroindole-2-COOH, Pip: pipecolic acid).28 Together, it could be concluded that the residue in position 9 is not essential for molecular recognition of some analogs of NDP-αMSH at hMC1R and hMC5R, receptors found mainly in various peripheral sites.
Structure-function studies focused on the elucidation of the role of D-Phe7 in molecular recognition at hMC1b,3-5R, another aromatic residue of “the essential segment” of NDP-αMSH, will be described separately. In the present study, several analogs of the Pip9 or Pro9 peptides with the modified phenyl ring in position 7 were shown to be potent, full, or partial agonists at hMC1bR. The D-Nal(2′)7, D-Tyr(Me)7, Phe(pCl)7, Phe(pI)7, or D-4,4′Bip7 peptides were able to activate the other MC receptors to various degree. Of those compounds, the most interesting was D-4,4′Bip7,Pip9-NDP-αMSH. It was a 0.3 nM agonist at hMC1bR and a 3.2 nM full agonist at hMC5R. Apparently, the biphenyl side chain in position 7 of the Pro9 and Pip9 peptides was able to form stronger interactions with the aromatic side chains of the hydrophobic binding pocket of hMC1bR than the single, unmodified phenyl group of Phe in the same position. The pharmacological properties in vitro of other analogs of NDP-αMSH (and MTII) with 4,4′Bip in position 7 have been previously reported by us.28
In summary: the present structure-function studies have shown that residue 9 (Trp) of NDP-αMSH is not crucial for efficient interaction of this peptide with hMC1bR. The indole side chain in position 9 could be replaced with various aliphatic, aromatic, hydrophilic, or hydrophobic moieties. Also, the chirality of the residue 9 could be reversed or this residue could be omitted from the peptide structure without substantial loss of agonism at hMC1bR. However, these changers are deleterious to molecular recognition at hMC3-5R and render most of the new analogs inactive at those receptors. The role of Trp9 in the interactions of the cyclic analogs of αMSH, the MTII and SHU9119 peptides, with the MC receptors will be discussed in a separate manuscript.
Several compounds reported in this study were highly potent agonists at hMC1bR (EC50 = 0.5–10 nM) and of high selectivity with respect to the human MC receptors 3, 4, and 5 (>2000-fold). These analogs may be useful for the evaluation of the physiological role of hMC1bR in humans and rodents. They may also be valuable in studies geared towards the refinement of the proposed model of the complex between NDP-αMSH and hMC1R.
