Structure–Activity Relationships of the Tetrapeptide Ac-His-Arg-(pI)DPhe-Tic-NH2 at the Mouse Melanocortin Receptors: Modification at the (pI)DPhe Position Leads to mMC3R Versus mMC4R Selective Ligands

The five melanocortin receptors (MC1R–MC5R) are involved in numerous biological pathways, including steroidogenesis, pigmentation, and food intake. In particular, MC3R and MC4R knockout mice suggest that the MC3R and MC4R regulate energy homeostasis in a non-redundant manner. While MC4R-selective agonists have been utilized as appetite modulating agents, the lack of MC3R-selective agonists has impeded progress in modulating this receptor in vivo. In this study, the (pI)DPhe position of the tetrapeptide Ac-His-Arg-(pI)DPhe-Tic-NH2 (an MC3R agonist/MC4R antagonist ligand) was investigated with a library of 12 compounds. The compounds in this library were found to have higher agonist efficacy and potency at the mouse (m) MC3R compared to the MC4R, indicating that the Arg-DPhe motif preferentially activates the mMC3R over the mMC4R. This observation may be used in the design of new MC3R-selective ligands, leading to novel probe and therapeutic lead compounds that will be useful for treating metabolic disorders.


The melanocortin system
Since switching the Phe and Arg positions results in MC3R agonism and MC4R partial agonism/antagonism for the Ac-His-Arg-(pI)DPhe-Tic-NH2 scaffold, it was hypothesized that further substitutions at the para-position might result in decreasing MC4R efficacy while retaining MC3R agonism. Therefore, a library of 12 peptides was synthesized based upon the scaffold Ac-His-Arg-(pX)DPhe-Tic-NH2 (substitutions for (pX)DPhe can be found in Figure 1) and assayed at the mouse MC1R, MC3R, MC4R, and MC5R, in order to understand how the para-position within this scaffold influences melanocortin receptor selectivity, potency, and efficacy.

Peptide Synthesis and Pharmacological Evaluation
Peptides were synthesized manually with microwave irradiation using standard Fmoc synthesis techniques [44,45] and purified using semi-preparative reverse-phase high-pressure liquid-chromatography (RP-HPLC). Peptide molecular mass was confirmed through ESI-MS (University of Minnesota Mass Spectrometry Laboratory), and each peptide was assessed for purity (>95%) using analytical RP-HPLC in two different solvent systems (acetonitrile and methanol; Table  1). Agonist pharmacology was measured at the mMC3R, mMC4R, and mMC5R using a colorimetric β-galactosidase assay that measures cAMP production [46]. Agonist pharmacology was assessed at the mMC1R using the Amplified Luminescent Proximity Homogenous Assay Screen (AlphaScreen, PerkinElmer), as previously described [47][48][49]. The MC2R is only stimulated by ACTH, and was not examined in this study. For both assays, HEK293 cells stably expressing the mMCRs were used. For agonist assays, the peptide ligands NDP-MSH [50] and Ac-His-DPhe-Arg-Trp-NH2 [51] were used as positive controls. Ligands were considered full agonists if they stimulated the receptor to >90% of the maximal signal of NDP-MSH and were considered inactive if they did not stimulate the receptor to at least 20% of the signal of NDP-MSH at a 100 µM concentration. Compounds that did not possess at least 50% of the maximal NDP-MSH signal were assessed for antagonist pharmacology using a Schild assay design [52], with NDP-MSH as the agonist. Compounds that were within 3-fold potency range were considered equipotent and within the inherent experimental error of the assays.

Peptide Synthesis and Pharmacological Evaluation
Peptides were synthesized manually with microwave irradiation using standard Fmoc synthesis techniques [44,45] and purified using semi-preparative reverse-phase high-pressure liquid-chromatography (RP-HPLC). Peptide molecular mass was confirmed through ESI-MS (University of Minnesota Mass Spectrometry Laboratory), and each peptide was assessed for purity (>95%) using analytical RP-HPLC in two different solvent systems (acetonitrile and methanol; Table 1). Agonist pharmacology was measured at the mMC3R, mMC4R, and mMC5R using a colorimetric β-galactosidase assay that measures cAMP production [46]. Agonist pharmacology was assessed at the mMC1R using the Amplified Luminescent Proximity Homogenous Assay Screen (AlphaScreen, PerkinElmer), as previously described [47][48][49]. The MC2R is only stimulated by ACTH, and was not examined in this study. For both assays, HEK293 cells stably expressing the mMCRs were used. For agonist assays, the peptide ligands NDP-MSH [50] and Ac-His-DPhe-Arg-Trp-NH 2 [51] were used as positive controls. Ligands were considered full agonists if they stimulated the receptor to >90% of the maximal signal of NDP-MSH and were considered inactive if they did not stimulate the receptor to at least 20% of the signal of NDP-MSH at a 100 µM concentration. Compounds that did not possess at least 50% of the maximal NDP-MSH signal were assessed for antagonist pharmacology using a Schild assay design [52], with NDP-MSH as the agonist. Compounds that were within 3-fold potency range were considered equipotent and within the inherent experimental error of the assays.
Similar to prior reports [49,51], the Ac-His-DPhe-Arg-Trp-NH 2 peptide (KNS2-153) possessed agonist potencies of 10, 190, 12, and 5 nM at the mMC1R, mMC3R, mMC4R, and mMC5R, respectively and partial agonist efficacy at the mMC4R (40% of the NDP-MSH signal, EC 50 = 150 nM; Figure 3). An antagonist pA 2 value of 7.3 was observed for KNS2-22-4 at the mMC4R ( Figure 3). In prior studies, this compound was observed to possess nanomolar agonist potency at the MC3R (30-40 nM), partial agonist stimulation of the MC4R, and sub-micromolar antagonist potency at the MC4R (pA 2 of 6.6-7) [40,41], similar to the results in the present study. a HPLC retention time (min) for peptides in solvent system 1 (10% acetonitrile in 0.1% trifluoroacetic acid/water and a gradient to 90% acetonitrile over 35 min) or solvent system 2 (10% methanol in 0.1% trifluoroacetic acid/water and a gradient to 90% methanol over 35 min). An analytical Vydac C18 column (Vydac 218TP104) was used with a flow rate of 1.5 mL/min. The peptide purity was determined by HPLC at a wavelength of 214 nm. b Two peaks were observed for the (pBr)DPhe amino acid due to the approximately equal natural abundance of 79 Br and 81 Br. Table 2. Tetrapeptide pharmacology at the mouse melanocortin-1 receptor using the AlphaScreen cyclic adenosine monophosphate (cAMP) assay a .
The least potent compounds possessed a hydroxyl (KNS2-23-9) or nitrile (KNS2-23-8) group at the para-position. The substitution of DTyr (KNS2-23-9) resulted in potencies of 40 nM, 4300 nM, and 1000 nM at the mMC1R, mMC3R, and mMC5R, respectively, and did not possess agonist or antagonist activity at the mMC4R in the concentrations assayed. An 85% partial agonist response was observed at the mMC3R. Similar agonist potencies of 27 nM, 4000 nM, and 500 nM at the mMC1R, mMC3R, and mMC5R were observed for KNS2-23-8, with partial efficacy at the mMC3R (75%). At 100 µM concentrations, this ligand was able to partially stimulate the mMC4R (40% of the maximal NDP-MSH signal; Figures 2 and 3) and did not result in antagonist activity (Figure 3).

Discussion
Previous results exploring the DPhe para-position in the Ac-His-DPhe-Arg-Trp-NH 2 scaffold resulted in full MC4R agonists with different MC3R agonist and antagonist activities [42]. Select substitutions resulted in full MC3R agonist efficacy, while others resulted in partial receptor activation at 100 µM concentrations and micromolar to sub-micromolar antagonist potencies [42]. Thus, a DPhe-Arg motif resulted in full agonism at the MC4R and full to partial agonism at the MC3R accompanied by antagonist activity (dependent on the DPhe para-position). Due to the MC3R agonism and MC4R antagonism observed in the Ac-His-Arg-(pI)DPhe-Tic-NH 2 ligand [40,41], it was hypothesized that different substitutions at the DPhe para-position within this scaffold (possessing an Arg-DPhe motif and a Tic residue in position 4) may modulate MC4R agonist efficacy. The results in Table 3 demonstrate that the efficacy at the mMC4R was modulated by various para-substitutions. Full agonism was observed for the ligand Ac-His-Arg-DPhe-Tic-NH 2 (KNS3-10) at the mMC4R (Figure 2), and an additional four substitutions resulted in over 50% agonist efficacy at the mMC4R ((pBr)DPhe, (pCl)DPhe, (pF)DPhe, and (pMe)DPhe). Modest agonist efficacy (20-50%) was observed for four ligands (possessing the (pI)DPhe, (pCF 3 )DPhe, DBip, and (pCN)DPhe substitutions), and three substitutions ((3,4-diCl)DPhe, (ptBu)DPhe, and DTyr) resulted in compounds that did not produce >20% response of the maximal NDP-MSH signal at up to 100 µM concentrations at the mMC4R. A partial agonist response at the mMC3R was also observed for four of the ligands. Thus, the para-substitution at the DPhe position within the Ac-His-Arg-(pI)DPhe-Tic-NH 2 scaffold modulates agonist efficacy at both the mMC3R and mMC4R, with the Arg-DPhe motif in general resulting in a more efficacious response at the mMC3R.
Several compounds from this study may be useful lead ligands in the development of MC3R/MC4R-selective compounds. One compound (KNS2-23-9) possessed micromolar mMC3R agonist potency and did not possess agonist or antagonist activity at the mMC4R. An additional three compounds were at least 100-fold selective agonists for the mMC3R over the mMC4R (KNS2-23-4, KNS2-23-3, and KNS2-23-1), though these three ligands possessed micromolar to sub-micromolar mMC4R antagonist potencies. Further optimization to increase MC3R potency and efficacy, and to minimize MC4R pharmacology, may be required to develop selective MC3R ligands that can elucidate the roles of the MC3R. The use of MC3R KO and MC4R KO mice may also be used with the present ligands to begin to clarify the roles of the different melanocortin receptors in vivo. Alternatively, three compounds (KNS2-22-4, KNS2-23-6, and KNS2-23-3) possessed mMC3R agonist potencies of less than 100 nM and were sub-micromolar potent mMC4R antagonists. Further optimization of this dual pharmacology (increased MC3R agonism with increased MC4R antagonism) might result in novel tool compounds that can characterize the in vivo role of the MC3R and MC4R in the regulation of food intake.
While these substitutions had an effect on efficacy at the mMC3R and mMC4R, all compounds assayed were full agonists at the mMC1R, and only one compound was not a full agonist at the mMC5R (KNS3-10, stimulating the mMC5R to 65% of the maximal NDP-MSH response). It therefore appears that the Arg-DPhe position switch may only lead to mMC3R over mMC4R selectivity. Potency trends at the mMC1R and mMC5R were similar to that at the mMC3R. The two compounds that were micromolar potent mMC3R agonists (KNS2-23-9 and KNS2-23-8) were also the least potent mMC1R (40 and 27 nM, respectively) and mMC5R (1000 and 500 nM, respectively) agonists. While no compound was significantly more potent than the Ac-His-DPhe-Arg-Trp-NH 2 ligand at the mMC5R, five ligands resulted in at least a 10-fold potency increase at the mMC1R (KNS2-22-4, KNS2-22-3, KNS2-22-1, KNS2-23-7, and KNS2-23-1).
Another report investigated the para-position within the Ac-His-DPhe-Arg-Trp-NH 2 scaffold for MC1R selectivity [53]. In addition to pF, pCl, pBr, and pCF 3 substitutions, Arg was replaced with a neutral Nle residue due to hypothesized interactions with the Arg and basic residues in the MC3R and MC4R [53]. As a general trend, these substitutions increased binding affinity at the MC1R compared to the other melanocortin receptors, as well as increased agonist selectivity for the MC1R [53]. Intraperitoneal (i.p.) injection of the pCF 3 substituted ligand resulted in in vivo pigmentation effects when administered to Anolis carolinesis lizards [53]. Our results indicate that switching the Phe-Arg positions and substituting Tic for Trp may also increase MC1R potency. When combined with the Nle substitution at the Arg position, these substitution patterns may result in increased MC1R selectivity and/or potency. , and acetic anhydride were purchased from Fisher Scientific. All reagents were ACS grade or higher and were used without further purification.

Peptide Synthesis
Peptides were synthesized on a CEM Discover SPS manual microwave synthesizer using standard fluorenyl-9-methyloxycarbonyl (Fmoc) methodology [44,45]. The rink-amide resin was added to a fritted polypropylene reaction vessel (25 mL CEM reaction vessel). The resin was allowed to swell in DCM for 1 h. Deprotection of the Fmoc group consisted of two steps: (1) 20% piperidine in DMF at rt for two minutes, followed by (2) 20% piperidine in DMF using microwave irradiation for 4 min at 75 • C with 30 W. The resin was washed with DMF, and the presence of a free amine was assessed using the ninhydrin [54] or chloranil [55] (for the Tic residue) tests. Coupling reactions were carried out with 3.1 equivalents (eq) of the incoming Fmoc-protected amino acid, 3 eq HBTU, and 5 eq DIEA using microwave irradiation (5 min, 75 • C, 30 W). A lower temperature (50 • C) was utilized for His. For Arg coupling, higher equivalents of Arg (5.1 eq), HBTU (5 eq), and DIEA (7 eq) were used with a longer (10 min) microwave irradiation time. Following resin washing with DMF, the completeness of the coupling reactions was assessed with the ninhydrin or chloranil tests, and amino acids were recoupled if necessary. Following the coupling of the N-terminal His residue, the final Fmoc group was removed and the N-terminal was acetylated with 3:1 acetic anhydride:pyridine for 30 min at rt. Peptides were side-chain deprotected and cleaved from the resin for 2 h using a 91:3:3:3 TFA:thioanisole:TIS:H 2 O solution, except for KNS2-23-9 (Ac-His-Arg-DTyr-Tic-NH 2 ), which was cleaved in a 91:3:3:3 TFA:EDT:TIS:H 2 O solution. After cleavage, peptides were precipitated in ice-cold diethyl ether, and pelleted using a Sorvall Legend XTR centrifuge using a swinging bucket rotor (4000 rpm for 4 min at 4 • C). The peptide was washed with diethyl ether and pelleted at least three times before drying overnight in a desiccator.
The peptides were purified by RP-HPLC on a semipreparative C18 reverse-phase column (Vydac 2181010, 10 × 250 mm) using a Shimadzu UV detector (Shimadzu, Kyoto, Japan). The collected fractions were concentrated on a rotary evaporator and lyophilized. The purified compounds were characterized analytically by RP-HPLC on an analytical C18 reverse-phase column (Vydac 218104; Hichrom, Theale, UK) using two solvent systems-methanol and acetonitrile. Peptides were determined to be greater than 95% pure as assessed by peak area at 214 nm, and the correct average molecular mass was confirmed using ESI/TOF-MS (Bruker, BioTOF II, University of Minnesota Department of Chemistry Mass Spectrometry Laboratory, Minneapolis, MN, USA).

AlphaScreen Bioassay
Peptide ligands were dissolved in DMSO at stock concentrations of 10 −2 M. To assess the pharmacological activity of the tetrapeptides at the mMC1R, HEK293 cells stably expressing the mMC1R were stimulated with the ligands using the cAMP AlphaScreen assay (PerkinElmer) according to the manufacturer's instruction and as previously described [47,49,56].
Following stimulation, biotinylated cAMP (0.62 µM) and streptavidin-coated donor beads (0.5 µg) were added to the wells in a subdued light environment with 10 µL lysis buffer (0.3% Tween-20, 5 mM HEPES, and 0.1% BSA, pH = 7.4). Plates were incubated for an additional 2 h in the dark. Post incubation, the plates were read by an EnSpire plate reader (PerkinElmer, Waltham, MA, USA).

β-Galactosidase Assay
The peptide ligands were assessed for pharmacological activity at the mMC3R, mMC4R, and mMC5R using a β-galactosidase assay. Briefly, HEK293 cells stably expressing the MC3R, MC4R, or MC5R were plated into a 10 cm dish and grown to 40% confluency. The HEK293 cells were transfected with 4 µg of CRE/β-galactosidase using the calcium-phosphate method, as previously described [46]. Cells (5000 to 15,000) were plated on collagen-treated Nunclon Delta Surface 96-well plates (Thermo Fisher Scientific) and incubated at 37 • C with 5% CO 2 . Plates were stimulated 48 h post-transfection with 100 µL solutions of peptide (a seven-point dose response curve with concentrations between 10 −4 to 10 −12 M, depending on potency) or forskolin (10 −4 M) in assay media (DMEM containing 0.1 mg/mL BSA and 0.1 mM IBMX) for 6 h. The assay media was aspirated and 50 µL of lysis buffer (250 mM Tris-HCl, 0.1% Triton X-100, pH 8.0) was added to each well. Plates were stored at −80 • C for up to two weeks.
Thawed plates were assessed for protein content and assayed for β-galactosidase activity. Relative protein concentration was determined by adding 10 µL of cell lysate to 200 µL of a 1:5 dilution of Bio Rad G250 protein dye in a 96-well plate. Absorbance was measured using a 96-well plate reader (Molecular Devices) at λ = 595 nm. To determine β-galactosidase activity, 40 µL of 0.5% BSA in phosphate buffered saline (PBS) (37 • C) and 150 µL of the β-galactosidase substrate (60 mM Na 2 HPO 4 , 1 mM MgCl 2, 10 mM KCl, 50 mM 2-mercaptoethanol, and 660 µM 2-nitrophenyl β-d-galactosidase) were added to the remaining 40 µL of cell lysate. Plates were incubated at 37 • C and periodically read on the 96-well plate reader until the absorbance at λ = 405 nm reached approximately 1.0 relative absorbance units for the positive controls.

Data Analysis
The EC 50 and pA 2 values represent the mean of duplicate replicates performed in at least three independent experiments. The EC 50 and pA 2 values and their associated standard errors (SEM) were determined by fitting the data to a nonlinear least-squares analysis using the PRISM program (v4.0, GraphPad Inc., San Diego, CA, USA). The ligands were assayed as TFA salt and not corrected for peptide content.

Conclusions
The tetrapeptide Ac-His-Arg-(pI)DPhe-Tic-NH 2 , possessing a switched Arg-DPhe motif and Tic at the fourth position relative to the Ac-His-DPhe-Arg-Trp-NH 2 melanocortin agonist sequence, was characterized to be an MC3R agonist/MC4R antagonist ligand following a mixture-based positional scan to identify MC3R agonist-selective ligands. Previous characterization of the DPhe para-position within the Ac-His-DPhe-Arg-Trp-NH 2 scaffold indicated that substitutions influenced MC3R efficacy while maintaining full MC4R agonism. It was therefore hypothesized that different substitutions at the DPhe para-position in the Ac-His-Arg-(pI)DPhe-Tic-NH 2 scaffold might modulate MC4R efficacy while maintaining MC3R agonism. A range of MC4R efficacies was observed from the library of 12 compounds, including one full agonist and three ligands that possessed no agonist activity at concentrations up to 100 µM. Efficacy at the MC3R was also modulated, though all compounds maintained at least at 75% stimulation of the MC3R relative to NDP-MSH. Thus, the inversion of the Arg and DPhe positions within the melanocortin tetrapeptide sequences appears to result in preferential MC3R agonism over MC4R, a useful design motif for the development of MC3R-selective ligands that may serve as novel probe and lead ligands in the treatment of various disorders of altered energy balance.

Conflicts of Interest:
The authors declare no conflict of interest.