Development of hMC1R Selective Small Agonists for Sunless Tanning and Prevention of Genotoxicity of UV in Melanocytes

Abstract

Activation of the human melanocortin 1 receptor (hMC1R) expressed on melanocytes by α-melanocortin plays a central role in regulating human pigmentation and reducing the genotoxicity of UV by activating DNA repair and antioxidant defenses. For the development of a hMC1R-targeted photoprotection strategy, we designed tetra- and tripeptide agonists with modifications that provide the necessary lipophilicity and hMC1R selectivity to be effective drugs. These peptides proved to be superior to most of the existing analogs of the physiological tridecapeptide α-melanocortin because of their small size and high hMC1R selectivity. Testing on primary cultures of human melanocytes showed that these peptides are highly potent with prolonged stimulation of melanogenesis, enhanced repair of UV-induced DNA photoproducts, and reduced apoptosis. One of the tripeptides, designated as LK-514 (5), with a molecular weight of 660 Da, has unprecedented (>100,000) hMC1R selectivity when compared with the other melanocortin receptors hMC3R, hMC4R, and hMC5R, and increases pigmentation (sunless tanning) in a cultured, three-dimensional skin model. These new analogs should be efficacious in preventing skin cancer, including melanoma, and treatment of skin disorders, such as vitiligo and polymorphic light eruptions.

INTRODUCTION

α-melanocortin (α-melanocyte stimulating hormone [α-MSH]) is best known for regulating integumental pigmentation in various vertebrate species, including humans (Abdel-Malek et al., 1995; Castrucci et al., 1989; Geschwind et al., 1972; Hruby et al., 1987). In addition to stimulating the synthesis of the dark photoprotective pigment eumelanin (Tamate and Takeuchi, 1984), α-MSH activates antioxidant defenses and DNA repair pathways in human melanocytes (Kadekaro et al., 2010, 2005; Song et al., 2009; Swope et al., 2020, 2014). These effects are mediated by binding and activating the human melanocortin 1 receptor (hMC1R), the only melanocortin receptor expressed on melanocytes (Suzuki et al., 1996), and define the hMC1R as a melanoma predisposition gene (Kennedy et al., 2001; Palmer et al., 2000).

For decades, there has been an interest in developing analogs of α-MSH, mainly for stimulation of skin pigmentation (tanning) without sun exposure, and for the detection and therapeutic targeting of melanoma tumor cells (Froidevaux et al., 2004; Levine et al., 1991; Raposinho et al., 2008). α-MSH is the agonist of hMC1R, hMC3R, hMC4R, and hMC5R, expressed on different cell types, and regulates many physiological functions beside pigmentation, including exocrine gland function, food intake, sexual potency, memory, and learning, and it has anti-inflammatory and antipyretic effects (Abdel-Malek, 2001; Cai and Hruby, 2016). This has made it difficult to develop selective agonists for one of these receptors. The most investigated α-MSH analog is the potent full-length [Nle⁴, D-Phe⁷]-α-MSH (NDP-α-MSH) (Sawyer et al., 1980), which is highly efficacious in stimulating sunless tanning (Barnetson et al., 2006; Levine et al., 1991). It is now Food and Drug Administration approved for treatment of erythropoietic porphyria (Langendonk et al., 2015), and has been efficacious in repigmenting vitiligo skin (Lim et al., 2015). However, NDP-α-MSH lacks selectivity for hMC1R, and beside stimulating pigmentation, it increases libido, and reduces appetite. Another analog, Melanotan II, a Food and Drug Administration approved heptapetide for treatment of hypoactive sexual desire disorder in women (Dhillon and Keam, 2019), is also nonselective for melanocortin receptors and has pigmentary effect that is undesirable for some users. Our goal is to develop a strategy for skin cancer chemoprevention and treatment of vitiligo and photosensitivity disorders based on selectively targeting hMC1R on melanocytes with druggable small melanocortin analogs that can be delivered through the skin.

We previously reported on the efficacy of tetra- and tripeptide α-MSH analogs on human melanocytes (Figure 1a) (Abdel-Malek et al., 2009, 2006). We hereby report on newly designed tetrapeptide 2 and tripeptides 3, 4, and 5, which proved to be highly selective for hMC1R. We tested the potencies of these peptides on primary cultures of human melanocytes relative to α-MSH in activating hMC1R signaling via cAMP production, stimulating melanogenesis, enhancing repair of DNA photoproducts, and reducing apoptosis induced by UV exposure. The effect of the tripeptide 5, as well as the previously reported tetrapeptide 1, on pigmentation was confirmed in melanocyte-containing three-dimensional cultured human skin substitutes. Our tetrapeptide 2 is more potent than the recently reported pentapeptide (Jackson et al., 2019) and tetrapeptides (Almeida Scalvino et al., 2018; Mowlazadeh Haghighi et al., 2018). In addition, our tripeptides 3–5 are smaller, with molecular weight <900 Da, with remarkably higher selectivity for hMC1R. Our newly designed drug-like hMC1R selective agonists should be efficacious for prevention of skin cancers, mainly melanoma, by inducing sunless tanning and reducing UV-induced DNA damage, and treatment of vitiligo and photosensitivity disorders.

Figure 1. Design of tetrapeptide (2) and super-selective hMC1R tripeptide agonists His⁶-D-(PheX)⁷-Arg⁸ (3–5).

(a) Amino acid sequence of α-MSH, NDP-α-MSH, 1 and 6, and the newly designed peptides 2–5. Minimal core sequence of α-MSH and the preserved amino acids in the derived analogs are highlighted in red, and the modifications are in purple. (b) Structures of tetrapeptide 2 and tripeptides 3–5, and the published tetrapeptide 1 and tripeptide 6, with the modifications shown in red. (c) Effects of bulky hydrophobic substituents (in green) at the D-Phe⁷ site in 1. Red arrows point to amide bonds with non-natural amino groups. (d) Molecular volume of D-Phe with two adjacent amide bonds and its bulky analogs, with the non-polar hydrophobic surfaces (in green and/or yellow) and polar amide region (in red), calculated and rendered in Trident, Wavefunction (Irvine, CA). α-MSH, α-melanocyte stimulating hormone; hMC1R, human melanocortin 1 receptor; NDP-α-MSH, [Nle⁴, D-Phe⁷]-α-MSH.

RESULTS

Chemical design of tetrapeptide and tripeptide hMC1R selective agonists

Our goal was to design druggable chemically stable selective hMC1R agonists that do not activate hMC3R, hMC4R, or hMC5R, with molecular weight <1,000 Da (Keller et al., 2006; Lipinski, 2004). Structure-activity studies on α-MSH or its full-length D-Phe⁷ analog NDP-α-MSH revealed that the minimal sequence for melanogenic activity is the tetrapeptide Ac-His⁶-Phe⁷-Arg⁸-Trp⁹-NH₂ (Holder et al., 2002; Hruby et al., 1987). As we reported earlier, N-capping of the weak nonselective melanocortin tetrapeptide His⁶-D-Phe⁷-Arg⁸-Trp⁹-NH₂ with Ph(CH₂)₃CO (LK-184; 1) dramatically increased its agonist potency at hMC1R (Koikov et al., 2003). The removal of Trp⁹ from 1 resulted in tripeptide LK-394 (6) (Figure 1a and b), with reduced potency at hMC1R in transfected cells. Testing on cultured human melanocytes confirmed that compared with α-MSH, tetrapeptide LK-184 (1) was more potent and longer acting, whereas tripeptide LK-394 (6) was less potent (Abdel-Malek et al., 2009, 2006). The minimal effective dose (MED) of 6 was 3,000-fold higher than 1 in both cAMP and tyrosinase activity assays.

Replacement of D-Phe in the tetrapeptide 1 or the tripeptide 6 with bulky hydrophobic groups (Figure 1b–d) was intended to increase stability of the resulting novel peptides owing to steric shielding of both the neighboring amide bonds (Figure 1c), and, therefore, slower enzymatic cleavage of the nonnatural substrate. Replacing D-Phe in 1 with naphthyl resulted in the tetrapeptide 2, with only one amide bond between two natural amino acids (Arg⁸-Trp⁹). Substitution of D-Phe in 6 by biphenyl resulted in 3 and 4 (N-Me amide analog), and substitution by t-Bu resulted in 5 (Figure 1b). All three tripeptides lacked amide bonds connecting the natural amino acids. These modifications increased lipophilicity of 2 compared with 1 by an order of magnitude (LogD 1.33 vs. 0.60, respectively) and prevented the drop in logD of the tripeptides caused by the removal of the hydrophobic Trp⁹ (Table 1). The logD values of the tripeptides 3, 4, and 5 (0.94, 1.24, 1.09, respectively) were even higher than the values of the unsubstituted tetrapeptide 1 (0.6). The above modifications, particularly of the tripeptides, also appear to enhance interactions of these short peptides with a presumable hydrophobic pocket of hMC1R but not hMC3R, hMC4R, and hMC5R.

Table 1.

hMC1R Selectivity of the Tetrapeptides 1 and 2, and the Tripeptides 3–6, Compared with α-MSH and NDP-α-MSH, and lipophilicity (logD) of Peptides 1–6

	hMC1R	hMC3R	hMC4R	hMC5R	Selectivity vs. hMC1R
Peptide	EC50 (nM)	EC50 (nM)	EC50 (nM)	EC50 (nM)	MC3/MC1	MC4/MC1	MC5/MC1	Log D
α-MSH	0.20 ± 0.05	0.25 ± 0.05	0.38 ± 0.07	11.1 ± 0.74	1	1.5	55	-
NDP-α-MSH	0.078 ± 0.010	0.19 ± 0.03	0.11 ± 0.02	0.49 ± 0.02	2.4	1.4	6	-
LK-184 (1)	0.031 ± 0.003	4.75 ± 0.28	0.43 ± 0.02	564 ± 5.6	153	14	18,193	0.60
LK-467 (2)	0.13 ± 0.01	210 ± 38	7.1 ± 1.1	560 ± 30	1,615	55	4,308	1.33
LK-511 (3)	0.44 ± 0.05	3570 ± 170	9,145 ± 1050	>100 μM	8,113	20,784	>>100,000	0.94
LK-513 (4)	0.42 ± 0.10	4,840 ± 380	6,100 ± 560	>100 μM	11,524	14,524	>>100,000	1.24
LK-514 (5)	0.74 ± 0.11	>100 μM	>100 μM	>100 μM	>>100,000	>>100,000	>>100,000	1.09
LK-394 (6)	0.97 ± 0.065	2,950 ± 1,700	4,360 ± 104	>100,000	13,350	4,495	>100,000	−0.59

Abbreviations: α-MSH, α-melanocyte stimulating hormone; EC₅₀, concentrations required for 50% maximal effect; HEK, human embryonic kidney; hMC1R, human melanocortin 1 receptor; NDP-α-MSH, [Nle⁴, D-Phe⁷]-α-MSH.

EC₅₀ values (mean of values obtained in 3 experiments, with duplicate samples included in each data point ±SEM) for each peptide on hMC1R, hMC3R, hMC4R, and hMC5R expressed in HEK cells were determined by measuring cAMP levels, as described in Materials and Methods. The ratios of EC₅₀ values on hMC3R, hMC4R, or hMC5R to EC₅₀ values on hMC1R for each peptide were then calculated. LogD values for peptides 1–6 were determined as described in Materials and Methods.

The peptides 2–5 showed high selectivity for hMC1R, as measured by their efficacy to increase intracellular cAMP levels in human embryonic kidney cells expressing hMC1R, hMC3R, hMC4R, or hMC5R (Table 1 and Supplementary Figure S1). The two tetrapeptides, 1 and 2, had markedly lower concentrations required for 50% maximal effect in cells expressing hMC1R versus the other 3 receptors. However, 2 was more hMC1R selective than 1, with higher concentrations required for 50% maximal effect on hMC3R (44-fold) and hMC4R (16.5-fold). The tripeptides 3, 4, and 5 were even more selective for hMC1R, with remarkably lower activity on the hMC3R, hMC4R, and hMC5R, with 5 being the most selective. As expected, α-MSH and its full-length analog NDP-α-MSH were nonselective. The potency of the peptides was not tested on hMC2R, which only binds ACTH and not any other physiological melanocortin or melanocortin analog (Schiöth et al., 1996).

Effects of peptides on cAMP levels in melanocytes

Dose-response experiments were conducted to measure the effects of peptides 1–5 on the levels of intracellular cAMP, the second messenger for hMC1R, in human melanocytes (Suzuki et al., 1996) (Figure 2a and b). To test the potency of the tetrapeptides 1 and 2, melanocytes were treated with increasing doses of either peptide, ranging from 1 pM to 10 nM (Figure 2a). The MED of either tetrapeptide was 1 pM, compared with the previously reported MED of 0.1 nM for α-MSH (Suzuki et al., 1996). At 10 pM, 100 pM, and 1 nM, both tetrapeptides had equal effects on cAMP, but at 10 nM, 2 was less effective than 1. The effects of the tripeptides 3, 4, or 5, as compared with α-MSH, on cAMP levels were determined after treatment with concentrations of 1, 10, or 100 nM (Figure 2b). The MED of 3 and 5 was 1 nM, while that of 4 was 10 nM. At 10 or 100 nM, 4 and 5 were equally potent, but significantly more effective than 3, and comparable to 1 or 10 nM α-MSH, respectively. Therefore, despite their very small size, 4 and 5 are only 10-fold less potent than α-MSH in activating the hMC1R.

Figure 2. Dose-dependent effects of 1, 2, 3, 4, 5 on cAMP levels and tyrosinase activity in human melanocytes.

Dose-dependent effects on cAMP of (a) tetrapeptides 1 and 2, and (b) tripeptides 3–5, versus α-MSH. All concentrations of 1 and 2, 3–5 and α-MSH, were statistically different from control, except 1 nM 4 (P ≤ 0.05). (a) *= 1 and 10 pM 1 statistically different from 1 and 10 nM 1; ^= 100 pM 1 statistically different from 10 nM 1; #= 10 nM 1 statistically different from 10 nM 2 (P ≤ 0.05). (b) *= 1 nM α-MSH statistically different from 1 or 10 nM 3, and 1 nM 4 and 5; ^= 10 nM α-MSH statistically different from 1 or 10 nM 3, 4, and 5; #= 100 nM α-MSH statistically different from 1, 10, and 100 nM 3 and 5, and from 1 and 10 nM 4; += 100 nM 3 statistically different from 100 nM 5 (P ≤ 0.05). (c) Dose-dependent effects of 1 and 2 on tyrosinase activity, as compared with α-MSH. *= 10 pM α-MSH statistically different from 1 pM 1; ^= 100 pM α-MSH statistically different from 10 pM 1; #= 100 pM α-MSH statistically different from 100 pM 2. (d) Dose-dependent effects of 3–5 on tyrosinase activity, as compared with α-MSH. *= 100 pM α-MSH statistically different from 1 nM 3, 4 and 5; ^= 1, 10, and 100 nM α-MSH statistically different from 100 nM 3, 4, and 5. In (a–d), each data point represents the mean of triplicate determinations ± SEM. P ≤ 0.05.

The above effects of these peptides were similar in cultured melanocytes wild-type for hMC1R (Figure 2a and b), or heterozygous for one loss of function hMC1R variant, D294H or R160W, which increases melanoma risk (Supplementary Figure S2a) (Kadekaro et al., 2010; Kennedy et al., 2001; Palmer et al., 2000). In the latter melanocytes, 1 was more potent, whereas 5 was effective, yet less potent than α-MSH (Supplementary Figure S2a). Collectively, these data show that the efficacy of our peptides in activating hMC1R in melanoma-prone melanocytes. None of the tetra- or tripeptides had any stimulatory effect on cAMP levels in melanocytes expressing 2 loss of function hMC1R variants, which cause loss of hMC1R activity (Supplementary Figure S2b and c) (Kadekaro et al., 2010), confirming that their effects are mediated by activating hMC1R.

Effects of peptides on melanogenesis

The efficacy of the tetrapeptide and tripeptides in increasing melanogenesis was first determined by comparing the stimulation of the activity of tyrosinase, the rate-limiting enzyme for melanin synthesis, using melanocyte strains derived from dark skin (Supplementary Table S1). Dose-response experiments carried out on four different melanocyte strains showed that the MED of the tetrapeptides 1 and 2 was 1 pM, compared with the MED of α-MSH equivalent to 10 pM (Figure 2c and Supplementary Figure S3). Similar responses were obtained with all four melanocyte strains, demonstrating that 1 and 2 were at least 10-fold more potent than α-MSH. In each experiment, at 10 nM concentration, 1, 2, and α-MSH resulted in comparable maximal stimulation of tyrosinase activity. Compared with the full-length analog NDP-α-MSH, at the lowest concentrations of 1 or 10 pM, 1 was more potent in stimulating tyrosinase activity (Supplementary Figure S4).

The MED of the tripeptides 3, 4, and 5 was 10-fold lower than that of α-MSH (100 pM, compared with 10 pM α-MSH) (Figure 2d). In three additional melanocyte strains (Supplementary Figure S5), the tripeptides were at least 10-fold less potent than α-MSH, yet all four melanocyte strains responded with considerable and significant stimulation of tyrosinase activity, with the effect of 100 nM being less than or equal to that of 1 nM α-MSH. The tripeptide 5 seemed to have a greater effect than 3 or 4, resulting with a higher maximal stimulation of tyrosinase activity at 100 nM concentration. All peptides tested, increased melanocyte proliferation dose-dependently (Supplementary Figure S6).

The dose-dependent effects of the peptides on tyrosinase activity were quantified following treatment for a total of 6 days (Figure 2c and d and Supplementary Figures S3–S5). The kinetics of stimulation of tyrosinase activity by 1 nM 1 or 2, 100 nM 5 (representative of the 3 tripeptides), or 10 nM α-MSH (doses that resulted in maximal stimulation of tyrosinase activity in response to these peptides) was determined after 2, 4, or 6 days of treatment (Figure 3a and Supplementary Figure S7). In all five melanocyte strains tested, a statistically significant increase in tyrosinase activity was achieved after 2 days, and in three of the five strains, maximal increase was observed after 4 days of treatment with each peptide (Figure 3a and Supplementary Figure S7a and b). In the fourth melanocyte strain, maximal increase was achieved on day 4 of treatment with α-MSH or 1, and on day 6 of treatment with 2 or5 (Supplementary Figure S7c). In the fifth strain, maximal stimulation of tyrosinase activity by all peptides was achieved after 2 days of treatment (Supplementary Figure S7d).

Figure 3. Kinetics of the melanogenic effects of 1, 2, and 5, and their residual effects, as compared with α-MSH or NDP-α-MSH.

(a) Time-dependent increase in tyrosinase activity after 2, 4, or 6 days of treatment with α-MSH, 1, 2, or 5. All SEM were less than 10%. Increases from 2–4 d, and from 2–6 d were statistically significant (P ≤ 0.05). (b) Western blot analysis of MITF, tyrosinase and TRP-1, following treatment of melanocytes as in (a). (c) Residual effects of 1, 2, and 5, versus α-MSH and NDP-α-MSH. 0 d = end of 4 day-treatment; 2 d, 4 d, 6 d, and 8 d in absence of treatment. SEM were less than 10%. 1, 2, and 5, were significantly different from control at all time points. α-MSH and NDP-α-MSH were statistically different from control at 0 d, 2 d, and 4 d. The decline in tyrosinase activity in the α-MSH-treated groups was statistically significant on 2 d, 4 d, and 6 d, in the groups treated with 1 on 4 d and 6 d, and with 2 and 5 on 6 d. 1, 2, 5 were statistically different from α-MSH on 6 d and 8 d, and 5 was statistically different from 1 and 2 on d 8 (P ≤ 0.05). α-MSH, α-melanocyte stimulating hormone; d, day; NDP-α-MSH, [Nle⁴, D-Phe⁷]-α-MSH.

Western blot analysis was carried out using the same treatment protocol as illustrated in Figure 3a and Supplementary Figure S7 to detect the protein levels of the melanogenic regulators, the transcription factor MITF and its downstream targets tyrosinase and TRP-1, in three different melanocyte strains derived from lightly pigmented skin with a low melanin content (Fang et al., 2002) (Figure 3b and Supplementary Table S2). Densitometry analysis revealed marked increases in the protein levels of MITF after 2 days of treatment, except in response to 2 and 5 in one melanocyte strain (Supplementary Table S2b). Increases in tyrosinase and TRP-1 levels were evident in all three strains after 2 days of treatment, and TRP-1 levels remained considerably higher than control up to 6 days of treatment (Supplementary Table S2a, b, and c). These results further confirm the melanogenic effects of these peptides on melanocytes, regardless of constitutive pigmentation.

The residual effects of the peptides on tyrosinase activity was determined in darkly pigmented melanocytes treated for 4 days with 1 nM of 1, 2, or NDP-α-MSH, 100 nM of 5, or 10 nM of α-MSH, at 2, 4, 6, and 8 days after cessation of treatment (Figure 3c and Supplementary Figure S8). Tyrosinase activity diminished gradually after all of the above treatments, with the greatest and fastest decline close to baseline levels by day 6 observed in the groups treated with α-MSH in all five strains, and NDP-α-MSH, which was tested in four of the five strains. The residual effects of 1, 2, and 5 remained statistically higher than control until at least day 6. In three of the five strains tested, 5 had the greatest residual effect (Figure 3c and Supplementary Figure S8a and b).

Pigmentary effect of peptides on cultured human skin substitutes

The tetrapeptide 1 and the tripeptide 5 were tested for their effects on pigmentation of three-dimensional melanocyte-containing cultured human skin substitutes. After 10 days of daily treatment, both peptides induced a macroscopically visual increase in pigmentation (Figure 4a). Fontana-Masson staining for melanin revealed that 1 resulted in a dose-dependent increase in pigmentation, with 1 and 100 nM resulting in 60% and 2.5-fold increase, respectively, and treatment with 5 μM of 5 increased pigmentation by 50% above vehicle-treated control (Figure 4b and c).

Figure 4. Pigmentary effects of 1 and 5 on cultured skin substitutes.

Vehicle (DMSO), 1 at 1 or 100 nM, or 5 at 5 μM was added to the culture medium daily for 10 days. (a) Macroscopic visualization of the increase in pigmentation in paraffin-embedded sections in response to 1, 5, or vehicle. (b) Light microscopic images of Fontana-Masson staining of vehicle control-, 1-, or 5-treated sections. (c) Quantification of the increase in Fontana-Masson-positive area/total epidermal area, using Image J software (mean of three different experiments testing 1, and 4 experiments testing 5; (*)= statistically different from control at P < 0.05. (d) Histology of cultured skin sections depicted by H&E staining. (e) Immunostaining of melanocytes for TRP-1 to detect melanocytes. Bar = 100 μm.

Neither peptide had any adverse effect on the histology of the cultured skin, as detected by H&E staining and TRP-1 immunostaining for melanocytes (Figure 4d and e). Importantly the tetrapeptide 1 did not have a significant effect on the number of epidermal melanocytes (TRP-1 positive cells) (Supplementary Figure S9a and b).

Repair of UV-induced cyclobutane pyrimidine dimers and inhibition of apoptosis by peptides

Activation of the hMC1R enhances the repair of DNA damage and inhibits apoptosis in UV-irradiated melanocytes (Böhm et al., 2005; Jarrett et al., 2014; Kadekaro et al., 2010, 2005; Swope et al., 2020, 2014). Lightly pigmented melanocyte strains, which are highly susceptible to DNA damage (Hauser et al., 2006), were used to determine the effect of peptides on repair of cyclobutane pyrimidine dimers (CPDs), the major form of UV-induced DNA photoproducts, and apoptosis. Treatment of human melanocytes with 1 nM of 1 or 2 and 100 nM of 3, 4, or 5, significantly reduced the levels of UV-induced CPD by 20%, similar to those for 10 nM of α-MSH (Figure 5a), and apoptosis by 40% less than UV, similar to that for 1 nM of α-MSH (Figure 5b).

Figure 5. Enhanced repair of UV-induced CPD, and reduction of UV-induced apoptosis by analogs 1, 2, 3, 4, 5, as compared with α-MSH.

Melanocytes were pretreated with 10 nM or 1 nM α-MSH (in a and b, respectively), 1 nM 1 or 2, or 100 nM 3, 4, or 5 for 4 days prior, and immediately after UV exposure for 48 hours in (a), or 24 hours in (b). In (a) median fluorescence of CPD staining was quantified by flow cytometry. The data are the mean of five independent experiments using five lightly pigmented melanocyte strains. In (b), live melanocytes were stained for Annexin V, and analyzed by flow cytometry. The presented data are the mean of three different experiments, using three lightly pigmented melanocyte strains. * = statistically different from UV at P ≤ 0.05. α-MSH, α-melanocyte stimulating hormone; CPD, cyclobutane pyrimidine dimers.

DISCUSSION

α-MSH is the physiological agonist of the melanocortin 1 receptor, a G_s protein coupled receptor expressed on the cell surface of melanocytes and a central regulator of constitutive pigmentation, tanning ability, and DNA repair capacity, innate protective mechanisms against the photocarcinogenic effects of solar UV (Böhm et al., 2005; Box et al., 1997; DiGiovanna and Kraemer, 2012; Halder and Bridgeman-Shah, 1995; Jarrett et al., 2014; Kadekaro et al., 2010; Smith et al., 1998; Swope et al., 2020, 2014).

Despite the known photodamaging effects of sun exposure, in western cultures tanning is perceived as a sign for beauty and wellness. This sparked interest in developing safe sunless tanning agents. Given the significance of hMC1R in the UV response and maintenance of genomic stability of human melanocytes, we aimed at targeting this receptor with small agonists that can potentially be used topically for stimulating pigmentation without sun exposure, and additionally preventing UV-induced mutations by enhancing DNA repair. We have developed tetra- (2) and tripeptide analogs (3–5) of α-MSH that are highly selectivity for hMC1R, which should prevent off target effects resulting from binding other melanocortin receptors (Figure 1, Table 1, and Supplementary Figure S1). Particularly the tripeptide 5, with molecular weight of only 600 Da, is superior to the currently available α-MSH analogs in its impressive hMC1R selectivity, potency, and prolonged effects (Table 1; Figures 2b, d, and 3; Supplementary Figures S1, S5, S7, and S8; and Supplementary Table S2). Although our peptides are intended for topical application, their selectivity for hMC1R is important to insure lack of binding to other melanocortin receptors expressed on other skin cells, namely hMC5R, known to be expressed on sebocytes (Zhang et al., 2011). Also, our tetrapeptides 1 and 2 are more potent than the tetrapeptides recently reported by others (Almeida Scalvino et al., 2018; Mowlazadeh Haghighi et al., 2018).

Functional assays demonstrated that the tetrapeptide 1, as reported previously (Abdel-Malek et al., 2006), and the newly designed tetrapeptide 2 are more potent than α-MSH, with 10-fold lower MED for activating hMC1R and melanogenesis, in melanocytes representative of different pigmentary phenotypes (Figures 2a, c, and 3b; Supplementary Figure S3; and Supplementary Table S2). Tetrapeptide 1 is even more potent than the full-length analog NDP-α-MSH in stimulating tyrosinase activity (Supplementary Figure S5). Tetrapeptides 1 and 2, and more so 5, had longer lasting residual effects on tyrosinase activity than NDP-α-MSH or α-MSH, which is desirable for drug development (Figure 3c and Supplementary Figure S8). We speculate that the greater residual effects of our peptides 1, 2, and 5, as compared with α-MSH and NDP-α-MSH, are due to differences in their dissociation kinetics from the receptor (Haskell-Luevano et al., 1996), and/or their different binding sites (pockets) on the hMC1R. The expected stability of the peptides, based on their chemical structure, might contribute to their possible prolonged binding and activation of the receptor. Despite their small size, the tripeptides 3, 4, and 5 still had substantial melanotropic activity, with MED about 10-fold higher than that of α-MSH (Figure 2d and Supplementary Figure S5). That 1, 2, and 5 increased the protein levels of the transcription factor MITF, the master regulator of melanocytes, and its downstream targets tyrosinase and TRP-1 (Fang et al., 2002) (Figure 3b and Supplementary Table S2) strongly suggests stimulation of eumelanin synthesis, a known effect of α-MSH (Sakai et al., 1997). Despite individual genetic variability and pigmentary phenotype, all the tested melanocyte strains that expressed functional hMC1R, including those heterozygous for a hMC1R loss of function variant, responded similarly to our peptides with increase in cAMP levels (Figure 2a and b and Supplementary Figure S2a). The melanogenic effects of peptides 1 and 5 were further substantiated by their ability to increase pigmentation of cultured human skin substitutes (Figure 4a, b, and c). In addition, our peptides reduced UV-induced CPDs and apoptosis of cultured melanocytes (Figure 5). These results strongly support the efficacy of our peptides in reducing UV-induced genotoxicity in melanocytes by increasing pigmentation and reducing DNA damage, as determined by the multiple functional assays we used.

There are concerns about melanocortin analogs inducing eruption of nevi (Cardones and Grichnik, 2009). The increase in proliferation of cultured melanocytes after peptide treatment (Supplementary Figure S6a, b, and c) can be attributed to synergistic interaction with other mitogens present in the melanocyte growth medium. However, in three-dimensional cultured skin, a more physiologically relevant milieu, the number of melanocytes did not significantly change upon treatment with the potent tetrapeptide 1 (Supplementary Figure S9a and b).

Targeting the hMC1R with potent selective agonists for chemoprevention of skin cancers, including melanoma, will benefit millions, mainly 50% of all whites in the United States who are heterozygous for a hMC1R variant, a genotype that contributes to the risk for melanoma (Kadekaro et al., 2010; Kennedy et al., 2001; Schaffer and Bolognia, 2001). Since the effects of our peptides require binding and activating of hMC1R, they will not be efficacious for individuals expressing two loss of function hMC1R variants, which render the receptor inactive. Our long-term goal is to develop our novel peptides, particularly the tripeptides, into a topical application, for practical use. Their prosurvival effects can potentially enhance and sustain repigmentation of vitiligo skin, a disorder that impacts about 1% of the world population (Ezzedine et al., 2015; Lim et al., 2015). The unique attributes of our novel peptides, including high hMC1R selectivity, small size and molecular weight < 900 Da, high lipophilicity, potency, and prolonged residual effects, make them superior to other existing melanocortin analogs, and more amenable for drug development. To the best of our knowledge, our tripeptides are the only example where counterintuitive shortening of the native ligand resulted in an unanticipated increase in hMC1R selectivity, and retention of substantial melanotropic activity.

MATERIALS AND METHODS

Peptide synthesis

The peptides trifluoracetic acid salts (>95% purity) were obtained using solid-phase and Fmoc chemistry, as described previously (Koikov et al., 2003), purified by HPLC, and characterized by electrospray ionization mass spectrometry (Waters Corporation, Milford, MA).

LogD Determination

Lipophilicity of the peptides was determined by Analiza (Cleveland, OH), by measuring L\log D values in octanol-10 mM universal buffer (0.15 M sodium chloride and 0.01 each of phosphoric, boric, and acetic acids) adjusted to pH 7.4.

Determination of selectivity of peptides for hMC1R

The AlphaScreen cAMP Functional Bioassay (AlphaScreen cAMP technology; PerkinElmer Life Sciences) was utilized to measure cAMP signaling after ligand stimulation in human embryonic kidney 293 cells stably expressing the hMC1R, hMC3R, hMC4R, and hMC5R, as previously published (Lensing et al., 2019). Dose-response experiments (three independent experiments) were conducted using peptides at concentrations ranging from 0.1 pM to 100 mM, with duplicate data points per group. Data were analyzed using the PRISM program (v4.0; GraphPad, San Diego, CA), and expressed as the mean AlphaScreen signal (in c.p.m.).

Primary human melanocyte cultures

Each primary human melanocyte culture was established from a single neonatal foreskin, or surgically discarded adult skin. Collection of discarded human skin has been deemed exempt from approval by the University of Cincinnati Institutional Review Board. The peptides were tested under culture conditions previously described (Abdel-Malek et al., 1995). In case of any contaminating fibroblasts, primary melanocyte cultures were treated once or twice for 1 week with 100 μg/ml geniticin (G418 sulfate; Gibco Laboratories, Waltham, MA) that preferentially kills fibroblasts (Halaban and Alfano, 1984). The purity of the melanocyte cultures was confirmed by immunostaining for the melanocyte marker TRP-1. For all experiments, proliferating melanocyte strains between passage 3–10 were used. Total melanin content was determined as described (Lee et al., 1972), and hMC1R sequencing was performed as previously reported (Scott et al., 2002).

Quantitation of cAMP levels in human melanocytes

cAMP levels were measured after treatment of cultured melanocytes for 45 minutes with peptides or α-MSH, as described (Suzuki et al., 1996).

Quantitation of the effects of peptides on tyrosinase activity

To determine the effects of peptides and α-MSH one tyrosinase activity and proliferation, melanocytes derived from darkly pigmented skin were plated at a density of 0.15 × 10⁶ cells/well in 6-well plates (n = 3 wells per group), and 48 hours thereafter, fresh medium and treatment were added every other day for the indicated time periods. Tyrosinase activity was assayed as described previously (Abdel-Malek et al., 1992). The cell number in each well was counted using a Coulter Counter. Data were calculated as mean tyrosinase activity (in disintegrations/minute)/10⁶ cells, then expressed as percent of control ±SEM.

For dose-response experiments, four different melanocyte strains were tested in independent experiments, and melanocytes were treated for 6 days with increasing concentrations of α-MSH or peptides. To determine the kinetics of the effects of peptides and α-MSH on tyrosinase activity, melanocytes were treated for 2, 4, or 6 days with the concentration of peptide that resulted in maximal increase in tyrosinase activity (1 nM of 1, 2, and NDP-α-MSH; 10 nM of α-MSH; 100 nM of 3, 4, and 5). Five different melanocyte strains were tested in independent experiments (Supplementary Table S1). To compare the residual effects of 1e5, α-MSH, and NDP-α-MSH on tyrosinase activity, melanocytes were treated with the above doses of peptides, α-MSH, or NDP-α-MSH, for 4 days, then maintained in medium devoid of any treatment for 2, 4, 6, and 8 days thereafter. Tyrosinase activity was determined at each of these time points. Five different melanocyte strains were tested in independent experiments (Supplementary Table S1).

Western blot analysis of MITF, tyrosinase, and TRP-1

Western blot analysis was carried out to detect the protein levels of MITF, tyrosinase, and TRP-1. Three different melanocyte cultures from lightly pigmented foreskins with low melanin content (Supplementary Table S1) were used in independent experiments. Total cell lysates of melanocytes treated with 1 nM of 1 or 2, 100 nM of 5, or 10 nM of α-MSH for 2, 4, or 6 days, and antibodies specific for MITF (Cell Signaling Technology #97800 Boston, MA), tyrosinase (Santa Cruz Biotechnology #sc20035 Dallas, TX) and TRP-1 (Santa Cruz Biotechnology #sc166857) were used. Molecular weight markers (BioRad Precision Plus Protein Western C Standards, BioRad, Hercules, CA) were ran to ensure the authenticity of the detected bands. Densitometry analysis was carried out, comparing each band to its respective actin loading control. Data were expressed as percent of untreated control.

Determination of the pigmentary effect of peptides on cultured human skin

The pigmentary effects of the peptides were validated using three-dimensional cultured human skin, which was developed in our laboratory as described (Boyce et al., 2017). Cultures were treated daily with vehicle (DMSO), 1 or 100 nM of 1 or 5 μM of 5, with peptide or vehicle added to the culture medium. After 10 days of treatment, cultured skin samples were fixed in formalin, embedded in paraffin and sectioned. Sections were stained with H&E for histological examination, and with Fontana-Masson for melanin. Sections were also immunostained with TRP-1 antibody to detect melanocytes. Fontana-Masson stain was quantified using Image J by measuring the staining per unit area of the epidermis (3–6 sections per group per experiment, with 10–15 photographic images analyzed per section). Statistical analysis was performed using GraphPad and Student’s t-test.

Determination of the effects of peptides on repair of UV-induced CPD in human melanocytes

Lightly pigmented melanocytes were treated with 0, 1 nM of 1 or 2, 100 nM of 3, 4, or 5, or 10 nM of α-MSH for 4 days before, and after irradiation with 90 mJ/cm² UV emitted from FS 20 lamps (National Biological, Twinsburg, OH) with Kodacel filter (Kodak Eastman, Rochester, NY) to remove any UVC rays, harvested 48 hours post UV, and immunostained with a specific antibody for CPDs (Cosmo Bio, Tokyo, Japan; five melanocyte strains), or harvested 24 hours post UV, and stained for Annexin V (three melanocyte strains), as described (Kadekaro et al., 2010). After CPD or Annexin V staining, melanocytes were analyzed by flow cytometry (10,000 events per sample, with triplicate samples per group). Data are presented as percent of UV group ± SEM.

Statistical analysis

Unless otherwise stated, statistical analysis was carried out using one-way ANOVA followed by Tukey’s Multiple Comparison Test, using GraphPad statistical software.

Data availability statement

No datasets were generated or analyzed during this study.

Abbreviations:

α-MSH

α-melanocyte stimulating hormone

CPD

cyclobutane pyrimidine dimers

hMC1R

human melanocortin 1 receptor

MED

minimal effective dose

NDP-α-MSH

[Nle⁴, D-Phe⁷]-α-melanocyte stimulating hormone