Exposure of cultured human melanocytes to ultraviolet radiation (UV) results in DNA damage. In melanoma, UV-signature mutations resulting from unrepaired photoproducts are rare, suggesting the possible involvement of oxidative DNA damage in melanocyte malignant transformation. Here we present data demonstrating immediate dose-dependent generation of hydrogen peroxide in UV-irradiated melanocytes, which correlated directly with a decrease in catalase activity. Pretreatment of melanocytes with α-melanocortin (α-MSH) reduced the UV-induced generation of 7,8-dihydro-8-oxyguanine (8-oxodG), a major form of oxidative DNA damage. Pretreatment with α-MSH also increased the protein levels of catalase and ferritin. The effect of α-MSH on 8-oxodG induction was mediated by activation of the melanocortin 1 receptor (MC1R), as it was absent in melanocytes expressing loss-of-function MC1R, and blocked by concomitant treatment with an analog of agouti signaling protein (ASIP), ASIP-YY. This study provides unequivocal evidence for induction of oxidative DNA damage by UV in human melanocytes and reduction of this damage by α-MSH. Our data unravel some mechanisms by which α-MSH protects melanocytes from oxidative DNA damage, which partially explain the strong association of loss-of-function MC1R with melanoma.
This study demonstrates induction of oxidative DNA damage in melanocytes by UV and the role of α-MSH in protection from this damage. We show that α-MSH limited the UV-induced 8-oxodG, by at least partially, counteracting the reduction in catalase activity. Pretreatment with α-MSH upregulated catalase and ferritin protein levels which enhance cellular antioxidant potential. By activation of early antioxidant responses, α-MSH reduced oxidative DNA damage that might be mutagenic to melanocytes. Previously, we showed that in addition to stimulation of melanogenesis, α-MSH inhibited UV-induced hydrogen peroxide generation, and that pretreatment with α-MSH enhanced repair of DNA photoproducts. Here, we further demonstrate that α-MSH is an important participant in the early antioxidant responses to UV, providing further explanation for the higher risk for individuals with loss-of-function MC1R to develop melanoma.
The skin is equipped with a cellular antioxidant defense system that neutralizes ROS. This includes non-enzymatic (e.g., α-tocopherol and vitamin C) as well as enzymatic antioxidants [e.g., SOD, catalase, glutathione peroxidase (GPx)]. An increase in antioxidant defense system whether non-enzymatic (Maalouf et al., 2002; Morley et al., 2003) or enzymatic (Rezvani et al., 2007, 2008) prevents the deleterious effects of UV-induced ROS on normal skin, such as photocarcinogenesis and photoaging. Melanin, particularly eumelanin, synthesized by melanocytes, is a scavenger of ROS (Bustamante et al., 1993). A direct correlation between catalase activity and melanin content in melanocytes in in vitro and ex vivo models has been shown, suggesting that individuals with lower melanogenic activity are more susceptible to UV-induced oxidative stress (Maresca et al., 2006, 2008; Picardo et al., 1999).
It is established that the cutaneous response to UV is mediated by a paracrine network of cytokines and growth factors. We have demonstrated that the paracrine/autocrine factors α-MSH and ACTH are major participants in the response of melanocytes to UV, mediating the melanogenic response, and importantly enhancing nucleotide excision repair (Im et al., 1998; Kadekaro et al., 2005). Interestingly, α-MSH and by analogy ACTH, reduced the generation of hydrogen peroxide by UV-irradiated melanocytes, a significant finding, as oxidative DNA damage is implicated in melanomagenesis. Unlike other skin cancers that commonly have UV-signature mutations caused by unrepaired DNA photoproducts, melanoma tumors have mutations that are characteristic of oxidative DNA damage, such as the activating BRAF mutation, and somatic p16 mutations (Landi et al., 2006; Tanaka et al., 1999).
In this study, we have addressed the protective effect of α-MSH against UV-induced oxidative stress. Our results provide evidence for the critical role of α-MSH in photoprotecting melanocytes by activating early defense mechanisms that reduce ROS generation, and hence oxidative DNA damage. These antioxidant effects of α-MSH might be critical for its role in melanoma prevention, and explain, at least in part, the high risk for individuals with loss-of-function MC1R to melanoma.
UV increased H2O2 generation in a dose-dependent manner
Irradiation of three different human melanocyte cultures with UV resulted in similar dose (0, 35, 75, 105, and 135 mJ/cm2)- and time (T0 = 1–2, 15, 30, and 45 min)-dependent release of H2O2. The level of H2O2 in the group irradiated with the highest dose of 135 mJ/cm2 UV was more than threefold higher than that of the group irradiated with the lowest dose of 35 mJ/cm2 UV 45 min after irradiation. In response to all doses of UV, the highest levels of H2O2 were measured 45 min after UV irradiation (Figure 1).
α-MSH decreased UV-induced oxidative DNA damage
Using single cell electrophoresis (comet assay) we observed that pretreatment with α-MSH (1 nM) significantly decreased [P ≤ 0.05, anova, followed by Student Newman Keuls Multiple Comparisons Test (SNK)] total (including oxidative) DNA damage induced by UV (105 mJ/cm2) (Figure 2A,B). In this assay, the extension of each comet was analyzed using the tail moment, which is considered the variable most directly related to DNA damage, and is defined as the product of DNA in the tail and the mean distance of its migration. To unequivocally demonstrate that α-MSH inhibited oxidative DNA damage, we pulse treated melanocytes for 30 min with 100 μM H2O2, with or without 4-day pretreatment with 1 nM α-MSH. One hour after H2O2 treatment comet formation showed great variability (data not shown), within 2 h all the H2O2-treated cells presented measurable tail moments, which were drastically reduced in the groups treated with α-MSH. A marked difference between these two groups remained evident 24 h after H2O2 treatment. To demonstrate that activation of MC1R is required for mediating the effects of α-MSH on the reduction of oxidative DNA damage, the same experiment was repeated using melanocytes that express loss-of-function MC1R. As expected, α-MSH had no effect on H2O2-induced tail moment, but treatment with forskolin, a direct activator of adenylate cyclase, markedly reduced DNA migration (Figure 2C, D).
Induction of 8-oxodG, the major form of oxidative DNA damage, was detected by immunofluorescence staining, using FITC-conjugated 8-oxodG antibody, 2 and 24 h after UV exposure. Melanocytes irradiated with 105 mJ/cm2 UV exhibited more than twofold higher immunofluorescence than the unirradiated control, and pretreatment with 1 nM α-MSH reduced the fluorescence of the UV-irradiated group by 50% (Figure 3A, B). To further demonstrate that activation of MC1R mediates the effects of α-MSH on the reduction of oxidative DNA damage, we have used agouti signaling protein (ASIP)-YY, a synthetic analog of ASIP, the physiological MC1R antagonist (Suzuki et al., 1997, Abdel-Malek et al., 2001). We investigated how blocking the MC1R with ASIP-YY would alter the effects of α-MSH. In the presence of ASIP-YY, the inhibitory effect of α-MSH on the UV-induced 8-oxodG was reduced by only 20% (Figure 4A, B) in contrast to 50% in the absence of the antagonist.
α-MSH reversed the UV-induced decrease in catalase enzymatic activity
Catalase exerts protective effect against oxidative stress by converting H2O2 to H2O and O2. To study the acute effects of UV on this antioxidant enzyme, the activity of catalase was measured shortly after (10, 20, 30, and 40 min) UV irradiation. Results of preliminary experiments showed that catalase activity was not altered during the first 10–20 min, but showed reduction by 30 min after UV exposure (data not shown). Therefore, we used the latter time point to measure catalase activity in subsequent experiments (Figure 5). Results showed the dose-dependent down-regulation of catalase activity by 75, 105, and 135 mJ/cm2 UV (Figure 5A). There was a strong correlation between H2O2 generation at 45 min after UV exposure and reduction of catalase activity 30 min after irradiation (r2 = 0.95, Figure 5B). Importantly, pretreatment with 1 nM α-MSH significantly inhibited the effect of UV by increasing catalase activity (Figure 5C).
α-MSH promoted increase in enzymatic and non-enzymatic antioxidant defense in hMCs
We also observed the antioxidative effect of α-MSH on the protein levels of catalase and ferritin (Figure 6). Western blot analysis of protein extracts 24 h after UV irradiation of untreated, and melanocytes chronically treated with α-MSH for 4 days, showed that UV (105 mJ/cm2) decreased catalase (Figure 6A, B), but not ferritin protein levels (Figure 6C, D). Pretreatment with α-MSH increased the expressions of catalase and ferritin proteins in unirradiated as well as UV-irradiated melanocytes (Figure 6).
The main goal of this study was to investigate the role of α-MSH in reducing UV-induced DNA oxidative stress in human melanocytes, and to elucidate some of the mechanisms involved in the immediate antioxidant response. The increase in ROS after exposure of the skin to solar UV and the generation of H2O2 during the normal process of melanin synthesis may potentially have genotoxic effects on melanocytes (Mastore et al., 2005). Inability of melanocytes to deal efficiently with oxidative stress has been implicated in vitiligo (Giovannelli et al., 2004). Additionally, somatic mutations in p16 gene, a highly penetrant melanoma susceptibility gene, and the activating BRAF mutation commonly found in premelanoma lesions, are characteristic of erroneously repaired oxidative DNA damage (Tanaka et al., 1999). These reports highlight the importance of addressing the mechanisms by which melanocytes deal with oxidative stress.
The antioxidant effect of α-MSH has long been recognized, but the impact of this effect on the UV response of human melanocytes has not been amply investigated. It has been suggested that α-MSH or its peptide analogs have the ability to reduce inflammation, as well as peroxide-generated oxidative stress in keratinocytes and melanoma cells (Haycock et al., 2000), and suppress superoxide anion generation in B16 mouse melanoma cells (Valverde et al., 1996). Previously, we reported that α-MSH reduced UV-induced H2O2 generation (Kadekaro et al., 2005). Here we show that generation of H2O2 upon UV exposure resulted in induction of oxidative DNA damage, as determined by the comet assay (Figure 2) and confirmed by immunofluorencence staining for 8-oxodG (Figure 3A, B). We found that pretreatment with α-MSH reduced the tail moment of melanocytes treated with H2O2, and decreased the positive staining for 8-oxodG, 2 h after H2O2 or UV exposure, suggesting reduction in the induction, rather than enhancement of repair of 8-oxodG by α-MSH (Figures 2A, B and 3A, B).
The observed reduced induction of oxidative DNA damage by α-MSH was mediated by activation of the MC1R, as it was absent in melanocytes expressing loss of function MC1R (Figure 2C, D), and was abrogated by concomitant treatment with ASIP YY in melanocytes expressing functional MC1R (Figure 4A, B). Reduction of oxidative DNA damage in the former melanocytes by forskolin suggests the involvement of the cAMP pathway in this effect.
Reduction of UV-induced generation of oxidative DNA damage by α-MSH was partially due to the early reversal of the UV-induced inhibition of catalase activity. We speculate that the increase in the protein levels of catalase and ferritin after prolonged treatment with α-MSH, enhance the cellular antioxidant status that sustain the inhibition of ROS formation, thus protecting from oxidative damage. Others have demonstrated that low levels of catalase enzymatic activity and high levels of oxidative substrates increase the susceptibility to UV damage (Bessou-Touya et al., 1998; Maresca et al., 2006; Picardo et al., 1999). Overexpression of catalase was found to reduce oxidative DNA damage induced by UV, and thus prevent apoptosis of skin cells and the risk of photocarcinogenesis (Rezvani et al., 2008). In this study, we showed that UV irradiation suppressed catalase activity in a dose-dependent manner (Figure 5A) and reduced its protein level (Figure 6A, B). However, pretreatment with α-MSH reversed the suppressive effect of UV on catalase activity (Figure 5C) and partially restored the levels of catalase protein in irradiated melanocytes (Figure 6A, B).
The balance between the anti- and pro-oxidant molecules determines the extent of UV-induced oxidative DNA damage. UVB induces superoxide anion radical and H2O2 and catalase converts H2O2 to H2O and O2. Hydrogen peroxide can be converted in the presence of iron, to the highly reactive hydroxyl radical via the Fenton reaction (Arosio and Levi, 2002; Mates and Sanchez-Jimenez, 2000; Regan et al., 2008). In fact, it has been proposed that the first step in the pathogenic process of cutaneous melanoma is the photo-induced release of a pool of iron cations, suggesting the role of oxidative stress in melanomagenesis (Meyskens and Berwick, 2008). Here, we demonstrate that in addition to inhibiting the UV-induced reduction of catalase activity and protein level, α-MSH increased the level of ferritin, an endogenous iron sequestrant (Figure 6C, D). Our results confirm microarray data we recently obtained, which revealed upregulation of a group of antioxidant genes, including ferritin, by α-MSH in human melanocytes (data not shown). Increased levels of ferritin are expected to reduce ROS formation by removing free iron that acts as a catalyst for hydroxyl radicals (Arosio and Levi, 2002; Mates and Sanchez-Jimenez, 2000) and to contribute to reduction in oxidative stress in melanocytes, and consequently to melanoma formation.
We conclude that α-MSH has a protective role against UV irradiation, preventing DNA damage caused by oxidative stress. The immediate effect of α-MSH in reducing the formation of H2O2 indicates that α-MSH is important in the first line of antioxidant defenses. Studies are currently being carried out to further investigate the role of α-MSH in oxidative stress, particularly in modulating the activity of NF-E2-related factor 2 (Nrf-2) a major transcription factor involved in antioxidant responses, and its downstream targets, genes with antioxidant responsive element (ARE) in their promoters, such as NAD(P)H:quinine oxidoreductase (NQO1), heme oxygenase 1 (HO-1) and glutathione S transferase.
Primary hMC cultures were established from neonatal foreskins or discarded tissue derived from plastic surgery. Each foreskin or discarded tissue was processed individually to establish single primary cultures. The Institutional Review Board at the University of Cincinnati has deemed the protocol for obtaining these skin samples exempt from approval at the Institute. Otherwise stated, the 10 different primary cultures used in this study expressed functional MC1R, as determined by α-MSH dose-dependent increase in cAMP and/or tyrosinase activity. The cells were regularly maintained in culture as described previously (Abdel-Malek et al., 1995), using complete growth medium consisted of Minimum Essential Medium Eagle (MCDB) 153 supplemented with 4% fetal bovine serum, insulin (5 μg/ml), α-tocopherol (1 μg/ml), 1% penicillin/streptomycin/amphotericin, human basic fibroblast growth factor (bFGF, 0.6 μg/ml), phorbol 12-myristate 13-acetate (PMA, 8 nM), and bovine pituitary extract (BPE, 13 μg/ml). All the medium components were from Sigma (Sigma Aldrich, St. Louis, MO, USA) except for BPE (Clonetics, San Diego, CA, USA). In the experiments, early passage (<10) cultures were used to insure minimal genetic drift in vitro, and experiments performed in MCDB supplemented medium deprived of BPE (−BPE), a well known source of growth factors.
UV source and irradiation
The UV source consisted of a bank of FS20 lamps with 75% emission in the UVB, and 25% emission in the UVA range, with peak emission at 313 nm. The UVC emission was blocked by the use of a Kodacell filter. Cells were washed twice with, and irradiated, in 5 ml PBS. Except for the dose–response to UV experiment, hMCs were irradiated with a dose of 105 mJ/cm2 UV, previously determined to induce moderate cytotoxicity (20–30% of apoptosis observed 24 h after UV exposure). The amount of UV radiation emitted by the lamps were determined using a UVB meter (National Biological Corporation, Twinsburg, OH, USA) and the exposure time was calculated by the formula Exposure Time (seconds) = UV Dose/Meter Reading.
Determination of hydrogen peroxide generation
Human MCs were plated onto 100-mm dishes at a density of 1.0 × 106 cells and kept in medium −BPE for 4 days. Melanocytes were irradiated with UV (0, 35, 75, 105, and 135 mJ/cm2). The release of H2O2 was measured 1–2 min, 15, 30, and 45 min after UV exposure, by determining the luminescence of luminol, as described in detail previously (Kadekaro et al., 2005). To confirm the specificity of the reaction, one UV-irradiated group was treated with 300 units/ml catalase (data not shown), which degraded hydrogen peroxide to water and oxygen. Luminescence readings were plotted against a standard curve generated using known concentrations of H2O2, Data were expressed as mean average of H2O2 in pmols/ml/106 cells ± SE and data compared between two consecutive doses of UV at each time point using Student’s t test, *P ≤ 0.05, significantly different from the group irradiated with dose immediately lower.
Oxidative DNA damage determined by comet assay
To determine the effects of α-MSH on oxidative DNA damage, we have used hMCs treated with H2O2 as the stress agent. Human MCs expressing functional MC1R were plated onto 60-mm dishes at a density of 0.5 × 106 cells/dish and treated with 0 or 1 nM α-MSH for 4 days. Cells were then incubated with 100 μM H2O2 for 30 min, in the presence or absence of α-MSH. After the pulse treatment, medium was replaced by fresh medium with or without α-MSH and cells returned to the incubator for another 2 h (induction oxidative damage) and 24 h (repair of oxidative damage). Cells were detached with trypsin-EDTA, counted and resuspended at a density of 1 × 105/ml in PBS (Ca2+and Mg2+free). Single cell gel electrophoresis was performed according to the manufacturer’s instructions. Briefly, melanocytes were embedded in 1% low-melting point agarose gel and placed onto comet assay slides in triplicates. The slides were kept for 30 min at 4°C to harden the agarose and then kept for 1 h at 4°C in lyses solution. Slides were transferred to alkali solution (200 mM NaOH; 1 mM EDTA, pH > 13) and incubated for 20 min at room temperature to allow unwinding of the DNA. Electrophoresis was carried out in prechilled alkali solution for 30 min at 20 V. Under these conditions, undamaged DNA remains in the nucleus while, DNA containing strand breaks (alkali-labile sites) will move toward the anode. The DNA was then stained with SYBR Green. To demonstrate that the effects of α-MSH were mediated by activation of MC1R, we have used a primary culture of hMCs expressing non-functional MC1R, as previously determined by the lack of cAMP and tyrosinase activity stimulation by α-MSH. Cells were plated and treated as described for hMCs expressing functional MC1R and after UV exposure cells returned to the incubator for 24 h before being processed for comet assay. Cells were analyzed at 400× magnification using a fluorescent microscope. The microscope images showed cells with circular shapes, indicating undamaged DNA, and cells with comet-like shapes, indicating that the damaged DNA had migrated out from the head to form a tail. The extension of each comet was analyzed using a computerized image analysis system (TriTek CometScoreTM Freeware, TriTek Corp., Sumerduck, VA, USA) that provided the tail moment, which is defined as the product of DNA in the tail and the mean distance of its migration in the tail. Data were expressed as the mean value of 30–50 randomly selected tail moments (μm) ± SE. Data were analyzed using anova, followed by Student Newman–Keuls Multiple Comparison Test (SNK), and difference between groups considered statistically significant (*) for P ≤ 0.05.
Oxidative DNA damage determined by 8-oxodG immunofluorescence
Human melanocytes were plated onto cover slips at a density of 0.15 × 105 cells in medium −BPE. For the effect of α-MSH on oxidative DNA damage, cells were pretreated with 1 nM α-MSH for 4 days before UV irradiated. After irradiation PBS was replaced by fresh medium with or without α-MSH and cells returned to the incubator for another 2 h (induction oxidative damage) or 24 h (repair of oxidative damage). After the appropriate times, cells were fixed with 4% paraformaldehyde in PBS; dehydrated in increasing series of cold methanol and permeabilized in 99% methanol for 30 min at 4°C. Cells were incubated with blocking solution at 37°C for 1 h to block non-specific binding and then incubated with FITC-conjugated 8-oxodG antibody (OxyDNA Assay Kit; Calbiochem/EMD Biosciences, La Jolla, CA, USA) or mouse IgG as a negative control, at 37°C for 1 h. The IgG controls were then incubated with anti-mouse conjugated with FITC (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA). To further demonstrate that the effects of α-MSH were mediated by activation of the MC1R, primary culture of hMCs expressing functional MC1R were treated with α-MSH in the presence and absence of 50 nM ASIP-YY primary. Immunofluorescence was observed under microscope with FITC filters. Data were expressed as the mean average of the fluorescence intensity ± SE, determined using the densitometry values from 6–8 randomly obtained images, using the computer assisted program Alpha Innotech Imaging System and the AlphaEase FC StandAlone Software (San Leandro, CA, USA). Data were analyzed using anova, followed by SNK, and difference between groups considered statistically significant (*) for P ≤ 0.05.
Determination of catalase activity
Human MCs were plated onto 100 mm dishes at a density of 1.5 × 106 cells. For the effect of increasing doses of UV on catalase activity, cells were irradiated with UV (0, 35, 75, 105, and 135 mJ/cm2) and then disrupted in RIPA buffer containing a cocktail of protease inhibitors 30 min after irradiation. For the effect of UV and α-MSH on catalase activity, cells were pretreated with α-MSH as described for oxidative DNA damage determination and cell extracts were obtained 30 min after UV exposure. Catalase activity was determined using a commercially available Amplex Red Catalase assay kit (Molecular probes; Invitrogen, Eugene, OR, USA). Briefly, triplicate aliquots of 10–15 μg of each cell lysate in a total volume of 25 μl were placed into a 96-well microplate. About 25 μl of the 40 μM H2O2 solution was added to each well containing the samples. The microplate was incubated for 30 min at room temperature to allow for the reaction of H2O2 with the endogenous catalase. Additionally 50 μl of Amplex Red/HRP solution was added to each microplate well and the reaction was incubated for 30 min at 37°C. The absorbance was detected at 560 nm emission using a microplate reader Bio-RAD model 550 and the catalase activity estimated against a stand curve generated with known amounts of catalase. Catalase activity (U/μg protein) was expressed as mean average of percent of control ± SE and analyzed using anova, followed by SNK, with difference between groups considered statistically significant (*) for P ≤ 0.05.
Western blot analysis of catalase and ferritin
Human MCs were plated onto 100-mm dishes at a density of 1.5 × 106 cells and were treated with α-MSH as described above. Cell extracts were collected 24 h after UV exposure for detection of catalase and ferritin. Western blot analysis was carried out using catalase antibody (1 : 200, mouse monoclonal; Santa Cruz Biotechnology, Santa Cruz, CA, USA) and ferritin heavy chain antibody (1 : 1000, rabbit polyclonal; Santa Cruz). The membranes were then incubated with horseradish peroxidase-conjugated anti-mouse IgG (1 : 1000; Calbiochem). Membranes were also reacted with actin antibody conjugated with horseradish peroxidase (1 : 1000; Santa Cruz Biotechnology) for loading control. The respective bands were visualized using enhanced chemiluminescence (Amersham, Arlington Heights, IL, USA). Densitometry analysis of the bands was carried out using Alpha Innotech Imaging System and the AlphaEase FC StandAlone Software (San Leandro) in three independent experiments using three different primary cultures. Data were expressed as the mean average of densitometry values (arbitrary units) ± SE. Data were analyzed using anova, followed by SNK, and difference between groups considered statistically significant (*) for P ≤ 0.05.
The authors would like to thank Dr. Glenn Milhauser (Department of Chemistry and Biochemistry, University of California) for the kind supply of the agouti signaling protein analog ASIP-YY. This study was supported by the Dermatology Foundation Research Grant and Ohio Cancer Research Associates Grant (A.L.K.) and RO1 ES 009110 (Z.A.M.).