α-Melanocyte-stimulating Hormone Reduces Impact of Proinflammatory Cytokine and Peroxide-generated Oxidative Stress on Keratinocyte and Melanoma Cell Lines*

We have previously shown that α-melanocyte-stimulating hormone (α-MSH) can oppose tumor necrosis factor α activation of NF-κB (1–2 h) and intercellular adhesion molecule 1 up-regulation (mRNA by 3 h and protein by 24 h) in melanocytes and melanoma cells. The present study reports on the ability of four MSH peptides to control intracellular peroxide levels and glutathione peroxidase (GPx) activity in pigmentary and nonpigmentary cells. In human HBL melanoma and HaCaT keratinocytes tumor necrosis factor α and H2O2 both activated GPx in a time- and concentration-dependent manner (by 30–45 min). α-MSH peptides were found to inhibit the stimulated GPx activity and had biphasic dose-response curves. MSH 1–13 and MSH [Nle4-d-Phe7] achieved maximum inhibition at 10−10 and 10−12 m, respectively. Higher concentrations (10–100 fold) of MSH 4–10 and MSH 11–13 were required to produce equivalent levels of inhibition. α-MSH was also capable of reducing peroxide accumulation within 15 min, and again this inhibition was biphasic. The data support a role of α-MSH in acute protection of cells to oxidative/cytokine action that precedes NF-κB and GPx activation. The rapidity and potency of the response to α-MSH in pigmentary and nonpigmentary cells suggest this to be a central role of this peptide in cutaneous cells.

␣-Melanocyte-stimulating hormone (␣-MSH) 1 is a 13-amino acid peptide that arises by proteolytic processing of the proopiomelanocortin precursor molecule and is produced in several vertebrate tissues, including the pituitary, gut, and skin (1,2). It is best known for its role in the control of melanogenesis in pigmentary cells. However, more recent work has demonstrated potent and broad acting roles as an antipyretic, antiinflammatory, and immunomodulatory peptide (2)(3)(4)(5)(6)(7). The ac-tions of ␣-MSH are transmitted via a family of specific MC G-protein receptors. Five different MC receptors (MC-1R to MC-5R) have been cloned and are located on pigmentary cells (e.g. melanocytes and melanoma cells) but also on several dissimilar nonpigmentary cell types (e.g. monocytes, keratinocytes, fibroblasts, endothelial cells, neural cells, and adipocytes) (8).
The mechanism by which ␣-MSH acts as an anti-inflammatory peptide is not completely understood, but a number of studies suggest inhibition of proinflammatory cytokine production/action (e.g. TNF-␣ and IL-1 (5)) or an increase in the synthesis of anti-inflammatory cytokines (e.g. IL-10 (9)). Consistent with these reports, our group has previously suggested ␣-MSH to have a role in cutaneous immunomodulation. Thus ␣-MSH was demonstrated to inhibit the ability of the proinflammatory cytokine TNF-␣ to up-regulate ICAM-1 expression on melanocytes and melanoma cells (10,11). Surface ICAM-1 expression is a necessary requirement for T-lymphocyte binding to target cells. It is known that melanoma cells are able to produce ␣-MSH/ACTH, and therefore it is possible that ␣-MSH may prevent immune system recognition of melanoma cells (10,11). ␣-MSH autocrine control may similarly extend to regulating inflammation in skin (e.g. following UV exposure) and may include autocrine/paracrine regulation on other cutaneous cells, such as keratinocytes and Langerhans cells in addition to melanocytes.
In investigating the mechanism of action of ␣-MSH inhibition of TNF-␣-stimulated up-regulation of ICAM-1 (which was evident after a 24-h exposure to TNF-␣ and ␣-MSH), we have recently reported that ␣-MSH acts as a potent inhibitor of TNF-␣-stimulated NF-B transcription factor in melanoma cells and melanocytes (12). NF-B is responsible for expression of several inflammatory and immune system genes, and hence inhibition of its activity by ␣-MSH is likely to be one important signaling pathway by which ␣-MSH exhibits anti-inflammatory behavior. Maximum inhibition of NF-B activity by ␣-MSH occurred after 2 h. The rapid activation of NF-B is triggered by several stimuli, in particular, cytokines, lipopolysaccharides, and UV light, but also by ROS or "oxidative stress" (13)(14)(15). Previous studies have suggested that the transient intracellular generation of ROS, especially as peroxide, has the potential to act as second signaling messengers although the precise identity of a particular ROS species remains to be identified (15). TNF-␣ signaling is known to generate intracellular peroxide species because of "leakage" of electrons via complexes I and III of the mitochondrial electron respiratory chain. Release of superoxide anions and rapid dismutation to peroxide by manganese-superoxide dismutase is thought to occur, and a number of studies suggest that this oxidative generation is an important step in the TNF-␣ postreceptor signaling pathway (16 -18). Levels of intracellular peroxide are controlled by the GPx family of antioxidant enzymes (19). Cellular, extracellular and a phospholipid hydroperoxide GPx isoforms exist, and each contains a seleno-cysteine catalytic center. Together, these have the ability to metabolize H 2 O 2 and organic and lipid peroxide species (catalase is also able to scavenge hydrogen peroxide but is present at lower levels than GPx in most cell types). Accordingly, GPx is thought to be of primary importance in ROS metabolism in most tissues and has therefore been suggested to have a role in controlling transcription factor activity and gene expression by regulating the availability of those peroxide derivatives reported to have second messenger potential (15).
The aim of the present study was to investigate the mechanism of how ␣-MSH assists cells to resist pro-inflammatory cytokines and oxidative stress. We examined TNF-␣ and hydrogen peroxide activation of GPx and generation of intracellular peroxide (as GPx substrate) in a cutaneous melanoma cell line and a keratinocyte cell line, investigating to what extent ␣-MSH is able to oppose these early events in the response of cells to proinflammatory and oxidative species.

EXPERIMENTAL PROCEDURES
Cell Culture-HBL is a human cutaneous melanoma cell line established in one of our laboratories (20). Cells were cultured in Ham F-10 (Life Technologies, Inc.) supplemented with 5% fetal calf serum, 5% newborm calf serum (Advanced Protein Products, West Midlands, UK), 2 mM L-glutamine (Life Technologies, Inc.), 100 units/ml penicillin, and 100 g/ml streptomycin sulfate. HaCaT is a human keratinocyte cell line obtained as a generous gift from Professor N. E. Fusenig (Institute of Biochemistry, German Cancer Research Center, Heildelberg, Germany). Cells were cultured in Dulbecco's modified Eagle's medium (Sigma, Poole, UK), supplemented with 5% fetal calf serum and 2 mM L-glutamine. Cells were incubated in a humidified atmosphere of 5% CO 2 and 95% air at 37°C.
Measurement of Glutathione Peroxidase-Melanoma cells and keratinocytes were seeded into T75 flasks at 2 ϫ 10 6 cells/flask and grown to ϳ60% confluence. Medium was changed 18 h prior to experimentation and TNF-␣ (100 units/ml to 1200 units/ml) or H 2 O 2 (100 M to 1200 M) were added for 15, 30, 45, 60, or 120 min (note that H 2 O 2 at 1200 M for 120 min was not found to be cytotoxic). ␣-MSH and related peptides, forskolin (10 Ϫ5 and 10 Ϫ4 M) or isobutylmethylxanthine (IBMX) (10 Ϫ4 and 10 Ϫ3 M), were added to the cells 15 min prior to the addition of TNF-␣ or H 2 O 2 . The peptides used were: (i) ␣-MSH 1-13, (ii) ␣-MSH 4 -10, (iii) ␣-MSH [Nle 4 -D-Phe 7 ], and (iv) ␣-MSH 11-13 at 1 ϫ 10 Ϫ13 to 1 ϫ 10 Ϫ8 M. Incubations were terminated by medium removal and washing (ϫ2) of cells in PBS. Cells were then removed by scraping and processed for glutathione peroxidase activity measurement. Glutathione peroxidase was measured according to the method in Ref. 21. Briefly, cells were removed by scraping and resuspended in 100 l of 50 mM K 2 HPO 4 /KH 2 PO 4 , pH 7.0, and placed on ice. Cell lysis was obtained by repeated needle aspiration, and the membrane fraction was then removed by centrifugation at 11,000 ϫ g for 5 min. 20 l of supernatant was used for each assay. Test mixtures for the determination of GPx contained 1 mM GSH, 0.15 mM ␤-NADPH, 0.24 units/ml glutathione reductase in 0.1 M K 2 HPO 4 /KH 2 PO 4 , pH 7.0, 1.0 mM EDTA. Reactions were started by the addition of t-butyl hydroperoxide (0.12 mM final concentration), and oxidation of ␤-NADPH was followed for 5 min at 340 nm using a Labsystems iEMS 96-well plate reading spectrophotometer at 25°C. Enzyme activities were calculated from the linear part of the kinetic curve. One unit was defined as the amount of enzyme oxidizing 0.5 mol of ␤-NADPH (corresponding to 1 mol of GSH)/min.
Measurement of Intracellular Peroxide Accumulation-To assess levels of intracellular peroxide formation, flow cytometric analysis was carried out using an oxidation-sensitive fluorescent probe DCFH-DA ( ex ϭ 488, em ϭ 510) (22). Briefly, cells (HBL or HaCaT) were cultured to ϳ60% confluence in 6-well plates (ϳ5 ϫ 10 4 cells). Cell medium was removed, and cells were washed with PBS (without Ca 2ϩ or Mg 2ϩ ions) two times, and then PBS (without Ca 2ϩ or Mg 2ϩ ions) containing 5 M DCFH-DA was added to the cells for 10 min. DCFH-DA was removed, and cells were washed thoroughly with PBS (without Ca 2ϩ or Mg 2ϩ ions) five times. Medium was added back to the cells for experimentation. Cells were incubated with ␣-MSH 1-13 (varying from 10 Ϫ11 to 10 Ϫ6 M), forskolin (10 Ϫ5 and 10 Ϫ4 M), or IBMX (10 Ϫ4 and 10 Ϫ3 M) for 5 min prior to the addition of hydrogen peroxide (300 M for the HaCaT cell line, or 600 M for the HBL cell line). Reactions were stopped after 15 min by medium removal; cells were washed three times with PBS and detached by incubation with 200 l of PBS/EDTA (0.02%, 5 min). Cells were collected in a final volume of 1 ml FacsFlow solution (Becton Dickinson) for analysis.
Protein Determination-Protein concentration was determined by use of the bicinchinonic acid kit (Pierce), using bovine serum albumin as standard.
Calculations and Statistics-Data for the GPx enzyme activity were calculated using standard unit values (as described above under "Measurement of Glutathione Peroxidase"). Data for inhibition of GPx activity were expressed as percentage values, where unstimulated enzyme activity was 0% (equivalent to 100% inhibition) and TNF-␣/H 2 O 2 stimulated activity was 100% (equivalent to 0% inhibition). Data were analyzed by the Student's t test, and results were expressed as mean Ϯ S.E.
Using the incubation times where maximum GPx activity was achieved, a concentration-dependent increase in activity was observed when cells were stimulated with TNF-␣ or hydrogen peroxide. This was maximum for HBL cells when stimulated with 900 M hydrogen peroxide (Fig. 1c Effect of ␣-MSH and Related Peptides on TNF-␣ and Hydrogen Peroxide Stimulation of Total GPx-We investigated the ability of four ␣-MSH peptides to inhibit TNF-␣ or hydrogen peroxide-stimulated increases in GPx activity, using the respective incubation times determined for each cell type. The four MSH peptides investigated were: , and (iv) ␣-MSH 11-13. Cells were preincubated for 15 min with the MSH peptides prior to the addition of either TNF-␣ or hydrogen peroxide. MSH peptides alone did not affect unstimulated cell GPx activity (data not shown), but they did inhibit GPx enzyme activity stimulated by TNF-␣ or hydrogen peroxide. In general terms, the ability of the MSH peptides to inhibit enzyme activity was similar irrespective of whether the cells were stimulated with TNF-␣ or hydrogen peroxide. However, the two cell types differed slightly in their response to the peptides ␣-MSH 4 -10 and ␣-MSH 11-13. The inhibitory data for all MSH peptides are shown in Fig. 2.
Effect of Forskolin and IBMX on Total GPx Activity-As ␣-MSH is known to operate through the G-protein activation of cyclic AMP, we substituted forskolin (which directly activates adenylate cyclase independent of the presence of the receptor) or IBMX (an inhibitor of the adenylate cyclase-generating phosphodiesterase, which degrades cyclic AMP) for ␣-MSH. Forskolin at 10 Ϫ4 M and IBMX at 10 Ϫ3 M both achieved inhibition of 60 -80% against TNF-␣-and peroxide-stimulated GPx activity in both cells (Table II).
Effect creases in intracellular fluorescence after exposure, interpreted as an increase in the level of peroxide species (potentially as lipid peroxide, organic peroxide, or hydrogen peroxide species) within the cell (Fig. 3a) (we were unable to detect any significant effect of TNF-␣ on peroxide accumulation in either cell line). When cells were preincubated with ␣-MSH for 5 min, followed by a 15-min incubation period with hydrogen peroxide, intracellular fluorescence compared with hydrogen peroxide alone stimulation was found to decrease (Fig. 3b). Fig. 4 shows a comparison of the sensitivity of HaCaT and HBL cells to ␣-MSH. For HaCaT cells a decrease of ϳ50% in fluorescence was observed using concentrations ranging from 10 Ϫ11 to 10 Ϫ8 M ␣-MSH (Fig. 4a); concentrations greater than 10 Ϫ8 M were less effective. In HBL cells the ␣-MSH inhibition of peroxide accumulation was concentration-dependent in that increasing concentrations of ␣-MSH increased the inhibition to a maximum of 50% achieved at 10 Ϫ8 M (Fig. 4b). Again, higher concentrations of ␣-MSH were less effective. When forskolin or IBMX were substituted for ␣-MSH, both were effective in inhibiting the peroxide-initiated increase in intracellular fluorescence in a manner similar to that observed for ␣-MSH (Fig. 3c). As with the inhibition of stimulated GPx activity, the concentrations of forskolin and IBMX required for effective inhibition were 10 Ϫ3 and 10 Ϫ4 M, respectively (Table III), for both HBL and HaCaT cell lines. Forskolin at 10 Ϫ4 M achieved greater than 60% inhibition of the response to hydrogen peroxide. DISCUSSION Our objective in this study was to learn more about acute signaling events that ␣-MSH initiates enabling it to oppose the actions of proinflammatory cytokines. ␣-MSH is produced in the pituitary gland, brain, and various extraneural tissues including immunocompetent cells (Langerhan's and murine Th1 dendritic epidermal cells), melanocytes, keratinocytes, and melanoma cells (23,26,27,28). Generation arises via proteolysis of proopiomelanocortin, which is also a precursor for ␤-en-dorphin, ␤-lipotrophin, and ACTH (24,25). MSH signaling is transmitted via the MC receptor family. Of the five different MC receptors (MC-1R to MC-5R) MC-1R has high affinity for ␣-MSH and is expressed on surface melanocytes, melanoma cells, monocytes, keratinocytes, fibroblasts, and endothelial cells (8, 29 -31). This receptor activates adenyl cyclase and elevates cAMP. MC-2R is found in adrenal cortical cells, and MC-3R is found in placenta, gut, and brain (30,32,33). MC-4R is found in brain and gut, and MC-5R is found in thymus, spleen, bone marrow, skeletal muscle, adipocytes, adrenal glands, and brain (34,35).
The different tissues in which ␣-MSH is released and where melanocortin receptors are expressed have led a number of investigators to suggest alternative extrapigmentary roles for MSH in controlling host reactions such as the control of fever, inflammation, and secretion of C reactive protein (2). In support of this, in vivo studies demonstrate ␣-MSH to have potent anti-inflammatory and antipyretic activities (3, 4, 6, 36 -39) which are suggested to act by two major routes: (i) in the brain via neural melanocortin receptors, which generate efferent signals that act at distant tissue sites by a ␤ 2 -adrenergic pathway and (ii) extraneurally, where ␣-MSH release at a distant tissue site acts locally via proximal melanocortin receptors (e.g. in the skin). Inflammatory inhibition is not thought to arise by a glucorticoid-dependent mechanism but by inhibiting production/action of proinflammatory cytokines (e.g. TNF-␣ (11), IL-1␤ (5), and IL-6 (4)) and by stimulating production of antiinflammatory cytokines (e.g. IL-10 (9)).
Exposure of skin tissue to stress (e.g. UV) will increase ␣-MSH production (40 -42), together with the release of proinflammatory cytokines (e.g. TNF-␣/IL-1), which normally upregulate expression of cell adhesion molecules (e.g. ICAM-1 (43)) on adjacent cells. Expression of ICAM-1 is obligatory for interaction of infiltrating T-lymphocytes with target cells, and we have previously shown that ␣-MSH strongly opposes TNF- HaCaT keratinocytes were exposed to 300 M hydrogen peroxide or 400 units/ml TNF-␣, and HBL melanoma cells were exposed to 600 M hydrogen peroxide or 400 units/ml TNF-␣. ***,p Ͻ 0.001.  ␣-stimulated up-regulation of ICAM-1 in melanocytes (10) and melanoma cells (11). Expression is thought to be related to the metastatic potential of cutaneous melanoma cells, where metastatic cells express more ICAM-1 than primary melanoma (44). However, ICAM-1 expression is also relevant for lymphocytic and macrophage interactions at primary tumor sites. Thus, the ␣-MSH attenuation of TNF-␣ may also extend to melanoma progression, especially as melanoma cells produce ␣-MSH, raising the possibility of autocrine/paracrine control of melanoma cell interaction with immune cells.
As ICAM-1 expression is central to many aspects of inflammation, we next investigated the ability of ␣-MSH to inhibit the NF-B transcription factor, which regulates expression of ICAM-1 and several other inflammatory mediators (45,46). ␣-MSH significantly reduced the activation of NF-B by TNF-␣, and p50-p65 was identified as the complex predominantly inhibited, implying functional inhibition of transcription (12). Concentrations of around 10 Ϫ9 M were most effective with higher and lower concentrations proving less effective. This biphasic concentration dependence has previously been noted by ourselves (10, 11) and by others (2) with respect to anti-inflammatory responses.  Activation of NF-B requires phosphorylation, ubiquitination, and proteosomal degradation of the inhibitor B subunit (47). Inhibitor B release allows the p50-p65 complex to migrate to the nucleus and induce transcription. Several studies indicate that activation of NF-B can be controlled by cellular redox status (13,18,48,49). Evidence is based on observations that most agents activating NF-B trigger the formation of ROS, or are oxidants themselves (e.g. superoxide, lipoxygenase products, or H 2 O 2 ). For instance, TNF-␣ is known to increase the generation and leakage of superoxide species between complexes I and III of the mitochondria (17,18). Superoxide is then rapidly dismuted to H 2 O 2 by resident manganese-superoxide dismutase (50). Rotenone (a mitochondrial complex I inhibitor) can decrease TNF-␣-stimulated NF-B activity, whereas antimycin A (a complex III inhibitor) can amplify NF-B activity (18). Activation may also arise directly by organic hydroperoxides or be induced by the addition of H 2 O 2 (used as a nonphysiological stimulus). NF-B activation is also inhibited by a number of antioxidants, either as exogenous agents or by transfection with overexpressing antioxidant enzyme vectors (15). The GPx family may therefore have a role in the control of transcription factor/NF-B activity as H 2 O 2 and organic/lipid hydroperoxides reduce intracellular glutathione levels (19), and many common antioxidants (e.g. N-acetyl cysteine and ␣-lipoic acid) elevate these levels (51,52). A wide range of substrates for GPx exist including H 2 O 2 , hydroperoxides, products of multiple lipooxygenases, and 15-lipooxygenase products, and many of these have NF-B-activating potential (19).
In this study we focused specifically on the events relevant to the immunomodulatory activities of MSH pre-NF-B activation, as we have previously shown ␣-MSH to be capable of attenuating the response to TNF-␣ on NF-B by 1-2 h (12). Thus, we looked specifically at total GPx and the generation of intracellular peroxide as factors arising before and strongly implicated in NF-B activation. We found that TNF-␣-and H 2 O 2 -stimulated total GPx activity in both the HBL and Ha-CaT cell line and that ␣-MSH and related peptides inhibited this activation for both cell types.
We investigated full-length and MSH peptide derivatives to learn more about the nature of the peptide receptor interaction. Full-length ␣-MSH has previously been reported to have pigmentary (1) and anti-inflammatory actions (2), whereas ␣-MSH [Nle 4 -D-Phe 7 ] is similar in this respect but differs from ␣-MSH in that a higher receptor binding affinity and cAMP response (up to 24-fold) is observed. ␣-MSH 4 -10 contains the essential peptide core sequence necessary for cAMP activation in melanoma cells and is the minimum sequence required for pigmentary action. However, the anti-inflammatory properties of ␣-MSH 4 -10 have not been fully investigated. Results obtained for ␣-MSH 11-13 contribute to a somewhat confused under-standing on how this peptide works. Present data show that this tripeptide responds in a similar manner to the other three MSH peptides in terms of inhibiting GPx activity, albeit at higher concentrations. Studies have previously reported ␣-MSH 11-13 to be ineffective in the lizard skin assay and adenylate cyclase assay in Cloudman S91 mouse melanoma cells (53). Also, melphalan-conjugated ␣-MSH 11-13 is reported not to bind to the MSH receptor in HBL cells, although data here are for a conjugated peptide (54). However, other studies report on MSH 11-13 being effective at opposing inflammation in vivo (36), albeit at higher concentrations than ␣-MSH. The latter is consistent with results from the current study. Indeed, very low concentrations of MSH 11-13 (10 Ϫ13 M) have been reported to stimulate IL-10 production in human peripheral blood mononuclear cells, although less effectively than ␣-MSH (9).
The predominant features of MSH inhibition were: (i) it was effective at nanomolar concentrations, (ii) the extent of the inhibition was partial rather than complete (although inhibition in the order of 70 -80% was often seen), and (iii) the response to ␣-MSH and related peptides was often biphasic (with higher concentrations proving less effective than lower concentrations) in inhibiting the stimulated GPx activity. The HBL cutaneous melanoma cells (with pigmentary potential) and the keratinocyte HaCaT cell line (which lack pigmentation ability) showed very similar responses to the full-length peptide with 10 Ϫ10 M proving the most effective concentration for inhibition of GPx activity stimulated by either TNF-␣ or H 2 O 2 . Cells were extremely sensitive to ␣-MSH [Nle 4 -D-Phe 7 ] with concentrations of 10 Ϫ12 M achieving greater than 70% inhibition. In the HBL cell line ␣-MSH 4 -10 was equipotent with ␣-MSH 1-13, and ␣-MSH 11-13 was only 10-fold less potent than ␣-MSH 1-13. In contrast, the HaCaT cell line required 100 times more of ␣-MSH 4 -10 or ␣-MSH 11-13 to achieve equivalent inhibition to that seen with the full-length peptide. The observed differences may be attributable to the number of MC-1Rs expressed/cell (which ϳ2000 -3000 for HBL cells (20)). There are, as yet, no reports of MC-1R number for HaCaT cells to the best of our knowledge.
Inhibition of total GPx was also achieved using receptorindependent methodologies for elevating cAMP, namely forskolin or IBMX, strongly suggesting the role of cAMP in the inhibitory actions of ␣-MSH. We suggest that the reduced GPx response to TNF-␣ or H 2 O 2 when MSH was present is explained by MSH acting upstream of GPx on substrate availability, rather than direct enzyme inhibition as the MSH peptides alone had no effect on basal GPx activity. This finding was substantiated by the reduction in DCFH intracellular fluorescence when MSH was present in H 2 O 2 -stimulated cells. cAMP elevation also reduced peroxide detection. In this system the TABLE III Inhibition of intracellular peroxide/oxidative stress by forskolin and IBMX The influence of forskolin and IBMX on intracellular peroxide/oxidative stress generation in HaCaT keratinocytes and HBL melanoma cells, as determined by DCFH fluorescence and flow cytometric analysis. Cells were preincubated with forskolin or IBMX for 5 min prior to stimulation with exogenous hydrogen peroxide for 15 min. HaCaT keratinocytes were exposed to 300 M hydrogen peroxide, and HBL melanoma cells were exposed to 600 M hydrogen peroxide. Values in parentheses show percentage inhibition for respective median and geo-mean values. Results shown are typical of three repeat analyses. peroxide detected is total peroxide, which is the resultant of the rate of peroxide generation less the rate of removal. As ␣-MSH, forskolin, or IBMX alone had no effect on peroxide levels, it is unlikely that their action prevents peroxide generation. Thus, we suggest that ␣-MSH acts rapidly (within 15 min) via a cAMP-dependent mechanism to remove peroxide species generated in the cell. In summary, the present investigation clearly shows that MSH and related peptides are extremely potent in opposing the generation of peroxide and inhibiting GPx activation. The rapidity of action of the peptides and their potency in pigmentary and nonpigmentary cells suggest that these events are central to MSH biology rather than peripheral to the activity of these peptides. We conclude that ␣-MSH is likely to have a major role in assisting keratinocyte and melanocytic cells to cope with oxidative stress.