Role of UV in Cutaneous Melanoma†
This paper is part of a special issue dedicated to Professor Hasan Mukhtar on the occasion of his 60th birthday.
*Corresponding author email: email@example.com (Vijayasaradhi Setaluri)
Malignant melanoma arises from epidermal melanocytes, the cells responsible for the production of the skin pigment melanin. The photoprotective role of melanin, which is transferred to neighboring keratinocytes, in UV-induced skin carcinogenesis, specifically in nonmelanoma skin cancers, has been well documented. Although melanocyte-resident melanin is expected to offer similar protection to melanocytes from UV-induced damage, UV radiation has long been suspected to have an etiologic role in cutaneous melanoma. However, nearly three decades of efforts using a variety of in vitro and in vivo models of human skin and mouse genetic models have produced conflicting data. Epidemiologic studies have also failed to establish a definitive association between UV exposure and risk of melanoma. In this review, we evaluate the dual role of the melanin pigment as a photoprotector as well as a photosensitizer and examine the evidence for association between melanin levels (constitutive and induced) and melanoma risk. We also discuss possible reasons for the lack of signature UV mutations in melanoma oncogenes known to date and potential alternative mechanisms to explain the role of UV in melanomagenesis.
The incidence of skin cancer has been increasing dramatically (1,2) and more than 1 million cases of skin cancer are reported annually in the United States. Melanoma is the most serious form of skin cancer. Melanoma incidence and mortality rates have increased dramatically in the past few decades in the United States (3,4). In the year 2007, about 59 940 persons are expected to be diagnosed with melanoma resulting in the death of an estimated 8110 individuals (5). Also, alarmingly, the incidence of melanoma is increasing rapidly in children (6). Exposure to solar radiation, especially the UV region (wavelengths 280–400 nm) of the solar spectrum, appears to play a role in the development of skin cancers (7,8). While some sunlight is needed to synthesize vitamin D, which is necessary for human health, increased exposure to UV radiation is harmful to human health. UV is genotoxic and UV exposure damages a wide range of organic molecules including DNA and proteins, and is considered to be the major and ubiquitous carcinogen reaching the earth’s surface from the sun. Luckily, UV light does not penetrate the body any deeper than the skin. Within the spectrum of UV radiation, UVB is well documented as the cause of nonmelanoma skin cancer. However, the role of UV radiation in melanoma tumorigenesis is somewhat controversial. This is partly due to the ability of melanocytes, the cells from which melanoma originates, to produce melanin pigment, which is thought to be both a photoprotector and a photosensitizer. In this review, we provide a comprehensive overview of studies on the role of UV in melanoma and evaluate the evidence that support or question the relationship between UV and cutaneous melanoma.
Sunlight is the fundamental energy for life on earth. Solar radiation is a continuous spectrum of electromagnetic radiation that consists of electrical, radio, microwave, infrared, visible, UV, X-rays and γ-rays. Due to its effects on both biotic (living organisms) (9,10) and abiotic (physical and chemical aspects of an organism’s environment) factors (11), the UV region of sunlight has received more attention. UV radiation is also the most significant region of solar radiation with respect to skin cancer. The UV region of the spectrum is subdivided into the long wavelength UVA (320–400 nm) and short UVB (280–320 nm) and shorter wavelength UVC (200–280 nm). UVC which is generally absorbed by oxygen and ozone in the atmosphere does not reach the earth’s surface and is less harmful. Therefore, nearly all studies on the effects of UV on human skin have been focused on UVA and UVB.
UV-Induced Cellular Damage
Most of the UVB radiation from the sun is absorbed in the ozone layer of the stratosphere. The depletion of the ozone layer in the stratosphere has steadily increased the solar UVB reaching the earth’s surface (12,13). Energy from UV radiation is commonly absorbed by the cellular proteins and DNA. UV light is known to induce several types of mutagenic DNA lesions. Major photoproducts induced by carcinogenic UV radiation are the cyclobutane pyrimidine dimers (CPDs), pyrimidine-pyrimidone (6-4) photoproducts (6-4PPs) and their Dewar valence isomers (reviewed in Refs. [14–16]). Regions of DNA containing 5-methylcytosine have been identified as hotspots for UVB-induced mutation (17–20). More recently, it has been shown that UVB can also cause significant DNA degradation (21), presumably via the p38 pathway through the generation of oxidative stress and merging with the p53 pathway (22).
UVB is considered more carcinogenic than UVA; however, data are conflicting. UVA is more abundant in sunlight and can penetrate deeper into the skin than UVB. UVA wavelength is not significantly absorbed by the native DNA and is less efficient in inducing direct DNA damage. UVA might indirectly damage DNA through the absorption by non-DNA endogenous sensitizers and by the formation of reactive oxygen radicals (23,24). UVA exposure can also produce secondary photoreactions of the existing photoproducts or damage DNA by indirect photosensitizing reactions (25). UVA is also known to induce photoproduct 8-oxo-7,8-dihydro-2′-deoxyguanosine and CPDs in human skin (26). UVA contributes more to the formation of recurrent or hotspot mutations at methylated CpG sites in the mammalian genome than UVB (27). Thus, UVA clearly contributes to the biologic effects of solar radiation.
Melanocytes, Melanin and Protection from UV
The ability to cope with the harmful effect of UV radiation depends on the ability of cells to produce UV-absorbing compounds and repair DNA damage caused by UV (28). Organisms have adopted various mechanisms to overcome the harmful effects of UV radiation, including producing UV-absorbing pigments such as melanin in animals and humans (29,30) and other pigments (25,31–33). Human skin, which provides a barrier between the host and the physical, chemical and biologic environment, acts as a natural UV-protecting shield. Although human skin is adapted for UV stress, UV radiation penetrates the upper layer of the skin and prolonged exposure to UV induces a variety of responses in keratinocytes and melanocytes, the two major cell types in the skin. Melanocytes comprise approximately 1–2% of epidermal cells and are the second most numerous epidermal cell types (34). Melanocytes are derived from the neural crest and migrate into the epidermis late in the first trimester of development (35). Melanocytes reside in the basal layer in contact with neighboring keratinocytes. Melanocytes produce and distribute photoprotective melanin and are therefore key players in UV protection. Melanocytes synthesize melanin in response to UV radiation and protect the epidermal keratinocytes from UV-induced DNA damage. However, melanocytes or melanin cannot completely prevent UV radiation reaching the DNA in the superficial tissue. DNA damage caused by UV induces mutations. If left unrepaired, these mutations can lead to apoptosis or cancer. Therefore, DNA repair mechanism plays an important role in correcting UV-induced DNA damage. Nucleotide excision repair (NER) is the primary mechanism to remove the UV-induced dipyrimidine DNA photoproducts in human cells (36). Defects in NER genes result in sun-sensitive, cancer-prone genetic disorders such as xeroderma pigmentosum (XP), Cockayne’s syndrome and trichothiodystrophy in humans (37–39). Can melanocytes, which by virtue of their production of melanin provide protection to the surrounding epidermal keratinocytes, become themselves targets for UV-induced carcinogenesis?
UV and Melanoma: Is There an Etiologic Connection?
Despite the clear evidence for UV-induced DNA damage and nonmelanoma skin carcinogenesis and the long-held belief that solar radiation also plays a role in cutaneous melanoma in humans, the exact relationship between sun exposure and melanoma formation is not fully understood. It has been suggested that the risk of nonmelanoma skin cancer is related to cumulative lifetime UV exposure whereas the risk of melanoma skin cancer is related to annual UV exposure (40). Retrospective epidemiologic data seem to support a link between melanoma risk and UV exposure early in life, particularly during childhood (41,42). A systematic review of epidemiologic studies reveals that exposure to high levels of sunlight in not only childhood but also adulthood may contribute to this risk (43). This association has also been examined experimentally in an in vivo experiment using human newborn foreskin grafts in Rag1 knockout mice (44) and full-thickness skin grafts from adults in SCID and Rag1 knockout mice (45). In support of the risk associated with UV exposure in early life, in a transgenic hepatocyte growth factor/scatter factor (HGF/SF) transgenic mouse genetic model for melanoma, UV exposure of neonatal transgenic mice has been shown to shorten the latency of melanoma development, whereas UV irradiation of adult mice increased the multiplicity of melanoma tumors and also resulted in nonmelanoma skin cancer (46). In conflict with the retrospective UV epidemiologic and experimental data, Pfahlberg et al. did not find supporting evidence for the presence of the “critical period” during childhood in which there is a higher risk of melanoma because of sunburns. This was confirmed in a case–control study using 603 melanoma cases and 627 population controls from seven European countries, where the same magnitude of melanoma risk elevation was found with an increasing number of sunburns during both childhood and adulthood (47,48). A more recent study with 386 melanoma patients between ages 35 and 74 and 727 controls of the same age group and geographic area did not support childhood exposure of UV toward melanoma initiation (49). However, lifetime UV exposure was associated with melanoma risk in both men and women; in women significant trends of increased risk with UV exposure during ages 1–10, 11–20 and 31–40 were noted, but no significant associations at any time period was noted. Interestingly, Kennedy et al., in their studies using a cohort of 966 individuals (all from The Netherlands and mostly skin Types I and II) who participated in a case–control study investigating environmental and genetic risk factors for skin cancer, concluded that lifetime sun exposure appeared to be associated with a lower risk of malignant melanoma (50). Thus, the role of UV exposure in determining the risk of melanoma initiation remains controversial and a much debated topic.
Increased risk associated with melanoma might be linked with amount of UV radiation at a particular location (49,51,52) or the amount of time exposed to UV radiation (53,54) or latitude of residence (55) or level of pigmentation (56,57). As expected, cutaneous melanomas are more common in whites than in blacks or in other groups (58). Pennello et al. suggested that UV exposure increases melanoma risk even in blacks (59). However, Eide and Weinstock showed no evidence to support the association of UV exposure and melanoma incidence in black or Hispanic whites and they found melanoma incidence is associated with increased UV index and lower latitude only in non-Hispanic whites (60). Similarly, although time spent in outdoor activities had no significant effect on melanoma risk in any age strata, when adjusted for UV exposure based on residential history, time spent outdoors during young age significantly increased the risk for melanoma (61). However, melanoma in Asian and blacks occurs at sites that are not exposed to sun such as the sole of the foot (62). The genetic alterations identified in melanomas at different sites and with different levels of sun exposure indicate that there are distinct genetic pathways in the development of melanoma (63). As a defect in repair UV-induced DNA damage contributes to cancer-prone syndromes, it is reasonable to hypothesize that defects in NER machinery could contribute to the risk of melanoma development. However, there is no definitive association between defects in NER and risk of cutaneous melanoma (64–66).
The specific contribution of UVA and UVB exposure toward the risk of melanoma is also controversial. Chronic UVB irradiation with or without an initiating carcinogen can induce human melanocytic lesions, including melanoma (44). Melanocytes appear to be more resistant to physiologic doses of UVB with a potential to initially survive DNA damage that may be relevant to the subsequent malignant transformation (67). It has been shown that exposure to UVB in combination with increased expression of a combination of growth factors (such as basic fibroblast growth factor, stem cell factor and endothelin-3) may contribute to malignant transformation of melanocytes (68). However, a hospital-based case–control study showed that UVB-induced mutagen sensitivity may not play a role in susceptibility in cutaneous malignant melanoma (69). In the HGF/SF transgenic mouse model of melanoma, UVB, but not UVA, has been shown to induce melanomas (70).
UVA has a longer wavelength than UVB; it can penetrate deeper into the skin than UVB. UVA is thought to contribute 10–20% of carcinogenic effects of UV radiation (71). Up to 95% of the total UV exposure received UVA by humans comes from UVA, and UVA is found to contribute approximately 14% of the erythemal UV (72). UVA can induce melanomas in Xiphophorus hybrid fish (73) and if the action spectrum (range of wavelength that is most effective) for human melanoma were similar to that observed in fish, then approximately 90% of the melanoma-inducing effects in solar radiation could be from UVA and visible light. Epidemiologic data and a few animal experiments indicate that UVA might play a significant role in the induction of melanomas in humans (74). UVA induces cell cycle delay (G1 arrest) in human melanocytes (75). In vitro studies on the effects of UVA on melanocytes cultured from fair-skinned Caucasian individuals showed that simulated solar UVA radiation at a relatively high dose induces G2-M arrest and genes involved in DNA repair and oxidative stress response. These data suggest that UVA could contribute to sunlight-induced genotoxicity and melanocyte transformation (76).
Melanin as a Photoprotector
The controversy over the role of UV in melanoma tumorigenesis is partly due to the dual properties of melanin—both as a photoprotector and as a photosensitizer. Here, we will critically evaluate the available data that argue for and against these two seemingly opposite roles of melanin in the etiology of cutaneous melanoma.
Melanocytes produce two major types of melanin—eumelanin, the dark black and the major photoprotective pigment, and pheomelanin, the lighter pigment present in fair skin, red hair and freckled individuals. Melanosomes containing the accumulated melanin are transferred to the surrounding keratinocytes via dendritic extensions. Melanin in the keratinocytes is generally distributed as a supranuclear cap, localized between the keratinocyte nucleus and the external environment, protecting DNA in the nucleus from UV-induced DNA damage.
Melanin in the skin is an important chromophore that acts as a filter by absorbing UV photons and visible light. Melanin also acts as a free radical scavenger which scavenges molecular oxygen and hydroxyl radicals and preserves the DNA from pyrimidine base formation (77). Melanin is known to transform the absorbed light energy into heat and disperse it between the hairs and capillary vessels. Melanin has a sun protection factor (SPF) value between 2 and 3, based on its protective value from erythema and DNA damage (78). Whereas skin Type I has an SPF value of 1.0, SPF values of Type II, III, IV–V are 1.67, 2.5 and 4.0, respectively (79). Thus, differences in the content and nature of melanin pigment in the skin among different ethnic groups contribute to skin color variation, phototypes and sensitivity to UV radiation.
Role of Melanin in Melanoma Tumorigenesis
As melanin has good photoprotective properties, it is reasonable to assume that melanoma incidence will be higher in individuals and populations with fair skin (less melanin) and higher photosensitivity. This appears to be generally the case, as seen by (a) low incidence and mortality of melanoma in African Americans (80), and (b) occurrence of cutaneous melanomas in patients with albinism and other hypopigmentary disorders (81). However, several lines of evidence also indicate that there is no strict correlation between melanin pigment in the skin and melanoma incidence. For example, overall there is no evidence of higher incidence of melanoma among albino individuals lacking melanin pigmentation (82). Albinos in Africa present a unique opportunity to study the effect of absence of pigment protection without other confounding hereditary influences (83). In 18 African albinos, Yakubu and Mabogunje found that squamous cell carcinoma was the most common tumor type, but did not find any cutaneous melanomas (84). Similar observations were made in 164 albinos among Tanzanian black populations (85) and a population of 111 albinos in Johannesburg (86).
Additionally, animal experiments also appear to question the relationship between pigmentation and melanoma tumorigenesis. Exposure of three inbred mouse strains with different coat colors and albino mice to UV showed that albino mice were susceptible and developed skin tumors earlier than agouti or black mice. Interestingly, in these mice the most common histologic types of tumors observed were fibrosarcomas followed by squamous cell carcinomas, but no melanomas were found (87). These observations suggest that complete absence or less melanin content does not increase or has any relation to malignant transformation of melanocytes. However, these observations on the incidence of melanoma in mice should be interpreted with caution because melanocytes in mice are present primarily in the hair follicles in contrast to epidermal melanocytes in human skin. In light of the proposed roles for epidermal melanin as a photoprotective agent (88–90), natural sunscreen against UV and also a neutralizer of reactive oxygen species (ROS) generated by UV (56,91) these observations between the relationship between epidermal melanin and melanoma warrant reexamination of interactions of melanin and/or reaction products of melanogenic enzymes such as tyrosinase-related proteins 1 and 2 (TYRP1 and TRP2) with other genetic factors in rendering melanocytes susceptible to malignant transformation.
Melanin and Photoprotective Responses of Melanocytes
Melanocytes exposed to UV increase their melanin content (92) by increased expression of the master transcription factor MITF and its downstream melanogenic proteins, including Pmel17, MART-1, TYR, TRP1 and DCT (57). In human skin, UV exposure upregulates the levels of both eumelanin and pheomelanin in tandem rather than independently (30,57,93,94). In a study of the effects of UV exposure in situ on normal human skin of different phototypes and UV sensitivities, levels of melanin were found to correlate inversely with amounts of DNA damage induced by UV (30). These observations suggest that intracellular melanin plays an important role in preventing UV-induced cell killing by reducing the DNA damage as measured by the amount of CPDs and 6-4PPs. In sum, melanin can reduce DNA photo-damage induced by UV radiation and act as photoprotective agents in the skin.
In a more recent study, Yamaguchi et al. examined DNA damage in the upper and lower epidermal layers in various types of skin before and after exposure to UV and observed that UV-induced DNA damage in the lower epidermis (including keratinocyte stem cells and melanocytes) is more effectively prevented in darker skin, suggesting that the pigmented epidermis is an efficient UV filter and noted that UV-induced apoptosis is significantly greater in darker skin, which suggests that UV-damaged cells may be removed more efficiently in pigmented epidermis (95). The authors suggested that the low doses of UV absorbed by melanin may cause selective damage to pigmented structures enhancing their removal. Thus, decreased DNA damage and efficient removal of UV-damaged cells indicate that melanin plays an important protective role in photocarcinogenesis. Miyamura et al. measured melanocyte density in skin biopsies after a single and repeated UV exposure using antibodies to melanocytic marker proteins (57). Melanocyte densities began to increase at 3 weeks, and especially at 5 weeks the melanocyte density increased significantly to levels about three-fold higher than in unirradiated skin. Interestingly, melanin contents did not increase significantly within 1 or after 3 weeks of repetitive UV exposure. However, there was a significant increase in melanin after 5 weeks of exposure. Increase in melanocyte density, but not melanin content, to repetitive UV exposures suggests that melanocyte density may also be involved in protecting the skin from UV-induced damage. Thus, even constitutive melanin pigmentation may provide a protective effect from carcinogenesis in the face of UV-induced proliferation of melanocytes. It is therefore reasonable to speculate that, in addition to the endogenous and inducible melanin, other as yet unknown cellular mechanisms protect melanocytes from UV damage.
Barker et al. cultured human melanocytes derived from different pigmentary phenotypes (skin Types I–VI) and irradiated them once with different doses of UVB. After UVB irradiation, heavily pigmented melanocytes had the same percent survival but a greater capacity to resume proliferation than their lightly pigmented counterpart. In the lightly pigmented melanocytes, the increase in melanin content was significantly less than in darkly pigmented cells with a correspondingly higher number of CPDs. These differences might be due to increased susceptibility of individuals with lightly pigmented skin compared with individuals with dark skin to the photodamaging and photocarcinogenic effects of UV (90). In another similar study, cultured melanocytes from light-skinned individuals which synthesize less melanin accumulated more CPDs and 6-4PPs upon UVB exposure than did melanocytes from black skin. Moreover, a measurement of DNA damage and the quantity and quality of melanin revealed that eumelanin concentration correlated better with DNA protection than pheomelanin (89). In another in vitro study, after UVB irradiation, a stronger induction of endonuclease-sensitive sites and lower rate of survival was found in melanocytes with lower total melanin (but higher pheomelanin) content (96).
In sum, the evidence suggests that melanocytes, melanin and pigmentation are involved in protection of human skin from UV radiation, but other factors may also be involved in initiation of cutaneous melanoma. Vitiligo is an interesting experiment of nature to address the relevance of melanin in melanoma tumorigenesis. Although there are incidences of co-occurrence and association of vitiligo and malignant melanoma (97–102), the frequency is very low. This suggests that lack of UV-absorbing melanin alone does not increase the risk of melanoma.
Melanin as Photosensitizer
Melanin and the process of melanin synthesis are known to act as photosensitizers because of generation of ROS, which causes DNA damage (103–106). If this were the case, UV exposure could be considered doubly dangerous by causing direct DNA damage as well as photosensitization via increased melanin synthesis. Irradiating melanin in vivo in congenic mice of black, yellow and albino coat colors induced DNA lesions (CPDs) and apoptosis equally in all the three strains (105). However, UV-irradiated pheomelanin (yellow and red melanin) was thought to photosensitize adjacent cells more than irradiated eumelanin (brown and black). The contribution of photosensitization of melanocytes, by melanin, to malignant transformation is not well understood.
UV radiation, especially UVA, is also known to cause photosensitization in vitro in cultured cells by generating ROS which can indirectly induce oxidative base lesions and single strand DNA breaks (SSB) (107,108). Examination of cultured human melanocytes derived from skin Type I and VI showed that skin Type I melanocytes have low sensitivity to UVA. However, increasing melanin content by culturing the cells in medium containing high tyrosine resulted in a three-fold increase in their sensitivity to UVA as measured by SSB (109). These data raise the possibility that UVA-irradiated cultured human melanocytes are photosensitized by their own chromophores, most likely pheomelanin and/or melanin intermediates. Similarly, in normal human Caucasian melanocytes in culture exposed to simulated solar UV (SSUV, 300–400 nm), even at relatively high doses (12 kJ m−2 UVB and 110 kJ m−2 UVA), there was limited cell death and lower p53 accumulation than in unpigmented fibroblasts. These data also point to a role for melanin in melanocyte photosensitization, sunlight-induced genotoxicity and presumably in melanocyte transformation (76). Thus, melanin can be viewed both as a photoprotector and as a photosensitizer in human skin.
Tanning is a process of induction of skin pigmentation following UV exposure. Exposure to sunlight causes tanning. Sunbeds and sunlamps are the two main artificial sources of indoor tanning. Artificial tanning is popular and is a fast growing industry in the United States (110). Case–control studies on the association of tanning lamp or other UV lamp use with melanoma revealed no significant association (111). There seems to be a significant relationship between greater frequency and duration of artificial tanning with increased melanoma risk. Consistent, but not strong, evidence was also found that points to greater risk of melanoma with younger age of exposure (112). The published body of research has several critical limitations. More recent evaluation of these data based on 19 informative studies showed that even one time ever use of sunbeds was positively associated with melanoma with consistent evidence for a dose–response relationship (113). However, detailed studies on the nature and amount of pigment synthesized during tanning response are required to causatively associate melanin with photosensitization of melanocytes during tanning.
Differences in the overall structure of the skin between darkly and lightly pigmented individuals could also contribute to sensitivity to UV radiation and risk for melanoma development. For example, it appears that five times more UV radiation reaches the upper dermis of Caucasian skin than that of black skin, and the main site of UV filtration in Caucasians is the stratum corneum, whereas in blacks it is the malpighian layers (114). It has been proposed that the superior photoprotection of black epidermis is due not only to increased melanin content but also to other factors related to packaging and distribution of melanosomes. Melanogenesis appears to account for the increased photoprotection after 2 weeks of exposure to UV radiation, but after 4 weeks other protective mechanisms come into play in human skin (115). Exposure of vitiligo and adjacent normally pigmented skin to UV showed, by minimal erythemal dose test, that stratum corneum was the main photoprotective factor not only in vitiligo but also in normally pigmented skin, and the effect of pigmentation in normal skin was less prominent (116). In another study, stratum corneum thickness was found to be of minor importance for the constitutive UV sensitivity, and epidermal thickness was independent of skin type and was not correlated with constitutive skin pigmentation (117). In case–control studies, no differences were found in the constitutive skin pigmentation in the non–sun-exposed buttock skin in control and melanoma patients, but facultative skin pigmentation in UV-exposed sites was higher in men (118). These observations clearly show that induction of melanoma is determined by multiple factors that include melanin pigmentation, UV sensitivity and other genetic factors (119).
UV-Induced Oncogenic Alterations in Melanocytes
As melanoma risk is thought to be associated with increased UV exposure including intermittent and childhood exposure (120) and increased number in melanocytic nevi (121), the key to understanding the process by which melanocytes are transformed into malignant melanoma lies in deciphering the interplay between genetic factors and the UV spectrum of sunlight. To date, the nature of this relationship remains unclear (14). Oncogenic mutation in either NRAS or BRAF is commonly associated with melanoma (122–124) and this mutation is also reported in 20–80% of melanocytic nevi (125–128). BRAF activation is the most frequent molecular event in melanoma development occurring in nearly 60% of all sporadic melanomas. Activating mutation in oncogenic BRAFV600E induces tumorigenicity of cultured melanocytes in nude mice (129). Transgenic zebrafish expressing mutant BRAF form fish (f)-nevi, and in p53-deficient fish, mutant BRAF induces melanocytic lesions that develop into invasive melanomas (130). However, the most common BRAF T1796A mutation is not a characteristic UV signature mutation (131). Thomas et al. suggested the possibility that BRAF mutation in melanocytic lesions in sun-exposed areas of the skin could arise by error-prone replication of damaged DNA (132). Based on the observation that the activating mutations in BRAF are common in cutaneous melanomas while they are completely absent in mucosal melanomas arising in sun-protected sites, it was suggested that despite the absence of the characteristic C>T or CC>TT mutation signature associated with UV exposure, mechanisms other than pyrimidine dimer formation are important in UV-induced mutagenesis (131). Additional studies are warranted to understand these potentially novel mechanisms. Interestingly, there seems to be an association between BRAF mutations and childhood sun exposure (133) similar to a two-fold higher mutation frequency of RAS genes in skin tumors of XP patients (134). Earlier and repeated exposure to UV might facilitate UV-induced mutations to initiate melanocyte transformation. NRAS-activating mutation at a dipyrimidine site was found in a melanoma tumor cell line derived from an XP patient, raising the possibility that it is initiated by UV-induced dipyrimidine mutation (135). NRAS mutation near dipyrimidine sites has also been found in melanomas arising at sites exposed to sun in non-XP patients, suggesting a role for UV in oncogenic activation of N-RAS (136,137). Thus, while a role for UV in induction of RAS-BRAF mutation in the melanocytes is plausible, increased sensitivity of RAS–BRAF-mutant melanocytes to UV exposure might also contribute to melanoma development.
INK4A-ARF (CDKN2A) gene mutations in sporadic melanoma show UV signature mutation (138), suggesting that the CDKN2A locus is a direct target for UV irradiation (139). Investigation into the role of INK4A in induction of malignant transformation of melanocytes suggested that loss of either p16 or p19ARF reduces the ability of the cells to process UV-induced DNA damage (140). Additionally, interaction of p16ink4a with JNK has been found to inhibit c-Jun phosphorylation induced by UV exposure, which, in turn, has been found to interfere with malignant transformation promoted by the H-Ras-JNK-c-Jun signaling pathway (141). Thus, UV exposure and sunburn is a highly significant risk factor for developing melanoma in families harboring germline mutations in CDKN2A. This is supported by studies using transgenic HGF/SF mouse models, where neonatal UV irradiation dramatically reduced (by nearly five-fold) the latency of tumor development in the transgenic mice on ink4a/arf-null background (142). The evidence suggests that ink4a/arf tumor suppressor plays a critical role in UV-induced melanomagenesis.
Conclusion and Perspectives
Despite years of focused efforts, the question “what is the role of UV in melanoma?” remains unanswered or answered only partially. While in vitro studies on human melanocytes and skin, and mouse models continue to refine our knowledge on the specific effects of UV on normal and genetically susceptible melanocytes, epidemiologic data still remain controversial. Among the factors that contribute to this controversial role of UV in melanoma are: (a) the dual, photoprotective and photosensitizing, properties of melanin, (b) gaps in our understanding of the complete spectrum of effects of UV on melanocytes and (c) inherent weaknesses in the design of epidemiologic studies. Ongoing efforts to catalog the genome-wide alterations in melanomas arising in sun-exposed and non–sun-exposed skin are expected to soon provide valuable new clues to answer this important question. The answers obtained will have a significant impact on public health by way of opening novel avenues for prevention and treatment of malignant melanoma.