SOX9 and the tanning response: something new under the sun

The ability of human skin to respond to the ultraviolet (UV) component of solar radiation by increasing its melanin content, i.e. the tanning response, provides one of the most striking and visible examples of an adaptative response to the environment. UV radiation has a profound impact on human skin as it affects the survival, proliferation and differentiation of cells in the epidermis and the dermis. Among the beneficial actions of UV, the best known is the stimulation of vitamin D metabolism, with insufficient exposure to the sun causing rickets. Unfortunately, excess UV radiation is responsible for skin photoaging and photocarcinogenesis. Thus, UV radiation can be considered as a two-edged sword, but it appears that the negative effects of UV outweigh the positive responses, hence the need for effective protection mechanisms. In human skin, photoprotection is mainly due to the presence of melanin. Melanin granules concentrate in the perinuclear area of melanocytes and keratinocytes, where they form supranuclear caps that protect DNA from the mutagenic action of UV by absorbing or scattering part of the impinging radiation. This photoprotective action of the melanin pigments, mainly the eumelanins, is well documented by a series of studies showing an inverse correlation between pigmentation and sensitivity to UV-induced skin damage or skin cancer. Therefore, increased pigmentation following UV exposure prepares the epidermis to better withstand further insults. The tanning response is local rather than systemic. This is strikingly evidenced, in summertime, by the sharp boundary between the tanned sun-exposed areas of the skin and the zones protected by clothing. Tanning is also a highly complex process, as it involves not only the stimulation of tyrosinase activity and melanin synthesis in melanocytes but also an increased melanocyte dendricity and a stimulation of melanosome transfer to the surrounding keratinocytes. Importantly, it appears that, during tanning, the number of melanogenically active melanocytes also increases. Most if not all these responses result from changes in gene expression in melanocytes or keratinocytes. Given the variety of effects of UV radiation and their local nature, it can be anticipated that exposure to UV radiation will trigger several signaling pathways and should involve the coordinate regulation of various transcription factors within UV-exposed skin cells. The involvement of at least two UV-responsive transcription factors has been convincingly shown. In melanocytes, upstream stimulating factor-1 (USF-1) enhances Tyrosinase gene expression following phosphorylation and activation by the stress-responsive p38 kinase (Corre and Galibert, 2005). Interestingly, the p38 kinase/USF-1 pathway also mediates the UV-induced activation of the proopiomelanocortin (POMC) and melanocortin 1 receptor (MC1R) promoters in melanocytes, melanoma cells and human skin in vivo, thus giving rise to an autocrine/paracrine loop that may further enhance melanogenesis. More recently, a role for p53 in the tanning response has also been demonstrated (Cui et al., 2007), although in this case the main effect appears to be at the level of keratinocytes. In these cells, UV radiation stimulates p53 gene expression. In turn, p53 stimulates the POMC promoter, and the absence of p53 in knockout mice impairs the tanning response. In this case, increased POMC gene expression would lead to higher levels of melanocortins in the skin, with a paracrine stimulation of melanogenesis. Very recently, Passeron et al., (2007) added another piece to the tanning puzzle by reporting the involvement of SOX9 in UV-induced melanocyte differentiation and pigmentation. SOX9 is a transcription factor of the SRY type High Mobility Group (HMG) box family which plays multiple roles during development. Although its best known actions are in sexual determination and chondrogenesis, several lines of evidence suggest that SOX9 may also be a relevant melanogenic transcriptional regulator. For instance, ectopic expression in the neural crest drives melanocyte differentiation and SOX9 expression increases in differentiated melanoblasts (Cook et al., 2005). Moreover, SOX9-directed antibodies have been detected in the sera of vitiligo patients, again suggesting the involvement of SOX9 in the regulation of melanocyte biology. Passeron et al. now show that SOX9 is expressed in adult skin melanocytes in vivo, both at the mRNA and protein levels. Importantly, SOX9 expression seems to correlate with basal pigmentation, as suggested by comparison of SOX9 protein levels in neonatal lightly versus darkly pigmented skin. Moreover, SOX9 expression increases rapidly following exposure to UVB radiation. These data suggest a role of SOX9 in the regulation of both constitutive and UV-induced skin pigmentation. The authors also analyzed downstream events linking SOX9 and pigmentation as well as upstream events responsible for regulation of SOX9 expression. Concerning downstream events, they show that SOX9 binds and activates the MITF promoter and increases Tyrosinase and DCT expression. The regulation of MITF expression by SOX9 has been confirmed in cultured normal human melanocytes by chromatin immunoprecipitation assays as well as by immunocytochemistry, after silencing SOX9 expression. Concerning upstream events, Passeron et al. report that Agouti signaling protein (ASP) downregulates SOX9 expression in normal human melanocytes. Moreover, they provide evidence of the involvement of the cAMP pathway in the regulation of SOX9 expression. The cAMP elevating agent forskolin increases SOX9 expression and binding to the MITF promoter, and the induction of SOX9 in UV-treated cells is blocked by the PKA inhibitor H89. Interestingly, these results are confirmed in cultured normal melanocytes and in a reconstructed skin model, by immunocytochemical techniques. Based on these and other data, the authors propose that UV radiation increases SOX9 expression and cAMP-responsive element binding protein (CREB) activity, at least partially through the cAMP-dependent activation of PKA. Then, SOX9 and activated CREB would co-ordinately regulate the MITF promoter. Increased MITF activity would ultimately lead to activation of the genes encoding for the melanogenic enzymes, and hence to an enhanced production of melanin pigments. Within this scenario, the inhibitory action of ASP on SOX9 expression in melanocytes might be related with a modulation of basal cAMP levels within those cells. Current evidence suggests that ASP acts primarily by binding to the MC1R. MC1R is constitutively active and signals through Gs protein-dependent activation of the cAMP pathway, and ASP can be considered an inverse agonist decreasing constitutive MC1R signaling rather than a typical competitive antagonist of α-melanocyte stimulating hormone and other activatory melanocortin peptides. Therefore, ASP is expected to decrease basal levels of cAMP, which may account for its effects on SOX9 expression. An intriguing question raised by the work of Passeron et al. concerns the nature of the pathway linking UV radiation and PKA activation. If PKA activation is indeed involved in UV-dependent activation of SOX9 expression, then this pathway must be very rapid as increased SOX9 mRNA levels can be detected as soon as 1 h after UV irradiation. This rapid kinetics appears to exclude an autocrine/paracrine mechanism dependent on increased expression of MC1R on one hand, and POMC and its derived melanocortin peptides on the other. Nevertheless, the effects of ASP do point to an involvement of MC1R in the regulation of SOX9, which may be consistent with the fact that variant MC1R alleles with impaired signaling are associated with increased sun sensitivity and a poor tanning response. As at least two other transcription factors, USF-1 and p53, have been shown to be involved in the tanning response, another important question concerns the contribution of each one of them. It would be interesting to assess the relative contribution of these pathways, maybe by selective silencing each one of them and assessing the resulting loss of tanning potential. It would also be very interesting to investigate the possible crosstalk of these pathways, and the effect of the genetic background provided by the MC1R haplotype on the induction of SOX9 by UV radiation. As it usually happens with relevant papers, the report by Passeron et al. is not only interesting in the light of the new information it provides but also because of the important questions that it raises.