Vitamin D Receptor, UVR, and Skin Cancer: A Potential Protective Mechanism

Abstract

More than 1 million skin cancers occur annually in the United States—of which 80% are basal-cell carcinoma (BCC), 16% are squamous-cell carcinoma (SCC), and 4% are melanomas—making skin cancer by far the most common cancer (Greenlee et al., 2001). UVR is the major etiologic agent. UV wavelengths shorter than 280 nm (UVC) are absorbed by the ozone layer and do not reach the earth. UV wavelengths longer than 320 nm (UVA) have limited ability to induce the characteristic mutations in DNA seen in epidermal cancers. Thus, UVB, with a spectrum between 280 and 320 nm, is the major cause of these cancers (Freeman et al., 1989), but this is the same spectrum required for vitamin D production in the skin. Is there a link?

The principal genotoxic lesions induced by UVR are cyclobutane pyrimidine dimers and pyrimidine (6–4) pyrimidone photoproducts, which, if not repaired, result in C-to-T or CC-to-TT mutations, the UVB “signature” lesion (Hussein, 2005). Mutations in p53 are common (50–90%) in both BCC and SCC (Brash et al., 1991; Daya-Grosjean and Sarasin, 2005; Ziegler et al., 1993, 1994), as well as in actinic keratoses, the precursor lesions to SCC (Bito et al., 1995). Precursor lesions for BCC have not been identified, but BCCs are thought to arise from interfollicular basal cells, hair follicles, and sebaceous glands. Mutations in ras are much more common in SCC than in BCC (Reifenberger et al., 2005), whereas mutations in the hedgehog (Hh) signaling pathway—in particular, in patched 1 (Ptch1)—characterize BCC (Aszterbaum et al., 1998, 1999; Hahn et al., 1996; Johnson et al., 1996). However, loss of heterozygosity in the region including the Ptch1 gene has been reported in a high percentage of both BCCs and SCCs (Danaee et al., 2006), and polymorphisms of the Ptch1 gene have altered the resistance to SCC formation in mice expressing an activated ras oncogene (Wakabayashi et al., 2007). Thus, alterations in the Hh signaling pathway contribute to the formation of both SCC and BCC.

1,25-Dihydroxyvitamin D₃ (1,25 (OH)₂D₃) has been evaluated for its potential anticancer activity for approximately 25 years (Eisman et al., 1979). The list of malignant cells that express the receptor for 1,25(OH)₂D₃—vitamin D receptor (VDR)—is now quite extensive and for the purposes of this Commentary includes BCCs and SCCs (Kamradt et al., 2003; Ratnam et al., 1996), as well as melanomas (Colston et al., 1981). The accepted basis for the promise of 1,25(OH)₂D₃ in the prevention and treatment of malignancy includes its antiproliferative, prodifferentiating effects on most cell types. Epidemiologic evidence supporting the importance of adequate vitamin D nutrition (including sunlight exposure) for the prevention of a number of cancers, including those of the colon, breast, and prostate (Bostick et al., 1993; Garland et al., 1985, 1990; Hanchette and Schwartz, 1992; Kearney et al., 1996), is strong. However, several large epidemiologic surveys have not demonstrated such a correlation with skin cancer (Hunter et al., 1992; van Dam et al., 2000; Weinstock et al., 1992). One potential complication is that UVB radiation has the dual effect of promoting vitamin D₃ synthesis in the skin (which can be further converted to 1,25(OH)₂D₃) and increasing DNA damage, leading to skin cancer. Thus, although UVR may be an efficient means of providing the nutritional requirement for vitamin D, the advantage to the skin may be countered by the increased risk of mutagenesis. However, recent studies indicate that the skin has evolved a vitamin D signaling mechanism whereby the harmful effects of UVR may be mitigated.

The study by Ellison et al. (2008) in this issue demonstrates the protective role of VDR for UVR-induced skin tumors. Earlier studies indicated that mice null for VDR were highly susceptible to epidermal tumor formation induced by either the oral administration of the carcinogen 7,12-dimethylbenzanthracene (DMBA) (Zinser et al., 2002) or its topical application (Indra et al., 2007). In the former study, the tumors were mostly papillomas, but several BCCs were observed. No SCCs were reported. In the latter study, only benign tumors were seen in the VDR null mice, unlike RXRα null mice, which developed both BCCs and SCCs (Indra et al., 2007). Ellison et al. (2008) confirmed the report by Zinser et al. (2002) demonstrating that oral administration of DMBA induced skin tumors in all of the VDR null mice and none of the wild-type mice. Furthermore, most of the tumors that formed appeared to originate from the folliculosebaceous unit. Importantly, Ellison et al. (2008) did not report increased susceptibility to DMBA-induced tumor formation in mice lacking the enzyme for 1,25(OH)₂D₃ production (CYP27B1). The authors thus concluded that, at least for DMBA induction of tumors, VDR protection does not require its principal ligand, 1,25(OH)₂D₃.

At this point the authors turned their attention to UVR-induced skin tumors comparing VDR null mice with wild-type controls. The mice were of C57BL/6 background, a strain relatively resistant to UVR-induced skin tumors. Even at the very high doses of UVB used in these experiments, only 4 of 23 wild-type mice developed tumors by 45 weeks after the start of UVR, 1 of which was classified as an SCC. In contrast, all 22 of the VDR null mice developed tumors, and 10 of the 27 tumors analyzed were classified as SCC, although markers to document this classification were not provided. Taken at face value, therefore, it would appear that UVR-induced tumors in VDR null mice differ in type compared with DMBA-induced tumors, with SCC predominating following UVR and BCC and folliculosebaceous tumors predominating after DMBA.

Potential mechanisms were then explored. The skin of VDR null mice appeared to have greater cyclobutane pyrimidine dimer formation following UVR, suggesting a DNA repair deficiency. Previous studies with cultured keratinocytes have demonstrated that 1,25(OH)₂D₃ and analogs not thought to act via the nuclear configuration of the VDR were also protective against UVR-induced cyclobutane pyrimidine dimer formation (or accelerated their repair) (Dixon et al., 2005; Gupta et al., 2007). Ellison et al. (2008) did not test whether CYP27B1 null mice were also predisposed to UVR-induced tumor formation, so the results in their article are silent as to whether this is a 1,25(OH)₂D₃-independent action of VDR. Apoptosis following a single exposure to UVR was reduced in VDR null mice, although no differences between wild-type and null mice were shown for p53 induction by UVR. These observations are discordant with earlier observations (De Haes et al., 2003) that 1,25(OH)₂D₃ increases p53 levels and protects keratinocytes from UVR-induced apoptosis (Gupta et al., 2007), suggesting that VDR and 1,25(OH)₂D₃ might have opposite effects in these responses to UVR. This possibility could be tested by comparing the responses in CYP27B1 null mice to those in VDR null mice. The proliferative response to UVR was also slightly reduced in the VDR null epidermis, which, despite the decrease in apoptosis, seems to have resulted in a thinner epidermis in the VDR null compared with wild-type mice, although UVR induced epidermal thickening compared with the nonirradiated controls of either genotype. This result is consistent with earlier studies (De Haes et al., 2003; Gupta et al., 2007) demonstrating that 1,25(OH)₂D₃ enhances cell survival in keratinocytes after UVR, but it again raises the question of whether these are 1,25(OH)₂D₃-dependent or -independent actions of VDR. The finding of the thinner epidermis in the VDR null mouse relative to controls is surprising, however, considering the higher proliferation rates in the epidermis and underlying folliculosebaceous unit of the VDR null mouse compared with controls under basal conditions (Bikle et al., 2006; Xie et al., 2002).

What, then, might be the underlying mechanisms? The DMBA studies producing tumors primarily of folliculosebaceous origin suggest that Hh signaling might be altered in VDR null skin. Appreciation of the pivotal role of the Hh signaling pathway in epidermal carcinogenesis began with the identification of the Ptch1 gene as the site of mutations underlying the rare autosomal dominant heritable basal-cell nevus syndrome (Gorlin’s syndrome), one cardinal feature of which is a high susceptibility to the development of BCCs (Aszterbaum et al., 1998; Hahn et al., 1996). The BCCs in these subjects frequently lose function of the inherited wild-type Ptch1 allele, leaving the tumor cells functionally null of Ptch1. It has subsequently become clear that essentially all BCCs, whether arising in patients with basal cell nevus syndrome or sporadically, have mutations in Ptch1 or other alterations in Hh signaling (Aszterbaum et al., 1999). This appreciation has resulted in the development of the Ptch1^+/− (Gorlin) mouse as the first practical model of murine BCC (Regl et al., 2004a). Treatment of these mice (unlike treatment of Ptch1 wild-type mice) with UVR or ionizing radiation produces BCC as well as SCC (Regl et al., 2004a). Ptch1 is the membrane receptor for sonic hedgehog (Shh). In the absence of Shh, Ptch1 inhibits the function of another membrane protein, smoothened (Smoh). Shh reverses this inhibition, freeing Smoh to enable the activation of a family of transcription factors, Gli1 and Gli2. Gli1 and -2 overexpression in keratinocytes can increase the expression of each other as well as Ptch1, the anti-apoptotic factor bcl2, cyclins D1 and D2, E2F1, and cdc45 (all of which promote proliferation) while suppressing genes associated with keratinocyte differentiation such as K1, K10, involucrin, loricrin and VDR (Grachtchouk et al., 2000; Nilsson et al., 2000; Regl et al., 2002, 2004a,b). Mice overexpressing Gli1, Gli2, or Shh in their basal keratinocytes (Grachtchouk et al., 2000; Nilsson et al., 2000; Oro et al., 1997) or grafted with human keratinocytes overexpressing Shh (Fan et al., 1997) develop BCC-like lesions. Furthermore, BCCs demonstrate overexpression of Ptch1, Smoh, Gli1, and Gli2 (Bonifas et al., 2001; Tojo et al., 1999). A number of the Hh pathway components, including both Gli1 and -2, have consensus vitamin D response elements in their promoters (Palmer et al., 2008; Wang et al., 2005). We (A. Teichert, J. Welsh, and D. Bikle, unpublished data) reported that the epidermis and epidermal portion (utricles) of the hair follicles of VDR null animals overexpress elements of the Hh signaling pathway, suggesting that one of the causes of the increased susceptibility of the epidermis to malignant transformation is a loss of VDR regulation of Hh signaling.

The wnt signaling pathway may also be involved. In a recent paper, Palmer et al. (2008) reported the interaction of VDR and wnt signaling in the regulation of hair follicle differentiation and tumor formation. The authors identified various combinations of putative vitamin D response elements and lymphoid-enhancing factor/T-cell factor response elements in a number of genes involved with epidermal and hair follicle differentiation, including components of the Hh signaling pathway such as Shh, Ptch2, and Gli1 and -2, as well as in LEF itself, the transcriptional partner for β-catenin critical for wnt signaling. At least in colon cancer cells, 1,25(OH)₂D₃ inhibits the tumor-promoting properties of β-catenin by inducing E-cadherin, to which β-catenin binds (Palmer et al., 2001), and promoting the binding of VDR to β-catenin (Shah et al., 2006), thus limiting its nuclear localization and transcriptional activity. Palmer et al. (2008) demonstrated in skin that lack of VDR predisposes the mouse to develop BCC when active β-catenin was overexpressed, although the link to increased Hh signaling was not explored.

In addition to suppression of Hh and wnt signaling, a third mechanism may also be at work, the DNA damage response (DDR). Many endogenous and exogenous factors, including oncogenes, tumor suppressors, radiation, chemical carcinogens, and viral infections, impact a common, but complex, set of pathways that regulate cellular survival, apoptosis, and genomic stability: DDR and the cell cycle checkpoints (Bartkova et al., 2005, 2006). DDR involves a cascade of damage recognition, repair, and signal transduction that coordinates the response of the cell cycle to DNA damage. DDR activates checkpoints that delay the cell cycle, providing time for repair, and directs damaged cells into senescent or apoptotic pathways. DDR demonstrates wide variation in components, activation signals, and downstream consequences according to the cell types, species, and damaging agents. DDR is tightly controlled and highly accurate in normal primary cells such that the spontaneous mutation rate is very low and changes in copy number are negligible (Bielas et al., 2006; Tlsty, 1990; Tlsty et al., 1989). During malignant transformation, a DNA repair “barrier” is abrogated (Bartkova et al., 2005, 2006), DDR becomes less controlled, and mutation rates and copy number changes increase by orders of magnitude (Bielas et al., 2006; Tlsty et al., 1989).

DDR from UV-induced photoproducts has distinct G1 and S phase responses. Nucleotide excision repair (NER) in G1, for example, excises UV photoproducts before onset of DNA replication and thereby eliminates mutagenic lesions completely (Maher et al., 1979). The main substrates for NER are the UV photoproducts between adjacent pyrimidines: cyclobutane pyrimidine dimers and pyrimidine (6–4) pyrimidone photoproducts (Cleaver et al., 2005), lesions produced by UVB. Mutations in NER genes are associated with several human cancer and neurodegenerative diseases, especially xeroderma pigmentosum (XP) and Cockayne’s syndrome (Hoeijmakers, 2001; Wood et al., 2001). Damage in transcribed genes is recognized through the arrest of RNA Pol II, which is relieved through the action of two proteins, CSA and CSB. Damage in the non-transcribed or global regions of the genome is recognized by the binding of two heterodimeric XP proteins, XPE (DDB1/DDB2) and XPC/HR23B, which may act in concert to bind to damaged DNA (Hoeijmakers, 2001). These converge on a common pathway through which the damage is verified (XPA/RPA), the DNA is unwound by the XPB and XPD helicase components of the transcription factor TFIIH, cleaved by structure-specific nucleases (XPG, XPF/ERCC1), and the damage is excised and replaced. The global branch of the NER (GGR: XP-A through G) pathway is strongly influenced by transactivation of the XPC and XPE genes by p53 through DNA damage-dependent increases in p53 expression. GGR defects increase carcinogenesis (Berg et al., 2000).

As noted above, recent studies have shown that vitamin D may have a strong regulatory effect on the capacity of cells and skin to carry out DDR. Treatment of the skin with 1,25(OH)₂D₃ rapidly increased photo product excision in human and mouse cells and tissues (Dixon et al., 2005; Gupta et al., 2007), but the mechanism of interaction with NER remains unclear. However, when Moll et al. (2007) performed microarray studies of keratinocytes treated with 1,25(OH)₂D₃ for 24 hours, they observed an upregulation of two genes important for DDR: XPC and DDB2/XPE. Accordingly, 1,25(OH)₂D₃ and VDR may be protective against UVR-induced epidermal tumor formation by stimulating DDR as well as by suppressing Hh and wnt signaling.

Ellison et al. (2008) make the important point that VDR serves as a tumor suppressor not only for chemically induced epidermal carcinogenesis but also for the more physiologically relevant UVR-induced epidermal carcinogenesis. Although these authors have convincingly demonstrated that the former is a 1,25(OH)₂D₃-independent action of VDR, this conclusion cannot yet be extended to UVR-induced carcinogenesis. Earlier studies indicated several mechanisms by which VDR, with or without its ligand 1,25(OH)₂D₃, might be protective. These mechanisms include altered Hh and wnt signaling and increased DDR. Future investigations are required to determine which, if any, of these mechanisms are involved and the degree to which VDR acts through these mechanisms in a 1,25(OH)₂D₃-dependent or -independent manner. However, this publication establishes an important fact: the skin has evolved a way of protecting itself from the deleterious actions of UVR using a key component of the pathway promoted by UVR, namely, the VDR.