Role of the vitamin D3 pathway in healthy and diseased skin – facts, contradictions and hypotheses

Authors


Bodo Lehmann, PhD, Department of Dermatology, Carl Gustav Carus Medical School, Dresden University of Technology, Fetscherstraße 74, D-01307 Dresden, Germany, Tel.: +49 351/458 2692, Fax: +49 351/458 4338, e-mail: bodo.lehmann@mailbox.tu-dresden.de

Abstract

Abstract:  Irradiation of human keratinocytes with UVB (280–320 nm) in vitro and in vivo activates the metabolism of 7-dehydrocholesterol to hormonally active calcitriol. The production of calcitriol in the skin strongly depends on the photosynthesis of vitamin D3 which is biologically inactive in the first instance. Vitamin D3 serves as the starting substrate for two subsequent enzymatic hydroxylation steps in epidermal keratinocytes. Both the amount of vitamin D3 and the activity of anabolic and catabolic vitamin D hydroxylases determine the cutaneous level of calcitriol. The hormonally active metabolite of vitamin D3 regulates a huge number of genes in keratinocytes, and thus acts in an autocrine and/or paracrine manner. This local pathway of vitamin D3 is unique, but its relevance for healthy and diseased skin is widely unknown, yet. Experimental findings implicate several questions: (1) Is UVB-induced formation of calcitriol involved in regulation of growth and differentaition of epidermal cells as well as immunological and skin protective processes? (2) What endogenous and exogenous factors including drugs affect the cutaneous vitamin D3 pathway? From a therapeutical point of view, it has been known for a long time that topical application of calcitriol and its analogs can improve hyperproliferative skin diseases like psoriasis. In spite of many encouraging studies in recent years, the fields of the routinely therapeutical application of calcitriol or vitamin D analogs in dermatology (e.g. treatment of immunological, inflammatory, malignancies and infectious skin diseases) have not been intensified. Why is that?

Abbreviations:
UVB

ultraviolet B

7-DHC

7-dehydrocholesterol

DBP

vitamin D-binding protein

VDR

vitamin D receptor

VDRE

vitamin D response element

Introduction

Calcitriol (1α,25-dihydroxyvitamin D3, 1α,25(OH)2D3), the hormonally active vitamin D3 metabolite is mainly produced in the kidney after a cascade of reactions prefixed in liver and skin. It is well known for a long time that 1α,25(OH)2D3 maintains calcium homeostasis primarily by promoting the intestinal absorption of calcium and phosphorus, decreasing the clearance of these minerals from the kidney, and promoting bone mineralization (1). It is common knowledge today, that calcitriol exerts additional physiological functions including regulation of growth and differentiation in a broad variety of normal and malignant cells (2–5).

Human epidermal keratinocytes contain the complete machinery needed to produce the hormone 1α,25(OH)2D3 (calcitriol) from its initial precursor 7-dehydrocholesterol (7-DHC) (6–9). These cells also express the nuclear vitamin D receptor (VDR) that mediates the effects of this hormone on keratinocytes (10,11). Thus, it is reasonable to assume that calcitriol acts in an autocrine and paracrine manner in the epidermis. Recent evidence suggests that calcitriol produced in the skin may induce a number of biological actions which will be discussed in this review.

The vitamin D3 metabolism

Photochemical reactions in the skin

A photochemical reaction with maximum spectral effectiveness from 297 to 302 nm (UVB: 280–320 nm) results in the generation of previtamin D3 from 7-DHC (provitamin D3) in basal and suprabasal layers of the skin (Fig. 1) (1). By comparison, the shortest terrestrial solar wavelengths reaching the surface are ≈295 nm with the 295–320 nm range comprising ≈4% of the solar UV irradiance. UV radiation above 315 nm is unable to produce previtamin D3 in human skin. Photochemical conversion of 7-DHC to previtamin D3 in the skin proceeds rapidly. Previtamin D3 then undergoes thermal isomerization over a few hours and generates vitamin D3 (cholecalciferol). The vitamin D-binding protein (DBP) preferentially translocates vitamin D3 into the circulation, and the sequential activation of vitamin D3 happens in the liver and kidney. The serum level of 25OHD3 produced in the liver widely accounts for the ‘vitamin D status’ of the the body. In contrast, the serum level of 25OHD2 which derives from nutritional vitamin D2 is commonly very low in the majority of the European population (<3 ng/ml) (12).

Figure 1.

 Vitamin D3 pathway in epidermal keratinocytes. (7-DHC, 7-dehydrocholesterol; CHOL, cholesterol; preD3, previtamin D3; 25OHD3, 25-hydroxyvitamin D3; 1α,25(OH)2D3, 1α,25-dihydroxyvitamin D3; 24R,25(OH)2D3, 24R,25-dihydroxyvitamin D3; 1α,24R,25(OH)3D3, 1α,24R,25-trihydroxyvitamin D3; 1α,25(OH)2-3-epi-D3, 1α,25-dihydroxy-3-epi-vitamin D3; CYP27A1, (27)25-hydroxylase; CYP27B1, 1α-hydroxylase; CYP24A1, 24-hydroxylase; Δ7-R, 7-DHC-Δ7-Reductase).

The minimal UVB radiation level that produces a significant increase of serum vitamin D3 was determined as being 18 mJ/cm2 for skin type II (13). Of relevance to public health is the finding that the threshold value of 18 mJ/cm2 is not generally reached during spring, fall and winter in North America and Europe (13). During prolonged exposure to the sun, both previtamin D3 and vitamin D3 undergo reversible photoisomerization to biologically inert sterols or previtamin D3 is subject to a back reaction leading to the regeneration of 7-DHC (Fig. 1) (14).

The 7-DHC level in normal human adult skin ranges between 1.9 and 75 μg/cm2 (15–19). An inverse relationship between the cutaneous 7-DHC concentration and age has been ascertained (20,21). Consistently, it was found by Holick et al. (21) that the amount of vitamin D3 in elderly subject’s circulation over the ensuing 72 h after exposure to simulated sunlight was about 30% of healthy young volunteers (skin type III, each). Other reports did not confirm such an inverse relationship, because responses in old and young volunteers (white-skinned) to artificial ultraviolet irradiation were quite similar (18). It should be noted, however, that spectral energy distribution of the UV lamps used in these studies was quite different, und may explain divergent findings. It seems, that only a small pool of 7-DHC in the upper layers of the epidermis is available for immediate conversion into previtamin D3 in response to ultraviolet irradiation, and this pool is only slowly replenished (18).

There is a biochemical equilibrium between 7-DHC and cholesterol which is adjusted by the activity of Δ7-reductase in epidermal keratinocytes. The conversion of 7-DHC into previtamin D3 in the skin depends on several individual and environmental factors including skin pigmentation (22), the solar zenith angle which depends on latitude, season and time of day (23–25) and usage of sunscreens (26–28). Some studies, however, have demonstrated that application of SPF 15 over a long time interval did not cause significant changes of the serum level of vitamin D3 (29–31).

The quantities of vitamin D3 synthesized by the skin are very small compared with the concentration of the precursor 7-DHC (assumed ≈2.000 ng/cm2). Human skin subjected to ultraviolet radiation in vivo produces up to 25 ng (1 IU) vitamin D3 per cm2 according to a conversion rate 1.3% (18,32). Other studies using foreskin exposed to summer sunlight have shown that formation of previtamin D3 may level off at approximately 7% of 7-DHC concentration (24). After whole-body UV irradiation of young adults with 1 MED the serum concentration of vitamin D3 reaches a maximum within 12–24 h and amounts to 20–30 ng/ml (800–1200 IU/l) (21,22,32,33).

Metabolism of vitamin D3

Vitamin D3 is stepwise hydroxylated in liver and kidney. The final product is hormonally active 1α,25(OH)2D3. This hormone acts locally in the kidney but is also transported by DBP to other target tissues that express VDR to operate in a genomic or nongenomic manner. However, the physiological serum concentration of calcitriol (10−11 to 10−11 m) is most likely too low to induce VDR-mediated hormonal effects in the skin (34,35). More than 99% of serum 1α,25(OH)2D3 is tightly bound to carriers such as DBP and albumin, and only 0.4% of the circulating 1α,25(OH)2D3 is free (35). Thus, the free plasma concentration of calcitriol approximates only around 6 × 10−13 m. Several studies have shown that inhibition of growth of human keratinocytes in vitro requires unbound calcitriol at concentrations higher than 10−8 m (equivalent to a highly unphysiological concentration of approximately 2.5 × 10−6 m total calcitriol in the circulating blood).

In vitro, many nonrenal cells, including keratinocytes, bone, placenta, prostate, macrophages, T-lymphocytes, dendritic cells and several cancer cells (e.g. from lung, prostate, and skin) can convert 25OHD3 to 1α,25(OH)2D3 (36–39). Several cell species including keratinocytes, macrophages, prostate epithelial cells and osteoblasts even express both 25-hydroxylase and 1α-hydroxylase activity which enables them to directly convert vitamin D3 to 1α,25(OH)2D3 (38,40–43). The efficient catalysis of substrate vitamin D3 to product 1α,25(OH)2D3 in these cells is dependent on several local factors in addition to the enzymes (CYP27A1 and CYP27B1) involved. These include the availability of: (1) molecular oxygen; (2) a source of electrons and accessory proteins used to deliver electrons to the mixed function oxydases CYP27A1 and CYP27B1 (i.e. NADPH-ferredoxin reductase and ferredoxin); (3) substrates (vitamin D3 and 25OHD3) localized at the enzyme’s catalytic sites; and possibly (4) intracellular vitamin D binding proteins which can regulate synthesis of 1α,25(OH)2D3 (44).

Circulating 25OHD – a substrate for further metabolism in epidermal keratinocytes?

The cutaneous metabolism of circulating 25OHD3 to 1α,25(OH)2D3 should be reconsidered as the amount of free 25OHD3 that permeates the cell membrane of epidermal keratinocytes is maybe too low to induce formation of sufficient amounts of 1α,25(OH)2D3. 25-Hydroxyvitamin D3 is very tighly bound to DBP (Kd = 5 × 10−8 m) in circulating blood (45). Due to this tight binding and the high plasma concentration of DBP (0.3–0.5 mg/ml) virtually all 25OHD3 molecules in the circulatory system are present in a complex with DBP. Only approximately 0.03% of the metabolite is found in free form (46). According to the free hormone hypothesis (47), steroids including 1α,25(OH)2D3 and 25OHD3 enter target cells by passive diffusion. It seems to be unlikely that epidermal keratinocytes can incorporate 25OHD × DBP or 1α,25(OH)2D x DBP complexes in vivo by megalin/cubilin-mediated endocytosis like mammary cells (48) or renal proximal tubule cells (49). The deeper layers of the epidermis are not vascularized, which additionally impairs the passage of 25OHD from blood to epidermal keratinocytes.

Specifics of the cutaneous vitamin D3 pathway

Human keratinocytes have an autonomous vitamin D3 pathway (6–9,38,50). This pathway includes UVB-induced synthesis of vitamin D3 and its stepwise hydroxylation to calcitriol (Fig. 1). A five-step inactivation pathway from calcitriol to calcitroic acid in epidermal keratinocytes is attributed to the catabolic enzyme CYP24A1, which is transcriptionally induced by the action of calcitriol in a very sensitive manner (35,51). It has been shown that already picomolar concentrations of free 1α,25(OH)2D3 can induce the CYP24A1 in cultured keratinocytes (35). The physiological importance of a second catabolic pathway which includes the conversion of 1α,25(OH)2D3 to the A-ring diastomer 1α,25(OH)2D-3epi-D3 is less clear (52).

Addition of exogenous 1α,25(OH)2D3 at supraphysiological doses (12 nm) to cultured keratinocytes reduce keratinocyte 7-DHC/vitamin D3 levels, suggesting that calcitriol regulates vitamin D3 production via an intracellular feedback loop (53). Additional endocrine factors including Ca2+, hydrocortison, and EGF seem to regulate the production of vitamin D3 in human keratinocytes. Some steroidal hormones (testosterone, androstendione, progesterone, oestradiol and 11-deoxycorticosterone) can potently inhibit the CYP27B1 of the vitamin D3 pathway in human keratinocytes (54). Moreover, It is well known that the epidermis is a site of high androgen production.

In vitro investigations have shown that dermal fibroblasts express one of the potential 25-hydroxylases (CYP27A1), but not the 1α-hydroxylase (CYP27B1) (Fig. 1). Therefore, fibroblasts might play an important role in the supply of calcitriol precursors (vitamin D3 and 25OHD3) for keratinocytes and possibly for circulating blood, too (55).

Conclusions, open questions and hypotheses

The main steps of the cutaneous vitamin D3 pathway are well known. Several subleties and fine tuning mechanisms of this pathway wait for final clarifying, yet. That implies questions: Why do some people exhibit a low vitamin D status despite abundant sun exposure? (56) Why do chronic hemodialysis patients exhibit defective photoproduction of vitamin D3, despite normal epidermal content of the substrate 7-DHC? (57). What endogenous and exogenous factors including drugs affect the cutaneous vitamin D3 pathway? Other questions arise: Are there interactions between cutaneous calcitriol production or external vitamin D analogs and other hormones/cytokines within the skin? The physiological role of the cutaneous vitamin D3 metabolism to calcitriol remains still unknown. Of paramount interest is the role of this pathway in the UVB-(and PUVA-photo(chemo)therapy? (58,59)) of psoriasis. Finally, it is noteworthy that strict sun protection combined with low nutritional intake of vitamin D over a long time does not result in absolute vitamin D3 deficiency and absence of 25OHD in the circulation. This fact implies the question: Can vitamin D3 be produced in ‘emergency situations’ by UVB-independent pathway(s)? Four nonphotochemical enzymatic mechanisms have been proposed as possible reactions for the generation of vitamin D3 in appropriate biological systems (60). There is, however, presently no experimental evidence to support their existence.

Hormonal effects of calcitriol in the skin

The particular role of VDR in gene activation

Skin cells (keratinocytes, epithelial cells of the epidermal appendages, melanocytes, Langerhans cells, CD11b+ macrophages, CD3+ T-lymphocytes and dermal fibroblasts) express VDR (10,61), an absolute prerequisite for regulation of genomic effects of calcitriol and other synthetic analogs. The VDR activates transcription by binding to vitamin D response elements (VDREs) within the promoter of vitamin D reponsive genes either as a homodimer (62,63) or a heterodimer with retinoid acid X receptor-α (RXR) (63) or thyroid hormone receptor (TR) (64,65). VDR RXR heterodimers form more stable complexes on VDERs as compared to VDR homodimers. Therefore, it has been generally considered that VDR RXR heterodimers play a dominating role in mediating calcitriol action. Suprisingly, it was shown that the ligand-dependent interaction of VDR with several coactivators (SRC-1 and TRAM-1) facilitates homodimer formation and that VDR homodimers may mediate ligand-induced transactivation, as well as VDR RXR heterodimers (66). Other studies using TR provided similar results and indicated that TR may function through its liganded homodimer (66,67). In contrast, it was demonstrated in another study that VDR diplays DNA binding and transactivation as a heterodimer with RXR but not with the TR (68).

That implies that RXR/TR selective retinoids and thyroid hormones, respectively, may influence targeting of calcitriol, in addition to the RXR and TR responsive ones. On the other hand, calcitriol and other nuclear hormones may also regulate RXR and TR responsive genes. Hence, combinations of RXR selective retinoids, thyroid hormones and calcitriol/vitamin D analogs might excel the expected therapeutical result and reduce unwanted side-effects/toxicity of each single agent (69).

Interestingly, selected transcriptional effects of VDR may be uncoupled from the 1α,25(OH)2D3 ligand in keratinocytes (70). It has been demonstrated that VDR activates the 24-hydroxylase (CYP24A1) promoter independently of 1α,25(OH)2D3 in primary keratinocytes (70). Overexpression of VDR in keratinocytes increases not only CYP24A1 expression but also a number of markers of differentiation such as keratin 1, loricrin and filaggrin whereas suppression of VDR expression markedly reduces the expression of these markers in the absence of 1α,25(OH)2D3 (71).

Finally, the functions of VDR are not limited to the binding to VDREs and the regulation of gene expression. VDR has also been found to bind β-catenin, a key transcriptional factor in the Wnt pathway. This pathway plays a role in a number of malignancies, because it can induce hyperproliferation. VDR binds to, and thus blocks the transcriptional activity of β-catenin (72). The ability of VDR to bind β-catenin may account for at least some of the antiproliferative actions of 1α,25(OH)2D3 in a variety of tissues including skin cancer.

Of note, UVB which is essential for the induction of the cutaneous vitamin D3 pathway suppresses VDR expression in cultured keratinocytes in vitro (73) and retinoid acid receptors (RAR-γ and RXR-α) in human skin in vivo (74). However, observed effects of UV radiation in human skin were oppositional in terms of the retinoid and calcitriol responsiveness: UV irradiation caused an almost complete loss of retinoid acid induction of distinct RAR/RXR target genes but did not affect calcitriol-induction of the VDR/RXR-regulated gene CYP24A1 (74). Thus, in contrast to the observed impairment of retinoid responsiveness, UV radiation had no apparent effect on 1α,25(OH)2D3-responsiveness in human skin.

Conclusions, open questions and hypotheses

It is well accepted that calcitriol exerts a number of genomic and non-genomic effects in keratinocytes. However, a more detailed dissection of the pathways leading to 1α,25(OH)2D3-independent VDR transactivation would be of utmost interest. A clear separation between ligand-dependent and ligand-independent VDR-mediated processes appears to be necessary with regard to genomic effects originally imputed to calcitriol. It also remains to be clarified whether lack of VDR indeed entailes deleterious effects on skin function contributing to skin photo-aging and carcinogenesis (75).

Targets of 1α,25(OH)2D3 in keratinocytes

A vast number of genes in primary human keratinocytes and squamous carcinoma cell lines are regulated by calcitriol and its analogs (76–78). Important candidate target genes related to 1α,25(OH)2D3-induced effects on cellular growth, differentiation and inflammation/woundhealing in keratinocytes are listed in Table 1.

Table 1.   Regulatory activities of 1α,25(OH)2D3 on genes related to growth, differentiation and inflammation/woundhealing in keratinocytes
Effect of 1α,25(OH)2D3 onmRNAProteinVDRERef.
Proliferation-associated genes
c-myc  (79,80)
c-fos +(81)
Cyclin D1  (77)
TGF-β1/2+(β2)(82,83)
p21WAF1/CIP1+(79)
p27KIP1  (79,84)
PTHrP (85)
EGF  (86)
GADD45α  (78,87)
Insulin like GF  (88)
IGFBP-3  (89)
17β-OH-Steroiddehydrogenase (90)
IEX-1 (91–93)
TRPV6 (87,94)
Differentiation related genes
Involucrin+(11)
Transglutaminase I (11)
u- and t-plasminogen activator+(95)
PLC (β, γ, δ)+(γ1)(96–98)
Integrin α7B  (87)
Peptidylarginine deiminases  (77)
Kallikrein  (77)
Serine proteinase inhibitors  (77)
PPARγ (99)
Caspase-14 (psoriatic skin)  (100)
Inflammation and woundhealing-related genes
TNFα+(101,102)
IL-1α  (103)
IL-6 (76)
IL-8  (103,104)
IL-10 (IL-10 receptor)↑(↑)  (105)
PDGF  (106)
RANTES  (104)
i-NOS  +(107)
5-Lox  +(107)
CAMP/cathelicidin+(108,109)
DEFB2/Defensin 2 +(109)

Regulation of cell functions of keratinocytes by 1α,25(OH)2D3

Growth and differentiation

Low concentrations of 1α,25(OH)2D3 stimulate growth of keratinocytes in vitro, whereas higher doses (≥10−8 m) inhibit keratinocyte proliferation (110). The former effect of calcitriol in keratinocytes may be based on enhancement of the growth-promoting activity of EGF ligands (86). Consistently, pharmacological doses of calcitriol and vitamin D analogs are effective in the treatment of the hyperproliferative skin disease psoriasis.

Although the mechanisms that underlie the antiproliferative and prodifferentiating effects of vitamin D analogs on keratinocytes are far from being understood, it is well known that these effects are largely of genomic nature and are mediated by VDR as well as several coactivators (71,111,112). VDR expression is associated with undifferentiated, proliferating keratinocytes, whereas RXR-α expression appears to be related to the differentiated phenotype. Thus, proliferating and differentiating keratinocytes may be differentially affected by active vitamin D compounds (113).

Growth inhibitory effects of biological inactive vitamin D3 on cultured keratinocytes?

Vitamin D3 by itself is not only biologically inactive but also nearly insoluble in aqueous culture media because of its extreme lipophilic character. This problem can be circumvented by adding exogenous bovine serum albumin (BSA) to the culture medium which improves the solubility of vitamin D3 and, therefore, its availability for the metabolism to 1α,25(OH)2D3 in cultured keratinocytes (7). BSA has, however, the disadvantage of changing the cell morphology and increasing cell aggregation (data not shown). It is more convenient to apply Pluronic® F127, a bifunctional triblock copolymer surfactant (poloxamer) composed of polyoxyethyle-ene-polyoxypropylene-polyoxyethylene (MW = 12.500) (114,115), to improve of the solubility of vitamin D3 in aqueous culture media. This agent considerably improves the uptake of vitamin D3 by keratinocytes without adverse effects on cell viability. Consistently, the conversion rate of vitamin D3 (1 μm) to 1α,25(OH)2D3 is 1,7 fold higher in the presence of Pluronic® F127 (100 μg/ml medium) than in the presence of BSA (10 mg/ml medium) (data not shown). The inhibition of proliferation of keratinocytes at increasing concentrations of vitamin D3 (250, 500, 1000 nm) in presence of Pluronic® F127 correlates well with the production of calcitriol (Fig. 2a), and highly significant inhibition of 3H-thymidine incorporation can be observed at 250 nm vitamin D3 (equivalent to 114 fmol calcitriol per well) already. In contrast, both the conversion rate of vitamin D3 to calcitriol and incorporation of 3H-thymidine in cellular DNA are clearly delayed in the absence of Pluronic® F127 (Fig. 2b). There was neither inhibition of cell growth in the absence of vitamin D3 nor in the presence of Pluronic® F127 alone (Fig. 2c).

Figure 2.

 Inhibition of keratinocyte growth by 1α,25(OH)2D3 formed after metabolization of exogeneous added biologically inactive vitamin D3. Keratinocytes were seeded in culture dishes (5 × 104 cells/dish, diameter: 32 mm) and cultivation was carried out in 1.5 ml KGM (0.15 mm Ca2+) at 37°C in a humidified atmosphere of 5% CO2 in air. (a) Three days after seeding (degree of confluency: 40–60%) KGM was supplemented with vitamin D3 (final concentration: 250, 500 and 1000 nm) and Pluronic® F127 (100 μg/ml KGM) that was solubilized in ethanol (final concentration of ethanol: 0.5%). Incorporation of 3H-thymidine as percent of control and formation of calcitriol in keratinocytes was determined 3 days after addition of vitamin D3; (b) KGM was supplemented with vitamin D3 (final concentration: 250, 500 and 1000 nm) without Pluronic® F127; (c) Control experiments were carried out in absence of vitamin D3. The calcitriol formed was extracted and determined as described previously (7). The effect of calcitriol on DNA synthesis was determined by incorporation of [3H]-thymidine in cellular DNA. Cells were pulse-labeled with 1 μCi of [methyl, 1′,2′-3H]thymidine (126 Ci/mmol, Amersham) for 3 h. Cultures were washed three times with PBS and twice with 10% (w/v) TCA. Cells were solubilized with 1 M NaOH and radioactivity was determined – with a scintillation counter. Concentrations of calcitriol (fmol/well) and 3H-thymidine incorporation (percent of control) are shown as mean ± SD of four independent experiments (Student–Newman–Keuls multiple comparisons test: ns not significant, *P < 0.05 significant, **P < 0.01 very significant, ***P < 0.001 extremely significant versus untreated control).

These findings confirm and extend results of Chen et al. (116) and Popadic et al. (117), who demonstrated that vitamin D3 is effective in inhibiting 3H-thymidine incorporation into DNA of cultured human keratinocytes at concentrations between 10−8 and 10−6 m as well as 1.25 – 5.0 × 10−6 m, respectively. Our results provide the first experimental evidence that the antiproliferative effect of vitamin D3 significantly correlates with the amount of hormonally active calcitriol intracellularly produced by hydroxylation of the biologically inactive substrate vitamin D3. Thus, our findings indicate that cutaneous metabolism of vitamin D3 to calcitriol may contribute to the normalization of the hyperproliferative condition of keratinocytes in vitro, and possibly in psoriatic skin, too. This hypothesis is supported by the fact that the UVB-induced synthesis of vitamin D3 is unimpaired in psoriatic individuals (27). It remains to be clarified, however, whether enzymatic conversion of vitamin D3 to calcitriol is unimpaired in psoriatic skin, too. Notably, in psoriatic keratinocytes 3H-1α,25(OH)2D3 is catabolized approximately 2.5 times faster than in normal keratinocytes (118). This finding suggests for a disturbed balance between UVB-induced synthesis of calcitriol and its catabolism in psoriatic keratinocytes.

Conclusions, open questions and hypotheses

Biological effects of calcitriol in keratinocytes and other skin cells require a sufficient intracellular level of this hormon. It is not clear from in vitro trials whether the extracellular calcitriol concentration induces an identical intracellular concentration. Concentrations of calcitriol created within keratinocytes may succeed the circulating concentration by far (8). Consistently, we have demonstrated that biologically inactive vitamin D3 developes antiproliferative activity on keratinocytes in vitro, however, only after its metabolization to 1α,25-dihydroxyvitamin D3. It has to be verified now, whether this growth-inhibitory effect of vitamin D3in vitro can also be observed in psoriatic skin. In addition, it would be of considerable interest whether cutaneous calcitriol generated in psoriatic skin after UVB exposure develops a growth-inhibitory effect on proliferating epidermal keratinocytes similar to topical applicated calcitriol.

Apoptosis

Physiological concentrations of calcitriol do not initiate apoptosis in cultured keratinocytes, but rather, cause resistance against proapoptotic ceramides, UV radiation and tumor necrosis factor-α (TNF-α) (119). The cytoprotective/antiapoptotic effect of calcitriol is seemingly linked to the generation of sphingosine-1-phosphate (119). In contrast, pharmacological concentrations of calcitriol (≥10−6 M) exert a proapoptotic effect on keratinocytes.

Conclusions, open questions and hypotheses

Regulatory effects of calcitriol on both cellular growth and apoptosis are characterized by a dual mechanism of action. Low calcitriol concentrations stimulate cell proliferation and suppress apoptosis; higher (pharmacological) concentrations dose-dependently decrease proliferation and initiate apoptosis of keratinocytes. The underlying cause of this different behavior is presently unknown.

Skin appendages: targets for cutaneous calcitriol?

Sebaceous glands

It was shown by Sato et al. (120) that calcitriol suppresed the accumulation of intracellular lipid droplets in hamster sebocytes whereas the cellular DNA content was not altered. Hence, 1α,25(OH)2D3 may be involved in the suppressive regulation of lipogenesis in hamster sebocytes and possibly in human sebocytes, too (120). More recently, it was reported that sebocytes express functional cathelicidin antimicrobial peptide (CAMP) which acts as an agent against propionibacterium acnes (121). This result might be of prime importance because strong upregulation of cathelicidin synthesis at the mRNA- and protein level by calcitriol and calcipotriol (MC903) has been found in human keratinocytes in vitro and in vivo (108,122). Thus, it seems entirely possible that calcitriol also regulates cathelicidin production in sebocytes.

Recently, a study has demonstrated that calcitriol dose-dependently suppressed the growth of the cell line SZ95, derived from human sebaceous gland (123,124) The mRNA of VDR, CYP27A1, CYP27B1 and CYP24A1 were stronly expressed in these cells (123). Hence, biochemical prerequisites for local synthesis of 1α,25(OH)2D3 from vitamin D3 or 25OHD3 and signal transduction of active vitamin D3 are available in these cells.

Conclusions, open questions and hypotheses

Metabolism of vitamin D3 and 25OHD3 to 1α,25(OH)2D3 in human sebocytes might be of relevance to both regulation of growth and production of antimicrobial peptides in these cells. On the other hand, it is well known that UVB radiation necessary for photosynthesis of vitamin D3 worsens acne. The question arises: Does the vitamin D3 pathway properly work in lesional skin of acne patients?

Hair follicle

Originally, it was assumed that 1α,25(OH)2D3 plays a fundamental role in hair follicle biology. This is probably wrong and should be reconsidered. The control of hair follicle development and cyling depends on the VDR expression; the presence of its ligand 1α,25(OH)2D3 is apparently not necessary (125,126). VDR expression in the hair follicle is increased during late anagen (hair growth phase), and catagen (regression phase), correlating with decreased proliferation and increased differentiation of the follicle keratinocytes. It is assumed, however, that repressive effects of the VDR on the hair cycle involve additional co-modulators or factors (127).

More recently the corepressor hairless (Hr) has been found in the epidermis and brain. Hr binds to the VDR in absence of 1α,25(OH)2D3 and blocks its ability to stimulate gene transcription (128). The ability to regulate the transcriptional activity of VDR suggests that such interactions are involved in hair follicle cycling. Loss of Hr just like loss of VDR results in alopecia. Interaction between these two molecules may control normal hair follicle cycling. Lack of VDR results in increased expression of Hr suggesting that VDR may act directly or indirectly to regulate Hr expression (129).

Results of Li et al. (130,131) have demonstrated that the retinoic acid receptor RXR-α plays a key role in anagen initiation during the hair follicle cyle. RXR-α ablation results in epidermal interfollicular hyperplasia with keratinocyte hyperproliferation and aberrant terminal differentiation, accompanied by an inflammatory reaction of the skin. Hence, RXR-α/VDR heterodimers play a major role in controlling hair cycling.

Data obtained from animal models indicate that topical calcitriol may protect against chemotherapy-induced alopecia (132). On the other hand, it was found that calcitriol treatment results in accelerated and qualitative improved hair regrowth but does not prevent chemotherapy-induced alopecia (133,134).

Conclusions, open questions and hypotheses

Hair follicles may be targets of cutaneos calcitriol. On the other hand, the control of hair follicle development and cyling needs particularly unliganded VDR; the presence of its ligand 1α,25(OH)2D3 is seemingly not necessary. Both Hr and VDR are required for normal hair follicle cyling. Whether Hr acts as a corepressor for ligand-independent actions of VDR in hair follicles has yet to be determined.

Immunomodulating effects of calcitriol in the skin

More recently, it was shown that both dendritic cells (DCs) and cocultures of DCs and T cells efficiently metabolize the biologically inactive prohormone vitamin D3 to hormonally active calcitriol, providing a mechanism for the local regulation of T cell ‘epidermotropism’ (135). Calcitriol induces the transcription and surface expression of the skin T-cell-associated chemokine receptor CCR10, conferring attraction to the epidermal chemokine CCL27 to T cells (135). It has been demonstrated that a single topical application of a physiological relevant amount of 1α,25(OH)2D3 to the shaved dorsal skin of mice enhances the suppressive capacity of CD4+ CD25+ regulatory T cells in the draining lymph nodes (136). Calcitriol inhibits differentiation and maturation of DCs and induces a phenotype that promotes tolerance and inhibits immunity after stimulation with antigen (137,138). Calcitriol and vitamin D analogs suppress IgE-production in vitro and IgE-mediated reactions in the skin (139,140). On the other hand, it has been reported that topical application of calcitriol or of its low-calcemic analog calcipotriol induces the expression of the cytokine thymic stromal lymphopoietin (TSLP) in the skin of mice resulting in an atopic dermatitis-like syndrome, which is known from transgenic mice overexpressing TSLP (141). Of note, TSLP is also produced in epidermal keratinocytes of AD patients (142). It is well known that most common side effect of topical treatment of psoriasis with calcipotriol is skin irritation (red, dry, and itchy skin) (143). Accordingly, the data of Li et al. (141) have shown that administration of calcitriol may exacerbate AD and very likely, asthma by promoting expression of TSLP in skin and lungs. Consistently, it has been concluded from these results that topical administration of calcitriol antagonists might be beneficial for treatment of AD.

Conclusions, open questions and hypotheses

Immunomodulating effects of calcitriol and its analogs on keratinocytes as well as monocytes, macrophages, T lymphocytes and, in particular DCs, have been convincingly demonstrated. However, light and shade are close together from the therapeutic point of view. Novel vitamin D anlogs are needed for highly specific and effective regulation of immunological processes. The role of the cutaneous vitamin D3 pathway is not clear yet. It also remains to be seen whether the so called vitamin D hypothesis can explain increased incidence of astma and allergies (144).

Calcitriol induces antimicrobial activity in human skin

There is convincing evidence that 1α,25(OH)2D3 and its analog calcipotriol directly regulate antimicrobial peptide gene expression in human skin, revealing the potential of these compounds for the treatment of opportunistic infections (122,145,146). The promoters of the human cathelicidin antimicrobial peptide (CAMP) and defensin 2 (defB2) genes contain consensus VDRE that mediate 1α,25(OH)2D3-dependent gene expression (108).

Using microdialysis, we have demonstrated that irradiation with UVB induces the synthesis of calcitriol in human skin (38,50). Recently, it was shown by Mallbris et al. (147) that under comparable experimental conditions both the VDR and the precursor of the antimicrobial peptide cathelicidin (CAMP) (also known as hCAP18, LL-37 or FALL-39), were significantly upregulated. The effect was not observed following UVA exposure, ruling out the cutaneous synthesis of vitamin D3 and calcitriol. In vitro, both 1α,25(OH)2D3 and 25OHD3 exhibited stimulatory activity on expression of CAMP (122). Induction of CAMP expression in keratinocytes and monocytes is mediated by Toll-like receptors (148,149). A link between the cutaneous vitamin D3 pathway and induction of CAMP might be of importance in reduction of inflammation, wound repair and healing of burns (146,148–150). AD also has been associated with depressed expression of LL-37; therefore, an appropriate therapeutical use of vitamin D analogs and/or UVB should be considered (146). On the other hand, this strategy would be conflicting because application of these agents might lead to exacerbation of AD and possibly astma by promoting expression of TSLP in skin and lungs (141).

Conclusions, open questions and hypotheses

Topical treatment of human skin with 1α,25(OH)2D3 has been shown to enhance cathelicidin peptide expression. However, molecular mechanisms controlling the expression of CAMP are still poorly understood. The capacity of 1α,25(OH)2D3 to alter the antimicrobial function of keratinocytes and other skin cells is unknown yet. Furthermore, it is not clear why cathelicidin production is initiated upon injury an incident not associated with alterations in vitamin D signalling. The relevance of AMP’s for several skin diseases like AD, psoriasis and infectious skin diseases with bacterial, viral or fungal background has to be clarified. It is attractive to speculate that acute exposure of the skin to sunlight and subsequent activation of the cutaneous vitamin D3 pathway might turn the immune balance in favor of innate immune defense.

The cutaneous vitamin D3 pathway and skin cancer

UV radiation, essential for triggering of the cutaneos vitamin D3 pathway, is a well-documented carcinogen, indisputably linked to the current continued increased rate of skin cancer (151). These cancers include melanomas and two types of malignant keratinocytes: basal-cell carcinomas (BCC) and squamous – cell carcinomas (SCC). 1α,25(OH)2D3 in turn stimulates the differentiation of epidermal keratinocytes, raising the hope that 1α,25(OH)2D3 may prevent the development of malignancies in these cells (98). In contrast to normal cells, malignant transformation may cause a resistance to 1α,25(OH)2D3 (152). SCCs fail to respond to the prodifferentiating effects of 1α,25(OH)2D3 (153,154). Considered for itself, this is suprising because these cells have normal expression of VDR and normal binding of VDR to vitamin D response elements. However, they overexpress coactivator DRIP205 such that the p160/SRC coactivator complex is blocked from binding to VDR (98,155).

Conclusions, open questions and hypotheses

Unfortunately, we do not exactly know how malignant cells can evade normal growth-arresting and prodifferentiating signals. By all appearances, clearing up of molecular alterations and disturbed control mechanisms underlying skin cancer will take some time for final clarification.

Skin protecting effects of calcitriol

Skin photocarcinogenesis is caused largely by DNA damage, most importantly mutations at sites of incorrectly repaired DNA photoproducts, of which the most common are the cyclobutane pyrimidine dimers (CPDs) (156). It has been reported that 1α,25(OH)2D3 protects primary human keratinocytes against the induction of CPDs by UVB (157,158). When applicated topically to mouse skin, calcitriol or analogs decrease UV-induced DNA damage, sunburn cells and immune suppression (159). Interestingly, 1α,25(OH)2D3 exerts its photoprotective effects via the rapid non-genomic pathway (160). Moreover, calcitriol produced in the skin may enhance UV-induced p53 protein expression and suppress nitric oxide (NO) products resulting in increased DNA repair (161). Other studies have demonstrated genoprotective effects of calcitriol or its analogs against accumulation of mutations which underly the cellular transformation and cancer progression (162,163).

Conclusions, open questions and hypotheses

It is tempting to speculate from these findings that the UVB-induced cutaneous production of calcitriol represents a hormonal feedback mechanism that protects the skin from the hazardous effects of solar UV-radiation. Therefore, endogenous calcitriol or active vitamin D compounds may represent promising candidates for the chemoprevention of UVB-induced skin cancer.

Outlook

A broad variety of skin diseases such as psoriasis, vitiligo, ichthyosis, Darier’s disease, ILVEN, lichen amyloides, localized scleroderma, morphea, bullous pemphigoid, acanthosis nigricans, acrodermatitis continua of Hallopeau, erythema annulare centrifugum, Grover’s disease and other disorders can be more or less improved by application of calcitriol or vitamin D analoga (164–166). Until now, however, therapeutical usage of calcitriol and its analogs in the dermatology is rather limited and only established for treatment of psoriasis. It remains to be seen whether calcitriol and other synthetic vitamin D analogs are sufficiently effective for clinical treatment of nonpsoriatic skin diseases, malignancies included (167,168). It should also be mentioned that calcipotriol a well known antipsoriatic vitamin D analog has been found to have no therapeutical effect in the following nonpsoriatic diseases: alopecia totalis, alopecia areata, acne vulgaris, ichthyosis bullosa of Siemens, palmoplantar keratoderma, and keratosis pilaris (Darier’s disease) (165).

We conclude from our own and other studies that locally produced 1α,25(OH)2D3 may in fact suppress growth and induces differentiation of keratinocytes (34,116,117,169). Altogether, our knowledge about regulation of cellular functions by the cutaneous vitamin D3 pathway or by vitamin D analogs is insufficient yet. It is necessary to characterize the underlying signalling pathways and continue with the identification of further calcitriol-responsive genes in the skin. From the therapeutical point of view, the inhibition of the enzymatically regulated catabolism of calcitriol and its analogs by selectively acting CYP24A1 inhibitors remains an interesting therapeutical approach (170).

Acknowledgements

I would especially like to thank Dr Wolfgang Eicheler (Clinic of Radiotherapy, Technische Universität Dresden) for critical proof-reading the paper. This paper was gracefully supported by Prof. Dr Michael Meurer (Professor and Chairman, Department of Dermatology, Technische Universität Dresden).

Ancillary