NOVEL NON-CALCEMIC SECOSTEROIDS THAT ARE PRODUCED BY HUMAN EPIDERMAL KERATINOCYTES PROTECT AGAINST SOLAR RADIATION

Abstract

CYP11A1 hydroxylates the side chain of vitamin D3 (D3) in a sequential fashion [D3→20S(OH)D3→20,23(OH)₂D3→ 17,20,23(OH)₃D3], in an alternative to the classical pathway of activation [D3→25(OH)D3→1,25(OH)₂D3]. The products/intermediates of the pathway can be further modified by the action of CYP27B1. The CYP11A1-derived products are biologically active with functions determined by the lineage of the target cells. This pathway can operate in epidermal keratinocytes. To further define the role of these novel secosteroids we tested them for protective effects against UVB-induced damage in human epidermal keratinocytes, melanocytes and HaCaT keratinocytes, cultured in vitro. The secosteroids attenuated ROS, H₂O₂ and NO production by UVB-irradiated keratinocytes and melanocytes, with an efficacy similar to 1,25(OH)₂D3, while 25(OH)D3 had lower efficacy. These attenuations were also seen to some extent for the 20(OH)D3 precursor, 20S-hydroxy-7-dehydrocholesterol. These effects were accompanied by upregulation of genes encoding enzymes responsible for defence against oxidative stress. Using immunofluorescent staining we observed that the secosteroids reduced the generation cyclobutane pyrimidine dimers in response to UVB and enhanced expression of p53 phosphorylated at Ser-15, but not at Ser-46. Additional evidence for protection against DNA damage in cells exposed to UVB and treated with secosteroids was provided by the Comet assay where DNA fragmentation was markedly reduced by 20(OH)D3 and 20,23(OH)₂D3. In conclusion, novel secosteroids that can be produced by the action of CYP11A1 in epidermal keratinocytes have protective effects against UVB radiation.

1. Introduction

The ultraviolet radiation (UVR) spectrum of solar light induces significant damage to the epidermis, which is tightly connected to the production of reactive oxygen species (ROS). These ROS, which include hydrogen peroxide (H₂O₂) and nitric oxide (NO), can cause oxidative damage and reduction of the important antioxidant, glutathione (GSH), as well as inhibition of DNA repair and genotoxic and mutagenic effects [1]. The mutagenic and genotoxic effects of UVB are related to its absorption of wave-lengths in the range of 280–320 nm, particularly by the pyrimidine bases [2], resulting in the formation of cyclobutane pyrimidine dimers (CPD) [3]. CPDs are DNA lesions produced in human skin mainly from exposure to UVB [4]. Tumor suppressor protein p53 mediates the cellular response to stress by activating cell-cycle arrest or apoptosis [5]. Thus, gamma and UV irradiation, hypoxia and some chemicals induce DNA damage and the activation of p53 [6]. The extent of DNA damage and the elevation in p53 levels are proportional [6]. Thus, p53 expression increases with UVB exposure and it initiates growth arrest and repair of damaged DNA [5, 7]. Phosphorylation plays an important role in determining the activity of p53 including its ability to bind to DNA [8, 9]. Phosphorylation of p53 at Ser-15 and Ser-20 promotes accumulation and activation of p53 and DNA repair, while p53 phosphorylation at Ser-46 closely regulates apoptosis following DNA damage [10]. Some endogenous regulators such as melatonin promote phosphorylation of p53 at Ser-15, thus activating p53 and consecutively inhibiting cell growth, preventing accumulation of damaged DNA and promoting antitumor activity [11, 12].

The human epidermis is the main site of photo-induced formation of D3 from 7-dehydrocholesterol (7DHC), representing the most fundamental reaction in photobiology [13–15]. D3 is hydroxylated in positions C25 and C1 at both systemic (liver and kidney) and local (epidermis) levels to produce 1,25(OH)₂D3 [14, 16, 17]. 1,25(OH)₂D3, in addition to regulating calcium metabolism, has important pleiotropic effects that include stimulation of differentiation and inhibition of proliferation of cells of different lineage, anti-cancer properties, stimulation of innate immunity, and inhibition of adaptive immunity and inflammation [14, 16–21]. In the skin it plays a significant role in the formation of the epidermal barrier and the functioning of the adnexal structures including hair follicles, and has a wide variety of ameliorating effects on skin cancer, and proliferative and inflammatory cutaneous diseases [14, 16, 18, 20, 22–24]. Most recently, it was reported that active forms of D3 prevent, attenuate, or even reverse UVB-induced cell death and DNA damage in skin cells [25–31]. Unfortunately, due to its calcemic (toxic) effect, chronic therapeutic use of D3 at pharmacological doses is severely limited, which forms a significant barrier to the use of classical forms of D3 including 1,25(OH)₂D3. However, the discovery of an alternative pathway of vitamin D activation initiated by CYP11A1 which produces novel secosteroids which are biologically active but non-calcemic [32], offers promise for therapeutic applications.

The novel pathways of vitamin D metabolism initiated by the action of CYP11A1 are: D3→ 20S(OH)D3 → 20,23(OH)₂D3 → 17,20,23(OH)₃D3 and D2 → 20S(OH)D2 → 17,20(OH)₂D2 →17,20,24(OH)₃D2 [32–40]. Our studies indicate that these pathways operate in vivo with CYP 27B1 being capable of hydroxylating some of the products in the 1α-position [41, 42]. In tissues with high expression of CYP11A1, 20S(OH)D3 is the main metabolite and is more abundant than 25(OH)D3 [41]. Similarly, tissues expressing low levels of CYP11A1 such as skin [43, 44] can also produce 20S(OH)D and its hydroxyderivatives [41, 42]. The initial metabolite, 20S(OH)D3, is the major product of the pathway, and recently we detected it in human serum based on its mass spectrum and identical HPLC retention time to authentic standard [41]. 20,23(OH)₂D3 is the second major metabolite of the pathway [41]. Both secosteroids demonstrate biological potency, equal or higher than that of 1,25(OH)₂D3, with anti-proliferative, pro-differentiation and anti-inflammatory activities on epidermal keratinocytes, melanocytes, melanoma cells and dermal fibroblasts [32, 45–49]. Importantly, 20S(OH)D3 is noncalcemic and non-toxic at the highest pharmacological doses tested in rats (3 μg/kg), and in mice (30–60 μg/kg)[50, 51], which are up to 100-fold higher than doses of 1,25(OH)₂D3 or its precursor, 25(OH)D3, that are toxic.

In this presentation we provide evidence that novel secosteroidal compounds produced by CYP11A1 can act as protective agents against UVB-induced damage.

2. Materials and Methods

2.1. Source of steroids and secosteroids

7-Dehydrocholesterol, vitamin D3, 25(OH)D3, and 1,25(OH)2D3 were purchased from Sigma-Aldrich (St. Louis, MO). 20(OH)D3 was produced either by chemical synthesis of 20-hydroxy-7-dehydrocholesterol (20(OH)7DHC) with subsequent photolytic transformation to secosteroidal structures [52], or biochemically from vitamin D3 using bovine CYP11A1 [33, 34]. 20,23(OH)₂D3 was also produced from vitamin D3 using bovine CYP11A1 [34]. These compounds were purified by TLC and reverse-phase HPLC. Their structures have previously been determined by NMR [33, 34, 52]. Ethanol was used as a dilutent and a vehicle, and stock solutions of secosteroids were prepared as described previously [49, 53].

2.2. Human skin and human primary skin cells

The use of human tissue and cells was approved by the UTHSC Institutional Review Boards as an exempt protocol #4. Human skin samples were obtained from the Regional One Health Center (foreskin) and the Methodist University Hospital (adult skin), Memphis, TN. Human epidermal neonatal keratinocytes or melanocytes were isolated from foreskin of African-American donors as previously described [38, 48, 49].

2.3. Cell culture

HaCaT keratinocytes were grown in Dulbecco’s minimal essential media (DMEM) supplemented with 5% charcoal-treated fetal bovine serum (FBS) (Atlanta Biologicals, Lawrenceville, GA, USA) [47, 54]. Epidermal keratinocytes were grown in keratinocyte growth media (KGM) supplemented with keratinocyte growth factors (KGF) (Lonza Walkersville Inc, Walkersville, MD), while epidermal melanocytes were grown in MBM-4 with MGM-4 medium containing 0.5% charcoal-treated FBS [38, 55]. Keratinocytes and melanocytes in their third passages were plated either on 96-well plates (ROS assays) or 24-well plates (other experiments, unless indicated below) to reach confluence or semiconfluence, respectively [56]. Cultures were treated with secosteroids or other listed compounds for 24 h before UVB exposure and for various times after UVB exposure. For UVB exposures the media were replaced with phosphate buffered saline (PBS) [38].

2.4. Exposure of cells to UVB

The UVB irradiation was performed with a Biorad UV transluminator 2000 (Bio-RAD Laboratories, Hercules, CA, USA) as previously described [57, 58]. Cultures were irradiated (from the bottom of the plates) with doses of 0, 25, 50, 75, or 200 mJ/cm², respectively. The irradiance spectrum of the UV incidence upon the cells is presented in Figure 1.

Figure 1.

Spectral irradiance of the Biorad Model 2000 ultraviolet transilluminator. (A), irradiance operating in the normal (—) and preparative (^·····) modes. (B), the spectrum of UV passed through the plastic culture petri dish. The UV irradiance spectra were measured using an Optronic spectroradiometer, model 754.

2.5. Determination of Cell viability assay

HaCaT keratinocytes were plated in 96-well plates and after reaching confluence they were exposed to different doses of UVB. Following this, graded concentrations of the steroids and secosteroids being tested were added from ethanol stocks with the corresponding ethanol concentrations serving as controls. After 44 h of incubation with the compounds, 20 μl of MTS/PMS solution (Promega, Madison, WI) were added to the cells. Four hours later, absorbance was recorded at 490 nm using an ELISA plate reader. The number of viable cells was measured in six replicates

2.6. Measurement of Reactive oxygen species (ROS)

The assays were performed as described previously [56]. Briefly, cells were plated onto 96-well plates and treated with secosteroids, or ethanol vehicle (dilution ≤1:1,000) for 1 or 24 h. After preincubation with CM-H2DCFDA dye, cells were irradiated with UVB as above. After 3 or 4 h of incubation cells were washed with PBS and the generation of ROS was determined by measuring the fluorescence with excitation at 480 nm and emission at 528 nm, using a SpectraMax M2e (Molecular Devices, Sunnyvale, CA).

2.7. Measurement of nitric oxide (NO⁻) production

Nitrite (NO⁻) levels were measured using Griess reagent (1% sulfanilamide-0.1% N-1-naphthyl-ethylenediamine dihydrochloride in 2.5% phosphoric acid) (Sigma, St. Louis, MO) as previously described [56, 59]. Confluent (keratinocytes) or semiconfluent (melanocytes) cultures were treated with secosteroids for 24 h then irradiated (see above). After post-irradiation incubation for 5 or 30 min cells were harvested by scraping and further centrifuged at 3000 g for 5 min as described [56]. Supernatants were collected and mixed with equal amounts of Griess reagent. The generation of NO was determined by measuring the absorbance at 540 nm of purple azo compound formed from the reaction between nitrates formed in samples and Griess reagent, with calculations performed as described previously [56, 59].

2.8. Measurement of H₂O₂ concentrations

Hydrogen peroxide (H₂O₂) levels were measured using a luminescence method [56, 60]. Cells were treated as described above for 24 h, media were changed to PBS and cultures were irradiated with UVB of different intensities (see above). The generation of H₂O₂ was determined by measuring the luminescence of luminol (Sigma, St. Louis, MO) by H₂O₂ that is released by cells from 5 min until 1 h after UV exposure, as described before, using a Turner Luminometer (TD20/20) (Promega, Madison, WI). The specificity of the reaction was determined by treating separate UV-irradiated cells with 300 units/mL of catalase (Sigma, St. Louis, MO), which degraded H₂O₂ to H₂O and O₂. Data are presented as concentration of H₂O₂ (pmol) per 0.1 million cells and further as a percentile of control (ethanol treated cells) [56].

2.9. Measurement of Reduced glutathione (GSH) concentrations

Reduced glutathione (GSH) levels were measured using a fluorometric method [56, 59]. Briefly, cells were treated with secosteroids for 24 h, irradiated with the UVB of different intensities (see above), further incubated for 1 h in culture media then harvested following trypsinization [56]. Cells were washed once with 1x PBS-EDTA, pH 8.0, centrifuged and cell pellets were resuspended in ice cold 5% meta-phosphoric acid (MPA) (Sigma, St. Louis, MO), sonicated, and centrifuged again at 12,000 rpm. After collection of the supernatant, aliquots were taken for protein determination. The remaining supernatant was mixed with 1x PBS-EDTA buffer and OPAME (o-phthaldialdehyde (Sigma, St. Louis, MO) in methanol (Fisher, Pittsburgh, PA) and borate buffer (potassium-tetraborate, Sigma, St Lois, MO). These samples were aliquoted into 96-well plate, incubated for 15 min at RT and reduced GSH levels were determined by measuring the fluorescence with excitation at 350 nm and emission at 420 nm, using a SpectraMax M2e (Molecular Devices, Sunnyvale, CA). Data are presented as a percentile of control (EtOH- treated cells).

2.10. Measurement of CPD dimers levels

Cyclobutane pyrimidine dimers (CPDs) were measured using immunofluorescence as described previously [56]. Briefly, keratinocytes or melanocytes were plated onto chamber slides, treated with secosteroids or ethanol vehicle (dilution 1:1,000) for 24 h prior UVB exposure, then a further 3 h after the exposure. Following the manufacturer’s protocol, cells were stained with anti-CPD antibody (clone TDM2) (Cosmo Bio Co Ltd., Tokyo, Japan) (dil 1:100) and subsequently processed as described previously. Images of the immunofluorescence were recorded and analysed using ImageJ software (NIH free download).

2.11. Immunofluorescent measurement of p53

Keratinocytes or melanocytes were plated onto chamber slides in duplicates. Cells were treated with secosteroids at a range of concentrations (see figure legends) for 24 h prior UVB exposure (see above) then for 3 h after exposure for detection of p53 or phospho-p53 Ser-46, or 12 h after exposure for detection of phosphor-p53 Ser-15 [11, 56]. Immunofluorescence analysis was performed as described previously [26, 56]. Briefly, cells were fixed in 4% paraformaldehyde (PFA) for 10 min at room temperature and washed three times with 0.1% Triton X-100 (BioRad, Hercules, CA) in PBS to permebealize membranes. Blocking was performed in 10% FBS in PBS for 1 h at RT after quenching of endogenous peroxidase with 1% H₂O₂ (BioRad, Hercules, CA) in PBS. Cells were incubated with primary antibodies diluted in blocking buffer (1:100), overnight at 4°C. Primary antibodies were either anti-p53 (ab4060, Cell Signaling, Danvers, MA), anti-phospho p53 Ser-15 (9284 Cell Signaling, Danvers, MA) or anti-p53 (phosphor S-46) (ab131348, Abcam, Cambridge, MA). The following day cells were washed and incubated with secondary antibody, Alexa-Fluor 488 goat anti-rabbit IgG (Invitrogen Molecular Probes, Eugene, Oregon, USA) diluted in blocking buffer (1:100), for 1 h at RT. After washing with PBS, nuclei were stained red with Vectashield mounting media with propidium iodide (Vector Laboratories, Burlingame, CA). Images of stained cells were recorded with a fluorescence microscope and analysed using ImageJ software (NIH free download).

2.12. Comet assays

Cells were plated in 12-well plates and cultured until confluent. Then the medium was replaced with PBS and cells were exposed to UVB at 200 mJ/cm². PBS was removed and replaced with fresh medium containing 100 nM secosteroids and cells further incubated for 3 h at 37°C. After detaching, cells were counted and used for the comet assay. The comet assay was performed following the manufacturer’s protocol (Trevigen, Gaithersburg, MD) as previously described [56, 61]. Cell suspensions at a density of 1x10⁵/ml were combined with molten 1.2% low-melting-point agarose, diluted 1:10, placed onto two frosted slides precoated with 0.6% normal agarose, and incubated at 4°C for 30 min. Cells were further digested in lysis solution at 4°C overnight. DNA fragments resulting from strand breaks were separated by electrophoresis using alkaline electrophoreses solution (200 mM NaOH, 1 mM EDTA, pH>13) in a horizontal gel electrophoresis slide tray (Comet-10, Thistle Scientific, UK). DNA fragments were exposed to alkaline unwinding for 20 min at room temperature after which electrophoresis was performed at 25 V for 35 min. Following neutralization with 0.4 M Tris, pH7, comets were visualized with propidium iodine (Sigma-Aldrich, St. Louis, MO). The slides were examined and images were captured using a Nikon fluorescent microscope and the Leica digital DM 4000b fluorescent microscope equipped with image analysis software. At least 60 comet images were taken for each condition. DNA damage was measured by the tail length [62] using TriTrek Comet Score version 1.5.

2.13. Reverse transcription (RT) and quantitative polymerase chain reaction (qPCR)

Human keratinocytes were isolated from neonatal foreskins, from donors of Afican-American descent, as described previously [48, 49]. Cells were cultured in KBM media supplemented with KGF (Lonza). Cells in their 3^rd passage were treated with 0.1 μM 1,25(OH)₂D3, 20S(OH)D3, 25(OH)D3, 20,23(OH)₂D3 or 20S(OH)7DHC, or ethanol vehicle, for 24 h. Cells were harvested and RNA isolated using the Absolutely RNA Miniprep kit (Stratagen) as previously described [63].

Reverse transcription was performed using the Transcriptor First Strand cDNA Synthesis kit (Roche) using 1 μg of RNA per reaction. Primers were designed using the Universal Probe Library software (UPL Roche, Germany or as previously described [64, 65] and are shown in Table 1. Real-time PCR (qPCR) data were generated from input cDNA using a Cyber Green Mix (n = 3) that was amplified using standard settings on a Roche LC480 LightCycler. Gene expression was normalized using β-actin by the DDCp method. Changes in gene expression are presented as a fold change ±SD, as indicated.

Table 1.

Sequences of the primers used for qPCR

Genes	Description	Primers
β-actin	β-actin	L 5′-CCAACCGCGAGAAGATGA-3′ R 5′-CCAGAGGCGTACAGGGATAG-3′
CAT	catalase	L 5′-CGTGCTGAATGAGGAACAGA-3′ R 5′-AGTCAGGGTGGACCTCAGTG-3′
Cu/Zn-SOD	superoxide dismutase	L 5′-GGCAAAGGTGGAAATGAAGA-3′ R 5′-GGGCCTCAGACTACATCCAA-3′
Mn-SOD	superoxide dismutase	L 5′-GCTCATGCTTGAGACCCAAT-3′ R 5′-CACCCGATCTCGACTGATTT-3′
NRF2	the transcription factor NF-E2-related factor 2	L 5′-TTCTGTTGCTCAGGTAGCCCC -3′ R 5′-TCAGTTTGGCTTCTGGACTTGG -3′

2.14. Statistical analysis

Data are presented as means ± SD and were analysed using the Student’s t-test or ANOVA with appropriate post hoc test (for more than two groups), Dunnett’s test using Microsoft Excel and Prism 4.00 (GraphPad Software, San Diego, CA). Statistically significant differences are denoted in the figures and corresponding figure legends. Data are presented as a percentile of the control ethanol-treated cells.

3. Results and Discussion

3.1. Overview on synthesis of 20S(OH)D3 and 20,23(OH)2D3 in epidermal keratinocytes

We have provided ex-vivo evidence that tissues expressing CYP11A1, including epidermal keratinocytes, are able to metabolize vitamin D3 producing 20S(OH)D3 as the major metabolite and 20,23(OH)2D3 the second major metabolite of the pathway [41]. Smaller amounts of other mon-, di- and tri-hydroxy-derivatives of D3 are also produced [41]. In the present follow-up study we have confirmed the ability of HaCaT keratinocytes to produce 20S(OH)D3 from D3 precursor (Fig. 2A and B). Using LC-MS/MS in the ESI mode we also detected the monohydroxy D3 transition for m/z 401[M+H]⁺→383[M+H-H₂O]⁺ at a retention time corresponding to 20(OH)D3, in samples of extracted human epidermis (Fig. 2C). While the first result is confirmatory, the latter indicates that 20S(OH)D3 is present in the human epidermis in vivo.

Figure 2.

Detection of 20S(OH)D3 in human keratinocytes. HaCaT epidermal were incubated with 50 μM vitamin D for 0 (A) or 18 h (B) and extracted for LC/MS analysis. Untreated human epidermis was similarly extracted and analysed (C). The production of 20S(OH)D3 in cultured HaCaT keratinocytes was determined by LC/MS-SIM with m/z = 383 [M+H-H₂O]⁺ (insert in B). The epidermal extract was analysed by LC-MS/MS by ESI with MRM for the transition m/z 401[M+H]⁺→383[M+H-H₂O]⁺ (C). The arrow shows the retention time corresponding to 20S(OH)D3 standard. The extraction procedures and conditions for LC/MS have been described previously [41].

We have previously established a chemical route for the synthesis of 20S(OH)D3 with 20S(OH)7DHC serving as an intermediary product [52], and synthesized the C-20 epimer with the R configuration (20R(OH)D3) [66]. Armed with these synthetic analogs and with the information listed above, plus the known ability of classical vitamin D3 metabolites to protect epidermal keratinocytes against UVB induced damage (reviewed in [25, 28]), we tested the protective effects of non-calcemic 20S(OH)D3 and 20,23(OH)2D3 against UVB-induced damage in comparison to 1,25(OH)₂D3 and related sterols, in human epidermal keratinocytes and melanocytes.

3.2. Anti-oxidative responses of cells to novel vitamin D analogs

In a preliminary experiment, we used the MTS test on HaCaT keratinocytes to compare the protective effects of 20S(OH)D3, its enantiomer 20R(OH)D3, and precursors 20S(OH)7DHC and D3 against UVB-induced cell death. Only 20S(OH)D3 attenuated cell death induced by UVB (Fig. 3). The lack of a detectable effect of 20R(OH)D3 at identical concentrations to 20S(OH)D3 is consistent with the reported lower potency of the R versus S enantiomer in previous biological testing [66]. Therefore, the naturally occurring 20S(OH)D3 enantiomer was selected for subsequent experiments.

Figure 3.

The effects of 20S(OH)D3, 20R(OH)D3, D3, 20S(OH)7DHC, and 20R(OH)7DHC on attenuation of HaCaT cell death induced by UVB. Confluent cultures of HaCaT keratinocytes were irradiated with 0, 25, 50, or 75 mJ/cm² doses of UVB and then treated with graded concentrations of the steroids and secosteroids. After 48 h, cell viability was measured with the MTS test. Data were analyzed using ANOVA, * p<0.05, ** p<0.01, and *** p<0.001.

Using HaCaT keratinocytes we tested the protective effects of steroids and secosteroids against UVB-induced production of ROS (Fig. 4). Initially we tested whether 20S(OH)D3 and 20,23(OH)₂D3, in comparison to 1,25(OH)₂D3, 20S(OH)7DHC, 7-hydroxycholestrol (7(OH)C) and cholesterol sulphate, could attenuate ROS production induced by 200 mJ/cm² of UVB (Fig. 4A and B). The last two cholesterol derivatives were included because of the recent detection of retinoic acid orphan receptors (ROR)α and γ in human keratinocytes [67], for which these compounds serve as ligands. With the exception of 7(OH)C and cholesterol sulphate, all the hydroxy-derivatives of vitamin D3 tested, as well as 20S(OH)7DHC, attenuated UVB-induced ROS production with significant protection being observed at concentrations of both 100 nM and 1 nM (Fig. 4A and B). The protective effects of 100 nM 20S(OH)D3, 20,23(OH)₂D3 and 20S(OH)7DHC were also examined in comparison to 1,25(OH)₂D3, 25(OH)D3 and D3 using lower UVB doses of 25 and 50 mJ/cm² (Fig. 4B and C). The vitamin D hydroxy-derivatives, but not D3 itself or 20S(OH)7DHC, attenuated ROS production with 25(OH)D3 having significant action only at 50 mJ//cm². The use of vitamin hydroxy-derivatives and their precursors at a concentration of 100 nM in this and following experiments was rationalized by our earlier observations that this concentration was most optimal for measuring their phenotypic effects in keratinocytes and melanocytes [38, 45, 47–49].

Figure 4.

The effects of 1,25(OH)₂D3, 25(OH)D3, 20S(OH)D3, 20,23(OH)₂D3, 20S(OH)7DHC, D3, cholesterol sulphate, or 7(OH)cholesterol on the attenuation of UVB-induced ROS generation. HaCaT keratinocytes were given dosed of UVB of 200 (A and B), 25 (C) or 50 (D) mJ/cm². Cells were treated with above compounds or the ethanol vehicle for 1 h (C, D) or 24 h (A, B) prior to the UV irradiation and for a further 4 h (A, B) or 3 h (C, D) post-UVB exposure, at the concentration 1 nM (A) or 100 nM (B,C,D). Data are presented as % of control [mean ±SD (n=6)] and analysed using the student t-test, * p<0.05, ** p<0.01, *** p<0.001.

To test the ability of vitamin D hydroxy-derivatives and 20S(OH)7DHC to scavenge H₂O₂ produced in response to UVR, HaCaT keratinocytes were incubated with these compounds for 24 h before UVB exposure (Fig. 5). Cells were irradiated with UVB doses of 25 or 50 mJ/cm². Supernatants were collected after 30 min of irradiation to measure the concentration of H₂O₂ released from the cells. While there was no H₂O₂ released by non-irradiated cells, keratinocytes produced H₂O₂ in response to UVB irradiation as reported before [1], and this production was UVB dose- and time- dependent with a peak after 30 min as described previously [56]. Treatment of cells with 20S(OH)D3, 20,23(OH)₂D3, 20S(OH)7DHC, 1,25(OH)₂D3 or 25(OH)D3 significantly reduced H₂O₂ concentrations in cell supernatants following UVB exposure, although the effect of 25(OH)D3 was not significant at the higher dose (50 mJ/cm²) (Fig. 5). At 25 mJ/cm², 20S(OH)D3, 20,23(OH)₂D3 and 1,25(OH)₂D3 exhibited a significantly stronger inhibition of H₂O₂ release than 25(OH)D3 (Fig. 5A), and at 50 mJ/cm² dose 20S(OH)D3 showed stronger inhibition than 1,25(OH)₂D3 (Fig. 5B).

Figure 5.

1,25(OH)₂D3, 25(OH)D3, 20S(OH)D3, 20,23(OH)₂D3 and 20S(OH)7DHC reduce levels of H₂O₂ produced by UVB-irradiated HaCaT keratinocytes. Keratinocytes were treated with the compounds above (100 nM) or ethanol vehicle for 24 h and then irradiated with 25 (A) or 50 (B) mJ/cm² UVB. H₂O₂ produced by keratinocytes was determined 30 min after UVB irradiation. The dose dependent stimulation of H₂O₂ production by UVB is shown as an insert. Data are presented as % of control [mean ±SD (n=3)] and were analysed using the t-test, * p<0.05, ** p<0.01, *** p<0.001.

The effect of UVB on H₂O₂ release by human melanocytes 5 and 30 min after UVB exposure is shown in Fig. 6A and B. There was a statistically significant inhibition of H₂O₂ release by 100 nM 20S(OH)D3 at 25 and 50 mJ/cm² 5 min after exposure (Fig. 6A), and at 50 and 75 mJ/cm² 30 min after exposure (Fig. 6B). Using the same passages of human melanocytes we also tested the protective effect of 20S(OH)D3 on UVB-induced production of NO (Fig. 6C and D). Cells were incubated with 100 nM 20S(OH)D3 for 24 h before replacing media with PBS and irradiating with UVB doses of 25, 50, or 75 mJ/cm². Treatment of cells with 20S(OH)D3 before UVB irradiation significantly reduced NO production at 25 and 50 mJ/cm² both after both 5 and 30 min of exposure (Fig. 6C and D).

Figure 6.

20S(OH)D3 reduces H₂O₂ and NO production induced by UVB irradiation of melanocytes. Melanocytes were treated with 100 nM 20S(OH)D3 or ethanol vehicle for 24 h and subsequently irradiated with 0, 25 50, or 75 mJ/cm² of UVB. H₂O₂ (A, B) or NO (C, D) production by cells was determined 5 min (A, C)) and 30 min (B, D) after UVB irradiation.. Data are presented as % of control [mean ±SD (n=3] and analysed using the t- test, * p<0.05.

Using the same experimental design we also tested the protective effects of other novel secosteroids against UVB-induced NO release. It is known that keratinocytes produce NO in response to UVB irradiation [1] in a UVB-dose dependent manner (Fig 7, insert). The production of NO occurs shortly after irradiation, reaching a peak at 30 min and returns to low levels after 1 h [56]. Treatment of cells with 20S(OH)D3, 20,23(OH)₂D3, 20S(OH)7DHC, 1,25(OH)₂D3 or 25(OH)D3 before UVB irradiation reduced NO production. Of note, the inhibition of NO release by 20(OH)D3, 20,23(OH)₂D3, 1,25(OH)₂D3 and 20(OH)7DHC was stronger in comparison to 25(OH)D3 at the higher dose of UVB, e.g., 50 mJ/cm² (p<0.05)(Fig. 7B).

Figure 7.

1,25(OH)₂D3, 25(OH)D3, 20S(OH)D3, 20,23(OH)₂D3, and 20S(OH)7DHC reduce levels of NO produced by UVB-irradiation of HaCaT keratinocytes. Keratinocytes were treated with the above compounds or ethanol vehicle for 24 h then irradiated with 25 (A) or 50 (B) mJ/cm² UVB. NO produced by keratinocytes was determined 30 min after UVB irradiation. The dose dependent stimulation of NO production by UVB is shown as an insert. Data are presented as % of control [mean ±SD (n=3)] and were analysed using the t-test, * p<0.05, ** p<0.01, *** p<0.001. The inhibitory effect of 20,23(OH)₂D3 on NO production at 50 mJ/cm² was stronger than that of 1,25(OH)₂D3 (p<0.05).

Exposure to UVB causes an immediate reduction of intracellular GSH [68]. Previously, we documented that melatonin and its metabolites can attenuate UVB-induced reduction of GSH in human epidermal keratinocytes [56]. Using a similar experimental design, we tested the effects of 20S(OH)D3, 20,23(OH)₂D3, 20S(OH)7DHC, 1,25(OH)₂D3 and 25(OH)D3 on the relative GSH content after treatment of keratinocytes for 24 h prior to UVB exposure (Fig. 8). The level of reduced GSH decreased with the exposure to UVB in a dose-dependent manner (Fig 8A, insert). At 25 and 50 mJ/cm², 20S(OH)D3, 20,23(OH)₂D3, 1,25(OH)₂D3 and 25(OH)D3, and 20S(OH)7DHC only at 25 mJ/cm², significantly attenuated the reduction of reduced GSH after 1 h of UVB exposure (Fig. 8).

Figure 8.

1,25(OH)₂D3, 25(OH)D3, 20S(OH)D3, 20,23(OH)₂D3, and 20S(OH)7DHC attenuate inhibition of GSH levels induced by UVB in epidermal keratinocytes. HaCaT keratinocytes were treated with the above compounds (100 nM) or ethanol vehicle for 24 h and then irradiated with 25 (A) or 50 mJ/cm² UVB (B). GSH produced in keratinocytes was determined 1 h after UVB irradiation. The differences in GSH production under different UVB intensities are shown as an insert. Data are presented as % of control [mean ±SD (n=3)] and analysed using the t- test, * p<0.05, ** p<0.01, *** p<0.001.

Since the above data clearly show that the secosteroids tested have a protective effect against UVB-induced oxidative damage, we tested whether they could stimulate the expression of genes encoding enzymes involved in protection against oxidative stress. Increases in the expression of genes for catalase and both forms of SOD were observed following the incubation of human epidermal keratinocytes for 24 h with 100 nM 20S(OH)D3, 20,23(OH)₂D3, 20S(OH)7DHC or 25(OH)D3, but not with 1,25(OH)₂D3, (Fig. 9A-C). In addition, the expression of NRF2, a master regulator of oxidative responses [69] and a protector of genomic stability [70], was also upregulated by 20S(OH)7DHC, 20S(OH)D3 and 1,25(OH)₂D3, but not by 25(OH)D3 or 20,23(OH)₂D3 (Fig. 9C)

Figure 9.

The effects of 1,25(OH)₂D3, 25(OH)D3, 20S(OH)D3, 20,23(OH)₂D3 and 20S(OH)7DHC on the expression of genes mediating the anti-oxidative stress response in human epidermal (HEKn) keratinocytes. Cells were treated with 100 nM of the above compounds for 24 h before RNA isolation. A: catalase (CAT), B: Cu/Zn superoxide dismutase (Cu/Zn-SOD), C: Mn superoxide dismutase (Mn-SOD), D: the transcription factor NF-E2-related factor 2 (NRF2). Gene expression was normalized using B-actin by the DDCp method. Changes in gene expression are presented as a fold change (mean ±SD, n=3) and analysed using the t- test, * p<0.05, ** p<0.01, *** p<0.001; **** p<0.0001.

UVB-induced pathology in the skin includes oxidative DNA damage and the production of reactive oxygen species (ROS), including hydrogen peroxide and nitric oxide, which change intracellular levels of glutathione [1, 2, 25]. UVB also induces transformation of 7DHC to vitamin D3 [13, 14], which in the epidermis is further hydroxylated to 1,25(OH)₂D3 [14, 19]. Experimental evidence is accumulating that 1,25(OH)₂D3 can protect epidermal keratinocytes against oxidative damage induced by UVB [25–28], as well as against epidermal skin cancers that are induced by UVB [18, 29–31]. The present data show that that CYP11A1-derived noncalcemic vitamin D hydroxy-derivatives, 20S(OH)D3 and 20,23(OH)₂D3, also inhibit UVB-induced production of ROS, H₂O₂, NO and attenuate the reduction of intracellular levels of GSH, being at least as effective as 25(OH)D3 and 1,25(OH)₂D3. These results imply that the novel secosteroids should play a role in protection against UVB induced skin cancer, that can be similar to the role proposed for 1,25(OH)₂D3 [25, 27, 29, 31]. The pharmacological advantage of novel CYP11A1-derived hydroxyderivatives of vitamin D over classical 1,25(OH)₂D3 is that they are non-calcemic and non-toxic at supra-physiological doses [50, 51, 53], while having inhibitory activity on melanomas [38, 45, 63].

An interesting finding of this study is the ability of 20S(OH)7DHC to inhibit H₂O₂ and NO production induced by UVB as well as cause significant stimulation of the expression of genes encoding CAT, both types of SOD and NRF2, which was unexpected. While the majority of protective effects against oxidative stress and attendant skin cancerogenesis by 1,25(OH)₂D3 [25, 29] and novel 20S(OH)D3 and 20,20(OH)₂D3 [38, 71] are expected to be mediated by the VDR, a target receptor for 20S(OH)7DHC has not been identified. Good candidates are ROR or LXR receptors; however, this hypothesis remains to be tested experimentally using transgenic mice. Also, non-receptor mediated induction of anti-oxidative actions should be considered, since our previous studies showed that 7DHC significantly attenuated oxidative protein modification more effectively than cholesterol, vitamin D3 and D2, and surprisingly 7DHC was more effective than GSH or melatonin [72]. The latter is recognised as a potent anti-oxidant also operating in the skin [73, 74]. Non-genomic receptors have been proposed for the UVB-protective effects of 1,25(OH)₂D3 [27]. For the CYP11A1 generated secosteroids, our molecular modelling predicts poor binding to the non-genomic (alternative) pocket of the VDR [71]. However, RORα and γ could represent alternative receptors to mediate some of these effects, since 20S(OH)D3 and 20,23(OH)₂D3 can also act through these receptors [67].

3.3. Attenuation of DNA damage induced by UVB

DNA serves as a chromophore for UVB [2] which causes DNA damage primarily by the formation of CPD [3]. P53 plays an important role in the repair of UV-induced DNA damage [5, 75]. After exposure to UVB or oxidative stress, phosphorylated p53 accumulates in the nucleus. The most important phosphorylation sites on p53 are Ser-15 which promotes accumulation and activation of p53 and DNA repair, and Ser-46 which regulates apoptosis following DNA damage [10]. Previously it was shown that hormonal factors such as melatonin can induce the p53-dependent DNA damage response through stimulation of phosphorylation at Ser-15 but not at Ser-46 [56, 76, 77].

We therefore tested whether post-treatment of normal epidermal keratinocytes and melanocytes with classical 1,25(OH)₂D3 and novel 20S(OH)D3 could increase phosphorylation of p53 at Ser-15 or Ser-46, and reduce CPD levels following UVB exposure (Fig. 10). Previously we have shown that phosphorylation of Ser-46 in keratinocytes occurs earlier after UVB exposure (3 h), while phosphorylation of Ser-15 occurs later (12 h after UVB exposure) [56]. Therefore, these times after exposure to UVB were evaluated in the current experiments. The vitamin D analogues tested did not significantly change the expression of p53 or p53 phosphorylated at Ser-46 (data not shown), which is similar to the lack of effect of melatonin on these markers after exposure to UVB [56]. However, 1,25(OH)₂D3 and 20S(OH)D3 stimulated p53 phosphorylation at Ser-15 in keratinocytes (Fig. 10A) and melanocytes (Fig. 10C), with significant effects seen at UVB doses of 25 and 50 mJ/cm² for keratinocytes (Fig. 10A) and 25 mJ/cm² for melanocytes (Fig. 10C).

Figure 10.

Treatment of keratinocytes or melanocytes with 1,25(OH)₂D3 or 20S(OH)D3 decreases CPD levels and increase p53 phosphorylation at Ser-15 following UVB exposure. Keratinocytes (A,B) and melanocytes (C,D) were treated with 100 nM 1,25(OH)₂D3 or 20S(OH)D3 for 24 h prior to UVB exposure. Cells were exposed to UVB intensities of 25, 50, or 75 mJ/cm² and immediately treated again with 100 nM 1,25(OH)2D3, 20(OH)D3, or vehicle control for 3 h for detection of CPDs (B,D) or for 12 h for detection of phosphorylated p53 at Ser-15 (S15) (A,C). Cells were fixed and stained with anti-CPD antibody (green) (B, D inserts) or with anti-phosphorylated p53S15 antibody (A, C inserts). Stained cells were imaged with a fluorescence microscope and fluorescence intensity was analysed using ImageJ software, and data are analysed using Graph Pad. Data are presented as % of control [mean ±SD (n=6)] and analysed using t- test, * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

In complementary experiments, UVB-irradiated keratinocytes (Fig. 10B) or melanocytes (Fig. 10D) were immediately treated after UV exposure with 1,25(OH)₂D3 or 20S(OH)D3 for 3 h, and CPD-positive cells were evaluated with CPD-specific antibody. CPD-positive cells were not detected in untreated cells (Fig. 10B and D). CPD-staining intensity of UVB-irradiated cells was weaker in cells treated with 1,25(OH)₂D3 or 20S(OH)D3 (Fig. 10B, D), as compared to untreated cells. These effects were again dependent on the UVB dose with significant inhibitory effects of 1,25(OH)2D3 or 20S(OH)D3 in keratinocytes at doses of 25 and 50 mJ/cm² (Fig. 10B), and in melanocytes at all does tested (25, 50 and 75 mJ/cm²) (Fig. 9D).

To further analyse the extent of the DNA damage, we employed the Comet assay which allows the DNA of damaged cells to migrate from intact DNA in agarose under an electric field, and uses fluorescent microscopy for visualisation of the DNA migration. We observed extensive DNA damage when melanocytes were exposed to UVB (Fig. 11, insert). There was a significant difference (p<0.0001) between UVB-induced DNA damage in control cells (non-treated) versus cells treated with 1,25(OH)₂D3, 20S(OH)D3 or 20,23(OH)₂D3, as measured by the mean comet tail moment (Fig. 11). Also, 20,23(OH)₂D3 and 20S(OH)D3 exhibited a statistically stronger inhibitory effect on DNA damage that 1,25(OH)₂D3, with 20S(OH)D3 showing the greatest efficacy.

Figure 11.

1,25(OH)₂D3, 20S(OH)D3, and 20,23(OH)₂D3 inhibit DNA damage induced by UVB as shown by the Comet assay. Epidermal melanocytes were treated with the above secosteroids for 3 h following UV-irradiation (200 mJ/cm²). Photographs (upper panel) show Comets for treated and non-treated cells. UVB caused strand breaks in DNA which migrated in the electrophoretic field to form comet tails. Tail moments wert used to measure DNA damage. Data were analysed using the student t-test and labelled as **** p<0.0001.

In summary, our studies using the immunofluorescent staining technique indicate that 20S(OH)D3, 20,23(OH)₂D3 and 1,25(OH)₂D3 (the latter serving as a positive control because of previous demonstrations of its protection against DNA damage in keratinocytes [25, 28]) can protect keratinocytes and melanocytes against UVB-induced damage as illustrated by CPD generation and by results of the Comet assay. The enhancement of the expression of p53 phosphorylated at Ser-15, but not at Ser-46 and down-regulation of CPD formation suggest that the vitamin D analogs may enhance the DNA repair capacity of melanocytes and keratinocytes. Thus, this study on the protective effects of secosteroids against UVB-induced DNA damage not only complements the excellent studies by Mason and coworkers [25–28], but also identifies novel non-calcemic CYP11A1-derived vitamin D derivatives, and the 20(OH)D3 precursor 20S(OH)7DHC, as protectors against UV induced DNA damage and oxidative stress, not only in human keratinocytes but also in melanocytes. The highest efficacy is seen for CYP11A1-derived 20S(OH)D3 and 20,23(OH)₂D3, and classical 1,25(OH)₂D3, with both precursors 25(OH)D3 and 20S(OH)7DHC having less efficacy or showing differential effects.

In conclusion, novel CYP11A1-derived and classical secosteroids which are endogenously produced in keratinocytes, protect both epidermal keratinocytes and melanocytes against UVB-induced oxidative stress and DNA damage.

4. Concluding remarks

Skin, being exposed to environmental stressors, has developed sophisticated systems to deal with noxious factors [78, 79]. The UVB spectrum of solar radiation represents a major cutaneous stressor, with the skin pigmentary system [80, 81] protecting against its deleterious effects. Most recently, it was proposed that the local melatoninergic anti-oxidative system (MAS) [82] and the cutaneous serotoninergic/melatoninergic system [74] protect the skin against UVB induced damage. The data coming from the laboratories of Dr Mason and Dr Bikle indicate that vitamin D3, generated through the action of UVB and hydroxylated to its active form, 1,25(OH)₂D3 [25–28], acts via the VDR [19, 29, 30] to protect the skin against the damage induced by UVB. We have discovered novel secosteroidogenic pathway(s) operating in vivo [41, 42, 83, 84], for which the major products also protect keratinocytes and melanocytes against UVB-induced oxidative stress and DNA damage, as shown in this study. Of note is the expression of the enzyme that initiates and regulates these novel secosteroidogenic pathways, CYP11A1, which is also be upregulated by UVB [85, 86]. Therefore, we propose that the activities of the classical (producing 1,25(OH)₂D3) and novel (producing 20(OH)D3 and 20,23(OH)₂D3) secosteroidogenic enzyme systems regulate the protection of the human epidermis against UVB exposure by attenuating DNA- and metabolic-damage via several mechanisms. These include the reduction in H₂O₂ and NO levels, elevation of GSH levels, enhancement of DNA repair, and promotion of keratinocytes and melanocytes survival in a context dependent fashion. Thus the secosteroidal molecules of which production in the skin is induced by solar radiation, are activated to secure skin homeostasis under the sun.

Highlights.

• We detected 20-hydroxyvitamin D3 in the human epidermis

• 20-Hydroxy- and 20,23-dihydroxyvitamin D3 attenuate oxidative stress in keratinocytes and melanocytes

• 20-Hydroxy-7-dehydrocholesterol attenuates oxidative stress in keratinocytes and melanocytes

• 20-Hydroxyvitamin D3 attenuates UVB induced DNA damage to a similar degree as 1,25-dihydroxyvitamin D3