High basal NF-κB activity in nonpigmented melanoma cells is associated with an enhanced sensitivity to vitamin D3 derivatives

Background: Melanoma is highly resistant to current modalities of therapy, with the extent of pigmentation playing an important role in therapeutic resistance. Nuclear factor-κB (NF-κB) is constitutively activated in melanoma and can serve as a molecular target for cancer therapy and steroid/secosteroid action. Methods: Cultured melanoma cells were used for mechanistic studies on NF-κB activity, utilising immunofluorescence, western blotting, EMSA, ELISA, gene reporter, and estimated DNA synthesis assays. Formalin-fixed, paraffin-embedded specimens from melanoma patients were used for immunocytochemical analysis of NF-κB activity in situ. Results: Novel 20-hydroxyvitamin (20(OH)D3) and classical 1α,25-dihydroxyvitamin D3 (1,25(OH)2D3) secosteroids inhibited melanoma cell proliferation. Active forms of vitamin D were found to inhibit NF-κB activity in nonpigmented cells, while having no effect on pigmented cells. Treatment of nonpigmented cells with vitamin D3 derivatives inhibited NF-κB DNA binding and NF-κB-dependent reporter assays, as well as inhibited the nuclear translocation of the p65 NF-κB subunit and its accumulation in the cytoplasm. Moreover, analysis of biopsies of melanoma patients showed that nonpigmented and slightly pigmented melanomas displayed higher nuclear NF-κB p65 expression than highly pigmented melanomas. Conclusion: Classical 1,25(OH)2D3 and novel 20(OH)D3 hydroxyderivatives of vitamin D3 can target NF-κB and regulate melanoma progression in nonpigmented melanoma cells. Melanin pigmentation is associated with the resistance of melanomas to 20(OH)D3 and 1,25(OH)2D3 treatment.

Melanoma is a neoplasm of melanocytic origin with an environmental aetiology that is linked to overexposure to solar radiation. Melanoma incidence depends on race and ethnicity, indicating that melanin may play a protective role (Slominski et al, 2004;Lin and Fisher, 2007). Skin exposure to ultraviolet B (UVB) radiation generates vitamin D 3 , a crucial hormone/prohormone, through photochemical transformation of 7-dehydrocholesterol (Holick, 2003). The sequential hydroxylation of vitamin D3 at positions 25 and 1 produces the active form, 1a,25-dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ; calcitriol), which regulates calcium body homeostasis and has immunomodulatory, antiproliferative, and anticancer activity (Holick, 2003;Deeb et al, 2007;Bikle, 2010). These pleiotropic activities are regulated by interaction with the vitamin D nuclear receptor (VDR) (Holick, 2003;Bikle, 2010). The novel vitamin D 3 hydroxyderivative (20-hydroxyvitamin (20(OH)D 3 )), a product of cytochrome P450scc action (Slominski et al, 2005), affects the proliferation and differentiation of human keratinocytes (Zbytek et al, 2008) and leukaemic cells , but also inhibits nuclear factor-kB (NFkB) activity in keratinocytes (Janjetovic et al, 2009). As 20(OH)D 3 is noncalcemic , it has a significant therapeutic potential for treating hyperproliferative disorders. Melanoma is the most aggressive form of skin cancer and is notoriously resistant to all current modalities of cancer therapy once the metastatic process has begun (Soengas and Lowe, 2003). Melanin pigment, which protects cells against the harmful actions of UV radiation attenuating melanomagenesis (Slominski et al, 2004), can paradoxically contribute to the resistance of melanoma to different forms of therapy (Slominski et al, 1998;Kinnaert et al, 2000;Meyskens et al, 2007;Brozyna et al, 2008).
Nuclear factor-kB (NF-kB) plays an important role in inflammation and cancer development (Van Waes, 2007), and is constitutively activated in many human cancers, including melanoma (Franco et al, 2001;Dolcet et al, 2005;Karin, 2006). In mammals, the NF-kB family includes NF-kB1 (p105/p50), NF-kB2 (p100/p52), Rel A (p65), Rel B, and cRel. In most cells, NF-kB is found bound to IkB as an inactive complex in the cytoplasm. Phosphorylation and subsequent degradation of IkB proteins result in NF-kB translocation into the nucleus, where it can bind to specific gene promoters and activate transcription. The activation of NF-kB is mediated through the activation of the IkB kinase (IKK) complex, which catalyses IkB phosphorylation. Nuclear factor-kB is a potential molecular target for cancer therapy and for steroid action (Van Waes, 2007). As increased NF-kB activity appears to play an important role in maintaining the aggressive nature of melanoma (Yang and Richmond, 2001), and melanin pigment affects the sensitivity of melanoma cells to chemotherapy (Slominski et al, 2004), we analysed the relationship between pigmentation and NF-kB activity in a well-defined cell culture model of inducible melanin pigmentation (Brozyna et al, 2008;Slominski et al, 2009). This relationship was also validated in clinical biopsies from melanoma patients. Furthermore, as active forms of vitamin D inhibit NF-kB (Riis et al, 2004;Janjetovic et al, 2009Janjetovic et al, , 2010, we examined whether pigmentation affects the antiproliferative activity of novel (20(OH)D 3 ) and classical (1,25(OH) 2 D 3 ) forms of vitamin D 3 and defined the mechanism of their action on NF-kB.

Cell proliferation assay
Melanoma cells were seeded in a 24-well plate (TPP, Trasadingen, Switzerland) and grown until reaching B80% confluence. Cells were serum starved overnight and then treated for 24 or 48 h with 0.1 -100 nM 1,25(OH) 2 D3 or 20(OH)D3 that was diluted in medium containing charcoal-stripped serum followed by the addition of 1 mCi ml -1 [ 3 H]-thymidine (Moravek Biochemicals Inc., Brea, CA, USA) to the concentration of 1 mCi ml -1 for 4 h. The excess unbound thymidine was removed by washing with 1 Â PBS. Cells were washed with 1 Â PBS and subjected to precipitation with 10% trichloroacetic acid (TCA) (Sigma, Munich, Germany) and then dissolved with 1 N NaOH. The precipitate was collected in vials and thymidine incorporation was determined in liquid scintillation counter (Beckman LS 6000, Santa Clara, CA, USA).

Gene reporter assays
Cells were co-transfected with the NFkB-Luc construct and a Renilla luciferase reporter gene (Promega, Madison, WI, USA), which served as normalisation control, using Lipofectamine Plus (Invitrogen, Carlsbad, CA, USA) in F-10 medium, as described previously (Zbytek et al, 2008). Cells were also transfected with p-Luc (Promega) that served as a vector control. At 24 h after transfection, cells were incubated with fresh medium containing 100 nM 1,25(OH) 2 D 3 , 20(OH)D 3 , or ethanol (as vehicle) for 1, 4 8, 16, 24, or 48 h. Firefly luciferase and Renilla luciferase activity was measured on a TD-20/20 luminometer (Turner Designs, Sunnyvale, CA, USA) following the protocol for DLR Assay (Promega). The resulting promoter-specific firefly activity was normalised to the Renilla signal and expressed relative to the activity obtained in control (untreated) cells.

Preparation of cell extracts
Nuclear extracts were prepared as previously described using a cell extraction kit (Active Motif, Carlsbad, CA, USA). Cells were harvested and resuspended in 1 Â hypotonic buffer. After incubation for 15 min on ice, the detergent was added and the suspension was centrifuged for 30 s at 14 000 g (Du et al, 2009). Supernatant was collected as cytoplasmic extract. The cell pellet was further resuspended in complete lysis buffer with DTT and protease inhibitor cocktail (Sigma). After 30 min of incubation on ice with shaking, the cell extract was centrifuged (14 000 g for 10 min at 4 1C). The resultant supernatant was considered the nuclear extract. Protein concentration was determined using a BCA protein assay kit (Thermo Scientific, Pierce, Rockford, IL, USA).
Whole-cell extracts were prepared by lysing cells in complete lysis buffer with DTT and protease inhibitor cocktail. After incubation for 60 min on ice, the suspension representing the whole-cell extract was centrifuged (13 226 g) for 20 min at 4 1C. The amount of protein was determined using a Bradford protein assay kit (Bio-Rad, Hercules, CA, USA).

Electrophoretic mobility shift assay (EMSA)
The NF-kB DNA-binding activity was determined by EMSA. Nuclear extract (20 mg) was incubated at 22 1C for 30 min with NF-kB 32 P-labelled oligonucleotide and gel shift binding buffer consisting of 2.5 mmol l -1 DTT, 0.25% Tween-20, and 0.25 mg ml -1 poly(dI):poly(dC). For supershift assays, 1 mg antibody against p50 and p65 (NCI Preclinical Repository, Rockville, MD, USA) or VDR (Santa Cruz Inc., Santa Cruz, CA, USA) were added to the nuclear extract before DNA binding and incubated for 20 min at 22 1C. Loading dye (2 ml of 10 Â ) was added to each sample and samples were separated on 5% TBE gels at 80 V for 3 h. The gel was dried and the bands were quantified by phosphoimage autoradiography (Cyclone, Packard, Palo Alto, CA, USA).

Immunoprecipitation analysis
Cells were lysed in RIPA buffer (Cell Signaling, Beverly, MA, USA) on ice for 1 h. After centrifugation (15 339 g for 20 min at 4 1C), the lysates were incubated with 25 ml of Protein A/G PLUS-Agarose beads (Santa Cruz Inc.) and 1 mg of normal mouse serum for 3 h with shaking at 4 1C, pelleted by centrifugation for 10 min at 4 1C, and incubated overnight at 4 1C on a rocking platform with 25 ml of Protein A/G PLUS-Agarose beads and 2 mg of anti-VDR antibody (D-6) (Santa Cruz Inc.). Immunoprecipitates were collected by centrifugation, washed four times with RIPA buffer, separated by SDS -PAGE, and subjected to immunoblotting. The primary antibody used for immunoblotting was mouse monoclonal antibody to VDR (Santa Cruz Inc.; 1 : 500 dilution) or the mouse monoclonal antibody to ERp57 (Santa Cruz Inc.; 1 : 50 000 dilution). The secondary antibody used was TrueBlot ULTRA HRP anti-mouse IgG (eBioscience, San Diego, CA, USA; 1 : 1000 dilution). The immunoblots were developed with the ECL reagent (Pierce, Thermo Scientific) and visualised on a Kodak Imager.

Enzyme-linked immunosorbent assay
To monitor p65 translocation into the nucleus, SKMEL-188 cells were grown to 80% confluency and then treated with 100 nM 1,25(OH) 2 D 3 or 20(OH)D 3 for 0, 0.5, 1, 4, or 24 h. The cells were harvested, nuclear and cytoplasmic extracts prepared, and nuclear p65 levels were determined using an enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's protocol (IMGENEX, San Diego, CA, USA). The amount of nuclear and cytoplasmic p65 was detected by adding a secondary antibody followed by alkaline-phosphatase-conjugated antibody. The absorbance value for each well was determined at 405 nm using a microplate reader.

Immunohistochemical analysis
Samples of melanoma were obtained from 79 patients (35 females and 44 males, age range from 25 to 90 years, median 59.8 ± 14.3) treated at the Oncology Centre -Prof. Franciszek Łukaszczyk Memorial Hospital, Bydgoszcz, Poland. Melanomas with different grades of pigmentation (amelanotic (n ¼ 27), moderately (n ¼ 32), and strongly pigmented (n ¼ 19)) were classified according to Clark's stage and Breslow's depth. Samples included 16 melanomas at Clark's stage I, 7 at Clark's stage II, 22 at Clark's stage III, 27 at Clark's stage IV, and 13 at Clark's stage V. Breslow's thickness of studied melanomas was 0 -1 mm, 17 cases; 1.1 -2.0 mm, 10 cases; 2.1 -3.0 mm, 9 cases; 3.1 -4.0 mm, 5 cases; and 44.1 mm, 32 cases. Melanin content was graded as 0 if melanin was absent, 1 if melanin was visible in up to 50% of cells, and 2 if melanin was present in 450% of cells. Sections were made as previously described with minor modifications (Janjetovic et al, 2009). In brief, formalin-fixed, paraffin-embedded sections were dewaxed with xylene and gradually hydrated. After antigen retrieval by heating in Tris/EDTA (pH 9.0), sections were blocked with 2% BSA and incubated overnight at 4 1C with anti-p65 antibody (Santa Cruz Biotechnology; dilution 1 : 150 in antibody diluent) (Dako, Glostrup, Denmark). AlexaFluor 488-conjugated secondary antirabbit antibody (dilution 1 : 500 in TBS) was added for 60 min (Invitrogen), cells were washed with PBS, and mounted in Vectashield mounting medium containing propidium iodide (PI) (Vector Laboratories, Burlingame, CA, USA) for nuclear visualisation. Primary antibody was omitted in negative controls. Consecutive sections were stained with H&E. Sections were viewed at Â 40 magnification with a BX-50 epifluorescence microscope (Olympus, Tokyo, Japan) equipped with FITC and TRITC filters and photo-documented using a Nikon DS (Melville, NY, USA) digital camera. The study was approved by the Committee of Ethics of Scientific Research of Collegium Medicum of Nicolaus Copernicus University, Poland. Sections were evaluated for the intensity, extent, and localisation of staining for the p65 subunit of NF-kB within the melanoma cells .
The proliferation activity of melanomas was assessed by Ki-67 immunocytochemistry. Deparaffinisation and antigen retrieval was achieved by boiling sections in high pH buffer in PT Link module (Dako). After quenching of the endogenous peroxidase activity with 3% H 2 O 2 , sections were incubated with anti-Ki67 primary monoclonal mouse antibody, clone MIB-1 (Dako; dilution 1 : 100) with EnVision FLEX Antibody Diluent (Dako) for 30 min at room temperature. Next, HRP-labelled secondary, anti-mouse antibody (EnVision þ System HRP Labelled Polymer Anti-Mouse; Dako) was added at room temperature for 30 min, followed by incubation with DAB (Dako) for 5 min. Then, sections were counterstained with haematoxylin, dehydrated, and mounted (Thermo Shandon Ltd, Rochester, New York, NY, USA/Thermo Fisher Scientific Inc., Kalamazoo, MI, USA). The positive controls for Ki67 staining were sections of lymph node. Ki67-labelled sections were evaluated under Nikon Eclipse 80i light microscope. The Ki67 expression was evaluated as percentage of Ki67-positive melanoma cells. Moreover, mitotic index was estimated and defined as number of cells in mitosis per 1 mm 2 . The VDR immunostaining was performed as previously described (Brozyna et al, 2011).

Immunofluorescence staining
The SKMEL-188 cells were seeded onto cover glasses in six-well plates and treated with 100 nM 20(OH)D 3 , 1,25(OH) 2 D 3 , or vehicle (0.1%. ethanol) for 1 h. Following treatment, cells were washed with PBS, fixed in 4% paraformaldehyde, permeabilised in 0.2% Triton X-100 in PBS for 5 min, and washed with PBS. After blocking in 2% BSA for 30 min, primary antibody (NFkB-p65 (1 : 100) in 1% BSA) was added and incubated overnight at 4 1C. Cells were washed with PBS and secondary antibody (Alexa Fluor 488 (1 : 500 dilution in PBS), Invitrogen) was applied for 1 h in the dark. After washing and mounting with Vectashield mounting medium containing PI (Vector Laboratories), cells were examined with a fluorescent microscope and photographed at Â 40 magnification (Janjetovic et al, 2009).

VDR translocation
To determine VDR translocation from cytoplasm to the nucleus induced by 20(OH)D 3 , SKMEL-188 cells were transduced with pLenti-CMV-VDR-EGFP-pgk-puro to express VDR and EGFP fusion protein . The cells were incubated with hydroxyvitamin D 3 derivatives for 2 h, followed by fixation with 4% paraformaldehyde and mounting with fluorescence mounting medium, and were analysed on a fluorescent microscope. Data are presented as a percentage of cells with fluorescent nuclei relative to the total cell number.

Statistical analysis
Data are presented as mean ± s.d. and analysed using Student's t-test (for two groups) and appropriate post hoc test (for more than two groups) using Prism 4.00 (GraphPad Software, San Diego, CA, USA). Statistically significant differences are denoted as *Po0.05, **Po0.01, and ***Po0.001.

Characterisation of NF-jB activity in nonpigmented vs pigmented melanoma cells
In previous studies (Janjetovic et al, 2009, we established that vitamin D derivatives target the NF-kB pathway in human keratinocytes. In this study, we examined the role of active forms of vitamin D in regulation of NF-kB activity in melanoma cells that depended on melanin pigmentation. For in vitro studies we used the human SKMEL-188 melanoma line that is amelanotic when cultured in F-10 medium relatively deficient in melanin precursors, but produces melanin pigment when cultured in a mixture of DMEM and F-10 (75 : 25) that contains an increased concentration of melanin precursors including L-tyrosine (Slominski et al, 1999). The medium rich in L-tyrosine induces melanin pigmentation and changes responsiveness of melanoma cells to different treatment protocols (Brozyna et al, 2008;Slominski et al, 2009).
First, we examined whether the induction of melanin pigmentation can affect the expression of the vitamin D receptor. Immunoprecipitation and subsequent western blot analysis ( Figure 1A) reveals that pigmented SKMEL-188 cells express less VDR protein than nonpigmented cells. In addition, blotting for anti-ERp57, which may represent a putative cell membrane vitamin D receptor (Khanal and Nemere, 2007), showed that the ERp57 protein was equally expressed in pigmented and nonpigmented cells. Therefore, melanin content is inversely proportional to VDR protein expression.
In general, the most common form of NF-kB in activated cells is a nuclear complex of the p50 and p65 subunits. To examine the subcellular localisation of the p65 NF-kB subunit in SKMEL-188 melanoma cells, we prepared cytoplasmic and nuclear extracts from pigmented and nonpigmented melanoma cells. Lysates from equivalent numbers of cells were analysed by western blotting with anti-p65. Results revealed that higher nuclear levels of p65 were present in nonpigmented SKMEL-188 melanoma cells than in pigmented cells. Conversely, higher cytoplasmic levels of p65 were observed in pigmented cells than in nonpigmented cells ( Figure 1B). To further characterise the activation of NF-kB in melanoma cells, we performed immunofluorescent staining for p65 in pigmented and nonpigmented SKMEL-188 cells. As shown in Figure 1C, prominent p65 nuclear staining was found in nonpigmented melanoma cells, whereas p65 was mainly localised in the cytoplasm of pigmented cells.

20(OH)D 3 reduces NF-jB DNA activation in melanoma cells
To determine whether 20(OH)D 3 or 1,25(OH) 2 D 3 affected NF-kBdriven transcriptional activity, pigmented and nonpigmented melanoma cells were transfected with a NF-kB-dependent luciferase reporter construct (NF-kB-p-Luc). The experiments revealed that constitutive NF-kB-driven transcriptional activity in nonpigmented melanoma cells was reduced by both secosteroids in a time-dependent manner (Figure 2). Treatment for 30 min with  Figure 2 The effects of treatment with 20(OH)D 3 or 1,25(OH) 2 D 3 on NF-kB-dependent transcriptional activity in nonpigmented melanoma cells. Cells were transiently co-transfected with NF-kB-Luc reporter plasmid and Renilla reporter for 24 h, and then treated with 100 nM 20(OH)D 3 or 1,25(OH) 2 D 3 , or vehicle (ethanol) for the indicated times. Luciferase activity was measured in four independent experiments and data are presented as percentile of control (mean ± s.d.). Transfection with p-Luc served as negative control (insert). Data are presented as a percent (%) of control.
1,25(OH) 2 D 3 and 20(OH)D 3 resulted in 50% and 80% reduction in NF-kB-dependent luciferase activity, respectively. After 1 and 4 h, the inhibitory effect was less prominent (B20% reduction), and after 8 h of treatment, NF-kB-dependent luciferase activity returned to baseline levels. In contrast, treatment of melanised melanoma with either 20(OH)D 3 or 1,25(OH) 2 D 3 was without any effect on NF-kB-driven transcriptional activity (data not shown). Transcriptional activity of NF-kB is usually associated with an increased DNA binding of NF-kB as determined by EMSA using a consensus NF-kB oligonucleotide probe. As shown in Figure 3A, high constitutive activation of NF-kB DNA-binding activity was observed in nuclear extracts prepared from nonpigmented SKMEL-188 melanoma cells, as compared with extracts prepared from pigmented cells. Treatment of nonpigmented cells with 20(OH)D 3 resulted in a time-dependent inhibition of NF-kBdependent DNA-binding activity. The presence of the p50 or p65 subunit within the NF-kB complexes was confirmed by supershift assays with the antibodies directed to these NF-kB subunits ( Figure 3B). The specificity of NF-kB binding was shown by competition with a 50-fold excess of unlabelled NK-kB oligonucleotide. Also, the presence of VDR within NF-kB complexes was confirmed by supershift assay with anti-VDR antibody ( Figure 3B). Treatment of pigmented cells with 20(OH)D 3 had no effect on NF-kB activation ( Figure 3A).
To further characterise the effects of secosteroids and pigmentation on NF-kB activity, we examined p65 nuclear translocation using a highly sensitive ELISA as well as by western blot and immunofluorescence. Treatment of nonpigmented SKMEL-188 melanoma cells with 20(OH)D 3 decreased nuclear levels of p65 ( Figure 4A) with a subsequent increase in cytoplasmic p65 levels ( Figure 4B). Translocation of p65 into the nucleus was greatly inhibited by treatment with 20(OH)D 3 or 1,25(OH) 2 D 3 (100 nM for 30 min). By 4 h, the inhibitory effect of the secosteroids on p65 translocation was no longer detected ( Figure 4C and D). Treatment of nonpigmented melanoma cells with 20(OH)D 3 decreased nuclear levels of p65 ( Figure 4A) with a subsequent increase in cytoplasmic p65 levels ( Figure 4B). In pigmented cells, treatment with 20(OH)D 3 or 1,25(OH) 2 D 3 had no effect on the intracellular localisation of p65 (data not shown). These results indicate that 20(OH)D 3 and 1,25(OH) 2 D 3 inhibit p65 nuclear translocation in nonpigmented but not in pigmented human melanoma cells. These data were further confirmed by immunofluorescent staining; that is, treatment with 20(OH)D 3 and 1,25(OH) 2 D 3 blocked p65 nuclear translocation in SKMEL-188 nonpigmented melanoma cells. Also, as shown in Supplementary Figure 1, secosteroid treatment (100 nM for 1 h) of nonpigmented cells resulted in the retention of p65 in the cytoplasm with little evidence of nuclear staining, whereas p65 nuclear staining was clearly evident in untreated cells. No effect of secosteroid treatment was seen in pigmented melanoma cells (Supplementary Figure 1).

P65 subunit is mainly localised in the nuclei of nonpigmented melanoma cells
To further validate our in vitro data, we evaluated whether differences in NF-kB activity in human melanoma samples was reflective of their pigmentation. We examined p65 expression by immunocytochemistry of melanomas obtained from 79 patients (35 females and 44 males, age range 25 -90 years, median 59.8 ± 14.3). Predominantly nuclear p65 NF-kB staining was observed in 23 out of 27 samples (85.2%) of nonpigmented melanoma, whereas in 18 of 32 samples (56.3%) of moderately pigmented melanomas, p65 was present in both the cytoplasm and nuclei. In contrast, in 15 out of 19 samples (78.9%) of strongly pigmented melanomas, p65 was found exclusively in the cytoplasm ( Figure 5). Moreover, the intensity of p65 staining in the cytoplasm was also found to increase with increasing melanoma pigmentation, with a significant difference being observed between nonpigmented (average relative value ¼ 1.01) and strongly pigmented melanomas (average relative value ¼ 1.37; Po0.01). However, the intensity of p65 staining in nuclei decreased with increased pigmentation (0.68 vs 0.11 for average relative values, in nonpigmented and strongly pigmented melanomas, respectively). The nuclear staining of p65 was significantly higher (Po0.005) in nonpigmented compared with moderately and strongly pigmented melanomas (Table 1). Thus, analysis of the large archival material (79 cases) demonstrated that pigmentation affects NF-kB activity as evidenced by changes in the intracellular distribution of p65 (predominantly nuclear p65 staining in nonpigmented cells as compared with predominantly cytoplasmic p65 staining in pigmented cells). This pattern was comparable with cell culture results showing higher nuclear levels of p65 in nonpigmented than in pigmented melanoma cells (Figure 1). Conversely, higher cytoplasmic levels of p65 were observed in pigmented than in nonpigmented cells (Figure 1).
In human melanoma samples, translocation of NF-kB into the cell nuclei was accompanied by a higher proliferation activity as assessed by both Ki67 immunostaining and the cellular mitotic index. Melanomas with o10% of cells exhibiting p65 nuclear staining showed significantly lower proliferation activity assessed by Ki-67 immunohistochemistry ( Figure 6) when compared with melanomas with p65 nuclear staining in 425% of cells ( Figure 7A). We also observed a correlation between percentage of NF-kB nuclear immunostaining and the percentage of Ki67positive cells (r ¼ 0.3144, P ¼ 0.0034). Similar results were observed when mitotic index was used for growth rate assessment, but Amelanotic melanoma Strongly pigmented melanoma Negative control Moderately pigmented melanoma Figure 5 Nuclear factor-kB expression in pigmented and nonpigmented melanoma tissue. Sections of human melanomas were subjected to immunohistochemical staining for the p65 NF-kB subunit. p65 staining with AlexaFluor 488 (green) was merged with nuclei staining with PI (red). Arrows indicate cytoplasm of melanoma cells, arrowheads indicate melanoma cell nuclei, and double arrowheads indicate melanoma cell with p65 localised in nuclei. Scale bars ¼ 50 mm. . The levels of p65 in nuclear (C) and cytoplasmic (D) extracts were quantified by ELISA using the same amount of proteins (determined by protein assay) and measuring OD at 405 nm. Following the manufacturer's recommendation, p65 standard curve was titrated to determine protein levels. statistically significant differences were only observed between melanomas with o25% and 475% of cells with p65 nuclear staining ( Figure 7B). However, the percentage of cells with NF-kB nuclear immunostaining also correlated with mitotic index (r ¼ 0.1977, P ¼ 0.0492).
The melanoma samples were also assessed for VDR expression by immunohistochemistry (Figure 8). First, we found a decrease of VDR expression in moderately and heavily pigmented melanomas in comparison with amelanotic ones (Figure 8A -C). Second, in melanoma patients, the highest nuclear VDR immunostaining was observed in melanomas with 475% of cells with p65 nuclear staining ( Figure 7C), and VDR immunostaining also correlated with the percentage of cells with nuclear NF-kB staining. There was significantly elevated VDR nuclear staining as compared with melanomas with o10% of cells with nuclear NF-kB. The relationship of cytoplasmic VDR immunostaining and translocation of NF-kB into nucleus was less clear, because the lowest cytoplasmic VDR immunostaining was seen in melanomas with o10% and 50.1 -75% of cells with nuclear NF-kB ( Figure 7D).

Inhibition of melanoma growth by 20(OH)D 3 and 1,25(OH) 2 D 3 is attenuated by melanotic phenotype
To further characterise the effects of secosteroids on melanoma, we examined their effects on SKMEL-188 cell proliferation. After growth in serum-free media for 24 h, nonpigmented and pigmented melanoma cells were treated with varying concentrations of 20(OH)D 3 or 1,25(OH) 2 D 3 , and cell proliferation rate was determined by thymidine incorporation. As shown in Figure 9, 20(OH)D 3 and 1,25(OH) 2 D 3 exhibited a statistically significant inhibitory effect on SKMEL-188 proliferation at 24 h (Po0.001; Figure 9A) and 48 h (Po0.05) after secosteroid addition ( Figure 9B). At the 100 nM concentration, the inhibitory effect on proliferation was significantly greater. Both 20(OH)D 3 and 1,25(OH) 2 D 3 had a greater inhibitory effect on the proliferation of nonpigmented melanoma cells as compared with pigmented cells (Figure 9C and D).
Using SKMEL-188 stably expressing VDR and GFP fusion protein , we also examined the effect of vitamin D 3 hydroxyderivatives on VDR translocation from cytoplasm to nucleus ( Figure 10). As shown in Figure 10, there is a dose-dependent ligand-induced translocation of VDR to the nucleus, with a similar effect for both 20(OH)D 3 (100 nM) and 1,25(OH) 2 D 3 at a concentration of 100 nM (Figure 10, insert). These results are consistent with the previously reported effect for 20-hydroxyvitamin D2 effect .

DISCUSSION
Malignant melanoma is a highly aggressive form of skin cancer that is extremely difficult to treat because it is highly resistant to conventional chemotherapeutic agents and radiation treatment (Soengas and Lowe, 2003). Nuclear factor-kB is constitutively activated in malignant melanoma and other cancers, and is Abbreviation: NF-kB ¼ nuclear factor-kB. a Data presented as mean value ± s.d. of percentage of cells with nuclear NF-kB localisation. b Data presented as number of cases (%) with NF-kB localised in nuclei/number of total cases: a (0 -10%), b (10 -25%), c (25 -50%), d (50 -75%), and e (75 -100%).

Amelanotic melanoma Moderately pigmented melanoma
Strongly pigmented melanoma Negative control Positive control Figure 6 The expression of Ki67 in pigmented and nonpigmented melanoma tissue. Sections of human melanomas were subjected to immunohistochemical staining for Ki67, visualised with DAB and counterstained with haematoxylin. Arrows indicate Ki67-positive cells, and asterisks indicate melanin. Lymph node was used as positive control. Scale bars ¼ 50 mm.
believed to confer resistance to cytotoxic therapies by suppressing the induction of apoptosis (Franco et al, 2001;Wang and Richmond, 2001;Yang and Richmond, 2001;Karin et al, 2002). Many mechanisms are responsible for the elevated level of NF-kB activity in malignant melanoma (Lopez-Bergami et al, 2008). In melanoma cells, the NF-kB pathway can be altered by upregulation of the p50 and p65 NF-kB subunits (Meyskens et al, 1999;McNulty et al, 2001) and downregulation of the NF-kB inhibitor, IkB (Yang  Dhawan and Richmond, 2002). The ability to synthesise melanin is an example of differentiation programme in normal and malignant melanocytes with usually decreased proliferative potential (Slominski et al, 2004). Using a human melanoma line in which melanin pigmentation can be induced by media with increased concentration of melanin precursors , we demonstrate that pigmentation affects NF-kB activity (Figure 1). The p65/p50 heterodimers of NF-kB are the predominant complex observed in activated cells and play an important role in regulating gene transcription after nuclear translocation. As determined by various assays, we found high nuclear levels of the p65 NF-kB subunit in nonpigmented cells, whereas p65 was localised in the cytoplasm of pigmented cells.
The high constitutive activity of NF-kB in nonpigmented melanoma cells is consistent with a role for NF-kB in the malignant characteristics of human melanoma (Yang and Richmond, 2001). Therefore, we next examined the status of NF-kB activation by immunohistochemical analysis of archival clinical material from 79 cases of melanomas. Consistent with our observations from the cell culture model of inducible melanogenesis, a predominant nuclear location for p65 was observed in specimens comprising amelanotic cells, whereas predominantly cytoplasmic location of p65 was observed in specimens containing melanotic cells. Furthermore, the translocation of NFkB into the cell nuclei was accompanied by higher proliferation activity as assessed by Ki67 immunostaining and mitotic index.
Melanin pigment not only protects melanocytes against the harmful effects of solar radiation (Slominski et al, 2004;Lin and Fisher, 2007), but can also increase the resistance of melanoma cells to various forms of anticancer therapy (Meyskens et al, 2007;Brozyna et al, 2008;Slominski et al, 2009). Vitamin D, an important nutritional factor, protects against melanomagenesis (Berwick et al, 2005;Egan, 2009;Newton-Bishop et al, 2009). In addition, 20(OH)D 3 , a novel product of vitamin D3 metabolism by CYP11A1 (Slominski et al, 2005), has strong antiproliferative activity in keratinocytes (Zbytek et al, 2008) and anticancer activity in leukaemia lines . The biological activities of novel 20(OH)D 3 and classical 1,25(OH) 2 D 3 are mediated by binding to the VDR (Deeb et al, 2007;Zbytek et al, 2008), which is ubiquitously expressed in cells. Therefore, we compared the sensitivity of nonpigmented and pigmented melanoma cells with these forms of vitamin D. First, we have shown that treatment with vitamin D3 derivatives causes VDR translocation to the nucleus. Second, both 20(OH)D 3 and 1,25(OH) 2 D 3 exhibited a concentration-dependent inhibition of melanoma cell proliferation, with nonpigmented cells being more sensitive to 20(OH)D 3 and 1,25(OH) 2 D 3 treatment than pigmented cells. The most logical explanation for these differences is reduced VDR expression in pigmented cells compared with nonpigmented cells (see Figure 1), a finding further confirmed by analysis of clinical material (see Figure 8). The NF-kB pathway is a promising anticancer target. In melanoma, gene transfer approaches have been used to inhibit the NF-kB pathway by inactivating RelA (McNulty et al, 2001) or by overexpressing IkB (Huang et al, 2000). Interestingly, we demonstrated that NF-kB activity was inhibited by 1,25(OH) 2 D 3 (Harant et al, 1997) and 20(OH)D 3 (Janjetovic et al, 2009). The latter effect was accompanied by decreased keratinocyte cell proliferation (Zbytek et al, 2008). In this study, using melanoma models, we found that 20(OH)D 3 selectively inhibited NF-kB activity in nonpigmented human melanoma cells, having little effect on melanised cells. A similar effect was observed with the classical vitamin D derivative 1,25(OH) 2 D 3 . Accordingly, this places a limitation on the use of this pathway in adjuvant treatment of pigmented melanomas. However, treatment of nonpigmented cells with the secosteroids (1,25(OH) 2 D 3 and 20(OH)D 3 ) resulted in a time-dependent inhibition of NF-kB activity that correlated with suppression of NF-kB activation, as measured by ELISA and NF-kB-dependent transcription reporter assays. Furthermore, secosteroid treatment of nonpigmented cells blocked the constitutive nuclear translocation of p65 subunit of NF-kB and resulted in p65 retention in the cytoplasm. Thus, downregulation of NF-kB represents a novel approach to treat amelanotic or poorly pigmented melanomas. In this context, application of nutritional factors (vitamin D) or its nontoxic 20(OH)D 3 derivative, which can be classified as a natural product (it is generated by an action of an enzyme or by isolated mitochondria; Slominski et al, 2005), can represent a new adjuvant therapy. This approach is also justified by recent clinical and epidemiological studies in melanoma, showing vitamin D levels to be associated with lower melanoma incidence (Egan, 2009;Newton-Bishop et al, 2009;Pinczewski and Slominski, 2010), and histopathological analyses showing that progression of melanoma is linked to reduction of VDR expression and lack of VDR expression is accompanied by shorter overall survival time (Brozyna et al, 2011).
In summary, this study highlights the importance of the NF-kB pathway in the progression and growth of melanoma cells.
Our data indicate that this pathway is a potential target for the treatment of malignant melanoma with vitamin D3 prohormone or its active hydroxyderivatives. However, the final outcome will depend on the pigmentary phenotype. Importantly, the noncalcemic vitamin D3 derivative, 20(OH)D 3 , can target NF-kB and regulate melanoma progression. We also show that the high melanin content in melanomas is associated with melanoma resistance to treatment by vitamin D3 derivatives.