Abstract
Background/Aim: Vitamin D receptor (VDR), activated upon binding of 1,25(OH)2D3, was described as a tumor suppressor in the skin. New biological functions of non-classical vitamin D derivatives were recently identified, that are mediated via binding to alternate receptors, including the aryl hydrocarbon receptor (AHR) and that indicate functional interaction between AHR and VDR signaling in various human tissues. We aimed to gain further insights into the cross-talk of VDR and AHR signaling in skin photo-carcinogenesis. Materials and Methods: Using real-time quantitative PCR, we analyzed in vitro effects of the complete carcinogen UVB and of 1,25(OH)2D3 on the expression of members of the AHR and VDR pathways in human keratinocytes revealing characteristics of different stages of skin photo-carcinogenesis. Results: In precancerous HaCaT keratinocytes, induction of a target gene of AHR-mediated transcription (CYP1A1) was markedly stronger after treatment with UVB, as compared to treatment with 1,25(OH)2D3. In contrast, in SCL-1 cells (that reveal the complete phenotype of malignant transformation), expression of CYP1A1 was higher after treatment with 1,25(OH)2D3 as compared to treatment with UVB. The classical VDR target CYP24A1 was up-regulated by 1,25(OH)2D3, but not by UVB, in both cell lines. However, the combined treatment with UVB strongly enhanced the 1,25(OH)2D3-mediated up-regulation of CYP24A1 exclusively in SCL-1, but not in HaCaT cells. Conclusion: There is a differential regulation of VDR and AHR target genes by UVB and 1,25(OH)2D3 in HaCaT and SCL-1 cells, that points to a complex and highly orchestrated network of vitamin D derivatives (and other photoproducts) and its relevance for photo-carcinogenesis.
- Squamous cell carcinoma
- keratinocytes
- 1,25-dihydroxyvitamin D3
- Ultraviolet-B
- 25-hydroxyvitamin D3
- CH223191
- aryl hydrocarbon receptor
- AHR
- cytochrome P450 1A1
- cyclooxygenase-2
- vitamin D receptor
- VDR
- CYP24A1
As the frontier of the human body to the environment, the human skin represents an important defense line against many different hazards, including infections, intoxications, and exposure to UV- and other types of radiation. It is well known that ultraviolet B radiation (UVB; wavelength range: 290-320 nm), found in solar radiation, is a potentially toxic and carcinogenic environmental factor. Whereas acute effects of skin exposure to UVB radiation are dose-dependent and include sunburns or immune modulation (1), long-term exposure is known to be a main risk factor for developing non-melanoma skin cancer including cutaneous squamous cell carcinoma (cSCC) (2, 3) and its precancerous skin lesion actinic keratosis (AK) (4, 5). The UVB-induced stress response in the human skin is called UV response (6-8). An important mechanism involved in this process is the activation of the Aryl hydrocarbon receptor (AHR) (9). This ligand-dependent receptor belongs to the basic helix-loop-helix/Per-Arnt-Sim (bHLH/PAS) family (10) and is located in the cell cytoplasm, bound to a chaperone complex (Hsp90/XAP2/p23) (11, 12). It is known for its role in the detoxification of harmful substances like 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) (12), polycyclic aromatic hydrocarbons (PAHs) (13-16), and natural flavonoids (17, 18) and is highly expressed in barrier organs like the skin (19, 20). The AHR activation caused by UVB activates two signaling pathways, one located in the nucleus and one in the membrane of the cell. In the nuclear pathway, UVB radiation triggers AHR translocation into the nucleus and causes induction of cytochrome P450 1A1 (CYP1A1) gene expression in a ligand-dependent manner (9). The xenobiotic metabolizing enzyme CYP1A1 is one of the most widely studied AHR target genes (14, 21) and is linked to various cutaneous immunologic processes (19), chemical carcinogenesis (22-24), and the development of non-melanoma skin cancer (25-27). Moreover, the DNA is able to absorb UVB, which results in formation of DNA photoproducts (28-32). In the cell membrane pathway, AHR activation causes internalization of the EGFR and activation of its downstream target MAP kinases ERK1 and ERK2. As a result, cyclooxygenase-2 (COX2) mRNA expression is up-regulated (9), which is linked with photocarcinogenesis and skin inflammation (33-39).
The hormonally active form of vitamin D3 [1,25(OH)2D3; calcitriol] is produced in keratinocytes of the epidermis in a UVB-dependent, 3-step process (40). It modulates important physiological and pharmacological processes like immunomodulation and bone metabolism (41, 42). In the skin specifically, it plays a vital role in the epidermal proliferation and differentiation (43, 44), apoptosis (45, 46), and barrier function (47-49). It also exhibits a protective effect against inflammatory skin diseases including psoriasis (50) and atopic dermatitis (51) and inhibits the growth of skin cancer such as melanoma (52, 53), basal cell carcinoma (BCC) (54, 55), and squamous cell carcinoma (SCC) (56). 1,25(OH)2D3 unfolds its biological function through binding to the vitamin D receptor (VDR; NR1I1) (57), a member of the nuclear receptor superfamily of transcription factors highly expressed in keratinocytes (58, 59). The VDR regulates gene expression by forming a heterodimeric complex with retinoid X receptors (RXRs), translocating to the nucleus and interacting with vitamin D responsive elements (VDREs) in the promoter of target genes (57, 60, 61). A major target gene of the VDR is cytochrome P450 24A1 (CYP24A1) (62, 63), which encodes for the 25-hydroxyvitamin D3 24-hydroxylase. By this enzyme, inactivation of 1,25(OH)2D3 through hydroxylation and termination of its biological activity is induced (64, 65). In some forms of cancer, it has been suggested to represent an oncogene (66).
Recent scientific findings indicate a functional interaction between AHR and VDR signaling in various human tissues. In this regard, new biological functions of non-classical vitamin D derivatives have recently been identified, that are at least in part mediated via binding to alternate receptors, including the AHR. In human naïve CD4+ T-cells, a suppressive effect of 1,25(OH)2D3 on AHR expression was found (67). In human oral keratinocytes (OKF6/TERT-2 cells) however, 1,25(OH)2D3 increased LPS-induced AHR and CYP1A1 expression (68). In human epidermal keratinocytes (HEKn and HaCaT cells), AHR was the major receptor target for vitamin D derivative 20,23(OH)2D3, (with VDR being the second signaling pathway identified) whereas weaker AHR activation was observed by 1,25(OH)2D3, 20(OH)D3 and 17,20,23(OH)2D3 (69). Matsunawa et al. (70, 71) demonstrated that combined activation of AHR and VDR enhanced CYP1A1 and CYP24A1 expression in breast cancer (MCF-7) and macrophage-derived (THP-1) cells. In the present study, we aimed to gain further insights into the cross-talk of 1,25(OH)2D3-induced VDR and UVB-induced AHR signaling in skin photo-carcinogenesis. We here found a differential regulation of VDR and AHR target genes by UVB and 1,25(OH)2D3 in precancerous HaCaT and malignant SCL-1 cells, which points to a complex and highly orchestrated network of vitamin D derivatives (and other photoproducts) and its relevance for photo-carcinogenesis.
Materials and Methods
Cell culture. Spontaneous immortalized human HaCaT (Human adult low Calcium, high Temperature) keratinocytes were purchased from CLS Cell Lines Service GmbH (Eppelheim, Germany) and cultivated in Dulbecco’s modified eagle’s medium (DMEM) (Gibco, Thermo Fisher Scientific, Dreieich, Germany). They exhibit a p53 mutation (p53mut) and represent initiated keratinocytes that express elements of early stage non-melanoma skin carcinogenesis (72, 73). In vivo animal studies have shown that HaCaT cells exhibit characteristic features of precancerous skin lesions (e.g., AK), including development of a stratified epithelium with dysplastic morphologic properties but no tendency to invasive or metastatic growth (74). SCC cell lines (SCL-1) were maintained in RPMI 1640 medium (Gibco). They represent malignant human keratinocytes that lack expression of the p53 protein (p53null) (75-79) and exhibit characteristic features of the non-melanoma skin cancer phenotype, including invasive and metastatic growth tendency (80). Both cell lines were supplemented with 1% L-glutamine (Thermo Fisher Scientific) and 10% fetal calf serum (Gibco). They were seeded in culture dishes (10 cm in diameter) and grown in a humidified atmosphere of 5% CO2 at 37°C. Cell culture medium was changed twice a week.
UVB irradiation. After the culture medium has been aspirated from the cell culture dishes, cells were washed with phosphate-buffered saline (PBS) and irradiated with UVB (50 J/cm2, midrange wavelength 302 nm) using Crosslinker CL-1000M (Ultra-violet products Ltd, purchased by Analytik Jena, Jena, Germany). Following irradiation, cells were provided with fresh medium and treatment substances.
Cell treatment. Cells were treated with 1,25(OH)2D3 (Sigma, Taufkirchen, Germany) in a final concentration of 10−7 M [5 μl of the 1,25(OH)2D3 of (10−4 M) stock solution solved in ethanol (EtOH) were added per 5 ml medium per culture dish]. Bovine serum albumin (BSA, 1%) was added to the medium when treating cells with 1,25(OH)2D3 to reduce unspecific binding of 1,25(OH)2D3 to the culture dish. Control samples were treated with EtOH (5 μl EtOH per 5 ml medium per culture dish) and BSA (1%, Sigma). In preliminary experiments, we first demonstrated that EtOH had no effect on gene expression, because cells treated with BSA alone showed similar results as compared to cells treated with BSA and EtOH. Cells were treated with AHR-Antagonist CH223191 (stock solution 10−4 M, solved in EtOH; final concentration 10−7 M, Sigma) and partly VDR-Inhibitor 25-Hydroxyvitamin D3 (25(OH)2D3, stock solution 10−4 M, solved in EtOH; final concentration 10−7 M, Sigma). Previous studies (81-83) confirmed that CH223191 and 25(OH)D3 in a final concentration of 10−7 M effectively block their corresponding receptors, AHR and VDR, respectively.
Cell harvesting. HaCaT-keratinocytes and SCL-1 cells were harvested (6 h intervals over 24 h) after irradiation and/or substance treatment.
RNA isolation. RNA isolation was performed with RNeasy Kit and QIA shredder (Qiagen, Hilden, Germany) according to the manufacturers’ manual.
Reverse transcription. Reverse transcription was performed with Omniscript RT Kit (Qiagen) according to the manufacturers’ instructions. Oligo-dT-primers, RNase inhibitors, and 1 μg mRNA were used in every reaction as templates.
Quantitative real-time PCR (RTqPCR) and analysis. Expression of the target genes AHR, CYP1A1, COX2, VDR, CYP24A1 was examined in 96-well plates using RTqPCR (120 cycles in StepOnePlus Real-Time PCR System, Thermo Fisher Scientific). The level of expression of each target gene was normalized against the mean of GAPDH and β-actin gene expression and shown as mean±standard deviation. Each sample was analyzed in duplicate. All gene-specific primers were purchased from Qiagen (Table I). The relative quantification method (RQ=2−ΔΔCt) was used in order to calculate the relative fold gene expression of the target genes (84). First, the relative amount of the target gene to each reference gene was determined for each sample (∆Ct). Then, the target/reference ratio of the treated sample was divided by the target/reference ratio of the control sample (∆∆Ct). To find out the N-fold target gene expression in treated samples relative to the control sample (final values), we calculated 2 to the power of the negative ∆∆Ct (2−∆∆Ct).
Gene-specific primers used in RTqPCR.
Statistical analysis. All data are represented as a mean±standard deviation (SD) of three experiments per cell line. The two-tailed, unpaired Student’s t-test was used to assess statistical significance and performed with the Microsoft Excel software (Microsoft Corporation, Redmond, WA, USA). Mean differences were considered to be significant when p≤0.05 (*), decisive (very significant) when p<0.005 (**) and conclusive (extremely significant) when p<0.0005 (***).
Results
Gene expression of AHR and CYP1A1 is elevated in untreated HaCaT and of COX2, VDR and CYP24A1 in SCL-1 cells. In untreated spontaneously immortalized HaCaT keratinocytes we observed higher AHR (p<0.0005) and CYP1A1 mRNA levels as compared to those in untreated SCL-1 cells (Figure 1A-D). In contrast, in untreated cancerous SCL-1 cells, mRNA expression of COX2, VDR, and CYP24A1 was higher than HaCaT (Figure 1E-J).
Relative mRNA expression (A, C, E, G, I) and AUC (B, D, F, H, J) of AHR, CYP1A1, COX2, VDR, and CYP24A1 relative to the mean of GAPDH and β-actin in untreated HaCaT and SCL-1 cells (mean 2−ΔΔCt). Cells were treated only with culture medium and harvested in 6 h intervals over 24 h. The mRNA expression was measured with RTqPCR and the relative fold gene expression was calculated with the 2−∆∆Ct method. HaCaT-cells harvested after 0 h were used as the internal control sample in the bar graphs (A, C, E, G, I). The “area under the curve” was calculated for each cell line from the respective time curve (data not shown). The values represent the means±SD of duplicate assays. The experiments were repeated thrice with similar results. *p≤0.05; **p<0.005; ***p<0.0005.
1,25(OH)2D3 and UVB radiation exert differential effects on the expression of key elements of the AHR signaling pathway in HaCaT-keratinocytes and SCL-1 cells. In non-malignant HaCaT keratinocytes, UVB radiation induced a strong increase in CYP1A1 mRNA (7.7-fold increase), that was markedly stronger as compared to treatment with 1,25(OH)2D3 (2.8-fold increase) (Figure 2C). Combined treatment increased CYP1A1 mRNA stronger than treatment with 1,25(OH)2D3 alone (3.7-fold increase, p≤0.05). In contrast, in malignant SCL-1 cells, expression of CYP1A1 was markedly higher after treatment with 1,25(OH)2D3 (6.9-fold increase, p<0.0005) as compared to treatment with UVB (2.4-fold increase, p<0.005) (Figure 2D). Combined treatment showed a synergistic effect, by conclusively up-regulating CYP1A1 mRNA to the highest extent (9.8-fold increase) (Figure 2D). UVB radiation up-regulated AHR mRNA in both cell lines (HaCaT: 2.2-fold increase; SCL-1: 1.6-fold increase). In contrast, lower levels of AHR expression were observed after treatment with 1,25(OH)2D3 (HaCaT: 17% decrease; SCL-1: 26% decrease). Combined treatment did not show any regulatory effect as compared to untreated controls (Figure 2A and B).
Relative mRNA expression (AUC) of AHR (A, B), CYP1A1 (C, D), COX2 (E, F), VDR (G, H), and CYP24A1 (I, J) relative to the mean of GAPDH and β-actin in treated HaCaT and SCL-1 cells (mean 2−∆∆Ct). After treatment, cells were harvested in 6 h intervals over 24 h. The mRNA expression was measured with RTqPCR and the relative fold gene expression was calculated with the 2−∆∆Ct method. Cells harvested after 0 h were used as the internal control sample for each time curve (data not shown). The “area under the curve” of every treatment condition was measured and set relative to the “area under the curve” of the EtOH-treated cells (vehicle). The values represent the means±SD of duplicate assays. The experiments were repeated thrice with similar results. All p-values are relative to cells treated with EtOH alone. *p≤0.05; **p<0.005; ***p<0.0005.
Induction of CYP1A1 mRNA expression by 1,25(OH)2D3 depends on AHR. Treatment with AHR-antagonist CH223191 alone or in combination with 1,25(OH)2D3 strongly suppressed expression of CYP1A1 mRNA in HaCaT and SCL-1 cells (HaCaT after 6 h: CH223191: 94% decrease, p<0.0005, CH223191+1,25(OH)2D3: 84% decrease, p<0.0005; HaCaT after 24 h: CH223191: 95% decrease, p<0.0005, CH223191+1,25(OH)2D3: 85% decrease, p<0.0005; SCL-1 after 6h: CH223191: 82% decrease, p<0.0005, CH223191+1,25(OH)2D3: 62% decrease, p<0.0005; SCL-1 after 24 h: CH223191: 88% decrease, p<0.0005, CH223191+1,25(OH)2D3: 74% decrease, p<0.0005) (Figure 3). In 25(OH)D3-treated cells, CYP1A1 expression was reduced (HaCaT after 6 h: 56% decrease, p<0.0005; HaCaT after 24 h: 19% decrease; SCL-1 after 6 h: 65% decrease, p<0.0005; SCL-1 after 24 h: 36% decrease, p<0.05). However, combined treatment with 1,25(OH)2D3 and 25(OH)D3 had a stronger induction in CYP1A1 mRNA compared to 1,25(OH)2D3 alone (except in SCL-1 after 24 h) (HaCaT after 6 h: 1,25(OH)2D3+25(OH)D3: 2.6-fold increase, p<0.005, 1,25(OH)2D3: 2.1-fold increase, p<0.0005; HaCaT after 24 h: 1,25(OH)2D3+25(OH)D3: 1.23-fold increase, 1,25(OH)2D3: 1.17-fold increase; SCL-1 after 6 h: 1,25(OH)2D3+25(OH)D3: 2.5-fold increase, p<0.0005, 1,25(OH)2D3: 2.4-fold increase, p<0.0005).
Relative mRNA expression of CYP1A1 in HaCaT and SCL-1 cells 6 h (A, B) and 24 h (C, D) after treatment relative to the mean of GAPDH and β-actin (mean 2−ΔΔCt). Cells were harvested 6 and 24 h after treatment. The mRNA expression was measured with RTqPCR and the relative fold gene expression was calculated with the 2−∆∆Ct method. Cells treated with solvent vehicle alone (EtOH) were used as the internal control. The values represent the means±SD of duplicate assays. The experiments were repeated thrice with similar results. All p-values are relative to cells treated with EtOH alone. *p≤0.05; **p<0.005; ***p<0.0005.
1,25(OH)2D3 induces COX2 mRNA in HaCaT but not in SCL-1 cells. Treatment with 1,25(OH)2D3 strongly increased COX2 mRNA expression in HaCaT cells (16.2-fold increase, p≤0.05), but barely altered it in SCL-1 cells (1.8-fold increase, p≤0.05) (Figure 2E and 2F). In both cell lines however, after treatment with 1,25(OH)2D3 and 25(OH)D3, up-regulation of COX2 gene expression was even stronger as compared to treatment with 1,25(OH)2D3 alone (HaCaT after 6 h: 1,25(OH)2D3+25(OH)D3: 4.2-fold increase, p<0.0005, 1,25(OH)2D3: 2.8-fold increase, p<0.0005; HaCaT after 24 h: 1,25(OH)2D3+25(OH)D3: 2.4-fold increase, p<0.0005, 1,25(OH)2D3: 2.3-fold increase, p<0.005; SCL-1 after 6 h: 1,25(OH)2D3+25(OH)D3: 1.4-fold increase, p<0.0005, 1,25(OH)2D3: 1.1-fold increase; SCL-1 after 24 h: 1,25(OH)2D3+25(OH)D3: 2.1-fold increase, p<0.0005, 1,25(OH)2D3: 1.4-fold increase, p<0.0005) (Figure 4). Treatment with UVB up-regulated COX2 gene expression in both cell lines to a similar degree (HaCaT: 5.4-fold increase; SCL-1: 3.9-fold increase, p≤0.05). Combination treatment with UVB and 1,25(OH)2D3 exerted a synergistic effect only in HaCaT (38.3-fold increase, p<0.05) and not in SCL-1 cells (4.2-fold increase, p≤0.05) (Figure 2E and F).
Relative mRNA expression of COX2 in HaCaT and SCL-1 cells 6 h (A, B) and 24 h (C, D) after treatment relative to the mean of GAPDH and β-actin (mean 2−ΔΔCt). Cells were harvested 6 and 24 h after treatment. The mRNA expression was measured with RTqPCR and the relative fold gene expression was calculated with the 2−ΔΔCt method. Cells treated with solvent vehicle alone (EtOH) were used as the internal control. The values represent the means±SD of duplicate assays. The experiments were repeated thrice with similar results. All p-values are relative to cells treated with EtOH alone. *p≤0.05; **p<0.005; ***p<0.0005.
1,25(OH)2D3, but not UVB, induces CYP24A1 mRNA in HaCaT and SCL-1 cells. CYP24A1 mRNA expression was increased in HaCaT and SCL-1 cells, after treatment with 1,25(OH)2D3, but not after treatment with UVB (Figure 2I and J). Co-treatment with 1,25(OH)2D3 and 25(OH)D3 induced CYP24A1 mRNA in SCL-1 cells even stronger than 1,25(OH)2D3 alone (after 6 h: 1,25(OH)2D3+25(OH)D3: 133.1-fold increase, p<0.0005, 1,25(OH)2D3: 82.7-fold increase, p<0.0005; after 24 h: 1,25(OH)2D3+25(OH)D3: 123.5-fold increase, p<0.0005, 1,25(OH)2D3: 94.9-fold increase, p<0.0005) (Figure 5). In contrast, expression of VDR was only altered in SCL-1 cells treated with 1,25(OH)2D3 (38% decrease, p<0.005), while other treatments with 1,25(OH)2D3 and/or UVB had only marginal effects in HaCaT or SCL-1 (Figure 2G and H). Combined treatment with UVB further enhanced the 1,25(OH)2D3-induced increase in CYP24A1 mRNA exclusively in SCL-1 [1,25(OH)2D3: 44,703.5-fold increase, p<0.0005, 1,25(OH)2D3+UVB: 119,233.4-fold increase, p<0.0005], but not in HaCaT cells [1,25(OH)2D3: 6,233,471-fold increase, p<0.05, 1,25(OH)2D3 +UVB: 5,127,778.3-fold increase, p<0.005].
Relative mRNA expression of CYP24A1 in HaCaT and SCL-1 cells 6 h (A, B) and 24 h (C,D) after treatment relative to the mean of GAPDH and β-actin (mean 2−ΔΔCt). Cells were harvested 6 and 24 h after treatment. The mRNA expression was measured with RTqPCR and the relative fold gene expression was calculated with the 2−ΔΔCt method. Cells treated with solvent vehicle alone (EtOH) were used as the internal control. The values represent the means±SD of duplicate assays. The experiments were repeated thrice with similar results. All p-values are relative to cells treated with EtOH alone. *p≤0.05; **p<0.005; ***p<0.0005.
Discussion
During the last decades, a continuously growing body of evidence has convincingly shown an important role of vitamin D in carcinogenesis and the progression of many malignancies (85-87). It can be speculated that during the next years, these new scientific findings in the vitamin D field, which include the identification of AHR, RORs, and LXR as alternative receptors for vitamin D compounds (69, 88, 89), will have a great impact on the prevention and therapy of cancer. It was the aim of this study to understand the role of the vitamin D endocrine system in the multistep process of skin photo-carcinogenesis (90), that shows characteristic early (e.g., initiated cells) and late (e.g., cells that express the complete malignant phenotype) stages (72-79). In particular, we investigated the molecular interaction of two different nuclear receptor pathways for vitamin D, which are activated either by binding of the classical biologically active vitamin D metabolite, 1,25(OH)2D3, to the VDR, or by binding of non-classical vitamin D hydroxyderivatives [e.g., 20,23(OH)2D3] to the AHR.
By analyzing the expression of AHR, CYP1A1, and COX2 as well as of VDR and CYP24A1, we showed that the expression of genes encoding for key elements of both VDR and AHR pathways are differentially expressed and regulated during different stages of skin carcinogenesis. For example, expression of AHR and CYP1A1 was much stronger in untreated HaCaT as compared to untreated SCL-1 cells, while in contrast, expression of VDR, CYP24A1, and COX2 was stronger in untreated SCL-1 as compared to HaCaT cells. It remains to be investigated in future studies, whether stage-dependent differences in the expression of key elements of these different nuclear signaling pathways for vitamin D compounds contribute to the carcinogenesis of non-melanoma skin cancer. It may be speculated that these findings are caused by functional changes associated to the p53 status in HaCaT (p53 mutation, p53mut) and SCL-1 (no p53 protein present, p53null) cells, as previous studies have reported a p53-mediated tissue-dependent regulation of AHR (91-93) and VDR (94, 95) signaling. Moreover, it can be speculated whether low basal levels of AHR and CYP1A1 in SCL-1, and of VDR, CYP24A1, and COX2 in HaCaT cells may point at a functional defect of AHR signaling in SCL-1 and of VDR signaling in HaCaT cells.
To further investigate the interaction between AHR and VDR signaling, we treated cells with UVB, and/or the VDR-ligand 1,25(OH)2D3, its precursor 25(OH)D3 [that has been described as a partial VDR-antagonist (83)], and the AHR-antagonist CH223191 (81, 82). It has to be noted that we did not succeed in obtaining the non-classical hydroxyderivatives [e.g., 20,23(OH)2D3] of vitamin D that were recently described as AHR-ligands.
This study also examined whether effects of UVB on expression of AHR target genes may be mediated via the UVB-induced cutaneous synthesis of 1,25(OH)2D3 or 25(OH)D3, indicating that oral supplementation with vitamin D could compensate for the effects of UVB both on AHR and VDR signaling pathways. Until now, only a few studies have analyzed the effects of 1,25(OH)2D3 on the expression of AHR target genes; however, studies in cutaneous SCC cells are lacking (68, 69). We here show that the complete carcinogen UVB and the anti-carcinogenic agent 1,25(OH)2D3 exert different effects on the expression of key elements of the VDR and AHR pathways. Although the results of our investigation do not allow definite conclusions, these findings do not support the assumption that effects of UVB on CYP1A1 expression are mediated via UVB-induced cutaneous production of 1,25(OH)2D3.
We showed that the expression of genes encoding for proteins that contribute to AHR signaling is regulated differentially by UVB in HaCaT and SCL-1 cells, representing keratinocytes that reveal phenotype characteristics for early and late stages of skin carcinogenesis, respectively. CYP1A1 mRNA was regulated differentially by 1,25(OH)2D3 and UVB in HaCaT and SCL-1 cells. In SCL-1, induction of CYP1A1 mRNA was stronger after treatment with 1,25(OH)2D3 (6.9-fold induction compared to control) as compared to treatment with UVB (2.4-fold induction compared to control); however, opposite effects were seen in HaCaT cells. In these cells, induction of CYP1A1 mRNA was stronger after treatment with UVB (7.7-fold induction compared to control) as compared to that after treatment with 1,25(OH)2D3 (2.8-fold induction compared to control).
It has been reported that CYP1A1 induction can be mediated via several independent mechanisms that include elevation of intracellular calcium and subsequent cell differentiation (96-98), or involve other nuclear receptors (NR). It is well established that 1,25(OH)2D3 plays a crucial role in calcium homeostasis (99-101) and promotes differentiation in cultured skin cells (102) and cancer cells like human colon cancer, CAFs, and CSCs (103). Thus, 1,25(OH)2D3 could regulate the CYP1A1 mRNA activity through its pro-differentiating effect. Interestingly, combination treatment with 1,25(OH)2D3 and UVB increased CYP1A1 mRNA activity even further. In vivo studies report, that even minimal UVB radiation levels of 18 mJ/cm2 are enough to activate 1,25(OH)2D3 synthesis in the skin (104). An enhancing effect of UVB radiation on 1,25(OH)2D3-induced CYP1A1 mRNA expression through additional endogenous 1,25(OH)2D3 production could therefore be considered. However, our results do not exclude the possibility that induction of CYP1A1 may be induced AHR independently via other mechanisms that may include the activation and involvement of other NR pathways. The Pregnane X receptor (PXR), a member of the nuclear hormone receptor family and regulator of xenobiotic and drug metabolism (105-107), that was linked to the development of SCC (108), was found to have similarities to both the AHR (109, 110) and VDR (111, 112), and was shown to regulate expression of CYP1A1 (113).
Interestingly, Wilkens et al. (114) were able to demonstrate that intravenous administration of 1,25(OH)2D3 up-regulates the mRNA expression of PXR in sheep renal tissue. Other CYP1A1-regulating NRs (115) including the Glucocorticoid receptor (GR), Estrogen receptor (ER) and Retinoid acid receptor (RAR) have also shown interactions with 1,25(OH)2D3 (69, 116, 117).
In agreement with previously published reports (82), exposure of HaCaT cells to a single dose UVB (50 J/cm2) induced expression of AHR and CYP1A1 mRNA, indicating that UVB-induced regulation of AHR signaling functions correctly in these cells.
1,25(OH)2D3 in a dose of 10−7 M had no conspicuous regulating effect on AHR expression in HaCaT and SCL-1 cells. These findings are in agreement with the results of Slominski et al. (69), who reported in epidermal keratinocytes a dose-dependent, 1,25(OH)2D3-induced AHR expression, that was detected after treatment with 1,25(OH)2D3 at a dose of 10−6 M, but not at a dose of 10−7 M.
In SCL-1 cells, UVB (50 J/cm2) induced expression of AHR and CYP1A1. CYP1A1 mRNA induction after treatment with 1,25(OH)2D3 was almost 7 times higher than that of the control group and 2.5 times stronger than that after UVB treatment. Until now, only a few studies have reported an effect of 1,25(OH)2D3 on the AHR target gene; however, none of them was carried out in cutaneous SCC (68, 69).
To analyze whether the induction of CYP1A1 mRNA in HaCaT and SCL-1 cells was AHR-dependent, we used CH223191 (AHR-antagonist). Treatment with CH223191 suppressed CYP1A1 mRNA expression both in HaCaT and SCL-1 cells under all experimental conditions.
It is well established that 1,25(OH)2D3 exerts its cancer-inhibiting activity in many cell types through various direct (e.g., regulation of the cell cycle, induction of apoptosis, inhibition of angiogenesis and tumor-invasiveness, - metastasis and -proliferation) and indirect (e.g., regulation of immuno-modulation, effect on tumor microenvironment) mechanisms (95, 118). Although AHR-activated CYP1A1 is associated with pro-carcinogen transformation and cancer development, some studies documented a contribution to cancer prevention (27). Therefore, key elements of AHR signaling may at least in part contribute to 1,25(OH)2D3-mediated anti-cancer mechanisms.
Notably, it has been shown that 1,25(OH)2D3 modulates the cell cycle through checkpoint regulation (118). Binding to the promoter region of genes encoding p21 and p27 results in cyclin dependent kinase (CDK) inhibition and cell cycle arrest in the G1 phase via decreased cyclin D1 expression (95, 118). Interestingly, ligand-dependent AHR activation was also found to increase p21 and p27 expression in addition to CYP1A1, resulting in G1 phase cell cycle arrest (119, 120). Another mechanism that may be involved in 1,25(OH)2D3-induced cell cycle regulation is executed through activation of distinct molecular pathways including intracellular kinase pathways (e.g., ERK, PI3K), pathways of transforming growth factor β (TGF-β) and of insulin-like growth factor-binding proteins (IGF-BP), which are found to interact with AHR signaling (121-127). Moreover, 1,25(OH)2D3 induces protein kinase C (PKC) activation, which plays an important role in the regulation of gene expression, cell differentiation, mobility, and metastasis. The subsequently induced mitogen-activated kinases 1 and 2 (MAPK1 and MAPK2) are regulators of cell growth (95) and of transcription factors as well as co-regulatory and chromatin proteins in malignant melanoma (128). Recent studies demonstrated, that PKC activity is required for classical AHR-mediated signaling in a tissue-dependent manner (129). MAPKs induced by TCDD were also found to be important for the induction of AHR-dependent gene transcription and CYP1A1 expression (130).
Inhibition of angiogenesis represents another anti-tumor mechanism exerted by 1,25(OH)2D3 (95). Through interaction with nuclear factor kappaB (NF-B), inhibition of Interleukin-8 (IL-8) transcription is achieved. Suppression of growth factors like vascular endothelial growth factor (VEGF) or platelet-derived growth factor (PDGF) and of hypoxia inducible factor 1 alpha (HIF1α) also seems to be an important part of this process (131). Tight interactions between NF-
B and AHR signaling have been studied in various immune cells contributing to xenobiotic metabolism and carcinogenesis. AHR has been found to modulate peptidoglycan (PGN)-induced expression of IL-8 in human sebocytes involving the NF-
B pathway (132). Another finding showed that after UVB irradiation, NF-
B preliminary suppressed CYP1A1 expression, indicating a role of NF-
B in UVB-dependent AHR signaling and potentially in a photo-protective cellular response (133).
The Hedgehog (Hh) signaling pathway, whose inappropriate activation is associated with cancer stimulation and progression (134, 135), represents another target of the cancer-inhibitory function of 1,25(OH)2D3. Different types of human cancer, including skin BCC, have been linked to deregulation of Hh signaling caused by gene mutations or uncontrolled Hh ligand production (135, 136). 1,25(OH)2D3 inhibits Hh-induced proliferation and signaling through modulation of Hh target gene GLI1 (137). Contrary to its previous described cancer-promoting properties, AHR signaling was found to inhibit the Hh pathway in vivo in medulloblastoma and was identified as a potent tumour suppressor (138). It remains to be clarified how CYP1A1 is involved in the regulation of Hh signaling and whether it participates in the execution of tumor-suppressive functions.
Indirect anti-cancer effects of 1,25(OH)2D3 mainly concern the tumor microenvironment. They include modulation of immune mediators [e.g., DNA methylation of CpG regions, production of Interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α)], cancer-associated metabolic cascades (e.g., Inhibition of estrogenic signaling through down-regulation of CYP19A1, suppression of 27-hydroxycholesterol (27HC) and CYP27A1), and homeostatic processes in surrounding tissues (e.g., down-regulation of pyruvate carboxylase, up-regulation of CYP3A4, reduction of AMP hydrolysis and adenosine production) (118). Interestingly, some of these effects may also be related to key players of AHR signaling. In human salivary cells, TNF-α significantly induced AHR along with CYP1A1 expression, whereas IL-1β did not affect AHR or CYP1A1 mRNA levels (139). Additionally, AHR inhibits ER activity in human breast cancer cells, rodent uterus, and mammary tumors (140). AHR activators PAHs and TCDD increased CYP3A4 mRNA expression in HepG2 cells (141). CYP3A4 induction by xenobiotics largely depends on PXR, which tightly regulates CYP3A4 expression (142). As there is ample evidence for an interaction between PXR and AHR (143-145), a possible role for AHR signaling in the regulation of the CYP3A4 gene is conceivable.
Additionally, cells were treated with 25(OH)D3 (partly VDR-inhibitor) to determine whether the CYP1A1 mRNA induction was VDR-dependent. Co-treatment with 25(OH)D3 induced and suppressed CYP1A1 gene expression in 1,25(OH)2D3- and in EtOH-treated cells, respectively. In conclusion, these findings indicate that 25(OH)D3 does not act as a partly VDR inhibitor, at least in HaCaT cells. As 25(OH)D3 is converted in epidermal keratinocytes to 1,25(OH)2D3 by the 1-alpha-hydroxylase (CYP27B1) (146-148), it could exert agonist activity and intensify the effect of 1,25(OH)2D3 on CYP1A1. Moreover, the mechanisms by which 25(OH)D3 suppresses CYP1A1 mRNA expression in these cells remain to be elucidated.
We here show that transcriptional activity of CYP24A1 is significantly up-regulated after 1,25(OH)2D3 treatment, while VDR mRNA expression is only marginally altered (HaCaT) or even down-regulated (SCL-1) when compared to the control group. It was reported that CYP24A1, the major metabolizing enzyme of 1,25(OH)2D3, is elevated in many human tumor tissues (149, 150) including non-melanoma skin cancer (151) and is associated with poor prognosis in various cancer types (152). Thus, CYP24A1 has been considered a possible oncogene (66). Consistent with these findings, our study demonstrated that mRNA expression of CYP24A1 was stronger in untreated SCL-1 cells, revealing the complete phenotype of malignant transformation, as compared to precancerous HaCaT cells. Increased levels of CYP24A1 mRNA could lead to rapid inactivation of 1,25(OH)2D3, resulting in abolition of its antiproliferative effects against cancer. It would therefore be conceivable, that the high transcriptional level of CYP24A1 is not due to up-regulation by the physiological 1,25(OH)2D3/VDR signaling pathway, but rather to over-expression of the gene. Identically to CYP1A1, the combination of 1,25(OH)2D3 and UVB increased CYP24A1 transcriptional activity in SCL-1 cells even more.
Notably, 1,25(OH)2D3 induced CYP24A1 expression 100 times stronger in HaCaT as compared to SCL-1 cells. The p53 status of HaCaT (p53mut) and SCL-1 (p53null) keratinocytes and the crosstalk of p53 with VDR may contribute to this finding. Under physiological conditions, the p53 protein protects cells from DNA damage by several mechanisms (e.g., apoptosis induction, cell cycle progression halting, cellular aging) (153). In several tumor types, the cancer-associated and mutated p53 (p53mut) has been found to exert new mechanisms that have been termed gain-of-function (GOF), enabling it to act at the molecular level in a similar way to 1,25(OH)2D3 (95). It has been shown that p53mut is able to interact with the VDR, modulate the expression of VDR-regulated genes and enhance the nuclear VDR translocation and accumulation (154). As these effects were even more prominent after 1,25(OH)2D3 supplementation, a mechanism leading to stronger VDR target gene expression in p53 mutated cells might be plausible. Importantly, p53mut reversed the impact of 1,25(OH)2D3 on cell death and converted it from a pro-apoptotic to an anti-apoptotic agent. Thus, p53 status may alter the biological function of 1,25(OH)2D3 in precancerous and cancerous skin cells and deregulate the anti-cancer effects of the VDR pathway. As p53 also modulates AHR target genes like CYP1A1 (93) and COX2 (155-157), it could be speculated that the AHR pathway may exert a similar effect in cooperation with 1,25(OH)2D3. However, this assumption could only be confirmed for COX2, as its mRNA expression after 1,25(OH)2D3 treatment was 9 times stronger in HaCaT cells than in SCL-1 cells.
In addition to its anti-apoptotic activity also induced by TNF-α, TNF-related apoptosis-inducing ligand (TRAIL) and Fas ligand (FasL), 1,25(OH)2D3 has been described to increase cell survival after UV damage and protect some cancer cell lines against cytotoxic drugs (154, 158). However, several other studies have shown contradictory results regarding the association between 1,25(OH)2D3 and skin tumorigenesis (95, 118), questioning the exclusivity of the anti-cancer properties of 1,25(OH)2D3 in the skin and speculating about possibly harmful and cancer-promoting effects.
In summary, we here show differential regulation of AHR- and VDR-mediated signaling in HaCaT as compared with SCL-1 cells and after treatment with UVB as compared with 1,25(OH)2D3. In conclusion, our data indicate that the complex network of AHR- and VDR-mediated signaling may contribute to the photo-carcinogenesis of non-melanoma skin cancer. Treatment of keratinocytes with UVB exerts additional biological effects in human skin cells as compared to treatment with 1,25(OH)2D3. These findings imply that oral uptake of vitamin D (e.g., by food or supplements) cannot compensate for all effects of UVB on human health, that include effects of non-classical, AHR-activating vitamin D derivatives. However, the exact mechanisms behind this are yet not fully understood. Further investigations are required to demonstrate the underlying pathophysiological relevance of our results. Advanced detection and assay methods, other malignant and non-malignant skin cell lines and CYP1A1-related signaling pathways, extended examination time points as well as multiple 1,25(OH)2D3 concentrations and UVB doses should be considered, in order to eventually open new perspectives regarding the prevention and treatment of skin cancer.
Acknowledgements
The Authors thank Prof. AG Römer of José-Carreras Center and Internal Medicine I at the University Hospital of Saarland, Campus Homburg, Homburg, Germany for his advice, guidance and technical assistance.
Footnotes
Authors’ Contributions
Christoforos Christofi: Study design, literature search, experimental implementation, data analysis, manuscript preparation; Leandros Lamnis: Study design, literature search, experimental implementation; Alexandra Stark: Experimental implementation, data analysis; Heike Palm: data analysis; Klaus Römer: data analysis; Thomas Vogt: Study design; Jörg Reichrath: Study design, literature search, manuscript preparation.
Presented at the Joint International Symposium “Vitamin D in Prevention and Therapy and Biologic Effects of Light”, May 4-6, 2022, Homburg/Saar, Germany.
Conflicts of Interest
Prof. Reichrath is member of the Arnold Rikli-Award Jury of the Jörg Wolff Foundation. The Saarland University, together with Prof. Reichrath as one of several responsible group leaders, has received a research grant from the Jörg Wolff Foundation, Stuttgart, Germany.
- Received June 10, 2022.
- Revision received July 5, 2022.
- Accepted July 6, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).