Antitumoral Effects of Calcitriol in Basal Cell Carcinomas Involve Inhibition of Hedgehog Signaling and Induction of Vitamin D Receptor Signaling and Differentiation

Published OnlineFirst August 30, 2011; DOI: 10.1158/1535-7163.MCT-11-0422

Molecular Cancer Molecular Medicine in Practice Therapeutics

Anja Uhmann1, Hannah Niemann1,Ber enice Lammering1, Cornelia Henkel1, Ina Heß1, Frauke Nitzki1, Anne Fritsch1, Nicole Prufer€ 1, Albert Rosenberger2, Christian Dullin3, Anke Schraepler7, Julia Reifenberger8, Stefan Schweyer4, Torsten Pietsch9, Frank Strutz5, Walter Schulz-Schaeffer6, and Heidi Hahn1

Abstract Activation of the Hedgehog (Hh)-signaling pathway due to deficiency in the Hh receptor Patched1 (Ptch) is the pivotal defect leading to formation of basal cell carcinoma (BCC). Recent reports provided evidence of

Ptch-dependent secretion of vitamin D3-related compound, which functions as an endogenous inhibitor of Hh signaling by repressing the activity of the signal transduction partner of Ptch, Smoothened (Smo). This suggests that Ptch-deficient tumor cells are devoid of this substance, which in turn results in activation of Hh-

signaling. Here, we show that the application of the physiologically active form of vitamin D3, calcitriol, inhibits proliferation and growth of BCC of Ptch mutant mice in vitro and in vivo. This is accompanied by the activation of the vitamin D receptor (Vdr) and induction of BCC differentiation. In addition, calcitriol inhibits Hh signaling at the level of Smo in a Vdr-independent manner. The concomitant antiproliferative effects on BCC growth are stronger than those of the Hh-specific inhibitor cyclopamine, even though the latter more efficiently inhibits Hh signaling. Taken together, we show that exogenous supply of calcitriol controls the activity of 2 independent pathways, Hh and Vdr signaling, which are relevant to tumorigenesis and tumor treatment. These data suggest that calcitriol could be a therapeutic option in the treatment of BCC, the most common tumor in humans. Mol Cancer Ther; 10(11); 2179–88. 2011 AACR.

Introduction Today, the inhibition of the Hh pathway is considered to be a promising strategy in the treatment of these tumors. The Hedgehog (Hh)-signaling pathway regulates cell Thus specific Smo inhibitors such as cyclopamine and differentiation, organ patterning, and cell proliferation (1). GDC0449 have been tested in several small, nonrando- In the absence of a ligand, the activity of the Hh-pathway is mized clinical trials (2). GDC0449 was recently shown to inhibited due to an interaction between the Hh receptor elicit antitumoral effects in 18 of 33 patients with locally Patched1 (Ptch) and its partner Smoothened (Smo). Bind- advanced or metastatic BCC (3). It remains to be eluci- ing of Hh to Ptch, inactivating Ptch mutations, or activating dated if these promising results will be confirmed in a Smo mutations may suspend this inhibition, resulting in prospective, randomized, and controlled study and if Hh the transcription of target genes including Gli1 (1). Thus targeting alone will be sufficient. In this context, a recent aberrant (e.g., mutation driven) Hh signaling results in report describes resistance to GDC0449 due to a therapy- tumor formation (2). One prominent example is basal cell associated Smo mutation (4). As with other tumors, ther- carcinoma (BCC), the most frequent tumor entity in apy of those associated with abnormal Hh signaling may humans, with active Hh-signaling due to mutations in Ptch. require targeting of additional signaling pathways. Vitamin D3 and its derivatives (e.g., EB1089; Paricalci- triol) are known to have antitumoral effects on different 1 Authors' Affiliations: Institute of Human Genetics; Departments of cancer types (5) including squamous cell carcinoma (6, 7) 2Genetic Epidemiology, 3Diagnostic Radiology, 4Pathology, 5Nephrology and Rheumatology, and 6Neuropathology; 7Central Animal Facility, Uni- or on hyperproliferative skin diseases such as psoriasis (8, 8 versity of Goettingen, Goettingen; Department of Dermatology, University 9). These effects comprise G0–G1 arrest, cellular differen- 9 of Duesseldorf, Duesseldorf; and Department of Neuropathology, Uni- tiation, induction of apoptosis, and modulation of inflam- versity of Bonn, Bonn, Germany mation or of different signaling pathways in tumor cells, Corresponding Author: Anja Uhmann, Institute of Human Genetics, University of Goettingen, Heinrich-Duker-Weg€ 12, 37073 Goettingen, as well as inhibiting tumor angiogenesis (5). Until now, the Germany. Phone: 49-551-3914100; Fax: 49-551-396580; E-mail: antitumoral effects of vitamin D3 were explained by bind- [email protected] ing of the biologically active form of vitamin D3, calcitriol a doi: 10.1158/1535-7163.MCT-11-0422 (1 ,25-dihydroxy vitamin D3), to the vitamin D receptor 2011 American Association for Cancer Research. (Vdr) and the subsequent regulation of Vdr-bound genes.

www.aacrjournals.org 2179

Uhmann et al.

Ptchflox/flox Calcitriol is produced from vitamin D3 by 2 hydrox- Animals and treatment of tumor-bearing þ/ ylation steps in the liver and kidney, respectively, and ERT2 mice with calcitriol þ to a lesser extent in other organs and in tumor cells (5). Conditional Ptchflox/flox ERT2 / mice were random- Through binding to Vdr, calcitriol regulates the tran- ized into 2 groups and BCC was induced in all animals scription of Vdr target genes (5). This so-called genomic by intramuscular injection of 100 mgtamoxifenasde- calcitriol/Vdr signaling regulates a variety of physio- scribed (13, 16). Tumors of this animal model lack the logic processes including cellular differentiation, espe- expression of wt Ptch alleles and are therefore deficient cially in the skin (10, 11), proliferation, and apoptosis, in Ptch (13, 17). Starting points of the daily intraperito- and can be monitored by measuring transcription of the neal treatment with 100 ng/kg calcitriol or vehicle were calcitriol metabolizing enzyme 24-hydroxylase (5). In day 0 or 60 after BCC induction. Treatment of each addition, calcitriol elicits rapid, so-called nongenomic cohort was conducted until day 90 after BCC induction. (i.e., transcription independent) effects such as calcium For RNA isolation and histologic examinations, skin influx (5). samples were collected from tails 24 hours after the last Recent data suggest a cross-talk between vitamin D3 calcitriol injection. and Hh signaling. Through medium transfer experi- Mice were fed with calcium- and phosphate-reduced ments Bijlsma and colleagues provided first evidence and vitamin D3-free food (ssniff Spezialdiaeten, E15312- for a Ptch-dependent secretion of vitamin D3 com- 14; ref. 18) 1 week before and during the injection pounds. They also showed that vitamin D3 inhibits period. the Hh-pathway at the level of Smo (12). This suggests All animals were treated and housed in accordance that, in addition to or instead of a direct protein–protein with the German animal protection law. interaction, Ptch may repress Smo via secretion of a vitamin D3 derivative (12). Measurement of tumor size The possibility of a Ptch-dependent secretion of a BCC size was measured on hematoxylin and eosin vitamin D3 derivative with Smo-inhibitory properties (H&E)-stained sections using the area calculation tool opens new perspectives for therapies of tumors that arise of the software CellF (Olympus Soft Imaging Solutions due to mutations in Ptch. One would expect that inacti- GmbH). vation of Ptch results in a disrupted secretion of this derivative. This should result in a potential deprivation Analysis of calcium blood serum values m of the vitamin D3 derivative, and in lack of Smo inhibi- Amounts of 100 L of blood were collected from the tion. If this hypothesis is true, the concerted action of retroorbital plexus. Serum calcium concentrations were active Hh and inactive Vdr signaling may be the driving measured using an O-cresolphthaleine–based assay force leading to enhanced cell proliferation, compro- (cobas, Roche Diagnostics GmbH). mised differentiation, and ultimately to tumor formation. It follows that it should be possible to revert or at least to Histopathology and immunohistochemistry impede these processes by administration of the respec- BCC and normal skin from the tail were embedded in tive vitamin D3 derivative. paraffin for histologic analyses or were used for isolation Here, we investigated the effect of the biologically of total RNA. The identity of BCC was confirmed by active vitamin D3 derivative calcitriol on Vdr- and Hh- examination of H&E-stained sections. Paraffin sections signaling, growth, apoptosis, and differentiation of Ptch- were stained using an anti-Ki67- and anti-active caspase deficient BCC cells in vitro and in vivo using the Ptchflox/flox 3-antibody as described (19). þ þ Rosa26CreERT2 / (Ptchflox/flox ERT2 / ) mouse model for BCC (13). In addition, we compared its effects with those Cell lines and primary cell culture of BCC þ of cyclopamine and sought to unravel the molecular The fibroblast cell line Ptchflox/floxERT2 / was estab- þ mechanisms underlying the calcitriol-mediated effects lished from dermis of a Ptchflox/floxERT2 / mouse. Ptch / on Hh signaling. cells are stable Ptch-deficient cells derived from tamox- þ ifen-treated Ptchflox/floxERT2 / fibroblasts. Smo / and Materials and Methods Vdr / fibroblasts have been described in Ma and colleagues (20) and Sun and colleagues (21), respectively. All Compounds fibroblast cell lines were maintained in Dulbecco’s Modi- Calcitriol (Sigma-Aldrich) and cyclopamine (Toronto fied Eagle Medium (DMEM), 10% fetal calf serum, and Research Chemicals Inc.) were dissolved in ethanol 1% penicillin/streptomycine (PS). (EtOH). Final concentrations for in vitro experiments The murine BCC cell line ASZ001 was established from þ are indicated in the respective experiments and corre- UV-induced BCC of Ptchneo12/ mice and maintained as spond to those normally used in cell culture (5, 14, 15). For described (22). All cell lines have been tested and au- in vivo use, calcitriol was diluted individually for each thenticated before using by genotyping PCRs on genomic animal in 20 mL EtOH/1,200 mL sterile sunflower oil DNA as described (13, 20, 21). (Sigma-Aldrich) to obtain a final concentration of 40 or Skin punches were isolated from BCC-bearing þ 100 ng/kg in 50 mL. Ptchflox/floxERT2 / mice 30, 40, and 60 days after tumor

2180 Mol Cancer Ther; 10(11) November 2011 Molecular Cancer Therapeutics

Calcitriol Inhibits Growth of Basal Cell Carcinoma

initiation by tamoxifen. Punches were maintained for and transglutaminase 1 (Tgm1) were 50-GCAAGGTCAA- 10 days in culture as reported (16). GACTGAGGCTGA-30/50-GCTGCTCCTGTGTGTCATA- CTCCTT-30;50-AGAACATGTGCTGCTCATGGC-30/50- Cell culture experiments TCATCTTGGCGTAGAGCTGGTGGCT-30;50-GTGTGG- For gene expression analysis or 5-bromo-20-deoxyuri- CAAGCGCACACGCT-30/50-CCGTGACAGCAGCGTA- dine (BrdU) incorporation and caspase assays, 100,000 or CAGT-30; and 50-GCAGTGGTGTAAATGCAGCTGG-30/ 4,000 cells/well were seeded in 6-well and 96-well plates, 50-ATGAGGAGCTCAAGGGCAATGC-30, respectively. respectively. For short interfering RNA (siRNA)-mediat- Amplification of 18S rRNA was done as an endogenous þ ed knockdown of Vdr expression, Ptchflox/flox ERT2 / control for the normalization of target gene expression. cells were transfected with Vdr-specific or control siRNA The amount of target and endogenous reference was as described below. After 24 hours, the cells were washed determined using the relative standard curve method. and incubated for additional 48 hours with medium Each sample was measured in triplicates. Graphs repre- supplemented with calcitriol, cyclopamine, or EtOH as sent the mean value of all measurements. indicated in the respective experiments. Glycerol-3-phosphate dehydrogenase (mGapd), wild- Cell proliferation was measured after BrdU-pulsing for type (wt) Ptch transcripts (derived from the nonrecom- the last 22 hours using a Cell Proliferation BrdU ELISA bined Ptchflox locus) and Ptchdel transcripts were (Roche Diagnostics GmbH). detected by RT-PCR as described in Zibat and collea- Activity of caspase 3 and 7 was measured using the gues (13, 24). Caspase-Glo 3/7 Assay (Promega) and a microplate reader (Biotek instruments Inc.) according the man- Protein extraction and Western blot ufacturer’s instructions. Ptch / fibroblasts were transfected with Vdr-specific Shh-N–conditioned medium (Shh-N-CM) or respec- siRNA or control siRNA as described. After 72 hours, the tive control medium were obtained from HEK293-Shh cells were harvested. Nuclear fractions of the transfected (HEK293-Shh express one of the 3 mammalian Hh pro- cells and of Vdr / fibroblasts were isolated using the kit teins Sonic hedgehog (Shh)) or HEK293 cells, respective- NE-PER Nuclear and Cytoplasmatic Extraction Reagents ly, as described (14). (Thermo Scientific) according to the manufacturer’s Knockdown of Vdr expression in Ptch / and Ptchflox/flox instructions. Protein concentrations were measured by þ ERT2 / cells was achieved by using a Vdr-specific Pierce BCA Protein Assay Kit (Thermo Scientific) and siRNA (50-CAGGCGGAGCATGAAGCGCAA-30). Scram- 16.5 mg of each protein sample of the nuclear fraction bled siRNA (AllStars negative, Qiagen) was used as con- were analyzed by Western blot using a rabbit anti-Vdr trol siRNA. 75 ng siRNA and 4.5 mL HiPerFect (Qiagen) antibody (1:100; clone 9A7; Abcam) and a mouse anti- were mixed with 100 mL DMEM and incubated for 10 heat shock 70 kDa protein 8 (HSC-70; 1:10,000; Santa minutes. Simultaneously, 15,000 cells/well were seeded Cruz) in combination with horseradish peroxidase-con- in DMEM/10% fetal calf serum/2% penicillin/strepto- jugated goat anti-rat IgG (1:10,000; Thermo Scientific) and mycine in 24-well plates. Subsequently, the siRNA/ rabbit antimouse IgG (1:5,000; GE Healthcare), respec- HiPerFect-mix was added to the cells. After 24 hours, tively. Signals were detected by using ECLþ reagent (GE the cells were incubated with the respective media (as Healthcare). indicated in the experiments) supplemented with vehicle or 10 nmol/L calcitriol. After additional 48 hours, the cells Statistics were collected and used for subsequent experiments. Mann–Whitney U testing was done to determine the For activation of Hh-signaling pathway in Smo / significance of the results. fibroblasts the cells were transfected with a plasmid- expressing hSMO (15, 23) using RotiFect following the Results manufacturer’s instructions. After 6 hours, the cells were incubated with media containing 10 nmol/L calcitriol, 5 Calcitriol activates the Vdr pathway, suppresses mmol/L cyclopamine, or EtOH for 48 hours. Hh signaling and inhibits proliferation of cultured Data shown are representative for at least 3 indepen- Ptch mutant BCC cells þ dent experiments each conducted in triplicate. As human BCC (5, 25–27), BCC from Ptchflox/floxERT2 / mice express elevated Vdr levels compared with normal Reverse transcription PCR and quantitative real- murine skin (Fig. 1A). time PCR analyses To test the response of Ptch-deficient murine BCC cells Total RNA was extracted using TriReagent (Sigma- to calcitriol in vitro, BCC-bearing skin punches were Aldrich). For skin samples, the RNeasy fibrous tissue mini collected 30, 40, or 60 days after BCC initiation in þ kit (Qiagen) was used. Synthesis of cDNA and primer Ptchflox/floxERT2 / mice (see Materials and Methods). combinations for amplification of 18S rRNA, Gli1 and The punches were cultured with 10 nmol/L calcitriol Keratin10 (K10) transcripts used for quantitative real-time or with 10 mmol/L cyclopamine for 10 days (the struc- PCR (qRT-PCR) were described previously (16, 19). Prim- tures of cyclopamine and calcitriol are provided er pairs used for qRT-PCR analysis of Gli2, Vdr, Cyp24a1, in Fig. 1B). Gli1 expression levels served to monitor

www.aacrjournals.org Mol Cancer Ther; 10(11) November 2011 2181

Uhmann et al.

Figure 1. Calcitriol inhibits proliferation and the Hh-signaling pathway of BCC-bearing skin punches and of the BCC cell line ASZ001. A, Vdr expression levels of BCC (n ¼ 3) of Ptchflox/floxERT2þ/ mice 90 days after tumor induction compared with NS (n ¼ 3). B, chemical structures of calcitriol and cyclopamine. C, Gli1 and Cyp21a1 expression levels of cultured BCC-bearing skin punches isolated from Ptchflox/floxERT2þ/ mice 30, 40, or 60 days after BCC induction. D, Ki67þ BCC cells of the respective punches. Gli1 and Cyp21a1 expression levels (E) and BrdU incorporation (F) in ASZ001 cells. G, caspase 3/7 activities of ASZ001 cells. Cells treated with 500 nmol/L staurosporin served as positive controls. The punches and the ASZ001 cells were incubated with vehicle (EtOH), calcitriol, or cyclopamine (CP) as indicated in the text. Gli1 expression and caspase 3/7 activities are shown in relation to the respective vehicle-treated controls. Ki67þ BCC cells and BrdU-incorporation are represented as percentage of respective vehicle-treated controls. **, P < 0.05; error bars: mean SD.

2182 Mol Cancer Ther; 10(11) November 2011 Molecular Cancer Therapeutics

Calcitriol Inhibits Growth of Basal Cell Carcinoma

Hh-signaling activity. Cyp24a1 (24-hydroxylase) tran- without causing toxicity. Tumor areas on H&E-stained scripts were measured to estimate activation of Vdr. skin sections in mice treated on days 0 to 90, but not on Calcitriol led to a significant induction of Cyp24a1 days 60 to 90, were significantly decreased when com- transcription (Fig. 1C), which was consistent with the pared with the vehicle-treated group (Fig. 2B and C). presence of Vdr in BCC (Fig 1A). In addition, calcitriol Furthermore, calcitriol inhibited tumor cell proliferation þ decreased Gli1 expression, thus indicating an inhibition as measured by Ki67 cells in tumors treated on days 0 to of the Hh-signaling pathway. As expected, cyclopamine 90 (Fig. 2D). This was accompanied by a significant did not induce Cyp24a1 transcription but repressed Gli1 decreased expression of Hh-pathway target genes Gli1 expression. The repressive effect was more pronounced and Gli2 (28–30) in BCC treated for days 0 to 90 (Fig. 2E). than that achieved with calcitriol (Fig. 1C). Consistent with the results from our in vitro studies no Next, we assessed the antiproliferative effects of calci- significant increase of caspase 3 positive BCC cells were triol. As revealed by anti-Ki67 antibody staining, calcitirol observed (data not shown). Moreover, calcitriol treatment inhibited proliferation of tumor cells in BCC-bearing skin resulted in a substantial activation of Vdr signaling as punches (Fig. 1D). Most interestingly, the antiprolifera- revealed by an enhanced expression of the immediate tive effect of calcitriol was more pronounced than that Vdr-target gene Cyp24a1 (Fig. 2E). Finally, a significantly achieved with the Hh-specific inhibitor cyclopamine increased expression of the keratinocyte differentiation (Fig. 1D). markers and Vdr-target genes Tgm1 and K10 were Similar results were obtained when the BCC-derived detected (Fig. 2F). Whereas both differentiation markers cell line ASZ001 was incubated with 10 nmol/L calcitriol can be induced by active Vdr signaling, K10 expression or with 10 mmol/L cyclopamine for 48 hours. As shown in also depends on Hh-signaling activity (10, 29, 31–34). Fig. 1E both substances efficiently inhibited Hh signaling. In summary, our data show that calcitriol significantly However, only calcitriol significantly inhibited BrdU- inhibits proliferation and induces cellular differentiation incorporation in these cells (Fig. 1F). Inhibition of prolif- of Ptch-associated BCC in vivo. Moreover, this response is eration apparently was not accompanied by an increase accompanied by activation of Vdr- and inactivation of the in apoptosis, as calcitriol had no effect on caspase 3/7 Hh-signaling pathways. activity (Fig. 1G). Together, these results show that calcitriol efficiently Calcitriol inhibits Hh signaling downstream of Ptch inhibits the Hh-signaling pathway and activates Vdr- but upstream of Gli1 signaling in BCC. In contrast to cyclopamine, only Vitamin D3 has been shown to inhibit Hh signaling at calcitriol mediated antiproliferative effects in BCC, the level of Smo (12, 35, 36). To investigate whether the even though cyclopamine more efficiently inhibits Hh- mechanism of Hh-pathway inhibition mediated by calci- signaling. triol is similar to that of Vitamin D3 we made use of þ Ptchflox/floxERT2 / and Smo / fibroblasts. þ Calcitriol treatment of Ptch mutant mice inhibits the Ptchflox/floxERT2 / fibroblasts normally express wt Ptch Hh-signaling pathway and growth of BCC and transcripts from the Ptchflox locus. Upon tamoxifen-in- stimulates BCC differentiation duced activation of ERT2 recombinase the exons 8 and 9 Next the in vivo antitumoral effects of calcitriol were of the floxed Ptch alleles are excised. This results in the þ tested in the Ptchflox/floxERT2 / mouse model for BCC. In expression of an aberrant, nonfunctional Ptch transcript this model, where all mice develop full-blown BCC (Fig. 3A). þ 90 days after activation of ERT2 recombinase by tamox- In Ptchflox/floxERT2 / fibroblasts both, the treatment ifen (13), treatment can be commenced at a specified time with Shh-N-CM or the tamoxifen-induced Ptch mutation after tumor induction and at a defined age of the animals. result in activation of Hh signaling as revealed by induc- Preliminary studies, in which mice were treated daily tion of Gli1 expression (Fig. 3B). Coincubation of Shh- or þ with 40 ng/kg calcitriol starting at the day of tamoxifen tamoxifen-treated Ptchflox/floxERT2 / fibroblasts with cal- injection for 90 subsequent days (days 0–90), did not citriol significantly inhibited Gli1 expression (Fig. 3B). result in an induction of Vdr signaling or differentiation Similar results were obtained with the validated Smo (data not shown). Therefore, we increased the daily inhibitor cyclopamine (Fig. 3B). These data show that calcitriol dose to 100 ng/kg. Calcitriol treatment was similar to cyclopamine, calcitriol inhibits Hh signaling started either immediately (day 0) or 60 days after downstream of Ptch. ¼ ¼ / BCC initiation (n60–90d 6; n0–90d 4). Vehicle-treated The effects of calcitriol were also tested in Smo cells, ¼ animals (nvehicle 17) served as controls. The treatment which express basal Gli1 levels. Due to lack of Smo, Hh- was continued until day 90, when all mice were sacrificed signaling cannot be activated in these cells by incubation (Fig. 2A). with Shh-conditioned medium (37). However, in these Calcitriol therapy led to a significant increase in serum cells Hh-pathway activity can be restored by transfection calcium concentrations (Fig. 2A) without causing weight with a hSMO expression plasmid. loss, hypercalcemia-driven kidney damage or signs of As shown in Fig. 3C, calcitriol did not influence basal nephrocalcinosis (data not shown). This indicates that the Gli1 expression levels in Smo / cells (Fig. 3C). How- treatment induced calcitriol-specific systemic effects ever, calcitriol efficiently inhibited Hh signaling after

www.aacrjournals.org Mol Cancer Ther; 10(11) November 2011 2183

Uhmann et al.

Figure 2. Calcitriol treatment of BCC-bearing mice inhibits tumor growth and the Hh-signaling pathway, and induces expression of differentiation markers. A, treatment scheme and calcium serum levels of BCC-bearing Ptchflox/floxERT2þ/ mice. BCC were induced in Ptchflox/floxERT2þ/ mice by tamoxifen. Treatment with 100 ng/kg/d calcitriol started either directly (0–90) or 60 (60–90) days after tumor induction and was continued until day 90 when all mice were sacrificed. Serum calcium levels of the respective BCC-bearing Ptchflox/flox ERT2þ/ mice were measured before (day 0) and during therapy. Serum

calcium levels of vehicle-treated mice served as controls (nvehicle ¼ 17; n60–90d ¼ 6; n0–90d ¼ 4). H&E stainings (B), relative tumor areas of H&E-stained BCC (C) and percentage of Ki67þ BCC cells (D) of vehicle- or calcitriol-treated mice. Gli1, Gli2, Cyp24a1 (E) and Tgm1 and K10 expression levels (F) from vehicle or calcitriol-treated BCC-bearing mice. Values of vehicle-treated controls for Gli1, Cyp24a1, Tgm1, and K10 expression were set to 1. *, P ¼ 0.06; **, P < 0.05; error bars: mean SD.

restoration of Hh-pathway activity upon transfection expression after calcitriol treatment in comparison with with hSMO. These results are similar to those obtained the controls (Fig. 4C, D). Whereas Vdr knockdown per se with cyclopamine (Fig. 3C) and show that calcitriol did not significantly attenuate Hh-pathway activity, normally inhibits Gli1 expression at the level of Smo. treatment with calcitriol resulted in a downregulation of Gli1 expression (Fig. 4C and D). These data show that Calcitriol inhibits Hh signaling in a calcitriol-mediated inhibition of the Hh-signaling path- Vdr-independent manner way occurs independently of the Vdr. Next, we tested whether calcitriol-mediated inhibition Taken together, these data show that calcitriol inhibits of the Hh-signaling pathway is independent of Vdr- Hh-pathway activity downstream of Ptch at the level of signaling. For this purpose, we used Vdr / fibroblast Smo in a Vdr-independent manner. (21). In these cells Vdr signaling is completely abrogated, since incubation with calcitriol does not result in the Discussion induction of the Vdr-target Cyp24a1 (Fig. 4A). As revealed by Gli1 expression, incubation of the cells with Shh-N- The active form of vitamin D3, calcitriol, is rapidly CM resulted in induction of Hh-pathway activity gaining importance in oncology due to its antiprolifera- (Fig. 4A), which was significantly inhibited by addition tive and differentiation-inducing effects combined with of calcitriol or by the control substance cyclopamine its low toxicity (5). (Fig. 4A). Similar results were achieved in Ptch / or Our work extends the range of calcitriol tumor targets þ Shh-N-CM–treated Ptchflox/flox ERT2 / cells after to those induced by an abnormal activity of the Hh- siRNA-mediated Vdr knockdown. Efficient downregula- signaling pathway (summarized in Fig. 5). Specifically, tion of Vdr expression was verified by Western blot we show that calcitriol inhibits growth of BCC induced by (Fig. 4B), qRT-PCR and by a significantly lower Cyp24a1 deletion of the Hh receptor Ptch. This is accompanied by

2184 Mol Cancer Ther; 10(11) November 2011 Molecular Cancer Therapeutics

Calcitriol Inhibits Growth of Basal Cell Carcinoma

the expression of relevant cell differentiation markers. Our data suggest that calcitriol could be a valuable supplement or even alternative to the established treat- ments of BCC, the most common tumor in humans, associated with aberrant Hh-pathway activity. As assessed by reduced Gli1 transcription, calcitriol inhibits canonical Hh signaling independently of Vdr signaling and downstream of Ptch (Fig. 5). An obvious molecular target of this Vdr-independent effect of calcitriol is Smo, because Smo-deficient cells (unlike those reconstituted with Smo or Ptch-deficient ones) show no decreased Gli1 transcription in response to this substance. A similar observation has been made for the inactive form of calcitriol, vitamin D3 (12). Nevertheless, whether calcitriol directly binds to Smo should be addressed in future studies. Besides inhibition of Hh-signaling pathway, calcitriol inhibits proliferation, and increases the expression of skin differentiation marker in BCC. The latter effects are also induced by calcitriol in skin of patients with hyperproliferative skin diseases such as psoriasis (6–9, 34, 38). Calcitriol treatment specifically results in inhibition of proliferation of psoriatic skin and induction of differentiation of keratinocytes (9, 39, 40). Similarly to BCC, psoriatic skin also expresses high levels of the Hh-target gene Gli1 (41). These facts raise the question whether calcitriol mediates its antiproliferative effects and differentiation stimuli via activation of Vdr signaling or rather via inhibition of the Hh-signaling pathway. A clue might come from our present study: Our in vitro study shows that calcitriol has a significantly stronger antiproliferative effect on BCC than the pure Smo inhibitor cyclopamine, even though the latter substance inhibits Hh signaling more efficiently (Fig. 1D and F). Together with the fact that both calcitriol (present study) and cyclopamine (42) inhibit BCC proliferation in vivo,itispossiblethatcalcitriolexertsitsantiproli- ferative effects via both signaling pathways (e.g., by inhibition or activation of Hh or Vdr signaling, respectively). On the other hand, differentiation of calcitriol- treated BCC is probably unrelated to inhibition of the canonical Hh-signaling pathway, because cylopamine never has been reported to induce the expression of keratinocyte differentiation markers in BCC. More likely, calcitriol induces BCC differentiation via Vdr signaling, which is supported by the increased expression of the Vdr-target genes Cyp24a1 and Tgm1 (Fig. 2F). Antiproliferative and Hh-signaling inhibitory proper- Figure 3. Calcitriol inhibits the Hh-signaling pathway downstream of Ptch ties have also been described for the inactive form of flox/flox but upstream of Gli1. A, Ptch ablation in tamoxifen-treated Ptch calcitriol, vitamin D3, on murine BCC in vitro and in vivo ERT2þ/ fibroblasts after incubation with 10 mmol/L tamoxifen for 48 (35). In contrast to calcitriol, vitamin D3 did not affect hours. Wt Ptch transcripts (derived from the nonrecombined Ptchflox locus) and Ptchdel transcripts were detected by PCR using the primer pair mPtc11/mPtc7R as described in Zibat and colleagues (13, 24). cDNA derived from 12.5 days old wild-type mouse embryos (E12.5 Ptchþ/þ) was cultured in either Shh-N-CM or control medium (CoM) with or without 10 used as a control. Gli1 and Cyp24a1 expression levels of Ptchflox/flox mmol/L tamoxifen as indicated. Smo / fibroblasts were either transfected ERT2þ/ (B) and Smo/ (C) fibroblasts after treatment with vehicle (EtOH), with a hSMO expressing plasmid or with empty vector as indicated. calcitriol, or cyclopamine (CP). Ptchflox/floxERT2þ/ fibroblasts were **, P < 0.05; error bars, mean SD; ntc, no template control.

www.aacrjournals.org Mol Cancer Ther; 10(11) November 2011 2185

Uhmann et al.

Figure 4. Calcitriol inhibits the Hh-signaling pathway in a Vdr-independent manner. A, Gli and Cyp24a1 expression of Vdr/ fibroblasts after treatment with vehicle (EtOH), calcitriol, or cyclopamine (CP) and Shh-N-CM or control medium (CoM). B, analyses of the Vdr protein level of nuclear extracts of si-Vdr or si- control transfected Ptch/ and Vdr/ fibroblasts by Western blot. Detection of HSC-70 protein served as control. C and D, Gli1, Cyp24a1, and Vdr expression of Ptch/ and Shh-stimulated Ptchflox/floxERT2þ/ fibroblasts after si-Vdr or si-control transfection and calcitriol treatment. Gli1 expression of Ptch/ fibroblasts and Vdr expression of Ptch/ and Ptchflox/flox ERT2þ/ cells are shown in relation to the respective vehicle-treated control. **, P < 0.05; error bars: mean SD.

differentiation in BCC, although it was topically applied This suggestion is based on a comparison with our study at high concentrations. Supposedly the treatment period in which a 30 days calcitriol application also had no using vitamin D3 (i.e., 30 days) may have been too short to significant effects on these processes (see BCC treated induce Vdr signaling and thus a differentiation response. for days 60 to 90, Fig. 2B and E).

Figure 5. Model for the dual function of calcitriol in Ptch-associated BCC. Normally Ptch inhibits its signaling partner Smo, thereby regulating the activity of the Hh-signaling pathway (normal cell). Mutations of Ptch lead to a constitutive activation of the Gli transcription factors, which resulted in cell proliferation and tumor formation (BCC cell). The known Smo-inhibitor cyclopamine inhibits Hh-pathway in the Ptch-mutant cells and thus Hh-related processes involved in tumor growth (cyclopamine-treated BCC cell). In contrast, calcitriol inhibits Hh-pathway activity and additionally activates Vdr signaling (calcitriol-treated BCC cell). Consequently, calcitriol not only inhibits tumor-relevant processes mediated by Hh signaling, but also induces antiproliferative effects and differentiation processes via the Vdr-signaling pathway.

2186 Mol Cancer Ther; 10(11) November 2011 Molecular Cancer Therapeutics

Calcitriol Inhibits Growth of Basal Cell Carcinoma

Finally, a new model of tumorigenesis driven by Ptch- Disclosure of Potential Conflicts of Interest deficiency may emerge from our study. According to a recent work, Ptch might function as an efflux pump for No potential conflicts of interest were disclosed. vitamin D -related compounds with Hh-inhibitory prop- 3 Acknowledgments erties (12). A deficiency of this compound due to Ptch inactivation would pathologically activate Hh-pathway We thank Stefan Wolf and Susan Peter for excellent animal care. We and reduce Vdr signaling. Vice versa application of this also thank J. Taipale, University of Helsinki, Finland for provision of compound should result in inhibition of Hh-pathway and Smo / fibroblasts, E. Epstein, UCSF, for the BCC cell line ASZ001, R. Toftgard, Karolinska Institute, Huddinge, Sweden for hSMO plasmid, activation of Vdr signaling (Fig. 5). Whether this vitamin Steven Johnsen, University of Gottingen,€ Germany for proofreading and D3-related compound is calcitriol remains to be analyzed for comments on the manuscript, and Leszek Wojnoswki, University of in the future. Mainz, Germany for comments on the manuscript. Taken together the application of calcitriol holds pro- Grant Support mises as an effective anticancer drug in the treatment of

BCC. Due to its dual effects on both Vdr and Hh signal- This work was supported by grants of the Deutsche Forschungsgemeinschaft ing, it may be superior to substances that solely target the UH 228/2-1 and UH228/2-2 to A. Uhmann and HA 2197/5-1 FOR942 to H. Hahn. Hh-signaling pathway. Calcitriol treatment may also be The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in superior to application of vitamin D3, which has to be accordance with 18 U.S.C. Section 1734 solely to indicate this fact. metabolized before activating Vdr signaling. The benefits of topical application of calcitriol in treatment of BCC Received June 8, 2011; revised August 3, 2011; accepted August 24, 2011; have to be tested in the future. published OnlineFirst August 30, 2011.

References 1. Hooper JE, Scott MP. Communicating with Hedgehogs. Nat Rev Mol 15. Taipale J, Chen JK, Cooper MK, Wang B, Mann RK, Milenkovic L, et al. Cell Biol 2005;6:306–17. Effects of oncogenic mutations in Smoothened and Patched can be 2. Yang L, Xie G, Fan Q, Xie J. Activation of the hedgehog-signaling reversed by cyclopamine. Nature 2000;406:1005–9. pathway in human cancer and the clinical implications. Oncogene 16. Nitzki F, Zibat A, Konig€ S, Wijgerde M, Rosenberger A, Brembeck FH, 2009;29:469–81. et al. Tumor stroma-derived Wnt5a induces differentiation of basal cell 3. Von Hoff DD, LoRusso PM, Rudin CM, Reddy JC, Yauch RL, Tibes R, carcinoma of Ptch mutant mice via CaMKII. Cancer Res 2010;70: et al. Inhibition of the hedgehog pathway in advanced basal-cell 2739–48. carcinoma. N Engl J Med 2009;361:1164–72. 17. Calzada-Wack J, Kappler R, Schnitzbauer U, Richter T, Nathrath M, 4. Yauch RL, Dijkgraaf GJ, Alicke B, Januario T, Ahn CP, Holcomb T, Rosemann M, et al. Unbalanced overexpression of the mutant allele in et al. Smoothened mutation confers resistance to a Hedgehog path- murine Patched mutants. Carcinogenesis 2002;23:727–34. way inhibitor in medulloblastoma. Science 2009;326:572–4. 18. Banach-Petrosky W, Ouyang X, Gao H, Nader K, Ji Y, Suh N, et al. 5. Deeb KK, Trump DL, Johnson CS. Vitamin D signalling pathways in Vitamin D inhibits the formation of prostatic intraepithelial neoplasia in cancer: potential for anticancer therapeutics. Nat Rev Cancer Nkx3.1;Pten mutant mice. Clin Cancer Res 2006;12:5895–901. 2007;7:684–700. 19. Ecke I, Rosenberger A, Obenauer S, Dullin C, Aberger F, Kimmina S, 6. Hershberger PA, Modzelewski RA, Shurin ZR, Rueger RM, Trump DL, et al. Cyclopamine treatment of full-blown Hh/Ptch-associated RMS Johnson CS. 1,25-Dihydroxycholecalciferol (1,25-D3) inhibits the partially inhibits Hh/Ptch signaling, but not tumor growth. Mol Carci- growth of squamous cell carcinoma and down-modulates p21 nog 2008;47:361–72. (Waf1/Cip1) in vitro and in vivo. Cancer Res 1999;59:2644–9. 20. Ma Y, Erkner A, Gong R, Yao S, Taipale J, Basler K, et al. Hedgehog- 7. Majewski S, Skopinska M, Bollag W, Jablonska S. Combination of mediated patterning of the mammalian embryo requires transporter- isotretinoin and calcitriol for precancerous and cancerous skin like function of dispatched. Cell 2002;111:63–75. lesions. Lancet 1994;344:1510–1. 21. Sun J, Kong J, Duan Y, Szeto FL, Liao A, Madara JL, et al. Increased 8. Stewart DG, Lewis HM. Vitamin D analogues and psoriasis. J Clin NF-kappaB activity in fibroblasts lacking the vitamin D receptor. Am J Pharm Ther 1996;21:143–8. Physiol Endocrinol Metab 2006;291:E315–22. 9. van de Kerkhof PC. An update on vitamin D3 analogues in the treatment 22. Xie J, Aszterbaum M, Zhang X, Bonifas JM, Zachary C, Epstein E, et al. of psoriasis. Skin Pharmacol Appl Skin Physiol 1998;11:2–10. A role of PDGFRalpha in basal cell carcinoma proliferation. Proc Natl 10. Bikle DD. Vitamin D and the skin. J Bone Miner Metab 2010;28: Acad Sci U S A 2001;98:9255–9. 117–30. 23. Xie J, Murone M, Luoh SM, Ryan A, Gu Q, Zhang C, et al. Activating 11. Haussler MR, Whitfield GK, Haussler CA, Hsieh JC, Thompson PD, Smoothened mutations in sporadic basal-cell carcinoma. Nature Selznick SH, et al. The nuclear vitamin D receptor: biological and 1998;391:90–2. molecular regulatory properties revealed. J Bone Miner Res 1998; 24. Uhmann A, Dittmann K, Nitzki F, Dressel R, Koleva M, Frommhold A, 13:325–49. et al. The Hedgehog receptor Patched controls lymphoid lineage 12. Bijlsma MF, Spek CA, Zivkovic D, van de Water S, Rezaee F, Peppe- commitment. Blood 2007;110:1814–23. lenbosch MP. Repression of smoothened by patched-dependent 25. Mitschele T, Diesel B, Friedrich M, Meineke V, Maas RM, Gartner BC, (pro-)vitamin D3 secretion. PLoS Biol 2006;4:e232. et al. Analysis of the vitamin D system in basal cell carcinomas (BCCs). 13. Zibat A, Uhmann A, Nitzki F, Wijgerde M, Frommhold A, Heller T, et al. Lab Invest 2004;84:693–702. Time-point and dosage of gene inactivation determine the tumor 26. Kamradt J, Rafi L, Mitschele T, Meineke V, Gartner BC, Wolfgang T, spectrum in conditional Ptch knockouts. Carcinogenesis 2009;30: et al. Analysis of the vitamin D system in cutaneous malignancies. 918–26. Recent Results Cancer Res 2003;164:259–69. 14. Chen JK, Taipale J, Young KE, Maiti T, Beachy PA. Small molecule 27. Reichrath J, Kamradt J, Zhu XH, Kong XF, Tilgen W, Holick MF. modulation of Smoothened activity. Proc Natl Acad Sci U S A 2002; Analysis of 1,25-dihydroxyvitamin D(3) receptors (VDR) in basal cell 99:14071–6. carcinomas. Am J Pathol 1999;155:583–9.

www.aacrjournals.org Mol Cancer Ther; 10(11) November 2011 2187

Uhmann et al.

28. Ikram MS, Neill GW, Regl G, Eichberger T, Frischauf AM, Aberger F, 34. Bikle DD. The vitamin D receptor: a tumor suppressor in skin. Discov et al. GLI2 is expressed in normal human epidermis and BCC and Med 2011;11:7–17. induces GLI1 expression by binding to its promoter. J Invest Dermatol 35. Tang JY, Xiao TZ, Oda Y, Chang KS, Shpall E, Wu A, et al. Vitamin d3 2004;122:1503–9. inhibits hedgehog signaling and proliferation in murine Basal cell 29. Regl G, Kasper M, Schnidar H, Eichberger T, Neill GW, Ikram MS, et al. carcinomas. Cancer Prev Res (Phila) 2011;4:744–51. The zinc-finger transcription factor GLI2 antagonizes contact inhibi- 36. Bruggemann LW, Queiroz KC, Zamani K, van Straaten A, Spek CA, tion and differentiation of human epidermal cells. Oncogene Bijlsma MF. Assessing the efficacy of the hedgehog pathway inhibitor 2004;23:1263–74. vitamin D3 in a murine xenograft model for pancreatic cancer. Cancer 30. Regl G, Neill GW, Eichberger T, Kasper M, Ikram MS, Koller J, et al. Biol Ther 2010;10:79–88. Human GLI2 and GLI1 are part of a positive feedback mechanism in 37. Varjosalo M, Li SP, Taipale J. Divergence of hedgehog signal trans- basal cell carcinoma. Oncogene 2002;21:5529–39. duction mechanism between Drosophila and mammals. Dev Cell 31. Stark HJ, Breitkreutz D, Limat A, Ryle CM, Roop D, Leigh I, et al. 2006;10:177–86. Keratins 1 and 10 or homologues as regular constituents of inner root 38. Bikle DD. Vitamin D and skin cancer. J Nutr 2004;134;3472–8S. sheath and cuticle cells in the human hair follicle. Eur J Cell Biol 39. Bikle DD. 1,25(OH)2D3-modulated calcium induced keratinocyte dif- 1990;52:359–72. ferentiation. J Investig Dermatol Symp Proc 1996;1:22–7. 32. Beck B, Lehen'kyi V, Roudbaraki M, Flourakis M, Charveron M, 40. Gottlieb AB. Therapeutic options in the treatment of psoriasis and Bordat P, et al. TRPC channels determine human keratinocyte atopic dermatitis. J Am Acad Dermatol 2005;53:S3–16. differentiation: new insight into basal cell carcinoma. Cell Calcium 41. Endo H, Momota Y, Oikawa A, Shinkai H. Psoriatic skin expresses the 2008;43:492–505. transcription factor Gli1: possible contribution of decreased neurofi- 33. Lee SC, Ikai K, Ando Y, Imamura S. Effects of 1 alpha, 25- bromin expression. Br J Dermatol 2006;154:619–23. dihydroxyvitaminD3onthetransglutaminase activity of trans- 42. Tas S, Avci O. Rapid clearance of psoriatic skin lesions induced by formed mouse epidermal cells in culture. J Dermatol 1989;16: topical cyclopamine. A preliminary proof of concept study. Derma- 7–11. tology 2004;209:126–31.

2188 Mol Cancer Ther; 10(11) November 2011 Molecular Cancer Therapeutics

Antitumoral Effects of Calcitriol in Basal Cell Carcinomas Involve Inhibition of Hedgehog Signaling and Induction of Vitamin D Receptor Signaling and Differentiation

Anja Uhmann, Hannah Niemann, Bérénice Lammering, et al.

Mol Cancer Ther 2011;10:2179-2188. Published OnlineFirst August 30, 2011.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-11-0422

Cited articles This article cites 42 articles, 8 of which you can access for free at: http://mct.aacrjournals.org/content/10/11/2179.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/10/11/2179.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/10/11/2179. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.